br termined as described previously For all
termined as described previously. For all experiments, female NOD.Cg-PrkdcscidIl2rgtm1Wjl/SzJ (NSG) skeletally mature mice
(6–8 weeks old) were obtained from the Virginia Commonwealth Uni-versity (VCU) Cancer Mouse Models Core Laboratory at Massey Cancer Steroids 150 (2019) 108447
Center (Richmond, VA). All animal experiments were approved by VCU’s Institutional Animal Care and Use Committee.
MCF7 tumors were created and treated with 24R,25(OH)2D3 or vehicle as previously described . HCC38 tumors were created by injecting 1 million HCC38 APTSTAT3-9R in 50 μL of 1:1 DPBS:BME (Cultrex Pathclear basement membrane extract Type 3, 14.5 mg/mL; #3632; R& D Systems, Minneapolis, MN) solution into the lactiferous duct of the right fourth mammary gland of a female 6–8 week old NSG mouse (N = 8). After 2 weeks, mice were injected intraperitoneally with 100 ng of 24R,25(OH)2D3 or a vehicle three times a week as previously described . After 10 weeks, tumors were harvested and measured along two perpendicular axes (length and width) with digital calipers beginning at 2 weeks post-cell injection, and tumor volume was calcu-
lated using the following formula: Volume = . Compar-
isons between vehicle and control treated tumors for both cell lines were made using a two-way analysis of variance (ANOVA). P-va-lues < 0.05 were considered significant.
2.5.1. Accessing CYP24A1 gene transcript expression data
CYP24A1 expression data were publically available from The Cancer Genome Atlas (TCGA) cohort (Project Id: TCGA-BC) (n = 525 patients) (https://portal.gdc.cancer.gov/projects/TCGA-BC) using University of California Santa Cruz (UCSC) Xena (http://xena.ucsc.edu/ ) [10,11]. In brief, gene expression data from estrogen-receptor-positive and estrogen-receptor-negative breast tumors were downloaded from Xena along with information about sample-specific CYP24A1 expres-sion. Overall survival (OS) was defined as the time of death by any cause and capped at five years. Kaplan-Meir curves of OS were gener-ated by GraphPad Prism v5.0 (GraphPad Software Inc.). Patients were grouped according to median gene expression as described previously [12,13].
2.6. Statistical analysis
Data are from 3 separate experimental repeats, each with an n = 6, and are presented as the mean ± standard error of an aggregate of all experimental repeats normalized to treatment/control. Data were analyzed by Mann-Whitney t-tests against the vehicle, or control group, which was normalized to 1.0 (dashed lines). Experiments with more than one independent variable (Fig. 3) were analyzed with two-way ANOVA with Bonferroni’s modification of Student’s t-test post-tests to determine the variance between row and column factors. TCGA data were analyzed with a survival curve log-rank analysis. All statistics were done in GraphPad Prism 6 ©.
3. Estrogen receptor profiles and their relationship with vitamin
Breast cancer can be broadly defined by the expression of three key molecular markers: estrogen receptors (particularly ERα), progesterone receptors, and human epidermal growth factor receptor 2 (HER2). The relative expression of these markers characterizes the four major types of breast cancer: luminal, normal breast-like, HER2 overexpressing, and basal breast cancer . Both luminal and normal breast-like cancers typically overexpress ERα and lack HER2, while basal-like cancers ty-pically do not express either receptor . HER2 overexpressing can-cers can be either ERα positive or negative. Luminal and normal-breast type cancers that overexpress ERα but lack HER2 expression have widely been considered the easiest cancers to treat with the best patient outcomes; while HER2 and basal cancers that do not express ERα have lower survival odds [14,16,17].