Implication of Long Non-coding RNA CCDC144NL-AS1/Micro RNA-143-3p Axis Expression Level as Novel Signature in Colorectal Cancer

1. Introduction 1.1. Background Colorectal cancer (CRC) is one of the malicious malignancies worldwide, accounting for nearly 8 % of all annual deaths \[1\]. It is considered Egypt's 7th most prevalent cancer, representing about 3.47 % of male tumors and 3 % of female tumors, respectively \[2\]. By 2030, there will be an estimated 60 % increase in incidence and mortality for CRC globally \[3\]. Early-stage CRC is usually asymptomatic, but when symptoms do manifest, timely detection is essential because any delay in the diagnosis may increase mortality rates \[4\]. 1.2. Problem Surgical resection could cure 90 % of CRCs in the early stages. Nevertheless, the majority of patients frequently have poor prognosis since they are detected at an advanced stage \[5\]. Although colonoscopy tissue biopsy is commonly used for CRC diagnosis, yet, it is an invasive high-risk test, not convenient to be implemented in routine medical examination for longitudinal monitoring or prognosis and is considered partially representative of inter-metastatic or tumoral genetic heterogeneity \[6\]. The classical circulating tumor biomarkers (TMs) as carbohydrate antigen 19-9 (CA19-9) and carcinoembryonic antigen (CEA) \[7\] are used as follow-up or prognosis markers, despite of their limited sensitivity and specificity \[8\]. Therefore, there is an urgent need for more efficient prognostic molecular biomarkers which could be "epi/genetic molecular marker(s)" that have 2 benefits, first, harbor potential therapeutic target(s) and second, augment classical circulating TMs, in order to improve CRC precision. Liquid biopsy has emerged as a minimally invasive diagnostic tool to analyze tumoral genetic and epigenetic molecular markers released into the circulation. Liquid biopsy captures, better, tumor genetic heterogeneity, reflecting the dynamic picture of tumor molecular landscape with lower processing time and lower cost than tissue biopsy \[9\]. Long non-coding RNAs (lncRNAs) are RNA molecules with \>200 nucleotides length, that regulate gene expression at the transcriptional, post-transcriptional, and translational levels, but cannot code for proteins synthesis \[10-12\]. Multiple cancer types exhibit deregulations of lncRNAs, which are involved in all cancer hallmarks, including cancer genesis, progression, and metastasis \[13,14\]. Emerging studies revealed several lncRNAs are implicated in CRC tumorigenesis, metastasis, and progression \[15,16\]. Beyond their potential for diagnosis, lncRNAs would also serve as possible therapeutic targets \[17\]. Coiled-lncRNA Coil Domain Containing 144 N-Terminal-Like antisense 1 (CCDC144NL-AS1) located on the 17p11.2 human chromosome, is a novel oncogenic lncRNA recently reported to be involved in carcinogenesis \[18\]. LncRNA CCDC144NL-AS1 was found to be upregulated in various cancers including gastric cancer (GC) \[18\], hepatocellular carcinoma (HCC) \[19\], non-small cell lung cancer (NSCLC) \[20\], ovarian cancer (OC) \[21\], osteosarcoma \[22\], and CRC \[23\]. However, its clinical role as biomarker for CRC needs to be elucidated. MicroRNAs (miRNAs or miRs) are single-stranded, small RNA molecules that are 18-25 nucleotides length \[24\]. They have a regulatory function in a variety of physiological processes, involving cell differentiation, growth, apoptosis, immunological response, hematopoiesis, and proliferation \[25\]. Several studies have shown miRs have crucial role in both initiation and progression of CRC, besides their potential as molecular biomarkers and possible therapeutic hits \[26-28\]. Hsa-miR-143 located on the human chromosome 5q32 \[29\] is a tumor-suppressor miR reported to be down-regulated via miR-mediated post-transcriptional gene silencing in several human cancers including prostate cancer \[30\], cervical cancer \[31\], OC \[32\], and B-cell lymphoma \[33\]. The 3' arm of the miR precursor product, hsa-miR-143-3p is down regulated in CRC and if it would contribute to CRC initiation \[34\] will be studied currently clinically. LncRNA-miR interaction plays an essential role during various cancer development \[35,36\]. CCDC144NL-AS1 was reported to act as competing endogenous RNA (ceRNA) for hsa-miR-143-3p during GC progression, via competing with the common binding regions of miRs, in order to sequester them and therefore, alter the expression of miRs downstream target genes or proteins \[18\]. Being approved experimentally by Fan et al. \[37\] in GC lncRNA CCDC144NL-AS1 upregulation in GC tissues and sponging hsa-miR-143-3p followed by upregulated expression of its direct endogenous target protein. Similarly, the clinical correlation between lncRNA CCDC144NL-AS1 and hsa-miR-143-3p in CRC patients' peripheral blood samples, that haven't been estimated previously, could further clarify our understanding of CRC molecular pathogenesis and proof the cancer findings documented in silico. High Mobility Group AT-hook 2 (HMGA2) gene encoded by 5 exons and an open reading frame of 330 base pair, is found at human chromosome band 12q13-15 \[38\]. The adult normal cell's HMGA2 protein concentration is minimal or absent under normal physiological conditions, but it is highly expressed during embryogenesis \[38\] and carcinogenesis \[39,40\]. Patients with CRC who have HMGA2 overexpression are experiencing worse prognosis \[41\]. HMGA2 takes part in almost every stage of biological activity of CRC, including cell division, proliferation, apoptosis, senescence, tumor invasion, epithelial-tomesenchymal transition (EMT), DNA repairing mechanism, and stem cell ability of self-renewal \[42\]. HMGA2 in osteosarcoma, was reported to be positively modulated by CCDC144NL-AS1 \[22\]. 1.3. Aim The oncogenic lncRNA CCDC144NL-AS1 if being involved in CRC pathogenesis, by acting as a ceRNA/sponging the tumor suppressor hsamiR-143-3p, and further, upregulating the expression of its endogenous target HMGA2 as an interaction arm will be highlighted in the current study. 1.4. Objectives Elucidate the role of lncRNA CCDC144NL-AS1, hsa-miR-143-3p, and HMGA2 protein as non-invasive epigenetic molecular biomarkers in liquid biopsy of CRC Egyptian patients, individually or as an interaction arm and in comparison, to the conventional protein TMs. In addition, we investigated the potential role of lncRNA CCDC144NL-AS1 as a mediator for development and/or progression of the cancer phenotype as well as CRC metastasis and its relation to both hsa-miR-143-3p and HMGA2, clinically and in silico. 2. Subjects 2.1. Sample size and power of the study Based on the previous study by Zhang et al. in 2019 lncRNA CCDC144NL-AS1 was normally distributed with a standard deviation (3.2) and large effect size (0.85) \[43\]. If the true differences between the CRC group and the control group means are 1 and 3.3, respectively, the study group sizes are 34 patients and 34 control subjects. Total sample size was 68 cases, which was increased by 25 % for expected losses, and total sample was finally 90 subjects, 60 CRC subjects and 30 control (2:1). This is to be able to reject the null hypothesis, that the population means of the experimental groups are equal with probability (power) of 0.8. The Type I error probability associated with this test of this null hypothesis is (0.05). Sample size estimation was performed by G power\* sample size online calculator http://www.gpower.hhu.de/en.html, depending on two-sided confidence level 95 %. 2.2. Study design Case-controlled retrospective mono-center study. 2.3. Institutional Review Board (IRB) statement The Research Ethics Committee of the Faculty of Pharmacy, Ain Shams University, granted ethical permission to conduct the study, 2022. All participants (controls or patients) were fully cognizant of the purpose of the study and signed a written, ethically-approved, informed consent (I⋅C) form. This study was conducted in accordance to the Declaration of Helsinki Guidelines approved in 2013 \[44\]. 2.4. Study participants 2.4.1. Patients group A total of 60 primary CRC treatment-naïve Egyptian patients admitted to the Dar Al Shefa Hospital, Cairo, Egypt, were enrolled in the study. 2.4.1.1. Patients' inclusion criteria. Patients visiting the Colonoscopy Unit for colorectal examination, suffering from variable colonic symptoms, including the CRC alarming symptoms, constipation, abdominal pain, rectal bleeding, and sudden weight loss. CRC diagnosis was clinically confirmed by colonoscopy, abdominal radio-imaging, and histopathologically. 2.4.1.2. Patients' exclusion criteria. Patients suffering from inflammatory disorders, hematological disorders, any cancer other than CRC, or those receiving chemotherapy, radiation, or have undergone surgery, patients with hematological disorders, or any cancer other than CRC. Individuals with inadequate data or missing histopathological diagnosis, as well as those with distant metastases at the time of diagnosis were excluded from the study. 2.4.2. Control group 30 age-matched and sex-matched apparently healthy volunteers, not receiving any medications or suffering from any disease, age range 30-60 years, were included as controls, male-to-female 1:1 (15/15). Control subjects were recruited randomly during routine check-up examinations for themselves or during blood donation. 2.4.3. Patients demographic, clinical, and pathological data The patients' demographic data including age, gender, smoking status, and the patient full history, retrieved from the hospital medical records. In addition to patients' colorectal surgery history, the complete family history of cancer, as well as history of diabetes mellitus (D.M) and hypertension (HTN) were recorded to determine the non communicable diseases status/impact. From patients' files, the following chemistry lab results were recorded, for statistical correlations, CEA, CA19-9 and routine biochemical testing of liver function profiling of alanine aminotransaminase (ALT), aspartate aminotransaminase (AST), and serum albumin, kidney function tests including serum creatinine and serum urea. Hemoglobin (Hgb) as well as prothrombin time (PT), platelet count, lymphocytes count, lactate dehydrogenase (LDH), and C-reactive protein (CRP) were all measured in blood at the Clinical Biochemistry Lab, Dar Al Shefa Hospital, Cairo, Egypt. Tumor site, tumor size, tumor type if mucinous or not, LN metastasis (LNM), tumor invasion or vascular invasion, tumor differentiation, tumor grade, tumor-node-metastasis (TNM) staging, inflammation status as inflammatory bowel disease (IBD), and CRC locations if colonic or rectal, as well as if transverse, sigmoid, rectosigmoid, and rectal, were collected from the patients' files at the Statistics Unit at Dar Al Shefa Hospital, Cairo. Pathological records according to the American Joint Committee (AJCC) on Cancer 2010 criteria \[45\] were collected as well. Patients were categorized into three stages from II to IV, Stage I/II is the local cancer with no LN involvement (N0), Stage III where LN involvement (N1-x) is there, and Stage IV with distant metastasis (M1). CRC staging was determined by the colonoscopy results, abdominal radiography, pathological findings, and clinical evaluation, where the early stage T2 indicates invasion of the muscularis propria by the tumor, T3 indicates the tumor penetration to the subserosa and the muscularis propria, while late tumor stage T4 indicates that it has penetrated the rectum or several colon layers. 3. Methods 3.1. In silico database(s) search and analysis (accessed on October 2021 and revised July 2023) 3.1.1. Identification of the investigated ncRNAs by bioinformatics The HUGO Gene Nomenclature Committee (HGNC) https://www. genenames.org/ supported by National Human Genome Research Institute grant, National Center for Biotechnology Information (NCBI) https://www.ncbi.nlm.nih.gov/, NCBI Genome Data Viewer (GDV) \[46\] (USA.gov) and Ensembl Databases https://www.ensembl.org/index. html for human ncRNAs genes characterization (Ensembl release 109) CCDC144NL-AS1 and MIR-143 are shown in Table 1. 3.1.2. LncRNA disease v2.0 expression The LncRNA and Disease Database (version 2.0) \[47\] http://www.rnanut.net/lncrnadisease/index.php/home to explore lncRNA CCDC144NL-AS1 expression in different cancer types retrieved from validated experimental results in publications or predicted http://www. rnanut.net/lncrnadisease/index.php/home/detail/LDA0044869. 3.1.3. Expression via ENCORI pan-cancer analysis platform \[48\] https://rnasysu.com/encori/panCancer.php of miR, lncRNA or genes across 32 types of Cancers. The expression box-plot values of genes from RNA-seq data were scaled with log2(FPKM +0.01), while the ones from miRNA-seq data were scaled with log2(RPM + 0.01). 3.1.4. Association via miREnvironment database \[49,50\] http://www.cuilab.cn/miren#fragment-1of curated and collected experimentally supported miRNA and various environmental factors (394 factor) interplay and their associated phenotypes (June-28, 2011, the original miREnvironment Database was released, Last update: Sep-9, 2012) for analysis for has-miR-143-3p as prediction of association between environmental factors and human disease. 3.1.5. Interaction via the lncRNASNP2-human \[51\] The lncRNA CCDC144NL-AS1 or HMGA2 as non-conserved targets of miRNA:hsa-miR-143 http://www.noncode.org/gene\_trans\_search.php search\_type=keyword\&keyword=CCDC144NL-AS1\&sbt=Search. 3.1.5.1. Binding targets interaction. Binding targets for hsa-miR-143-3p predicted via database RNAhybrid 2.2 https://mybiosoftware.com/rna 22-v2-microrna-target-detection.html \[52\], http://bibiserv.cebitec.uni -bielefeld.de/rnahybrid \[53\] and RNA22 v2 microRNA target detection https://cm.jefferson.edu/rna22/Interactive/ by Jefferson Computational Medicine Center (CMC) \[54\] to predict hsa-miR-143-3p interaction with either lncRNA CCDC144NL-AS1 or HMGA2 as target, and database RNA22 v2 microRNA target detection. 3.1.6. Functional enrichment analysis, targeted pathways and heatmaps 3.1.6.1. KEGG targeted pathways, clusters/heatmap using DIANA. Reverse Search \[55\] https://dianalab.e-ce.uth.gr/html/universe/index. php?r=mirpath to identify in KEGG pathways-involved miRs using Mirpath, using the DIANA-TarBase v7.0 Heatmap with cluster dendrogram with statistically significant results in red by a posteriori analysis method after an enrichment analysis is performed, p value threshold set at 0.05 and MicroT threshold set at 0.8 (Accessed on July 25th 2023). Functional enrichment analysis using STRING version 11.5 https://string-db.org/ \[56\]. Finally, using the Genome Browser (UCSC) \[57\] Dec. 2013 initial release; June 2022 patch release 14, selected top genes Targets of HMGA2 interactions and pathways from curated databases and textmining https://genome.ucsc.edu/cgi-bin/hgGateway (Accessed on July 25th, 2023). 3.2. Blood samples Five milliliters of peripheral venous blood were withdrawn from controls and CRC patients, under strict sterile conditions, following standard international biosecurity safety procedures, into polymer gel clot activator vacutainers (Greiner Bio-One GmbH, Australia). At room temperature (25 ◦C), completely coagulated samples were centrifuged at 4000 rpm for 10 min. Sera obtained were aliquoted into three DNase/ RNase free Eppendorf tubes and stored at -80 ◦C for molecular analysis. 3.2.1. Total RNA extraction Using the miRNeasy Mini kit (Cat. No.217004; Qiagen, Hilden, Germany) according to the manufacturer's protocol, from serum samples, RNA extraction was done. The isolated RNA was eluted in 40 μL of RNase-free water. 3.2.2. Quantitation of purified RNA including miRNAs Using a NanoDrop® 1000 spectrophotometer (Thermo Scientific, Wilmington, DE, USA) the concentration and purity of all RNA samples were determined. Absorbance at 260 nm was used to measure the conc. of RNA in the sample, whereas 260/280 and 260/230 nm ratios were used to evaluate RNA purity. After quantification, the isolated and eluted RNA was stored at -80 ◦C in aliquots. 3.2.3. Reverse transcription and measurement of ncRNAs expression 3.2.3.1. cDNA synthesis and measurement of lncRNA CCDC144NL-AS1 expression using qRT-PCR. Total RNA was reverse transcribed into cDNA using the Xpert cDNA Synthesis Kit (Cat. No. GK80.0100; Grisp Research Solutions, Rua Alfredo Allen, Portugal) containing Xpert Reverse Transcriptase (RNase H-), RNA-dependent DNA polymerase appropriate for cDNA synthesis from long RNA templates, following the manufacturer's protocol. The synthesized cDNA was then stored at - 20 ◦C till qRT-PCR. The expression of lncRNA CCDC144NL-AS1 was measured using the Xpert Fast SYBR (Cat. No. GE20.0100; Grisp Research Solutions, Rua Alfredo Allen, Portugal) in accordance with the manufacturer protocol. The primer RT2 lncRNA qPCR Assay for Human CCDC144NL-AS1 (Hs04941765 m1, Cat. No. 4426961) was used to assess the level of expression of the lncRNA CCDC144NL-AS1. The human GAPDH primer (LPH31725A-200, Cat. No. 330701) was used as endogenous control to normalize the expression of lncRNA CCDC144NLAS1. 3.2.3.2. cDNA synthesis and measurement of hsa-miR-143-3p expression using qRT-PCR. The miRCURY LNA RT Kit was used for cDNA synthesis (Cat. No.339340, Qiagen, Hilden, Germany) as proposed by the manufacturer's instructions. The resulting cDNA was kept at - 20 ◦C until quantification. qRT-PCR was used to measure the expression of the hsamiR- 143-3p using the miRCURY LNA miRNA PCR Assay (Cat. No. 339306, Qiagen, Hilden, Germany). The primer SNORD38B (hsa) miRCURY LNA miRNA PCR Assay (YP00203901, Cat. No. 339306) was used as endogenous control to normalize the expression of hsa-miR-143-3p. The reaction was carried out using the PCRmax Eco™48 qRT-PCR system (PCRmax, Staffordshire, USA). Primers sequences are listed in Table 2. All these primers were designed by Qiagen https://www.qiagen.com/workflow-configurator/ workflows intcmp=CM\_QF\_WFC\_1121\_OTHERS\_QB\_nav\_products, except for lncRNA CCDC144NL-AS1 that was obtained from Thermofisher https://www.thermofisher.com/taqman-gene-expression/pr oduct/Hs04941765\_m1 (Accessed on October 2021). The RNA relative expression was computed and normalized as fold change using the cycle threshold (Ct) method (2- ΔΔCt) with GAPDH or SNORD38B (hsa) as the internal control for lncRNA CCDC144NL-AS1 and hsa-miR-143-3p, respectively. ΔCt was determined by subtracting the Ct values of GAPDH and SNORD38B (hsa) from those of the lncRNA CCDC144NL-AS1 and hsa-miR-143-3p, respectively; where ΔΔCt = ΔCt cancer samples - ΔCt control samples \[58\]. 3.2.4. Quantification of HMGA2 protein concentration by ELISA HMGA2 protein concentration was measured in serum samples by commercially available ELISA kits from Bioassay Technology Laboratory (Cat.No. E7513Hu, Jiaxing, China) according to the manufacturer's instructions. The reaction is based on pre-coating the ELISA plate with Human HMGA2 antibody and HMGA2 present in added samples binds to antibodies coating the wells. Biotinylated human HMGA2 antibody was added to bind HMGA2 in samples. A second detector antibody was then added to bind the biotinylated HMGA2 antibody. A substrate solution was added that reacts with the enzyme-antibody-target complex to produce a measurable signal measured at 450 nm. 3.2.5. Indices and ratios 3.2.5.1. Body mass index (BMI kg/m2) was calculated in kg/m2 for each participant using the website https://www.nhlbi.nih.gov/health /educational/lose wt/BMI/bmicalc.htm. Normal weight individuals have BMI of 18.5-24.9 kg/m2, overweight BMI as 25-29.9 kg/m2, and 30 kg/m2 or more for obesity. 3.2.5.2. Platelets-to-lymphocytes ratio (PLR) was calculated by dividing the patient platelet count (x103 cell/μL) by the lymphocyte count (x103 cell/μL). PLR is an inflammation indicator and immune response-related predictor that has a stronger link with inflammatory diseases severity than the neutrophils-to-lymphocytes ratio (NLR) \[59\] or either individual cells alone. 3.3. Statistical analysis Data were collected and excel tabulated in Microsoft Excel 2019. Statistical package for social studies software SPSS 26.0 (IBM, Armonk, NY) (https://www.ibm.com/products/spss-statistics), and GraphPad Prism® version 8.01 (GraphPad Software, San Diego, USA) (htt ps://www.graphpad.com/scientific-software/prism/) were utilized for figures, while MedCalc Statistical Software version 19.2.6 of (MedCalc Software by Ostend, Belgium) (https://www.medcalc.org) was used for the receiver operating characteristic (ROC) curve analysis. Data were tested for normality using Shapiro-Wilk normality test for both groups and subgroups data. Since the patients' data were not normally distributed, data were expressed as median (interquartile range: IQR (25th percentile-75th percentile). Mann-Whitney (U) or Kruskal-Wallis (H) were conducted to compare between any two or more independent groups, respectively. The ROC curve was used to find the best cutoff, sensitivities (SNs), specificities (SPs), negative predictive values (NPVs), and positive predictive values (PPVs), with an area under the curve (AUC) calculated ranged from 0 to 1. In medical testing, negative likelihood ratios (LRs) are used to understand the diagnostic or prognostic test utility. Essentially, the LR indicates the likelihood that a patient has a condition or disease. The likelihood that they have the disease or condition increases with the ratio. A low ratio, on the other hand, indicates that they most likely do not. Therefore, a physician can use these ratios to either rule in or rule out an illness. The ratio, which expresses how likely it is for someone to have the disease or condition, supports the SNs and SPs identified by the ROC curve. An alternative definition of the LR is SN and SP, where negative LR = (100 - SN)/SP. Spearman's correlation coefficient r was used to evaluate the correlation between various variables. Additionally, the expression levels of the lncRNA CCDC144NL-AS1, hsamiR- 143-3p, and HMGA2 protein were set to Spearman correlation r, and the link among two variables-one continuous and one dichotomous- was assessed using point-biserial correlation. Significance level was set if the two-tailed statistical analysis test p-value is \<0.05.