Aggressiveness Potential of Spontaneous Canine Mucosal Melanoma Can Dictate Distinct Cancer Stem Cell Compartment Behaviors in Regard to Their Initial Size and Expansion Abilities

Abstract

Mucosal melanoma represents one of the most highly metastatic and aggressive subtypes of melanoma. The biology of mucosal melanoma is poorly documented, and the lack of experimental models makes it difficult to design and test new therapies. Dogs are frequently affected by melanomas of the oral cavity, making spontaneous canine melanoma a potentially predictable model for their human counterpart. We recently established and characterized two new canine mucosal melanoma cell lines named OCR_OCMM1 and OCR_OCMM2. Here, we identified quiescent cancer stem cell (CSC) subpopulations in both canine cell lines that displayed similarities to human quiescent CSCs: canine melanoma CSCs had the ability to self-renew, produced nonstem cell (SC) progeny, and formed melanospheres that recapitulated the phenotypic profile of the parental tumor. These CSCs also formed melanoma in immunodeficient mice, and the inhibition of PI3K/AKT signaling expanded the CSC pool. A subset of non-CSCs transitioned to become CSCs. OCR_OCMM1 and OCR_OCMM2 displayed different CSC compartment behaviors in regard to their initial size and expansion abilities. Collectively, this study showed that the OCR_OCMM1 and OCR_OCMM2 canine melanoma cell lines are powerful cellular tools to study melanoma SCs, not only for mucosal but also for the more common human cutaneous melanoma.

Keywords: comparative oncology; melanoma; stem cells.

FIG. 1.

Identification of a slow-cycling cell subpopulation with SC markers in canine melanoma cell lines. (A) OCR_OCMM1 (left) and OCR_OCMM2 (right) cells were plated at a clonal density of 1,000 cells/mL and were grown in sphere-forming conditions. After 7 days of culture, large tumor-like spheres formed, and images were taken with an inverted microscope. Scale bar, 100 μM. (B) Spheres divided slower than adherent cells. ***P < 0.005. The PDT was calculated as described in the Materials and Methods section for spheres (sph) and adherent cells (adh). (C) Immunostaining of adherent and spheroid OCR_OCMM1 and OCR_OCMM2 cells for Ki67 expression (green). Nuclei were stained with DAPI (blue). Scale bar, 50 μM. The arrows indicate Ki67-positive cells, and the red arrows indicate Ki67-negative (quiescent) cells. (D) Histograms showing the percentage of DiI^high-labeled slow-cycling/quiescent cells in OCR_OCMM1 and OCR_OCMM2 spheroids. (E) Spheroids were enriched for ABCB5^pos cells compared with adherent conditions. *P ≤ 0.05. (F) Representative flow cytometry graphs of ABCB5^posDiI^high cells (CSCs) in OCR_OCMM1 (left) and OCR_OCMM2 (right) spheroids. (G) Their quantification is represented with histograms. *P ≤ 0.05. CSC, cancer stem cell; DAPI, diamidino-2-phenylindole; PDT, population doubling times; SC, stem cell.

FIG. 2.

Identification of Rh123^low quiescent cells with stem-like properties. (A) After dissociation, OCR_OCMM1 and OCR_OCMM2 spheroid cells were incubated with 0.1 μg/mL Rh123 for 20 min (left panel). After 60 min of Rh123 exclusion, a subpopulation of low Rh123-retaining (Rh123^low) cells appeared (right panel). (B) Rh123^low and Rh123^high cells were sorted by FACS, and the cell cycle was assessed by flow cytometry after KI67/PI staining. A fraction of cells in the G0 phase of the cell cycle is shown in red. *P ≤ 0.05 and ***P ≤ 0.005. OCR_OCMM1 (C) and OCR_OCMM2 (D) Rh123^low and Rh123^high fractions were sorted by FACS, replated at a clonal density of 1,000 cells/mL, and grown under sphere-forming conditions. After 7 days of culture, large tumor-like spheres formed, mainly from the Rh123^low fraction (quantitative graph below). The self-renewal capacity of melanospheres was assessed by dissociating and replating the cells at a clonal density for three successive generations (G₁–G₃). **P ≤ 0.01; ***P ≤ 0.005; and ^###P ≤ 0.005 (* and # for the comparison of Rh123^low vs. Rh123^high and G_n Rh123^low vs. G_n+1 Rh123^low, respectively). FACS, fluorescence-activated cell sorting; PI, propidium iodide; Rh123, rhodamine 123; SFU, sphere-forming unit.

FIG. 3.

Time- and ABC transporter-dependent effects of LY294002 on the size of the Rh123^low pool. OCR_OCMM1 (A) and OCR_OCMM2 (B) sphere cells were treated with 10 μM of LY294002 (+) or growth medium (−) for different durations during the Rh123 dye exclusion assay (pretreatment-30 min; labeling-20 min; and exclusion-60 min) (n = 8). A final concentration of 50 μM verapamil (V) was added as a control where indicated; (0) indicates no exclusion. *P ≤ 0.05; ***P ≤ 0.005; and ****P ≤ 0.001.

FIG. 4.

Slow-cycling cells, DiI^high LRCs, from melanospheres induce tumor formation in immunodeficient mice. DiI^high LRC slow-cycling cells from OCR_OCMM1 and OCR_OCMM2 spheroids had a higher tumorigenic potential than their DiI^low highly proliferative counterparts. Nude mice were s.c. injected with 5,000 DiI^high or DiI^low cells. (A) The results are expressed as the percentage of mice bearing tumors relative to the total number of injected mice. The histograms represent the mean ± standard error of the mean of two independent experiments. Graph showing the tumor sizes at each time point for OCR_OCMM1 (B) and OCR_OCMM2 (C) conditions. Tumor size was evaluated as described in the Materials and Methods section. (D) M-CAM and Melan-A expression was evaluated by immunostaining of tumor sections generated from DiI^high spheres cells. Nuclei were stained with DAPI. Scale bar, 50 μM. LRCs, label-retaining cells; s.c., subcutaneously.