Epithelial cells use immune proteins to find and remove precancerous cells

Inside each person exists a quiet defense force just waiting for the signal to strike. Unlike the SWAT team that is the immune system, this protective force is more like a neighborhood watch. In their regular day-to-day lives, they are epithelial cells that line every inner and outer surface of the human body. But when a cell starts to look or act suspiciously, as if it may become cancerous, these regular epithelial cells put on their metaphorical police uniforms and remove the precancerous cell from the tissue in a process of cell competition called epithelial defense against cancer. How healthy epithelial cells recognize precancerous ones, however, remains a mystery.

“What's the initial difference between these cells that leads to an interaction?” asked Nicholas Baker, a cell competition researcher at Albert Einstein College of Medicine. It is unclear, he added, “whether it's an interaction between receptors and ligands on the different cells, or whether it's some other kind of interaction, like a mechanical interaction.”

In a new study published in Nature Immunology, a team of scientists led by senior author Takeshi Maruyama, a cell competition researcher at Waseda University, reported that precancerous epithelial cells upregulate expression of the adaptive immune system protein, major histocompatibility complex (MHC) class I, which healthy cells recognize via their expression of leukocyte immunoglobulin-like receptor subfamily B member 3 (LILRB3) protein (1). This recognition sets off a signaling cascade in the healthy cells neighboring the precancerous one, leading them to produce a mechanical force to remove the precancerous cell from the tissue.

The identities of the molecules that mediate this interaction are “really interesting and really important,” said Eugenia Piddini, a cell competition researcher at the University of Bristol who was not involved in the work. “We're looking at a phenomenon that could have very important implications for understanding what happens in tissues as well as for therapy.”

To identify the molecules involved in this precancerous cell recognition process, Maruyama and his team mixed healthy human skin epithelial cells in vitro with cells that express a constantly active form of the oncogene Ras. In healthy cells, Ras regulates cell proliferation, but in many cancers, Ras can become mutated in such a way that its expression never gets turned off, causing run-away cell proliferation that eventually leads to the development of a tumor. Thus, healthy cells would recognize a cells’ constant Ras expression as a sign that the cell could become cancerous.

After mixing the healthy and precancerous cells, the researchers noticed that the healthy cells upregulated their expression of a transcription factor called RUNX2. If Maruyama and his team deleted RUNX2 from healthy cells, the healthy cells could no longer eject precancerous cells, indicating that healthy cells need RUNX2 to recognize precancerous ones.

Because RUNX2 is a transcription factor, it must regulate the expression of a gene that healthy cells use to recognize precancerous cells. Maruyama’s team searched for genes that contained a RUNX2 promoter binding site. They further narrowed their search to genes that encoded a protein with a transmembrane domain because the researchers expected healthy and precancerous cells to recognize each other via the proteins on each of their cell surfaces.

From their search, the researchers identified the protein LILRB3. Healthy cells upregulated LILRB3 in the presence of precancerous cells, and healthy cells needed to express LILRB3 to remove neighboring precancerous cells. This finding points to LILRB3 as the protein that healthy epithelial cells use to recognize precancerous cells.

To find the protein that interacts with LILRB3 on the precancerous cell side, Maruyama and his team looked to prior research that reported that LILRB-family proteins can interact with MHC class I proteins, which are proteins important in the human adaptive immune response.

By deleting MHC class I genes in precancerous cells, the researchers found that healthy cells lost their ability to remove the precancerous cells, indicating that MHC class I expression on precancerous cells is important to mediate their removal from the group of healthy cells.

To confirm that MHC class I on precancerous cells interacted with LILRB3 on healthy cells, Maruyama’s team synthesized a peptide containing just the alpha 3 domain of the MCH class I protein. If the researchers treated precancerous cells with the alpha 3 peptide when there were no healthy cells around, nothing happened to the precancerous cells. On the other hand, when they treated precancerous cells with alpha 3 in the presence of healthy cells, the healthy cells removed more precancerous cells than when alpha 3 was not present.

Piddini was surprised at first that adding the alpha 3 peptide would lead to healthy cells removing more precancerous cells. She thought that the extra MHC class I peptides would compete with the endogenous MHC class I proteins expressed on the precancerous cells and result in decreased removal of precancerous cells.

But, she explained, by adding more MHC class I protein to the mix of healthy and precancerous cells, “they are overstimulating the ‘winners’ to extrude the ‘losers,’” she said. “When [alpha 3] reaches the cells where all of the signals are in place, it just potentiates that interaction, and… what they get is a good repression of tumorigenesis.”

Maruyama and his team next investigated what happens once MHC class I on precancerous cells interacts with LILRB3 on healthy cells. They saw that this interaction kicks off a signaling cascade that causes the protein filamin to accumulate in healthy cells at their borders with precancerous cells, leading to the production of a mechanical force that squeezes the precancerous cell out of the group. 

To investigate the interaction between MCH class I and LILRB3 in vivo, the researchers mixed healthy and precancerous cells in a basement membrane in vitro system and then injected the cell mixture into mice. When they treated the mice with the alpha 3 peptide, tumor growth slowed. If the researchers disrupted LILRB3, tumors grew larger, indicating that the MCH class I and LILRB3 interaction is important in the process of epithelial defense against cancer.

“At this stage, I think that it remains to be seen whether alpha 3 can be used for cancer prevention,” said Maruyama. “A normal cell extrudes the transformed cells from the epithelium just next to it, so if we can add a more powerful system to epithelial cells, maybe that can be available for a more powerful treatment.”

In addition to its treatment potential, Piddini wants to know how these molecular pathways activate in the first place.

“There are steps that are not well defined yet with respect to how mechanistically Ras activates this receptor in the ‘loser’ cells, as well as how that interaction with the wildtype cells leads to the upregulation of the receptor,” she said. “Part of that regulation may also help as an intervention.” 

Baker, who was not involved in the study, also wonders if the increased expression of MHC class I on precancerous epithelial cells could potentially act as an early-stage cancer diagnostic.

“MHC class I is not normally expressed in epidermis, but if you looked and you found occasional patches of MHC class I-expressing epidermal cells, could those be early-stage tumors?” he asked.

Moving forward, Maruyama plans to further investigate the molecular mechanisms involved in the MHC class I and LILRB3 interaction. He is also curious to learn how cancer cells escape this epithelial defense system. 

Overall, Piddini added, “this new study, now in a human system, bringing in molecules that are of clinical relevance like MHC, provides further amplification as well as the possibility of intervention for a pathway that could really revolutionize how we contain tumor growth.”

Reference

1. Ayukawa, S. et al. Epithelial cells remove precancerous cells by cell competition via MHC class I–LILRB3 interaction. Nat Immunol  22, 1391-1402 (2021).