THEMIS2 as a Novel Mediator of VEGFR2-driven Angiogenesis in Breast Cancer: Functional Suppression by miR-125b-5p

Materials and Methods

Cell culture, cell proliferation and sphere formation assay. Human breast cancer cell lines including MDA-MB-231 (ATCC number is HTB-26), MDA-MB-231-IV2, MDA-MB-468 (ATCC number is HTB-132), Hs578T (ATCC number is HTB-126) and Hs578T-FPI were cultured and used for assays as previously described (13, 14). Briefly, the cells were cultured in DMEM containing 10 % FBS for no longer than 3 months and were routinely examined to be mycoplasma-free and authenticated by the assessment of short tandem repeat loci profiling using the Promega StemElite ID System (Madison, WI, USA). Cell proliferation rates were determined by staining with 0.4% Trypan blue solution using the LUNA Automated Cell Counter (Logos BioSystem, Miami, FL, USA) at the indicated time points. For colony formation, cells (1×10⁴ cells/well) were seeded and cultured for 7 days before the colonies were stained with crystal violet. Experiments were performed in triplicates. Cancer stem cells derived from these cell lines were cultured in a stem cell selective condition to allow sphere formation as described previously (14). The experiments were repeated at least three times, and the results are presented as means±SEM. Invasive cell lines such as MDA-MB-231-IV2 and Hs578T-FPI cells were selected by injection of the MDA-MB-231 or Hs578T cells into tail vein of CB17-SCID mice and isolating lung metastasis to derive the highly metastatic subline named IV2 or FPI as previously described (15).

Ethics approval (animals). All animal experiments in this study adhered to the ARRIVE guidelines and conformed to the 3Rs principles (Replacement, Reduction, and Refinement). Our animal protocol number is CMUIACUC-2021-140-2.

The following describes our practices concerning animal care and use: 1) Housing environment: Experimental mice were housed in an Individually Ventilated Cage (IVC) system, ensuring controlled air quality and providing adequate space in accordance with IACUC guidelines. The animal facility maintained a constant temperature of 20-24°C and humidity of 40-70%, following a 12-hour light/12-hour dark cycle. Bedding and nesting material: Cages were provided with dust-free, absorbent bedding (e.g., sterilized wood shavings or paper pulp bedding), which was changed regularly (at least twice per week or as required by IACUC guidelines). For environmental enrichment, clean paper nesting material or species-appropriate shelters (e.g., PVC tubes, small cardboard boxes) were placed in the cages to meet the mice’s natural nesting and hiding behaviors. Diet: Mice had ad libitum access to autoclaved commercial rodent chow, ensuring a balanced and consistent nutrient source. Water: Sterilized drinking water was provided via water bottles or an automated watering system, with daily checks to ensure an uninterrupted supply. 2) During the experiment, the physical condition of all mice was monitored at least once daily (including weekends and public holidays) by trained personnel. Monitoring frequency was increased when necessary (e.g., in cases of abnormalities or after experimental procedures). 3) No animals became severely ill or died at any time prior to the experimental endpoint. As no animals died prior to the experimental endpoint and no animals became ill. 4) We strictly adhered to the IACUC-approved protocol for humane endpoints for experimental animals. This protocol ensures that when an animal exhibits irreversible pain or illness, euthanasia is performed promptly and decisively to minimize its suffering. Clinical signs used to determine when to euthanize animals (humane endpoints) included, but were not limited to, any one or more of the following: Significant Weight Loss: Body weight decreased by more than 15-20% of the starting research weight (or continuous decline for three consecutive days). Significant Deterioration of Body Condition Score (BCS): For example, a BCS score of 1 (emaciated, bony prominences visible). Persistent Signs of Pain or Distress: Manifested as hunching, piloerection (ruffled fur), semi-closed eyes, hunched posture, withdrawal, or slow/no response to stimuli. Severe Impairment of Mobility: Inability to walk, stand, or access food and water normally, or exhibiting lethargy and reluctance to move. Tumor-Related Symptoms: Extensive ulceration, necrosis, or severe infection of the tumor that cannot be effectively treated. Tumor size reaching the IACUC-stipulated limit (1.5 cm, 15 mm, in this study) or significantly affecting animal function even if below the limit. Tumor causing obvious compressive symptoms (e.g., dyspnea, paralysis). Inability to Eat or Drink Normally: Continuous lack of food or water intake for more than 24 hours. Respiratory Distress: Rapid, labored breathing, open-mouth breathing, or cyanosis. Other severe and irreversible disease states as assessed by a veterinarian. Once any of the above humane endpoint criteria were observed, immediate consultation with a veterinarian and euthanasia were performed. 5) Euthanasia for all animals in this study was performed in accordance with IACUC-approved standard operating procedures (SOPs). The primary method of euthanasia was inhalation of high-concentration carbon dioxide (CO₂) gas, following these steps: Mice were placed individually or in small groups (to avoid overcrowding) into a CO₂-pre-filled euthanasia chamber. CO₂ gas was introduced gradually, or directly into a pre-filled chamber, to ensure the animals did not experience distress from sudden gas exposure. After confirming complete unconsciousness, exposure to CO₂ was continued for at least 5 minutes to ensure complete death (cessation of heartbeat and respiration). To ensure complete death, a secondary method of confirmation was used after CO₂ exposure, such as cervical dislocation, pneumothorax/thoracotomy and organ exsanguination. All euthanasia procedures were performed by personnel trained in the method.

Cell chemotactic migration and invasion assay. Control, specific plasmid, siRNA or miRNA transfected cells (1×10⁵) were harvested from monolayer culture, resuspended in serum-free medium with 0.1% bovine serum albumin (Sigma-Aldrich, St. Louis, MO, USA), and then plated onto a Transwell Boyden chamber, with or without matrigel coating on the upper chamber, for invasion or migration assays, respectively. The chambers were incubated in DMEM for 24 h with the addition of 10% FBS to the lower chamber. Cells that did not move were removed using cotton swabs, and the chambers were stained with crystal violet. Photomicrographs of three random regions were captured from duplicated assay chambers. The cell numbers were counted and normalized to the control. All experiments were repeated independently at least three times.

Immunoblot analysis. Cell lysates (30 μg per lane) were separated using 10% SDS-PAGE and transblotted onto a polyvinylidene difluoride (PVDF) membrane (Millipore, Bedford, MA, USA), which was incubated at room temperature in a phosphate-buffered saline (PBS) blocking solution 5 % powdered skim milk and 0.02% sodium azide for 1 h. This was followed by further 2-h incubation with primary antibody including anti-THEMIS2 (A11109, Boster Biological Technology, Pleasanton, CA, USA), anti-p-VEGFR1 (AF4170, R&D Systems, Minneapolis, MN, USA), anti-VEGFR2 (55B11, Rabbit mAb#2479, Cell Signaling, Boston, MA, USA) anti-p-VEGFR2 (Tyr1175, 19A10, Rabbit mAb#2478, Cell Signaling), anti-p-VEGFR3 (CY1115, Cell Applications Inc., San Diego, CA, USA) or anti-β-actin (MAB1501, Chemicon, Temecula, CA, USA) at appropriate dilution (1:500 to 1:2,000). The membrane was washed three times and incubated with horseradish peroxidase conjugated secondary anti-mouse or anti-rabbit immunoglobulin (IgG) (1:5,000; Sigma-Aldrich) for 1 hour prior to signal development by chemiluminescence reagent (Amersham Pharmacia Biotech, Little Chalfont, UK).

Transfection and RNA inference. Non-specific and human THEMIS2-specific SMART-pool-targeting small interfering RNA molecules were obtained from Dharmacon (Lafayette, CO, USA), and the sequences were identical to previous study (16). Briefly, siRNAs were transfected into selected cells for 24 hours using a specialized transfection reagent (Catalog number sc-29528, Santa Cruz Biotechnology, Dallas, TX, USA) according to the manufacturer’s protocol. The pCMV6-puro plasmid containing full-length THEMIS2 cDNA was transfected into MDA-MB-231, MDA-MB-231-IV2, Hs578T or Hs578T-FPI cell lines, and stable THEMIS2-expressing cancer cell lines were selected by using puromycin (Sigma-Aldrich). miRNAs including miR-125a-5p, miR-125b-5p, anti-mi125a-5p and anti-miR125b-5p were transfected into cells using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA). Western blot was used to determine the effects of gene silencing or overexpression.

In vivo tumor growth and metastasis assay. For the tumor growth assay, 1×10⁶ cells of control, miR-125b-5p-transfected or miR-125b-5p-transfected/THEMIS2-overexpressing Hs578T cells were orthotopically implanted into the mammary fat pads of CB17-SCID mice. After tumors reached 0.8 to 1 cm³, tumor pieces were surgically implanted under the ovarian bursa of additional mice (n=5 for each group). In accordance with the Institutional Animal Care and Use Committee (IACUC) guidelines, the orthotopic tumors in this study’s mice are generally not to exceed 1.5 cm (15 mm). The tumor volumes were monitored twice per week using the IVIS imaging system for bioluminescence detection over a 3-week period. In vivo metastasis assay was performed, in which Hs578T-FPI cells (5×10⁵) transfected with both of miR-125b-5p and Anti-125b-5p, miR-125b-5p alone, or control miRNA were suspended in PBS and injected individually into the tail vein of 6 to 8-week-old female CB17-SCID mice, which were then monitored for 30 to 60 days before sacrifice. Lung tissues were removed, fixed, paraffin-embedded, serially sectioned, and subjected to hematoxylin and eosin (H/E) staining. Tumor volumes were measured once a week, and tumor growth was evaluated until 4 months after implantation. All procedures were performed in accordance with the Guide for Care and Use of Laboratory Animals issued by the Institutional Animal Care and Use Committee of China Medical University, Taiwan (CMUIACUC-2021-140-2).

Kaplan-Meier plotter analysis. Kaplan-Meier plots were generated through the automatic selection of the optimal cutoff values between lower and upper quartiles, which formed the high and low expression groups. This analysis was performed on patients with breast cancer who had either received effective chemotherapy treatment or no treatment (https://kmplot.com/analysis/). Clustering of the transcriptomic profiles of 78 triple-negative breast cancer (TNBC) cases was obtained from the ONCOMINE database.

Transcriptomic sequencing data analysis. Two transcriptomic RNA-seq datasets were obtained from Gene Expression Omnibus (GEO) hosted by National Center of Biotechnology Information (NCBI, https://www.ncbi.nlm.nih.gov/geo/). The first dataset (GSE268201) was generated using single circulating tumor and normal cells that were isolated using density gradient centrifugation and CD45-based immunomagnetic depletion from liquid biopsies of breast cancer patients. The second dataset (GSE7842) that contains miRNA expression data from 93 primary human breast tumors and five normal breast samples was utilized to assess expressions of miRNA-125a-5p and miRNA-125b-5p within normal breast samples versus primary breast tumors. Briefly, the read alignment was conducted using aligners HISAT2 (16), transcript abundance was quantified by counting mapped reads using Kalisto (17), followed by differential gene expression analysis and normalization on edgeR as previously described (15).

Statistical analysis. All of the in vitro experiments were reproducible and repeated independently at least three times, and the results are presented as means±SEM. Except where otherwise noted, data presented in figures are shown as mean±standard deviation (SD). Independent samples t-tests were conducted using SPSS Statistics for Windows, version 18.0 (SPSS Inc., Chicago, IL, USA) to determine differences in the frequencies and means of the categorical data, p-values less than 0.05 were considered statistically significant.