Development of skin-integrated melanoma organoids to study the tumor microenvironment-mediated mechanisms of melanoma progression and metastasis in vitro.

Abstract

e21547 Background: Resistance to targeted therapies remains a challenge in metastatic melanoma. The tumor microenvironment (TME) influences cancer progression and drug resistance, making it a promising therapeutic target. However, the lack of human in vitro 3D models that mimic the melanoma-TME crosstalk in a realistic 3D environment has limited the development of such therapies. To address this, we generated integrated skin-TME-melanoma organoids (mTMEOs) that incorporate key melanoma TME components and can serve to investigate and pharmacologically target the TME-mediated mechanisms of tumor progression in vitro. Methods: Primary human keratinocytes, melanocytes, fibroblasts, and pre-adipocytes were co-aggregated with GFP-labeled melanoma cells (SK-MEL-28 or A375) to form melanoma-skin organoids (mSOs), recapitulating melanoma anatomy, demonstrated via immunostainings. To generate the mTMEOs, mSOs were embedded in fibroblast-laden collagen gels, simulating the dermal stroma. mTMEOs were imaged in fluorescence using the Vireo Multi-Camera Array Microscope (MCAM) every other day, and a custom ImageJ-based pipeline was developed to track labeled GFP signals within the gels, allowing quantification of the number, size, and distance traveled of the invading melanoma cell clusters. Results: mSOs recapitulated key skin features, including a layered epidermis and dermal-hypodermal core, as well as melanoma-specific traits like epidermal spread, atypical melanocytes, and external migration. mTMEOs showed radial-to-vertical growth phase transition, enabling tumor invasion into the dermis-like stroma. Using the fluorescence images from the Vireo MCAM, we generated a Euclidean map that allowed the measurement of melanoma migration distances from the closest mSO. The analysis indicated that the presence of stromal fibroblasts in the surrounding gels did not influence melanoma invasion rates, but enhanced migration distances and led to fewer but larger melanoma cell clusters when compared to cell-free collagen gels. These findings align with literature suggesting that, while dermal invasion may be triggered by interactions with other cell types, stromal fibroblasts may play a tumor-supportive role by increasing melanoma spreading and shifting toward collective cell migration, associated with more aggressive metastases. Conclusions: Our novel mTMEO model replicates melanoma TME anatomical and pathological features, which can be altered by modifying individual organoid’s components. mTMEOs and the subsequent analysis workflow provide a versatile and tunable platform for studying TME-driven mechanisms of tumor progression and for drug development. By incorporating patient-derived cells, these constructs have the potential to serve as a personalized drug screening tool to improve patient outcomes.