Tracking Cell Competition in Colorectal Cancer

If tumorigenesis is akin to a marathon, not only are mutant cells intrinsically better equipped to stay the course and win, they also deliberately trip up normal cells early in the race. This was surmised in three recent, concurrently published studies probing the competitive dynamics between intestinal stem cells (ISC) harboring alterations in APC, KRAS, or PIK3CA, and their wild-type (WT) neighbors.

ISCs reside at the base of intestinal crypts, where they continuously self-renew. Although the cells in a given niche are mostly identical, those that function less well are replaced by their peers in an ongoing fitness contest to ensure maximum quality. Sometimes, though, this neutral process goes awry.

“We've been trying to understand how initial mutant cells become ‘fixed’ and are able to dominate in a tissue that's not quiescent, but rapidly renewing,” says Eduard Batlle, PhD, of the Institute for Research in Biomedicine in Barcelona, Spain. The usual brisk pace of ISC turnover, he points out, means “any mutants would have to really fight to overtake surrounding normal cells, which proliferate just as readily.”

As such, these three studies “are important, because they shed light on mechanisms by which neutral competition in the crypt niche can be bypassed,” adds Batlle, who was not involved in the research.

In one, Owen Sansom, PhD, of the Beatson Institute in Glasgow, UK, and his group focused on the tumor suppressor APC, whose loss of function is a key colorectal cancer trigger, by provoking unrestrained WNT signaling. Collaborating with Pekka Katajisto, PhD, at the University of Helsinki in Finland, the researchers found that APC-mutant ISCs, beyond utilizing hyperactive WNT for their own proliferation, also secreted WNT antagonists into the microenvironment, which “had this paracrine effect of inhibiting the growth” of nearby WT ISCs, Sansom explains (Nature 2021;594:430–5).

Of these WNT antagonists, the team reported that NOTUM was the most highly secreted, demonstrating—in vitro and in vivo—its essential role in enabling APC-mutant cells to steer their normal counterparts toward differentiation, not self-renewal. Nor were APC-deficient cells themselves vulnerable to NOTUM's growth-squashing effects; the APC mutation switches on constitutive WNT signaling downstream of NOTUM. The result was that APC-mutant cells gained a clonal advantage and could, in due course, “fix” the intestinal crypt with their progeny.

At Amsterdam University Medical Centers in the Netherlands, Louis Vermeulen, MD, PhD, and his research crew made a similar discovery. Their findings recapitulated Sansom's on how APC-mutant ISCs successfully outcompete WT cells: The latter “get exposed to WNT antagonists when they actually need WNT,” Vermeulen says. Although his group observed that NOTUM was a major culprit, “we think there are more players involved,” he adds—notably WIF1 and DKK2, which APC-deficient cells also secreted at high levels (Nature 2021;594:436–41).

“We concluded that the combined effect of all three antagonists is what most potently suppresses wild-type cell fitness,” Vermeulen says. “It makes sense—nature wouldn't bet on just one factor, but rather, a whole range of negative regulators to control” a pathway of WNT's importance to tissue homeostasis.

Meanwhile, Benjamin Simons, PhD, of Cambridge University, also in the UK, examined two well-known oncogenes, KRAS and PIK3CA. Mutations in either don't noticeably affect intestinal tissue architecture, Simons explains, whereas APC alterations “give rise to polyps, the first thing you see in a colorectal examination.” He and his team “wanted to profile tissue that wasn't physiologically disrupted to determine if early oncogenic mutations actually promote field cancerization”—a theory whereby, the thinking goes, mutant cells lacking visible abnormalities can form large mucosal patches that seed colorectal cancer (Nature 2021;594:442–7).

The researchers developed “a color-specific system to trace red-mutant and non–red wild-type clones in the same tissue,” Simons says. “We found that KRAS- or PIK3CA-mutant cells do two things: They secrete BMP ligands and other factors that degrade the crypt niche. They also exert long-range remodeling effects that perturb neighboring wild-type crypts, which they eventually outcompete.”

Together, these studies point to evidence in the small intestine “of a very intimate, reactive cross-talk not only between epithelial cells, but with the surrounding stroma too,” Simons observes. “It appears to be a rather ubiquitous pattern of behavior.”

With a better grasp of how mutant ISCs actively shape the microenvironment to increase their odds of “winning” in intestinal crypts, the question now is, “Can all this cross-talk we've uncovered provide direct therapeutic targets?” Vermeulen asks. His group is exploring the use of lithium chloride (LiCl), which activates WNT signaling downstream of the ligand–receptor level, by blocking GSK3β.

“The idea is to provide WNT where it's needed—in healthy cells—which may boost their fitness and hopefully render them insensitive to the suppressive effects of APC-mutant cells,” Vermeulen explains. Two other formulations, lithium bromide and lithium carbonate, are already used to treat patients with mood disorders, “so there's quite a bit of clinical knowledge about lithium, making it an attractive candidate to start a preventive study,” he adds. With funding from the Dutch Cancer Society, he plans to recruit individuals with familial adenomatous polyposis (FAP) who have a germline APC mutation that puts them at high risk for colorectal cancer.

“We're going for proof-of-concept, so polyp or adenoma formation is not the end point—that would require our trial running for some 20 years,” Vermeulen notes. “Instead, we'll evaluate participants' intestinal tissue to see if lithium reduces the rate at which mutant clones emerge and become ‘fixed’ in the crypt.”

“This is a clear therapeutic possibility,” Batlle agrees, “and definitely worth testing in those with FAP, especially if toxicity can be kept to a minimum.”

Sansom observes, too, that “there's an emerging class of synthetic WNT agonists, albeit their development focus has been tissue regeneration and repair.” As with lithium, these agents may have potential when it comes to priming WT cells in patients with high-risk FAP. However, he and his group are eyeing a different preventive approach: directly suppressing NOTUM.

Several candidate inhibitors are in the works for NOTUM, which is also expressed in the brain and thought to be associated with dementia, Sansom explains. None are clinical-stage agents, but “when one gets there, we'll be ready in the wings to emulate” the Dutch trial, he adds. Whereas lithium is “a broad tool that would hit quite a few things, we think targeting NOTUM is more specific.”

Although Simons's findings are less immediately actionable, “our focus now is the plasticity of epithelial cells, which naturally reprogram in response to injury or inflammation,” he says. “What happens if such reprogramming occurs on top of carrying an oncogene? We think it could ramp up the size of mutant clones and radically change their behavior.” The plan is to test this hypothesis in models of inflammatory bowel disease, he adds, “and we can envisage more preventive strategies down the road.”

In all, “the more we understand the biology of very early events in colon tumorigenesis, the better,” Batlle says, “as it may have implications for how patients are followed, even treated.”

Vermeulen agrees. “We should be smart enough to figure out interventions that disadvantage and make mutant ISCs less fit, while they're still few in number,” he says. “Then, through natural turnover, our bodies can take care of kicking them out.” –Alissa Poh