Maps of Immune Cell Connections Can Predict Cancer Survival

A growing cancer is shaped by more than just the tumor cells it contains; the tissue around a tumor also alters its biology. Now, researchers at The Jackson Laboratory (JAX) have combined advanced imaging techniques with a new computational method to probe how immune cells interact with each other in never-before-seen detail, revealing that the interactions between immune cells in the vicinity of breast cancer or melanoma can be used to predict immune responses to the cancers as well as patient outcomes.

The new approach, led by Jeffrey Chuang, a professor at JAX and senior author of the new study, showed that the more interactions between two specific immune cell types, for instance, the longer breast cancer patients tended to survive. The work, in collaboration with JAX professor Karolina Palucka, who is also director of the JAX Cancer Center and the Edison T. Liu Endowed Chair in cancer research, was published recently in the advanced online issue of Communications Biology.

"Researchers have long suspected that better characterizing this complex community, which includes immune cells, blood vessels, and signaling molecules, could shed light on how cancers grow, spread, and respond to treatment," said Chuang. "This new analysis lets us quantify the locations and interactions of cells and molecules in a way that has never before been possible using imaging."

Immune molecules

Immune cells communicate with each other by bringing critical signaling proteins to their surfaces, forming physical 'synapses' between cells. This movement of resources from the inner part of the cell to its surface helps coordinate immune responses against pathogens -- and cancers.

To understand how immune cells are interacting with each other in the tumor microenvironment -- the area around a tumor -- researchers often isolate those immune cells and measure what genes are active in each cell type. Alternatively, they can add fluorescent tags to particular proteins and use microscopy to visualize how much of those proteins are present in a cell by how bright the fluorescence is.

However, neither approach tells scientists whether the proteins are on the cell's surface at a synapse, contributing to interactions with other cells.

"There can be a number of reasons that an immune cell might contain a signaling protein but not be actively using that protein on its surface to interact with other cells," explained Zichao Sam Liu, a graduate student in the Chuang lab and co-first author, along with Drs. Victor Wang and Jan Martinek. "Just knowing that a protein is expressed by a cell only tells you a part of the story, not the whole story."

Chuang, Liu and their colleagues wanted to use existing microscopy data to see how signaling molecules were clumping together at immune synapses -- giving a fuller picture of the immune cells' interactions.

A new analysis technique

The team developed a method dubbed Computational Immune Synapse Analysis (CISA) which detects not only which cells within a tissue contact each other physically, but also whether key molecules are concentrated at those contact points.

The technique works by analyzing images of immune cells in a way that highlights the edges of the cells and potential immune synapses, then comparing to the localization of tagged molecules.

Focusing on the immune system's T cells, the researchers showed that CISA could be used to identify interactions between T cells and other immune cells in existing images of the tumor microenvironments of human melanoma samples.

"If we see a preferred localization of certain proteins toward these cell-cell interfaces, we think they are forming synapses and it implies that the cells are interacting," said Liu. "Our method lets us quantify this in the native tumor environment for the first time."

Further experiments revealed that synapses between T cells and macrophages, which engulf and digest pathogens and tumor cells, were associated with the increased proliferation of T cells.

Clinical lessons

The research team next tested whether the interactions between immune cells in breast cancer samples had any bearing on the progression of the cancers themselves. Indeed, they found that stronger connections between T cells and B cells, another type of immune cell, were associated with better survival rates for patients. This observation could eventually lead to new ways of predicting patient outcomes, selecting patients who will most benefit from immune treatments, or even the development of new immunotherapies.

This kind of result is the ultimate goal of CISA -- picking up new patterns in cell interactions with biological importance. Chuang's lab group has made the image analysis platform available online for other scientists, and say that it ultimately could be used to analyze contacts between any cell types. It also works to analyze different types of images; the melanoma samples were imaged with histocytometry, while the breast cancer samples were imaged with imaging mass cytometry (IMC).

"We've tried to make this a broadly applicable, easy-to-use package," said Liu.

Chang and Liu plan to apply the method to other tumor types and to other immune cell types to continue gaining insight into the tumor microenvironment and its impact on cancer.

-Note: This news release was originally published by the Jackson Laboratory. As it has been republished, it may deviate from our style guide.