Circulating Serum miRNAs as Diagnostic Markers for Colorectal Cancer

Aim The study was designed to assess the possibility of using circulating miRNAs (serum miRNAs) as diagnostic biomarkers in colorectal cancer (CRC) and to identify their possibility as candidates for targeted therapy. Methods The study involved two sample sets: 1- a training set which included 90 patients with colorectal related disease (30 with CRC, 18 with inflammatory bowel disease (IBD), 18 with colonic polyps (CP) and 24 with different colonic symptoms but without any colonoscopic abnormality who were enrolled as control group) and 2- a validation set which included 100 CRC patients. Serum miRNAs were extracted from all subjects to assess the expression profiles for the following miRNAs (miR-17, miR-18a, miR-19a, miR-19b, miR-20a, miR-21, miR-146a, miR-223, miR-24, miR-454, miR-183, miR-135a, miR- 135b and miR- 92a) using the custom miScript miRNA PCR-based sybergreen array. The area under the receiver operating characteristic curve (AUC) was used to evaluate the diagnostic performance of the studied miRNAs for colorectal cancer diagnosis. Results Data analysis of miRNA from the training set showed that; compared to control group, only miR-19b was significantly up-regulated in patients with IBD group (fold change = 5.24, p = 0.016), whereas in patients with colonic polyps, miR-18a was significantly up-regulated (fold change = 3.49, p-value = 0.018). On the other hand, miR-17, miR-19a, miR-20a and miR-223 were significantly up-regulated (fold change = 2.35, 3.07, 2.38 and 10.35; respectively and p-value = 0.02, 0.015, 0.017 and 0.016; respectively in CRC patients. However, the validation set showed that only miR-223 was significantly up-regulated in CRC patients (fold change = 4.06, p-value = 0.04). Conclusion Aberrant miRNA expressions are highly involved in the cascade of colorectal carcinogenesis. We have found that (miR-17, miR-19a, miR-20a and miR-223) could be used as diagnostic biomarkers for CRC. On the other hand, miR-19b and miR-18a could be used as diagnostic biomarkers for CP and IBD respectively.


Introduction
Colorectal cancer (CRC) is one of the most common malignant neoplasms worldwide, being the second in females and the third in males with 1.2 million annual new cases worldwide [1]. Despite the increased awareness as well as improved screening recommendations and techniques, CRC remains the second leading cause of cancer-related death in males and females [2] and being responsible for 10% of the cancer-related mortality world-wide [3].
It has been previously addressed that patients with inflammatory bowel disease (IBD) are usually associated with an increased risk of progression to epithelial dysplasia and CRC [4,5].
The miRNAs represent an interesting class of small (18-25 nucleotides long) noncoding RNAs that act as posttranscriptional regulators of gene expression through binding to the 3`untranslated regions (UTR) of the target mRNAs and promoting mRNA degradation or translational repression [6]. The fact that miRNAs play a role in cancer biology was supported by finding that more than 50% of the miRNA genes are located at the fragile sites and regions of deletion or amplifications, which are altered in different types of human cancer [7]. Cumulative evidence indicates that some miRNAs can behave either as oncomirs or tumor suppressor genes in the cascade of CRC.Therefore, they possess the potentiality to be used as diagnostic, prognostic or therapeutic tumor markers [8].
Different studies have reported significant changes of miRNA expression levels in CRC tissues compared to normal colonic epithelium, and identified groups of miRNAs that enable prognostic stratification of CRC patients [9]. In this context, increased expression of many miRNAs, which mediate cell growth and tumor progression, was reported in the blood and/or tissues of CRC cases using a miRNA microarray assay that included miR-20, miR-21, miR-17-5p, miR-15b, miR-181b, miR-191 and miR-200c [10,11]. The presence of miRNAs in serum, plasma and other body fluids such as urine, saliva, and amniotic fluid encouraged miRNAs research since this facilitates their detection and makes them ideal candidates as non-invasive biomarkers for early detection and monitoring disease progression [12].
The aim of the current study is to assess the role of aberrant miRNAs expressions in the development and progression of CRC cases. This was accomplished through studying the expression levels of 14 miRNAs at different stages of colorectal carcinogenesis cascade. The rational for selection of studied 14 miRNAs was based on prior references which illustrated their role in colorectal carcinogenesis (S1 Table).
the period from January 2011 to March 2012. Based on colonoscopic results and histopathological examination of the studied cases, patients were classified into four groups; 30 patients with CRC, 18 with inflammatory bowel disease (IBD), 18 with colonic polyps (CP) and 24 with different colonic symptoms but without any colonoscopic abnormality who served as a control group. The validation set included 100 CRC cases, which were diagnosed and treated the National Cancer Institute, Cairo University during the period from January 2013 to March 2014. A written informed consent was obtained from each patient after the approval of the ethical committees of the NCI (

Serum samples collection
From each patient and control subject, 5ml of venous blood were obtained, centrifuged and the serum was separated. All serum samples were stored at -80°C until used.

Statistical analysis
The relative expression of miRNAs was analyzed using SABiosciences software and the p-value was calculated based on a Student's t-test of the replicate 2^(-Delta Ct) values for each gene in the control and treatment groups. P-value less than 0.05 was considered significant. Receiveroperating-characteristic (ROC) curves and the area under the curve (AUC) were used to assess the diagnostic accuracy of serum miRNAs. The relevant clinicopathological features were determined by one way ANOVA for quantitative data and X 2 test for qualitative data. All tests were 2-sided and p-value <0.05 was considered statistically significant. Statistical analysis was relied on SPSS 17.0 software (SPSS Ltd., UK) and graphs were generated with Graph pad Prism 5.0 (Graph pad Software Inc, USA).

Clinical features of the studied groups
Regarding the training set, the mean age of the CRC patients (51.10± 10) was significantly higher than other groups (p-value<0.05). However, no significant difference was reported between the mean age of the control group (43.7 ±14.2) and those with CP (40.00 ± 15.0) and IBD (41.6 ± 15.4) or the sex for all groups (p-value >0.48). Additionally, there were statistically significant differences between the studied groups in relation to abdominal pain, mucus, loss of weight, and anemia (p-value <0.05). Regarding the validation set, there was no significant difference between the mean age of CRC group (46.7± 14.5) and the control group (43.7 ±14.2) or the sex between the two groups. However, there were statistically significant differences between the CRC group and the control group in relation to abdominal pain, mucus, loss of weight, and anemia (p-value <0.05) as shown in (Table 1).

Differential expression of the studied miRNAs in the investigated groups compared to the control group in the training set
Among all studied miRNAs, only miR-19b was significantly up-regulated in the IBD group (fold change = 5.24, p-value = 0.016) (Fig 1 and Table 2),whereas miR-18a was significantly up regulated in the CP group (fold change = 3.49, p-value = 0.018) (Fig 2 and Table 3) and miR-17, miR-19a, miR-20a and miR-223 were significantly up-regulated in the CRC group (fold change = 2.35, 3.07, 2.38 & 10.35 and p-value = 0.021, 0.015, 0.017& 0.0166 respectively) as shown in (Fig 3 and Table 4).
Upon comparing the CRC group (malignant) versus the other groups (non-malignant); only miR-223 was significantly up regulated (fold change = 4.49, p-value = 0.002) as shown in (Table 5).
However, only miR-135b and miR-454 were significantly down regulated in the CRC group compared to the IBD group (fold change = 0.24 & 0.44 and p-value = 0.023 & 0.018 respectively) as shown in S2 Table. While miR-135a and miR-135b were significantly down-regulated in the CP group compared to the IBD group (fold change = 0.12, 0.09 and p-value = 0.019 and 0.018 respectively) as shown in S3 Table. Differential expression of the studied miRNAs in the CRC group compared to the control group in the validation set Among all studied miRNAs, only miR-223 was significantly up-regulated in the CRC group (fold change = 4.06, p-value = 0.04) as shown in (  Table. On the other hand, only miR-17 was significantly up-regulated in the male CRC patients compared to the control group (fold change = 2.1, p-value = 0.03) as shown in S5 Table. ROC curve analysis of differentially expressed miRNAs For discriminating the CRC group from the control group; we have found that miR-223 (AUC 0.838), miR-17 (AUC 0.813), miR-19a (AUC 0.825) possessed higher diagnostic performance than miR-20a (AUC 0.788). On the other hand, miR-18a (AUC 0.8) showed good diagnostic performance in discriminating the IBD group from the control group. Also, miR-19b (AUC 0.875) showed good diagnostic performance in the discriminating the CP group from the control group as shown in (Table 7).

Discussion
Our study assessed the possibility of using different circulating miRNAs as serological biomarkers for detection of CRC. Our study was properly designed to include cases at different stages of colorectal related diseases i.e. patients with inflammatory bowel disease (IBD), adenomatous colonic polyps (CP) and CRC. Our results revealed that there was significant up regulation of miR-17, miR-19a, miR-20a and miR-223 in the CRC group compared to the control group. The miR-17, miR-20a and miR-19a are members of the polycistronic miR17-92 cluster (miR-17, miR-18a, miR-19a, miR-19b, miR-20a, and miR-92a) which located on 13q31.3; a locus that is frequently amplified in colorectal cancer [13,14]. Our finding regarding significant up-regulation of miR-20a came in accordance with the previous results by Xu et al., who found that up-regulation of miR-20 was associated with tumor invasion and lymph node metastasis of HCC patients [15]. So, miR-20a may possess a potential tumor oncogenic function in CRC. In addition, miR-20a promotes epithelial to mesenchymal transition (EMT), a key step in tumor metastasis, through targeting the tumor suppressor gene (tissue inhibitor matrix metalloproteinase 3 (TIMP3)); negative regulator of matrix metalloproteinase 2 (MMP2) and matrix metalloproteinase 9 (MMP9) that mediated extracellular matrix (ECM) degradation [15]. Also, miR-20a targets the activating members of the E2F family, including cell cycle and apoptosis regulatory genes, leading to G1-S phase transition. However, high levels of E2F proteins, especially E2F1, can induce apoptosis which acts as an emergency brake for over proliferating cells. The oncogenic function of miR-20a was proposed to hinder this emergency brake function of E2F factors and hence stimulate tumor growth [16]. Moreover, miR-20a controls cell cycle progression through targeting cyclin-dependent kinase inhibitors, CDKN1A/p21; important regulator of G2/S transition [17]. Furthermore, miR-20a targets tumor suppressor PTEN; which is a negative regulator of oncogenic prosurvival PI3K\AKT pathway [18,19]. Our data regarding the upregulation of miR-17a and miR-19a in CRC patients are comparable to previous reports in literature [9,[16][17][18][19]. Similarly to miR-20a, miR-17a was shown to promote cell cycle progression through targeting CDKN1A/p21 and E2F [16,17], whereas the oncogenic potential of miR-19a is achieved mainly through targeting the PTEN; thereby promoting the oncogenic pro-survival pathway [9,[18][19][20]. Our data regarding miR-223 over expression in CRC group are concordant with those of Huang et al who assessed the expression levels of miRNAs in the plasma of 100 CRC patients and in 37 patients with colonic adenomas compared to 59 healthy individuals. Among their panel, 12 miRNAs were up regulated, including miR-223 [21]. Similarly, Li et al. reported that the overexpression of miR-223 in the neoplastic colonic tissues compared to the adjacent noncancerous ones [22]. The oncogenic function of miR-223 in CRC was achieved through targeting FOXO1 [23] and RASA1 [24]. FOXO1 is a member of FOX transcription factors-O subfamily. It regulates the transcription activity of various genes including p21, and thus inhibits cell-cycle progression [25,26].On the other hand, RASA1, a RAS signaling terminator, is associated with tumor progression [27]. Moreover, Li et al. showed that miR-223 could be considered a poor prognostic marker since it's over expression correlates significantly with high TNM stage, nodal and distant metastases as well as with reduced overall survival rates [22]. Our data also showed that significant up regulation of miR-18a in patients with CP compared to the control group. Our results supported a previous report by Wu et al., [28] who demonstrated that miR-18a plays an oncogenic role in colorectal carcinogenesis through down regulation of the Ataxia telangiectasia mutated gene (ATM); a key enzyme in the repair of DNA double-strand breaks. In addition, it phosphorylates several key proteins that initiate activation of the DNA damage checkpoint, leading to cell cycle arrest, DNA repair or apoptosis [29].
As for miR-19b, significant up regulation was reported in patients with IBD compared to the control group. Jiang et al. showed that both miR17 and miR-19b, belonging to miR-17-92 cluster, inhibit Treg cell differentiation and promote Th1 responses [30]. Th1 is responsible for the pro inflammatory response via secreting of IL12, IL2, IL6, IFN and TNF [31]. The latter (TNF) plays a vital role in the inflammation process through activation of NF-kB pathway; a prototypical pro inflammatory signaling pathway [32]. The possible mechanistic role of significant miRNAs in the CRC, IBD and colonic polyps groups compared to the control group and their downstream targets were illustrated in the suggested cartoon (Fig 4). Red star indicates the significant miRNA in CRC while yellow star indicates significant miRNA in IBD and green star indicates miRNA in colonic polyps. The arrow indicates the activation while the circle at the end of line indicates inhibition.
Further analysis of the data using (ROC) analysis curve and the corresponding area under the curve were attempted for the significant miRNAs in each pairwise comparison to investigate their diagnostic accuracy. ROC curve analysis of individual miRNAs revealed that miR-17, miR-19a, miR-20a & miR-223 can effectively differentiate between the CRC group and the control group. However, miR-18a showed good diagnostic accuracy in discriminating the CP group from the control group. On the other hand, miR-19b showed good diagnostic accuracy in discriminating the IBD group from the control group.

In conclusion
Aberrant miRNA expressions are highly involved in the cascade of colorectal carcinogenesis. We have found that (miR-17, miR-19a, miR-20a and miR-223) could be used as diagnostic biomarkers for CRC. On the other hand, miR-19b and miR-18a could be used as diagnostic biomarkers for CP and IBD respectively.