Photo credit: CC0 Public Domain

Finding drugs that can kill cancer cells while leaving normal tissue intact is a holy grail of oncology research. In two new papers, scientists from UC San Francisco and Princeton University present complementary strategies to solve this problem with “intelligent” cell therapies – living drugs that remain inert unless triggered by combinations of proteins that only exist in cancer cells occur.

Biological aspects of this general approach have been studied for several years in the laboratory of Wendell Lim, Ph.D., and colleagues from the UCSF Cell Design Initiative and the Center for Synthetic Immunology sponsored by the National Cancer Institute. However, the new work adds a new dimension to this work by combining the latest therapeutic cell technology with advanced calculation methods.

For an article published in Cell Systems on September 23, 2020, members of Lim’s laboratory teamed up with the research group of computer scientist Olga G. Troyanskaya, Ph.D., from the Lewis Sigler Institute for Integrative Genomics at Princeton and the Flatiron of the Simons Foundation partnered institute. Using a machine learning approach, the team analyzed extensive databases containing thousands of proteins found in both cancer and normal cells. They then combed through millions of possible protein combinations to compile a catalog of combinations that could only target cancer cells precisely while leaving normal cells alone.

In another article published in Science on November 27, 2020, Lim and colleagues showed how this computationally derived protein data can be used to advance the design of effective and highly selective cell therapies for cancer.

“Right now, most cancer treatments, including cell therapies, are told to block or kill it,” said Lim, also a professor and chair of cellular and molecular pharmacology and a member of the UCSF Helen Diller Family Comprehensive Cancer Center. “We want to improve the nuance and sophistication of the decisions a therapeutic cell makes.”

For the past decade, chimeric antigen receptor (CAR) T cells have been in the spotlight as an effective method for treating cancer. In CAR T cell therapy, cells of the immune system are taken from a patient’s blood and manipulated in the laboratory to express a specific receptor that recognizes a very specific marker or antigen on cancer cells.

While scientists have shown that CAR-T cells can be very effective and sometimes curative in blood cancers such as leukemia and lymphoma, the method has so far not worked well on solid tumors such as breast, lung or liver cancer. Cells in these solid cancers often share antigens with normal cells found in other tissues, which carries the risk that CAR-T cells could experience side effects through targeted treatment to healthy organs. In addition, solid tumors often create suppressive microenvironments that limit the effectiveness of CAR T cells.

For Lim, cells are like molecular computers that can sense their environment and then integrate that information to make decisions. Because solid tumors are more complex than blood cancers, “you have to make a more complex product” to fight them, he said.

In the Cell Systems study, led by Ruth Dannenfelser, Ph.D., a former PhD student on Troyanskaya’s team at Princeton, and Gregory Allen, MD, Ph.D., a clinical associate in the Lim laboratory, the researchers examined public databases Examination of the gene expression profile of more than 2,300 genes in normal and tumor cells to determine which antigens can help distinguish them. The researchers used machine learning techniques to identify potential hits and determine which antigens are clumping together.

Based on this gene expression analysis, Lim, Troyanskaya, and colleagues applied Boolean logic to combinations of antigens to see if they could significantly improve T cell recognition of tumors while ignoring normal tissue. For example, using the Boolean AND, OR, or NOT tumor cells could be distinguished from normal tissue using the markers “A” OR “B” but NOT “C”, where “C” is an antigen that is unique to normal tissue is found.

To program these instructions into T cells, they used a system called synNotch, a customizable molecular sensor that synthetic biologists can use to optimize the programming of cells. Developed in 2016 in the Lim laboratory, SynNotch is a receptor that can be engineered to recognize a variety of target antigens. SynNotch’s output response can also be programmed so that the cell performs any range of responses whenever an antigen is detected.

To demonstrate the potential power of the collected data, the team programmed synNotch T cells to kill kidney cancer cells that express a unique combination of antigens called CD70 and AXL. Although CD70 is also found in healthy immune cells and AXL in healthy lung cells, T cells with a constructed synNotch AND logic gate only killed the cancer cells and spared the healthy cells.

“The field of big data analysis of cancer and the field of cell engineering have exploded in recent years, but these advances have not been merged,” said Troyanskaya. “The computational capabilities of therapeutic cells combined with machine learning approaches enable actionable exploitation of the increasingly available rich genomic and proteomic data on cancer.”

The work described in the new science paper, led by ex-UCSF graduate Jasper Williams, shows how multiple synNotch receptors can be chained to create a variety of complex cancer detection circuits. Because synNotch has the ability to “plug and play” expression of selected genes, these components can be linked in different ways to create circuits with various Boolean functions that enable precise detection of diseased cells and a range of responses when those cells are present identified.

“This work is essentially a cell engineering handbook that gives us blueprints for building different classes of therapeutic T cells that can recognize almost any kind of combinatorial antigenic pattern that could exist on a cancer cell,” Lim said.

For example, a synNotch receptor can be engineered so that when the cell recognizes antigen A, it makes a second synNotch that recognizes B, which in turn can induce the expression of a CAR that recognizes antigen C. The result is a T cell that requires the presence of all three antigens to trigger the killing. In another example, if the T cell encounters an antigen that is present in normal tissues but not in cancer, a synNotch receptor with a NOT function could be programmed to target the T cell that carries it , dies, protecting normal cells from attack and possible toxicities impact.

In the science paper, Lim and colleagues show, using complex synNotch configurations like this one, that they can selectively kill cells that carry various combinatorial markers for melanoma and breast cancer. When T cells equipped with synNotch were injected into mice that carried two similar tumors with different antigen combinations, the T cells efficiently and precisely located the tumor they had engineered for detection and reliably carried out the cell program designed by the scientists.

Lim’s group is currently investigating how these circuits in CAR T cells can be used to treat glioblastoma, an aggressive form of brain tumor that is almost always fatal with conventional therapies.

“You’re not looking for just one magical target. You’re trying to use all of the data,” Lim said. “We need to comb through all of the cancer data available to find unique combinatorial signatures for cancer. If we can do that, it could be using this usher in smarter cells that really harness the computational sophistication of biology and have real impact on combat. ” Cancer.”

New cell technology techniques can lead to precise immunotherapies

More information:
“Precise T-Cell Recognition Programs Developed by Transcriptionally Linking Multiple Receptors” Science (2020).… 1126 / science.abc6270 Provided by the University of California, San Francisco

Quote: Big Data supports the design of “intelligent” cell therapies for cancer (2020, November 26th), accessed on November 26th, 2020 from therapies.html

This document is subject to copyright. Except for fair trade for the purpose of private study or research, no part may be reproduced without written permission. The content is provided for informational purposes only.


Please enter your comment!
Please enter your name here