🧬 biology

Developing a melanoma-skin-on-a-chip integrated with vasculature using an edgeless skin reconstruction approach

This study presents a bioengineered, vascularized melanoma-skin-on-a-chip platform utilizing an edgeless 3D reconstruction strategy to accurately mimic the human tumor microenvironment, thereby enabling the investigation of melanoma invasion dynamics and the evaluation of therapeutic responses.

Original authors: Felikss Rumnieks, Deniz Ornek, Rolando Perez-Lorenzo, Hasan Erbil Abaci

Published 2026-06-29

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Felikss Rumnieks, Deniz Ornek, Rolando Perez-Lorenzo, Hasan Erbil Abaci

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

Imagine trying to understand how a fire spreads through a forest. If you study the fire in a flat, two-dimensional drawing, you miss how the wind, the soil, and the trees interact. If you try to study it in a different type of forest (like a jungle instead of a pine forest), the results won't match what happens in the real world.

This is the problem scientists have been facing with melanoma, the most dangerous type of skin cancer. Traditional lab models are like those flat drawings or the wrong type of forest—they don't look or act like real human skin.

Here is how the researchers at Columbia University built a better model, which they call the "edgeless skin-on-a-chip."

1. The "Edgeless" Forest (The Skin Model)

Most lab-grown skin models are like islands with open edges. In nature, skin is a continuous sheet with no edges. The researchers discovered that this "edge" changes how cells behave, much like how a garden bed with a fence behaves differently than a field that goes on forever.

To fix this, they built a continuous, "edgeless" skin model. Think of it as a perfect, endless sheet of skin that mimics the real thing. This model has two main layers:

The Epidermis (The Roof): The top layer made of skin cells (keratinocytes).
The Dermis (The Foundation): The bottom layer made of connective tissue, blood vessel cells, and support cells (fibroblasts).

2. The "Trojan Horse" (The Tumor)

Instead of just sprinkling cancer cells onto the skin, the researchers created melanoma spheroids. Imagine these as tiny, 3D "cities" or "bubbles" containing a mix of cancer cells, support cells, and blood vessel cells.

They dropped these "cities" into the foundation layer of their edgeless skin model. Because the skin has no edges, the tumor settled in naturally, pushing its way into the tissue just like a real tumor does in a human body. This allowed the tumor to grow in a realistic environment, surrounded by the same "neighborhood" (the microenvironment) it would have in a person.

3. The "City Planning" (Vascularization)

One of the biggest challenges in cancer research is understanding how tumors build their own blood supply. In this model, the researchers added blood vessel cells (HUVECs) to the mix.

They found that the "cities" (tumors) didn't all build roads (blood vessels) the same way:

The MeWo Tumor: This type of tumor was like a city that quickly built a dense, complex network of roads right inside its own borders. It had a lot of traffic and connections.
The A375 Tumor: This type of tumor was different. It barely built any roads inside itself. Instead, the roads stayed mostly on the outskirts, in the surrounding tissue.

This difference is crucial because it shows that different types of melanoma behave differently, and this model can capture those unique "personalities."

4. Testing the "Firefighters" (Drug Testing)

Once the model was built, the researchers tested it like a real-world simulation to see how the tumors reacted to medicine.

Testing Drug A (for A375): They used a drug called a BRAF inhibitor. The result was like sending firefighters to a fire that was easy to put out. The cancer cells stopped growing, and many of them died (apoptosis). The model successfully showed that this drug works on this specific type of tumor.
Testing Drug B (for MeWo): They used a drug called a MEK inhibitor. Here, the result was different. The cancer cells stopped growing (the fire was contained), but they didn't necessarily die off in large numbers. The model showed that this drug stops the cancer from spreading but works differently than the first drug.

Why This Matters (According to the Paper)

The paper claims that this "edgeless skin-on-a-chip" is a powerful new tool because:

It's Realistic: It mimics the mechanical and structural complexity of real human skin, which older flat models miss.
It's Specific: It can show how different types of melanoma (like MeWo vs. A375) build blood vessels and react to drugs in unique ways.
It's a Better Simulator: It captures the "neighborhood" effects—how the tumor talks to the surrounding skin cells and blood vessels—which is essential for understanding how cancer grows and survives.

In short, the researchers built a tiny, edgeless, 3D skin city that can host different types of cancer "cities," allowing scientists to watch how they grow, build roads, and react to medicine in a way that looks and feels much more like the human body.

Technical Summary: Edgeless Melanoma-Skin-on-a-Chip with Vasculature

Problem Statement
Melanoma is the most aggressive form of skin cancer, characterized by high metastatic potential and frequent resistance to targeted therapies. Current preclinical models, including 2D cultures and animal models, fail to accurately recapitulate the complexity of the human tumor microenvironment (TME). Specifically, animal models suffer from significant physiological differences in epidermal thickness, immune composition, and extracellular matrix (ECM) organization compared to humans. Furthermore, conventional in vitro 3D skin constructs often possess open lateral boundaries ("edges") that do not mimic the continuous nature of human skin. These boundary conditions significantly alter cell-ECM dynamics, tissue tension, and mechanical properties, thereby compromising the predictive power of drug screening and mechanistic studies for melanoma progression and therapeutic response.

Methodology
The authors developed an edgeless human skin-melanoma chip (eSM-chip) using a continuous 3D skin reconstruction strategy to eliminate lateral open edges. The platform integrates four key compartments:

Dermal Compartment: Composed of primary dermal fibroblasts and, in vascularized versions, human umbilical vein endothelial cells (HUVECs) embedded in a collagen type I matrix.
Tumor Integration: Melanoma spheroids were pre-formed using a mixture of melanoma cell lines (A375 or MeWo), fibroblasts, and HUVECs within collagen droplets. These spheroids were embedded into the dermal layer of the chip.
Epidermal Compartment: A stratified layer of primary keratinocytes was seeded on top of the dermis to form a reconstructed epidermis.
Culture Conditions: The constructs were cultured submerged initially, then transitioned to an air-liquid interface (ALI) to promote epidermal differentiation. Vascularized constructs were maintained in serum-free media supplemented with growth factors (VEGF, FGF-2).

The study utilized two melanoma cell lines: A375 (harboring the BRAF V600E mutation) and MeWo. The model was evaluated through histological staining (H&E, Picrosirius Red), immunofluorescence (for markers like MLANA, K14, FAP, $\alpha$ -SMA, Ki67, and cleaved caspase-3), and quantitative image analysis of vascular networks (vessel area fraction, length density, and junction density). Therapeutic relevance was tested by treating the constructs with BRAF inhibitor (PLX4720) for A375 and MEK inhibitor (trametinib) for MeWo.

Key Results

Structural Integration and Microenvironment: The eSM-chip successfully supported the integration of melanoma spheroids beneath the reconstructed epidermis, mimicking the vertical growth phase of melanoma. Histological analysis revealed heterogeneous multicellular masses with disorganized collagen architecture within the tumor compartment, contrasting with the aligned collagen in the surrounding dermis. Cyst-like metastatic structures enclosed in Collagen IV-positive membranes were observed, suggesting retained metastatic potential.
Stromal Activation: The model recapitulated the reprogramming of fibroblasts into cancer-associated fibroblasts (CAFs). Expression of CAF markers (FAP and $\alpha$ -SMA) showed spatial heterogeneity, with FAP enriched at the tumor periphery and $\alpha$ -SMA varying by cell line and distance from the tumor, indicating distinct activation states dependent on tumor proximity.
Vascularization: The presence of fibroblasts within the spheroids significantly enhanced HUVEC network formation. When integrated into the chip, MeWo constructs exhibited significantly higher vascular coverage (2.3-fold higher vessel area fraction) and network density (3.65-fold higher vessel length density) compared to A375 constructs. A375 constructs were largely avascular intratumorally, with vasculature confined to the distal stroma, whereas MeWo constructs maintained both intratumoral and stromal vascularity.
Therapeutic Response: The platform demonstrated sensitivity to targeted therapies consistent with known mutation profiles.
- A375 (BRAF mutant): Treatment with PLX4720 resulted in a concentration-dependent increase in cleaved caspase-3 (apoptosis) and a reduction in Ki67 (proliferation).
- MeWo: Treatment with trametinib (MEK inhibitor) reduced Ki67 proliferation but did not increase apoptosis; instead, cleaved caspase-3 levels decreased, suggesting G1 cell cycle arrest without parallel cell death, consistent with known MEK inhibition mechanisms.

Significance and Claims
The paper claims that the eSM-chip provides a human-relevant platform that overcomes the limitations of traditional 2D and animal models by incorporating mechanical boundary conditions (edgeless design) and a multi-compartmental TME (epidermis, dermis, tumor, and vasculature). The authors assert that this model successfully recapitulates key aspects of melanoma biology, including:

The mechanical and structural complexity of the TME, including ECM remodeling and CAF activation.
Cell line-specific differences in angiogenic capacity and vascular network organization.
Mutation-dependent responses to targeted therapies (BRAF and MEK inhibitors).

The study positions the eSM-chip as a tool for studying melanoma invasion, tumor-microenvironment dynamics, and preclinical drug screening. The authors suggest it offers a foundation for developing personalized therapeutic strategies, noting that the current model's lack of perfusion is a limitation that could be addressed by future integration into microfluidic platforms to better model tumor-vascular crosstalk and immune trafficking.