In mouse skin, patches of cells that lose p53 have more Wnt signaling at the center than at the edge

In Short. In mouse skin, patches of cells that have lost p53, a protein that can stop damaged cells from dividing, show a growth signal that is strongest at the center and weakest at the edge, and patches with that pattern grew larger than patches in which the signal was high everywhere, though this has not been shown in people.

THE PROTEIN p53 is made from a gene called TP53. When a cell's DNA is damaged, p53 can stop the cell from dividing or make it die. A 2017 review in the journal Cell calls TP53 the most frequently mutated gene in human cancer. This article is about what happens in mouse skin when cells lose p53. The work was done in mice, in the outer layer of the skin, and it has not been shown in people.

A commentary published alongside the study, by Dan Engelman and Cédric Blanpain, describes what one might predict. When p53 is activated by stress such as DNA damage, it turns on genes that cause cell cycle arrest, which is a pause in cell division, cellular senescence, which is a lasting stop to division in a cell that stays alive, or cell death. They write that one might expect the growth of cells that lose p53 to come from a failure of these usual functions. Whether that is what happens in skin is the question taken up below. First, some background about mutant cells and about skin.

Mutant cells in healthy skin

Cells pick up changes in their DNA as a person ages, and skin that is exposed to sunlight picks up more. A cell that carries such a change passes it to its descendants. A patch of cells that are all descended from one such cell is called a clone. Healthy human skin contains many clones, including clones of cells with a mutated p53 gene. A 1996 study of human skin samples found patches of cells with p53 mutations, and reported that in sun-exposed skin the patches were both more frequent and larger than in sun-shielded skin. A 2015 study in Science sequenced small biopsies of sun-exposed eyelid skin and found that several genes linked to cancer carried mutations in many skin cells, while the skin kept the normal functions of its outer layer.

According to the authors of a 2018 study in Cell Stem Cell, most mutant skin cells persist in normal tissue without forming tumors, and aging skin remains normal in appearance and function. In that study the researchers placed one specific p53 mutation in the skin of mice. The mutant cells first outcompeted their normal neighbors and then returned toward normal behavior, and the authors concluded that several mechanisms limit the growth of such cells. In that experiment, a p53 mutation alone did not make a clone grow without limit. The editor's summary of the new study begins from the same observation. It says that healthy epithelia contain cells with cancer-associated mutations and that most of these "remain growth restricted," while cells lacking p53 "can expand and occupy large tissue areas." Why those clones keep growing is the question that the rest of this article follows.

How skin renews itself

The epidermis is the outer layer of skin. It is made of stacked layers of cells. The cells in the lowest layer, which researchers call progenitor cells and which this article calls stem-like cells, are the ones that divide. Their descendants that mature stop dividing, move toward the surface and are shed there, so cells are continually lost from the surface and replaced by division in the lowest layer.

When a stem-like cell divides, there are three possible results. Both daughter cells can stay stem-like. Both can mature. One can do each. Staying stem-like is called self-renewal, and maturing is called differentiation. The 2018 Cell Stem Cell study describes a model of healthy mouse skin. In the model, the result of any single division cannot be predicted. The chance of two stem-like daughters equals the chance of two maturing daughters, so on average across the tissue equal numbers of each kind are made.

This has a consequence for a clone. If more of the daughter cells of its stem-like cells stay stem-like than mature, each generation leaves more stem-like cells behind, and the clone grows. Growth therefore depends on the balance between the two outcomes. The speed of division is a separate property. A 2018 study in Nature Cell Biology showed that the two can differ. In mouse skin with an activated mutation in a cancer-linked gene, the stem-like cells divided more, and they also matured more often. The authors reported that this skin had a growth disadvantage, and clones of the mutant cells were smaller than clones of normal cells.

What Wnt signaling does in skin

Cells in a tissue send chemical messages to their neighbors. Wnt signaling is one such message. In skin, it helps keep cells stem-like. A 2013 study in Science found that stem-like cells in mouse skin need Wnt signaling to divide, and that these cells make Wnt signals themselves, along with secreted proteins that inhibit Wnt over long distances. The authors suggested that this is how the cells renew themselves. A review published the same year describes Wnt signaling as the dominant pathway influencing the decisions of adult stem cells in the skin about which cell types to become.

The amount of a signal can differ from place to place. When the amount changes steadily with position, for example highest at one spot and lowest far from it, researchers call the pattern a gradient. In what follows, Wnt activity means how much Wnt signaling is occurring in a cell. A gradient of Wnt activity across a clone means that cells in different parts of the clone have different amounts.

What the mouse experiments found

The study, by Qiwen Gan, Wei Li, Rachel Lex, Zhe Ying and Slobodan Beronja, was published in Science on Oct. 1. The full text is behind a paywall, so the account below rests on the paper's structured abstract and editor's summary, the abstract of the commentary, and the earlier studies cited above. It reports no clone sizes or counts, because the sources used do not give them.

The authors used mouse skin. They combined lineage tracing, which follows the descendants of marked cells, with a quantitative analysis of the fate of daughter cells. They report that p53 loss shifted the stem-like cells toward self-renewal and away from maturing, "allowing the progenitor pool to expand over successive cell divisions." In their account, clones lacking p53 grew by enlarging the stem-like population, and they did not find that the usual stopping pathways had been disabled. The abstract states the result as "p53 suppresses clonal expansion by limiting progenitor self-renewal."

A second experiment points the same way. The authors used a form of p53 that cannot start cell cycle arrest, senescence or cell death. That form of p53 still restrained the renewal of stem-like cells. The authors conclude that targets of p53 other than the usual ones suppress clonal expansion. The commentary's abstract describes the study as reporting a gradient of Wnt signals that is linked to self-renewal of epidermal stem cells.

To find those other targets, the authors combined several kinds of work. They measured gene activity. They mapped where p53 attaches to DNA, using a method the abstract names as chromatin immunoprecipitation sequencing. They ran genetic screens in mouse skin, which test many genes in turn for an effect on cell behavior. And they imaged Wnt activity in single cells.

The work identified three genes. Their full names are secreted frizzled-related protein 1 (Sfrp1), low-density lipoprotein receptor-related protein 1 (Lrp1) and ubiquitin-specific peptidase 22 (Usp22). The abstract calls them direct targets of p53, meaning genes that p53 switches on by attaching to them, and "the most genetically upstream components" of the network that p53 controls, which means that in the genetic tests they act earlier in the chain of effects than the other genes found. The structured abstract says the three genes "suppress Wnt signaling at both the ligand and intracellular levels." A ligand is the signal molecule that cells release, and intracellular means inside the cell, so the three genes together reduce Wnt signaling both outside and inside cells. The sources used do not say which gene acts where, or what else each gene does. After p53 was lost, the abstract says, the expression of the three genes fell and Wnt activity rose.

How the authors tested the gradient claim

Because Wnt is suppressed at more than one level, the abstract says, the rise in Wnt activity across a p53-deficient clone was not even. It "was lowest at the edge of the clones and rose progressively toward their centers, creating a radial gradient that persisted across different stages of clonal expansion regardless of clonal density." Radial here means that the amount varies with distance from the center of the clone. In plain terms, the cells at the center of a mutant patch had the most Wnt activity, the cells at its edge had the least, and this pattern was seen at early and later stages and in patches of different density.

A gradient and a high level of Wnt everywhere are different conditions. A clone could have a gradient with a modest overall level, or no gradient with a high level throughout. The authors changed Wnt signaling genetically to separate the two. In the structured abstract they write that gradient Wnt activation "was associated with robust clonal expansion." They also made clones with uniformly elevated Wnt activity, including some in which the average Wnt activity was higher than in the p53-deficient clones, and those clones "expanded more slowly." The editor's summary states the comparison briefly: clones exposed to the gradient "expanded further than those with uniformly high Wnt activity."

Higher overall Wnt therefore did not predict greater expansion in these comparisons. The abstract states the result in careful words. It says that a gradient of Wnt activity, compared with uniform elevation, was "associated with" sustained clonal expansion. That is the authors' own wording, and it reports a link between the two. The Mount Sinai news release on the study describes the same comparison and the same screen of more than 1,000 candidate genes, and it agrees with the paper's account on the points above.

What the study does not show

All of the experiments described were done in mice, and in the epidermis. No source used for this article reports measuring Wnt activity across clones in human skin. The human studies cited above show that clones with p53 mutations exist in human skin. They report no measurement of a Wnt gradient in those clones.

The clones were followed in otherwise normal skin, and the sources used report no tumors in these experiments. The commentary notes that in humans, loss-of-function mutations in TP53 lead to expansion of a mutated cell and its progeny in epithelial tissues, including sun-exposed skin, and that this expansion can progress to invasive cancers. Whether the gradient has a role in that progression is not something the study reported.

Further points limit what can be said. The earlier mouse study described above used a single mutated copy of p53, and this one concerns loss of p53. The sources used do not say how the loss was produced in these experiments, so this article cannot say how closely the two situations match. This is one study. It is reported here from its abstract, editor's summary and the commentary's abstract, without the figures, clone sizes or counts in the full paper.

In mouse skin, clones that lost p53 grew into the surrounding normal tissue mainly because more of their cells stayed stem-like, a gradient of Wnt activity across each clone was associated with that growth, and the study does not show whether human skin has the same gradient.