Multiple breast cancer cell-lines derived from a single tumor differ in their molecular characteristics and tumorigenic potential

Abstract

Background: Breast cancer cell lines are widely used tools to investigate breast cancer biology and to develop new therapies. Breast cancer tissue contains molecularly heterogeneous cell populations. Thus, it is important to understand which cell lines best represent the primary tumor and have similarly diverse phenotype. Here, we describe the development of five breast cancer cell lines from a single patient's breast cancer tissue. We characterize the molecular profiles, tumorigenicity and metastatic ability in vivo of all five cell lines and compare their responsiveness to 4-hydroxytamoxifen (4-OHT) treatment.

Methods: Five breast cancer cell lines were derived from a single patient's primary breast cancer tissue. Expression of different antigens including HER2, estrogen receptor (ER), CK8/18, CD44 and CD24 was determined by flow cytometry, western blotting and immunohistochemistry (IHC). In addition, a Fluorescent In Situ Hybridization (FISH) assay for HER2 gene amplification and p53 genotyping was performed on all cell lines. A xenograft model in nude mice was utilized to assess the tumorigenic and metastatic abilities of the breast cancer cells.

Results: We have isolated, cloned and established five new breast cancer cell lines with different tumorigenicity and metastatic abilities from a single primary breast cancer. Although all the cell lines expressed low levels of ER, their growth was estrogen-independent and all had high-levels of expression of mutated non-functional p53. The HER2 gene was rearranged in all cell lines. Low doses of 4-OHT induced proliferation of these breast cancer cell lines.

Conclusions: All five breast cancer cell lines have different antigenic expression profiles, tumorigenicity and organ specific metastatic abilities although they derive from a single tumor. None of the studied markers correlated with tumorigenic potential. These new cell lines could serve as a model for detailed genomic and proteomic analyses to identify mechanisms of organ-specific metastasis of breast cancer.

Figure 1. Phase contrast photomicrographs of breast cancer cell lines at sub confluent and confluent stages.

(A) ARM-H (B) ARM-G (C) ARM-E (D) ARM-X cell lines (magnification x200).

Figure 2. Breast cancer cell lines induce tumors in nude mice and metastasize to different organs.

(A) Five female BALB/c nude mice received an injection of different breast cancer cells. #-Time when mice were sacrificed. Day 60 is the end of the experiment, mice were sacrificed and tumors were excised. (B) A comparison between different breast cancer cell lines in tumor growth in nude mice. The mean tumor size in each cell line group was compared. (C) Metastases of breast cancer cell lines in injected nude mice (D) Summary table of tumorigenicity and metastatic potential of breast cancer cell lines.

Figure 3. Photomicrographs of H&E stains of human breast cancer cell lines and metastatic tumors from mice.

(A) ARM-H (middle, 50×) shows focal residual skeletal muscle (arrow) almost completely replaced and infiltrated by invasive adenocarcinoma (between asterisks); (right, 200×) the same area shows sheets of markedly atypical and pleomorphic carcinoma cells with numerous mitotic figures in a haphazard growth pattern. (B) ARM-G (middle, 50×) lymph node with deposits of metastatic tumor cells within the peripheral sinuses and parenchyma (arrows); (right, 200×) same lymph node with numerous scattered metastatic tumor cells; tumor cells show enlarged nuclei, significant atypia and pleomorphism when compared to the surrounding normal lymphocytes, arrows indicate deposits of tumor cells. (C) ARM-C (middle, 50×) shows sheets of adenocarcnoma cells (between asterisks) invading into adjacent skeletal muscle (arrow); (right, 200×) view of the same area shows markedly pleomorphic and atypical tumor cells with areas of tumor cell necrosis (arrow). (D) ARM-X (middle, 50×) skin (arrow on epidermis), underlying adnexal structures and dermis completely replaced by invasive adenocarcinoma entrapping sweat glands and normal follicles; (right, 200×) the same section showing the markedly atypical tumor cells with mitoses, pleomorphic nuclei, and high nuclear to cytoplasmic ratio invading the dermis and surrounding adnexal structures (triangle); arrow indicates skin adnexal structures.

Figure 4. IHC stain of ERα in tumor isolated from nude mice.

(A) positive control human breast cancer tissue, (B) tumor from breast cancer cell line ARM-H derived from nude mice, arrows indicate negative mouse nodules.

Figure 5. Photomicrographs of IHC stain of ER and HER2 expression and western blotting analyses of ER-expression and response to E2 in breast cancer cell lines.

(A) SK-BR-3 (B) MCF-7 (C) ARM-H (D) ARM-G (E) ARM-E cell lines.

Figure 6. Photomicrographs of IHC stain of CK8/18 and mammaglobin for breast cancer cell lines.

(A) MDA-MB-231, (B) ARM-E, (C) ARM-X, (D) ARM-H, (E) ARM-G, (F) AMR-C; mammaglobin G) MDA-MB-231, (H) ARM-E, (I) ARM-X, (J) ARM-H, (K) ARM-G, (L) AMR-C (magnification 400×).

Figure 7. Expression of CD44/CD24 receptors on breast cancer cells.

The plots depict CD44, CD24 and an isotype control antibody staining of ARM-H, ARM-E and control MDA-MB-231. This is one representative experiment of three independent experiments.

Figure 8. Side population profile of breast cancer cell lines ARM-H, ARM-E and control MDA-MB-231.

The side-population cells are indicated in enclosed boxes and the percentage of the cells in this region is indicated in each panel. These data are representative of two independent experiments.

Figure 9. Representative metaphase (left) and interface (right) from four-breast cancer cell lines.

Each cell nucleus contains HER2 signals (red) and centromere 17 signals (green). (A) ARM-H (B) ARM-G (C) ARM-E and (D) SK-BR-3; (E) composite karyotype of chromosome 17 and derivatives in four cell lines; (F) summary table of FISH analyses; (G) number of chromosomes in breast cancer cell lines.

Figure 10. mRNA expression level and western blotting analyses of p53 and p21.

RNA was isolated from breast cancer cell lines, converted into cDNA, followed by q-PCR with (A) p53 or (B) p21 specific primers. Relative mRNA expression was calculated after normalization to the ribosomal protein (RPS11) control gene. Each experiment was done at least three times. Data represent mean ± SD. ANOVA one-way statistical analyses was used to compare results *p<0.05 considered statistically significant; (C) western blotting analyses of breast cancer cells treated with 5 uM of etoposide for 24 h followed by protein extraction, gel electrophoresis and antibody staining for indicated proteins.

Figure 11. Breast cancer cell lines express different levels of estrogen receptor and respond differently to estrogen and 4-OHT treatments.

(A) bottom. Western blotting analyses of breast cancer cells for ER expression and top RT-qPCR for cMYC gene expression in RNA extracted from indicated cells treated with 10 nM of E2 for 24 h. (B) Cytotoxicity assay of MCF-7, (C) ARM-E and (D) ARM-H breast cancer cells treated for 3 days with different concentrations of 4-OHT. Left axis represents the relative percentage of live cells treated with 4-OHT compared to medium-treated cells. Similar results were obtained in three independent experiments. Data represents mean ± SD. Student’s t test was used to compare means of treated versus untreated samples with *p<0.05 considered statistically significant.