More importantly Santagata et al have suggested a new breast
More importantly, Santagata et al. have suggested a new breast cancer subtyping method according to ER, androgen receptor (AR), and VDR with a higher prognostic value than the ordinarily used method based on ER, progesterone receptor (PR) and Her2 (Santagata et al., 2014). This implies that such complex interactive network between these hormonal receptors might affect the course of breast cancer (Al-Azhri et al., 2017). Although Santagata et al. reported significant concomitant expression of ER and VDR in breast tumors (Santagata et al., 2014), in ER positive breast cancer cells, treatment of breast cancer Agarose GPG/LMP low melt with vitamin D analogue has decreased ER expression and impeded estrogen induction of cell proliferation (James et al., 1994). Taken together, these data suggest a bidirectional interaction between VDR and ER signaling pathways in breast cancer. Noticeably, VDR and ER share evolutionary conserved domains (Lecomte et al., 2017) (Fig. 1).
In addition to protein coding genes, non-coding genes including long non-coding RNAs (lncRNAs) might be involved in such interactive network. Several independent studies have assessed contribution of lncRNAs in VDR or ER signaling pathways individually. For instance, VDR signaling alters the expression of some lncRNAs parallel with its protective effects against skin cancer (Jiang & Bikle, 2014a). On the other hand, the HOX Antisense Intergenic RNA (HOTAIR) increases ER signaling and contributes in tamoxifen resistance in breast cancer (Xue et al., 2016). However, the functional link between lncRNAs and VDR and ER signaling pathways is not studied yet. So we conducted an in silico study to find lncRNAs that modulate these two signaling pathways in breast cancer through identification of lncRNAs whose expression levels correlate with expression of VDR and estrogen receptor 1 (ESR1) genes.
Material and methods In the present study, we assessed expression pattern of VDR and ESR in breast cancer tissues in relation with expression of lncRNAs. The strategy used for such analysis is depicted in Fig. 2.
Correlation between expression levels of lncRNAs, ESR1 and VDR was assessed using Co-LncRNA tool (http://www.bio-bigdata.com/Co-LncRNA/) (Zhao et al., 2015a). Using this tool, lncRNA co-expressed ESR1 and VDR genes were pinpointed in TCGA datasets which included expression profiles of 619 patients with invasive breast cancer. Data were analyzed using linear correlation and a P value of less than 0.01 was considered as statistically significant.
Assessment of expression pattern of selected lncRNAs in breast cancer tissues Expression pattern of selected lncRNAs (obtained from the previous step) was analyzed in breast cancer tissues obtained from 873 breast invasive carcinoma and 105 matched normal samples using TANRIC (http://ibl.mdanderson.org/tanric/_design/basic/index.html) tool which has gathered large-scale RNA-seq datasets from TCGA (Li et al., 2015). P value of less than 0.01 was considered as statistically significant.
Assessment of genomic alterations of the selected lncRNAs in breast cancer Genomic alterations of lncRNAs in breast cancer tissues were extracted from the cBioPortal for Cancer Genomics (http://www.cbioportal.org/) (Cerami et al., 2012) containing data of somatic mutations obtained from targeted sequencing and copy number alterations of 2509 breast cancer samples and the Catalogue Of Somatic Mutations In Cancer (COSMIC) (Forbes et al., 2017).
Evaluation of interaction of selected lncRNAs with oncogenic/ tumor suppressor microRNAs (miRNAs) Interactions between selected lncRNAs and miRNAs were assessed using AnnoLnc (Hou et al., 2016) and miRcode (Jeggari et al., 2012) tools. In the former tool, the FASTA format of transcript sequence of each lncRNA was entered to receive the list of miRNAs with putative interaction with these lncRNAs. In the latter tool, the name of each lncRNA was searched to find the miRNAs which target 3’UTR, coding sequence (CDS) or 5’UTR of each lncRNA.