Combination of vitamins A, D2 and D3 reduce tumor load and alter the expression of miRNAs that regulate genes involved with apoptosis, tumor suppression, and the epithelial-mesenchymal transition in HCT-116 colon cancer cells

INTRODUCTION

​

Colorectal carcinoma (CRC) continues to be one of the primary causes of cancer deaths worldwide [1-3]. Numerous factors increase CRC risk, including genetic and racial differences, environment, diet, and lifestyle. These factors include hereditary components and predisposition to polyp formation, as well as alcoholism, high fat diet, large bowel inflammatory disease, obesity, smoking, and stressors [1–3]. Symptoms of CRC usually do not appear until late in the disease progression, thus annual screening is essential for long-term survival. Vitamins A and D have significant anticancer activities as have been shown in clinical trials, as well as in ecological, epidemiological, geographical, observational, and mechanistic studies [4-10]. Vitamin D deficiency significantly increases the risk of breast, colon, gastric, prostate, and other cancers, and patients with low vitamin D levels have a poorer prognosis in gastric cancers [11]. Conversely, patients with elevated serum vitamin D3 (Cholecalciferol, animal based) levels and colorectal cancer had a better prognosis, as well as reduced recurrence, metastasis, and mortality [12-18]. Similar data are available for ergocalciferol (vitamin D2, plant based), as well as vitamin A and all trans retinoic acid (ATRA) show that these nutrients inhibit the growth of breast, colorectal, and gastric cancers, as well as leukemia [19-28]. However, there are few studies that have investigated the effects of these vitamins in combination for the prevention and treatment of cancers.

​

Over the past few years, our group has been investigating the synergistic effects of the fat-soluble vitamins A and D combinations, as they are normally present together in many functional food products [27-29]. We have previously reported that vitamin A, D2 and D3 combinations induced apoptosis in HCT-116 and SW 480 colon cancer cells by increasing caspase 3/7 and 8 expression and activities, the expression of Bax and BIM (BcL2L11) and other tumor suppressors, including p53 [27-28]. Furthermore, RNA-seq analysis demonstrated that vitamin A and D combinations significantly altered the expression of numerous genes in the HCT-116 colon cancer cells including the expression of mRNAs involved in numerous canonical pathways associated with apoptosis, autophagy, cellular senescence, the epithelial mesenchymal transition (EMT), and immunity [29].

​

Along with mRNA, microRNAs (miRNAs) play a critical role in cell growth, differentiation, migration, and miRNAs are known to be dysregulated in cancer [30-33]. MicroRNAs (MiRNAs) are a class of small single-stranded non-coding RNAs that bind to messenger RNAs and alter their expression by post-translational modification [33]. By doing so, miRNAs control gene expression associated with many cellular functions including apoptosis, cell proliferation and differentiation, as well as progression of the cell cycle [30-33]. Up-regulation of miRNAs expression inhibits gene expression, while down-regulation of miRNAs increases mRNA expression. MicroRNAs play a central role in the initiation and progression of colon cancer and have now emerged as novel diagnostic and prognosticative tools for many types of cancer [30-33]. Since miRNAs are dysregulated in colon cancer, investigating the impact of specific treatments on miRNA expression can facilitate an understanding of novel mechanisms of action and potential treatments.

​

Here, we investigate the impact of vitamins A and D treatments in an HCT-116 xenograph mouse model of colon cancer, and further investigate potential mechanisms of action by transcriptomic profiling of HCT-116 colon cancer cells using mRNA and miRNA sequencing (mRNA-seq; miRNA-seq). Co-expression networks were developed using Ingenuity Pathway analysis to identify the interactions among the mRNA:miRNA pairs. The data showed the alterations in the expression of multiple miRNAs that were directly associated with differential gene expression (DGE) of tumor suppressors, and molecules associated with apoptosis and the EMT.

​

MATERIALS AND METHODS

​

Cell culture and maintenance: The HCT-116 and HCT-116-luc colon cancer cells were obtained from American Type Culture Collection (ATCC) and cultured as we have previously described [27-28]. Briefly, the HCT-116 cells were cultured in McCoy’s 5a medium supplemented with 10% FBS and 1% Penicillin/Streptomycin and incubated at 37°C in a humidified atmosphere of 5% CO2 and 95% air. The cells were harvested when they reached 80% confluence, after the addition of 0.25% trypsin/EDTA, and plated in 96 or 6 well plates for the assays. The HCT-116 cells were treated with ATRA+D2+D3 at the IC50 concentration (1µM), then harvested and RNA was isolated as we have previously described [27-28].

​

HCT-116-luc xenograph mouse model: Xenographs were established using fifteen Crl:NU(NCr)-Foxn1nu male mice (4-6 wks old, ~18-19 gms body weight; 3 arms x 5 mice per arm) obtained from Charles River Laboratories. All mice were housed in cages in a controlled environment in Rm 112 of the BRL (72ºC; 14/10-h light/dark cycle) and housed under barrier pathogen free conditions for 7 days prior to initiation of the study. The animals were fed a purified diet specific for imaging (Purina Test Diet AIN-93G #5800) from Purina Mills Inc., St Louis, MO, and given water ad libitum and cared for as detailed in ACC Protocol number 17-116. Group 1: After 7 days of acclimatization, N = 5 Crl:NU(NCr)-Foxn1nu mice were subcutaneously injected with HCT-116-luc cells (1 X 10⁶ in 100 µl 50% Matrigel and Sterile PBS) into the flank under 2% isoflurane anesthesia using a 26G needle and 1 ml syringe, and then on day 8 they were treated with the purified diet (containing PBS only negative control) for 21 days. Group 2: After 7 days of acclimatization, Crl:NU(NCr)-Foxn1nu mice (n = 5) were subcutaneously injected with HCT-116-luc cells (1 X 10⁶ in 100 µl 50% Matrigel and Sterile PBS) into the flank under 2% isoflurane anesthesia. On day 8 they were treated with the purified diet containing Vitamin A (retinol) 25,000 IU/Kg and vitamin D3 4,000 IU/Kg body weight for 21 days. Group 3: After 7 days of acclimatization, N = 5 Crl:NU(NCr)-Foxn1nu mice were subcutaneously injected with HCT-116-luc cells (1 X 10⁶ in 100 µl 50% Matrigel and Sterile PBS) into the flank under 2% isoflurane anesthesia and on day 8 treated the purified diet containing 35,000 IU Vitamin A (retinol) and vitamin D3 5,000 IU/Kg body weight for 21 days. The doses were based on the human dose equivalent and our preliminary in vitro data [34].

​

Imaging and data analysis: The mice were palpated (tumor measurement) and imaged once weekly. For the HCT-116-luc cells xenographs, we used fluorescence in vivo imaging of the animals to follow the progression and development of the tumors. Before image collection on day 8, all animals were anesthetized using 2% isoflurane and luciferin (intraperitoneal injection, dose 150 mg/kg body weight), as we have previously described [35]. Image collection was performed using a Xenogen IVIS 200 Series live animal imaging system (Xenogen Corp., Alameda CA) for the detection of the luciferase signal. Photon emitted per second (in photons/second/cm2/steradian) from each ROI (region of interest) was quantified using the Living Image® software. Imaging began on Day 8 after the tumor cell injections and then once weekly. At 21 days after treatment the animals were sacrificed using continuous flow CO2 from a bottled source and cervical dislocation. Tumors and tissues were resected, weighed, and measured using calipers. The body weight of the mice, as well as food and water ingestion were documented twice a week throughout the study. Tumor volume was measured by digital caliper and calculated using the formula: L1 x L2 x H 0.5238, where L1 = long diameter L2 = short diameter and Height = height of the tumor. After 21 days of treatment or if the tumor volume reached to ~ 1000 mm³ animals were sacrificed using CO2 inhalation. Excised tumors were weighed and frozen at − 80 ºC for later RNA extraction and qPCR. The results were calculated by dividing the mean tumor volume or weight of the treated group to that of vehicle control group.

​

RNA isolation and quality control: HCT-116 colon cancer cells were cultured at a density of 1.0 – 1.2 x 10⁶ cells in 1 ml per well of a 6-well plate and incubated overnight as we have previously described [29]. Total RNA and proteins were extracted using Trizol (ThermoFisher Scientific, Waltham, MA, USA) from cultured cells treated with vehicle solvent (0.01% DMSO) or the IC50 concentration of ATRA+D2+D3. RNA isolation from excised tumors was performed using Trizol, and the quality of the RNA (RIN number) was determined using an Agilent 2100 Bioanalyzer (Agilent Tech., Palo Alto, CA). Isolated RNA samples were used to generate parallel mRNA-Seq, miRNA-seq and qPCR data. RNA samples were quantified using NanoDrop™ One Spectrophotometer (Thermo Scientific) and analyzed for integrity using Agilent 4200 TapeStation and RNA Screen Tape (PN: 5067-5576). The relative levels of remaining DNA were checked using dual RNA/DNA measurements in a Qubit fluorometer (Invitrogen). The levels of DNA did not exceed 10%.

​

mRNA-seq data collection: Briefly, the sequencing libraries for Illumina sequencing were prepared using 250 ng of total RNA per sample exactly as we have previously described [29]. Library preparation was carried using a Universal Plus mRNA-Seq library preparation kit (Tecan/NuGen, 0520-A01), as described in the product manual (NuGen, M01485 v5). Final library purification was performed using Agencourt AMPure XP Beads (Beckman Coulter A63881). The final amplified libraries were measured using a Qubit 1X dsDNA HS Assay Kit (Invitrogen, Q33231). The final library pool concentration was confirmed by qPCR and test sequencing was performed to confirm sequencing efficiencies and to adjust proportions of individual libraries. The library pool was purified using Agencourt AMPure XP Beads (Beckman Coulter A63881) and quantified by qPCR using a KAPA Library Quantification Kit and sequenced on a NovaSeq 6000 S4 flow cell, 2x150 bp, approximately 30 M clusters per sample, at the University of Illinois Roy J. Carver Biotechnology Center High-Throughput Sequencing and Genotyping Unit.

​

miRNA-seq data collection: Total RNA was isolated from cell samples from HCT-116 colon cancer cells treated with vehicle solvent (Control, DMSO 0.01%) or ATRA+D2+D3 (IC50 1 µM). Total RNA was used for miRNA library construction with the QIAseq miRNA library prep kit for Illumina (Qiagen, 331505). Isolated total RNA quality was measured resulting in RNA Integrity Number (RIN) scores ≥ 9.6-10. Sequencing libraries were prepared using total RNA at 250 nanograms per sample. Library construction included the ligation of mature miRNA to specific 3' and 5' adapters [30, 36]. The ligated miRNAs were then reverse transcribed to cDNA using a reverse transcription (RT) primer with a UMI. cDNA cleanup was performed using a magnetic bead-based method. Following cDNA cleanup, library amplification was performed with a universal forward primer and indexing reverse primers where each sample was assigned a unique custom index. Following a final library cleanup using an Agencourt RNAClean XP Bead assay (Beckman Coulter, PN: A63987), each miRNA library was quantified using Qubit 2.0 Fluorometer (Invitrogen) and pooled in equimolar concentrations. Final sequencing pool was analyzed using Tape Station gel analyzer (Agilent) and re-quantified using the qPCR KAPA Library Quantification Kit (KAPA Biosystems). Sequencing was carried out on NovaSeq6000 SP flow cell, 1x100 nt reads, to achieve approximately 20 M reads per sample.

​

Bioinformatics Analysis and Database Annotation: Bioinformatic analysis of the mRNA-seq raw data was performed by the University of Illinois at Chicago Core for Research Informatics (UICCRI) exactly as we previously have described [29]. Briefly, quality-control for the mRNA sequencing data was obtained using FastQC and raw reads were aligned to human reference genome hg38 in a splice-aware manner using STAR and BWA MEM as we have described [29]. Analysis of the molecular functions and biological processes of differentially expressed genes (DEGs) detected from RNA-seq was performed using the ENSEMBL database (www.ensembl.org) using FeatureCounts [37-38]. EdgeR was used to compute differential expression statistics on raw expression counts with the exactTest function [39]. P-values were adjusted for multiple testing using the false discovery rate (FDR; q value) with the correction as described by Benjamini and Hochberg [40]. DEGs were categorized into functional clusters based on Gene Ontology (GO) function enrichment analysis using edgeR and DeSeq2 [29]. For miRNA, raw reads were trimmed to remove adapters sequence using cutadapt [41]. Reads with no adapter were discarded. Remaining trimmed reads were reads were aligned to mature known human microRNA database obtained from miRbase using BWA ALN [42]. miRNA expression levels where quantified by counting the number of reads mapped to each miRNA sequence and normalized to counts-per-million units for direct comparison between samples. Differential expression statistics were computed using edgeR [43] on raw expression counts with the exactTest function. P-values were adjusted for multiple testing using the false discovery rate (FDR) correction of Benjamini and Hochberg [40]. Bland plots were performed as described in edgeR [44,45].

​

Data and Statistical analysis: Data from in vitro assays were analyzed using the GraphPad Prism 9.0 software to determine IC50 concentrations. Statistical analysis was performed using the Student t test in GraphPad version 9.0 (San Diego, CA). The level of significance between the treatments groups as compared with control was estimated by one-way analysis of variance (ANOVA) followed by Tukey’s multiple comparison tests. For in vivo studies, the number of animals was calculated using a power/sample analysis, anticipating differences for the treated versus untreated groups (Instat Statistical software). The proposed sample size of 5 mice per group should be sufficient to detect a 50% reduction in outcomes (tumor size/volume) with a power of 80% using a t-test where P was considered statistically significant at < 0.05.

​

Ingenuity® Pathway Analysis (IPA): Ingenuity Pathway Analysis (IPA) was used to identify the canonical and molecular signaling pathways and gene networks that were correlated with observed changes in the transcription levels, and DEGs were further analyzed according to the predicted protein function using the Ensembl database [29, 46]. Differential gene expression and statistical data were uploaded into the Ingenuity® Pathway Analysis software (Qiagen, USA) and analyzed as we have previously described [29] using a false discovery rate (q value) < 0.05. The results were mapped to IPA’s knowledge bases, as well as the significant biological functions, networks and pathways.

​

Quantitative polymerase chain reaction to validate candidate genes: Quantitative PCR (qPCR) was used to validate the transcriptome levels of fragments per kilobase of transcript per million mapped reads (FPKM) generated by RNA-Seq as we have previously described [29]. Total RNA was reverse transcribed and amplified using a Power SYBR Green RNA-to-CT 1- step kit (Applied Biosystems, Foster City, CA, USA) on a Step One Plus Real Time PCR System (Applied Biosystems, Foster City, CA, USA) exactly as we have described [29]. Each reaction (performed in triplicate) contained Power SYBR Green RT-PCR Mix (2X), 200 nM of each primer, RT Enzyme Mix (125X) and 100 ng RNA. The cycling conditions were as follows: 48°C for 30 min, 95°C for 10 min, followed by 50 cycles of 95°C for 15 sec. and 60°C for 1 min. PCR reaction specificity was confirmed by melt curve analysis at 95°C for 15 sec, 60°C for 15 sec, 95°C for 15 sec. The qRT- PCR assays were conducted in triplicate and ΔCt values (Ct for GADPH or β-actin - Ct for the test gene) were calculated for each RNA sample. The Student t-test was used to analyze significance between the mean ΔCt for the control vs. the treated HCT-116 groups, with a threshold significance level p< 0.05. The fold change in gene expression was calculated as ΔΔCt (ΔΔCt = ΔCt of the treatment group - ΔCt of control group). The qPCR assays were conducted in triplicate and ΔCt values (Ct for GADPH or β-actin-Ct for the test gene) were calculated for each RNA sample. The specific primers used were (forward (F) and reverse (R) primers, respectively): GADPH: (F) 5′- ACCACAGTCCATGCCATCAC-3′ and (R) 5′- TCCACCACCCTGTTGCTGTA-3′; β-actin: (F) 5’TGACGTGGACATCCGCAAAG-3’ and (R) 5’- CTGGAAGGTGGACAGCGAGG-3’; ID2 (F)5’- ATGAAAGCCTTCAGTCCCGT-3’ and (R) 5’TTCCATCTTGCTCACCTTCTT-3’; PTEN (F) 5'-CCG AAA GGT TTT GCT ACC ATT CT-3' and (R) 5'-AAA ATTA TTT CCT TTC TGA GCA TTC C-3'; p53 (F) 5’-GTCACAGCACATGACGGAGG-3’ and (R) 5’TCTTCCAGATGCTCGGGATAC-3’; DDDX20(F) 5’- AGGCATAAAGAAGGCGCTAAC-3’ and (R) 5’- GGGTTCTGACTGTACTCGGTAAG-3’; BIM: (F) 5′- ATGTCTGACTCTGACTCTCG-3′ and (R) 5′- CCTTGTGGCTCTGTCTGTAG-3′; BAX: (F) 5’- GGCTGTGTTGCTGTGAATC-3’ and (R) 5’- GGTATCTGGGAGTAAGGAGGAG-3’, using published primers and NIH Primer BLAST [47-50]. GADPH and β-actin were used to control and calibrate cDNA synthesis.

​

Data sharing and availability: All of the raw datasets for the mRNA-seq and miRNA-seq supporting the conclusions of the work have been deposited and are currently available at the National Center for Biotechnology Information Gene Expression Omnibus repository (GEO), number GSE160109, https://www.ncbi.nlm.nih.gov/geo.

​

RESULTS

​

Vitamin A and D combinations reduce the viability of HCT-116 colon cancer cells in culture and the growth of HCT-116 induced xenograph tumors: Vitamin A and D combinations (ATRA+D2+D3) induced apoptosis and reduced proliferation of HCT-116 colon cancer cells (Figure 1A). As compared with ATRA alone, the combination treatment was more effective in reducing cell proliferation. In Crl:NU(NCr)-Foxn1nu mice, colorectal cancer developed after subcutaneous injection of HCT-116 colon cancer cells and after day 7, tumors were apparent, and the animals were administered vitamin A (25K or 35K IU) with Vitamin D (4 or 5K IU) in feed for 21 days. Figure 1B shows that the weight of the mice injected with HCT-116 colon tumor cells and then treated as follows: group 1 (negative control); group 2 (treated with 25,000 IU Vitamin A and 4,000 IU Vitamin D); or group 3 (treated with 35,000 IU Vitamin A and 5,000 IU vitamin D) for 21 days did not significantly differ. In Figure 1C, tumor volume of mice after 21 days of treatment with vitamin A & D 25/4K or 35/5K IU/Kg body weight was significantly reduced in a time-dependent manner. At the end of 21 days, comparison of Crl:NU(NCr)-Foxn1nu mice injected with HCT-116 colon tumor cells, treatment of the animals with 35,000 IU Vitamin A and 5,000 IU vitamin D human dose equivalents reduced tumor growth by ~58% (p < 0.001), while the lower dose reduced tumor growth by ~38% P <0.05 (Figure 2). Mouse xenographs treated with higher doses of vitamins A and D showed a tumor reduction of 85-90%, however the reduction in tumor burden was not statistically significant due to broad confidence intervals (data not shown).

​