However


However, these findings can only be clinically significant if clinical samples corroborate these findings. Therefore, data provided by TCGA in relation to miR-126-3p with gene expression, and Gleason score was analysed. The TCGA is formed in collaboration with National Cancer Institute (NCI) and National Human Genome Research Institute (NHGRI). The dataset was created from >11,000 patients and has been used to create genomic datasets for 33 forms of cancer, including PCa, to be used as a clinical data source for researchers (cancergenome.nih.gov, 2018; Leung et al., 2014). Figure 4 shows the statistically significant inverse correlation between miR-126-3p expression with SOX2 and BCL2, supporting the findings shown in in vitro analysis (Figure 3). The results from SOX2 analysis is in support with the sole other research article by Fujii et al. (2015) which finds a negative correlation between miR-126-3p and SOX2 in cell lines. However, this study is the first to analyse miR-126-3p regulation of SOX2 in clinical samples which validates the potential clinical use of these findings. Furthermore, miR-126-3p regulation on BCL2 has not been investigated previously in vitro or on clinical samples, indicating this as a novel finding and highlighting the need for further research into the role of miR-126-3p on BCL2 expression as a potential biomarker for PCa.
There was a lack of clinical data to analyse for a relationship between hsa-miR-126-3p and ADAM9 within the TCGA database. However, ADAM9 is part of the ADAM gene family which has 29 members. Of these 29 ADAM genes, 21 have metalloprotease functions like ADAM9, including ADAM22. There are however, a variety of ADAM-like genes including the ADAM-TS (A-disintegrin and metalloproteinase with thrombospondin motifs) genes, which ADAMTS5 belongs to. The connection between these two groups of genes can be seen in Figure 7. Both groups of genes share functions such as creation of proteins with peptidase and disintegrin activity.

Figure 7. A diagram indicating the similarities between ADAM and ADAM-TS protein products. CR = cysteine-rich domain; CT = cytoplasmic tail; DIS= disintegrin domain; EGF = epidermal growth factor-like domain; MP = metallopeptidase domain; PD = prodomain; SD = spacer domain; TM = transmembrane domain; TS = thrombospondin-like domain. (Taken from: Brocker et al., 2009.)

Consequently, similarly functioning ADAM genes, ADAM22 and ADAMTS5 have been included in Figure 5 to aid in prediction of clinical data for ADAM9. Figure 5 shows similar results to Figure 4, both with inversely correlated data. However, the results in Figure 3 cannot be dismissed and more research into possibilities for this increase in expression will need to be carried out.
Finally, the connection between miR-126-3p and existing diagnostic technique Gleason score was analysed to see the diagnostic potential of a miR-126-3p biomarker for PCa. Analysis of miR-126-3p expression and clinicopathological parameters of numerous cancers has been carried out, including cervical cancer and clear-cell renal carcinoma (Yang et al., 2014; Vergho et al., 2014). Sun et al. (2013) also carried out analysis of multiple clinicopathological parameters for PCa including; PSA, Gleason Score, pathological staging and lymph node status. From this analysis they discovered that miR-126-3p was associated with all clinicopathological parameters apart from age, Gleason score and surgical margin status. Interestingly, TCGA analysis from this study indicated a statistically significant positive correlation between miR-126-3p expression level and increasing Gleason score. However, only 2 samples were available for a Gleason score of 10 and miR-126-3p expression. It may have been helpful to have a larger sample number than 2 as it would help to create more reliable results. Nevertheless, it is still interesting that such a positive correlation is found between miR-126-3p and Gleason score and indicates that high miR-126-3p expression may therefore be a potential way to diagnose and prognose more aggressive PCa,
From the data obtained in this study, it has confirmed the hypothesis of ‘miR-126-3p overexpression downregulates expression of mRNA target genes SOX2 and BCL2’. However, the hypothesis of ‘miR-126-3p overexpression downregulates expression of ADAM9’ cannot be confirmed, but with further investigations, reasoning for why there was an increase in expression of ADAM9 could be gathered.
This study has established that there is an interaction between miR-126-3p with SOX2 and BCL2 within the PC3 cell line. Discovery of BCL2 interaction with miR-126-3p in PCa is a novel finding as there have been no previous studies discussing this. This novel discovery therefore indicates a new role of miR-126-3p and BCL2 in PCa and highlights the possibility of using miR-126-3p expression in relation to SOX2 and BCL2 as potential biomarkers for PCa.

Further Investigations
On reflection, there are limitations within this study and possible areas of improvement. Studies have shown that PCa presents in multiple heterogenous forms which can show varying types of aggressiveness. In addition, PCa can appear in an androgen dependant or independent form. The PC3 cell line used in this study is androgen independent and therefore, the data produced will only represent miR-126-3p action in this type of PCa. There is a lack of studies into the effects of miR-126-3p expression in PCa. However, Tai et al., (2014) chose to look at DU-145 and PC3 cell lines and found slightly different expression levels of their gene of interest following miR-126-3p overexpression. LNCaP cells possess androgen sensitivity and could provide a greater insight into the heterogenous nature of PCa. Currently there have been no specific studies using miR-126-3p with LNCaP. However, Musiyenko et al. (2008) used miR-126-3p complementary strand miR-126* which are both encoded by the same locus and therefore have similar internal functioning. They studied its effect in an LNCaP cell line and discovered in microscopical analysis, that it allows one to see the carcinogenic effects of miR-126* in PCa well (Figure 8). Within this figure, the increased invasiveness and motility of the miR-126* treated cells can be clearly seen and, therefore, offers a good cell line choice. Because of this research, if the experiment was to be carried out again, a combination of DU-145, PC3 and LNCaP may have provided better disease-specific results and give a better insight into expression of SOX2, BCL2 and ADAM9 in PCa cell lines (Figure 3).

Figure 8. A photograph taken of LNCaP cells before (no addition) and following (miRNA-126*) miR-126* addition. The increased motility and invasiveness can be clearly seen at the 12-hour mark, indicated by the downward progression of the black line. (Taken from: Musiyenko et al., 2008.)

In addition, only 2 replicates of the cell line analysis results seen in Figure 1 was carried out. If more repeats were carried out, this could help in obtaining higher accuracy and reliability of the experiment and therefore, increase the validity of conclusions made.