Cancer cells may evade the immune response thanks to a special type of collagen, scientists reported in a new study published in the July 21, 2022, issue of Cancer Cell. Researchers at The University of Texas MD Anderson Cancer Center have discovered oncogenic collagen secreted by cancer cells that differs from normal collagen. It modified the tumor microbiome and provided some resistance to the immune system. This collagen is exclusive to cancer, making it a potential therapeutic target.

Collagen is the most abundant protein in the human body. It is part of the extracellular matrix of bones (containing 90% collagen), tendons (80-90%), skin or cartilage (50-70%), arteries (10-25%), lung (10%) and liver (4%). There are different types of collagen. The most common is collagen type I (Col1), synthesized by fibroblasts and other cells. Col1 consists of a triple helix made up of three chains (two alpha1 chains, and one alpha2 chain) forming a heterotrimer. The COL1a1 gene encodes the alpha1 chains, while the COL1a2 gene encodes the alpha2 chain.

In their studies, the team found that oncogenic collagen was produced in small amounts and forms a distinct extracellular matrix. Cancer cells only expressed the COL1a1 gene. In these cells, different mutations, such as the Kras mutation, silenced the COL1a2 gene by hypermethylating the promoter region of the alpha2 chain. This epigenetic feature caused cancer cells to only produce the alpha1 chain, a helical homotrimer containing three alpha1 chains, which is not like normal collagen.

"This collagen is completely different. It is more resilient. It cannot be degraded easily. Normal collagen is good for us. This collagen has the opposite function. It doesn't allow immune cells to come near. It makes cancer cells proliferate, survive better and behave differently," Raghu Kalluri, chair of Cancer Biology and director of operations for the James P. Allison Institute, told BioWorld Science.

The scientists carried out key experiments in knockout mice for the COL1a1 gene in pancreatic cancer. This is a way to eliminate the synthesis of alpha1 in cancer cells and prevent them from forming oncogenic collagen (which needs a homotrimer of alpha1 chains to form).

Later, when analyzing the results, the researchers observed several effects. These included a lower proliferation of cancer cells, changes in the composition of the bacteria of the tumor microbiome; a corresponding decrease in MDSC myeloid-derived suppressor cells (by reducing expression of the CXCL5 gene, which attracts MDSC cells); an increase and a greater infiltration of T cells (by increasing the expression of CXCL16, which attracts T cells), and therefore, a greater elimination of cancer cells.

Kalluri and his collaborators have shown two additional effects: an increased formation of normal collagen (which is known to inhibit tumor progression), and the consequent alteration of the signaling pathways that interfere with the immune profile of the tumor.

The researchers found that oncogenic collagen disrupts alpha3 integrin-binding signaling pathways, leading to increased proliferation of cancer cells. By blocking the alpha3 integrin in vivo, they successfully inhibited Col1-promoted cancer cell adhesion, increasing T-cell infiltration and survival.

These results open the door to a long list for different approaches against cancer focusing on cancer collagen itself, on the recovery of the normal collagen, or the signaling pathways in which alpha3 integrin is involved.

"This discovery has opened up a whole new way of thinking about some therapies, like sending T cells or drugs to cancer cells specifically. If you can, somehow, find something that only binds homotrimer collagen, this will only go to cancer cells, because no other place has these homotrimers. It is a very unique environment," Kalluri explained. A second option, he said, would be to interfere with the signals in which this collagen makes cancer cells proliferate or increase their survival.

However, there is a third pathway that seems promising: the alpha3 integrin. "The most exciting thing is that one can make reagents, antibodies, or small molecules to target alpha3 integrin, disabling to bind the homotrimer. If you block that, the effect of the homotrimer is gone. That could be a great target, a monoclonal antibody for the alpha3 integrin." Although the mouse model of this study is for pancreatic cancer, oncogenic collagen can also be found in other types of cancer such as lung, colon, head and neck. In their next steps, the researchers could take several paths.

"We are going to first understand why oncogenes influence hypermethylation. If we can prevent the methylation, maybe the homotrimer will not be made, and the heterotrimer will be made," Kalluri remarked. "We can also try to target the homotrimer. For example, if you can generate CAR T cells that would go where the homotrimer is, next to cancer cells, maybe you could kill them." They could try different strategies, including the alpha3 integrin and other cancer types. "We want to see how common this process is in another cancer," he added.