This publication journey began with an opportunity to study real patient tumor samples directly from breast surgeries. With newfound access to tumor material, we aimed to establish a 3D model that would preserve the identity of the tissue ex vivo as it would be in the patient for subsequent drug studies. Most tumor samples we receive from the surgeries are of the luminal hormone receptor positive (ERa+) subtype, so we were interested in how well ERa expression was preserved within our model. Preserving this hormone receptor is key for developing new treatment strategies against hormone receptor positive breast cancers – easy peasy, right? Well not so much.

We started out by examining the luminal and basal phenotypes of breast tissue cells over time. We expected healthy human breast tissue to form a saclike cavity consisting of a hollow lumen surrounded by inner layer of luminal cells and outer layer of flat basal cells. However, within two days the structures switched to a basal phenotype. And within a week, luminal cells were nowhere in sight! The phenotype switch itself was less striking than the fact that it happened so quickly. Thanks to this we started to think that the hormone receptor expression might not be an intrinsic feature of cancer, but more a reflection of the microenvironmental conditions that are present in vivo.

This began our quest to prevent the cells from turning to a basal phenotype so that theoretically we could maintain estrogen receptor expression. Through a panel of different matrices, we were able to complete the quest – almost. It turned out that even though we managed to maintain a luminal phenotype in mouse and human cells, the actual expression of ERa was still missing! This means that luminal identity features and ERa expression are independently regulated in a 3D culture.

At this point we stepped outside our comfort zone and took a wider approach by taking into account the mechanical microenvironment with a little help from our new collaborators. We reached out to Professor Olli Ilkkala’s whose team at Aalto University is distinguished for their work in nature inspired material mimetics. They had the tools and know-how to measure additional mechanical properties of the matrices in question.

By combining our areas of expertise, we made an interesting finding. Mouse mammary epithelial cultures (MMECs) grown in the stiffest culture condition we had had the most similar gene expression profile to the original uncultured mouse tumor. To our great joy, the gene expression profile included estrogen signaling. Additional digging of the expression data revealed that the mechanism was likely due to the activation of the p38 stress-signaling pathway which was triggered by the stiff microenvironment (and later confirmed by directly activating this pathway). Stiffness = p38 stress pathway = ERa expression. Bam! Done! Or not…

To our great disappointment, the human tissue material did not produce similar results. We figured that the issue must have something to do with the fact that the human breast microenvironment is even MORE stiff than the mouse mammary gland. Therefore, we started to systematically search for materials that could form firm gels without harming the cells within them. Unfortunately, many of the materials that we tested contained toxic components, or the gelling conditions needed to create the culture would require unphysiological pH or temperature. At some point we realized the “perfect” polymer that would fit to our demands would be too challenging to create.

These were frustrating times, but in hindsight we really got to use our imagination. Even though making the ideal stiff 3D matrix was proving to be impossible, we persevered and figured out that the stiffness does not need to necessarily need to come from the gel itself. Since one of our candidates, agarose, forms a solid gel, we thought it might be able to handle external compression without breaking. This is when we started to think about using magnets to create pressure on the gel (thus increasing the pressure or “stiffness” within the gel).

Lo and behold! ERa was maintained in our human tissue cultures! Before celebrating too soon, we tested whether these tumors would react to anti-hormonal therapies that breast cancer patients would typically get. With promising results, we were finally ready to share our success and to start on our next adventure.