Genetic mutations can arise from imperfect DNA replication and failure to detect and repair DNA damage. While mutations are innocuous most of the time, alteration of key genes can lead to severe diseases including cancer. Cells have intrinsic failsafe DNA damage response programs leading to apoptosis or senescence, that eliminate the damaged cells. In addition, recognition and elimination of cells expressing mutant proteins is carried out by the immune system. However, tumors arise from cells that eventually escape these failsafe mechanisms.
Recently, it has been demonstrated that oncogenic mutations are present at a high rate in normal human skin epidermis, the tissue of origin of basal cell carcinoma and squamous cell carcinoma (SCC). The Beronja lab (Human Biology Division) hypothesized that a process other than apoptosis or senescence could be involved in the control of the tumorigenic process in skin cells. Dr. Ying and colleagues reported their methodology and findings in the November issue of the journal Nature Cell Biology.
The authors used a very elaborate and delicate technique to perform an in vivo screen to look for modifications affecting the proliferation and tumorigenic potential of skin cells. They performed ultra-sound guided in utero injections of lentiviruses in the mouse epidermis. These lentiviruses contained shRNA or Open Reading Frames (ORFs) that once expressed would reproduce alterations found in human SCC. By sequencing the epidermis 21 days after birth, they identified modifications promoting (enriched shRNA or ORF sequences) or inhibiting (depleted sequences) clonal expansion of infected cells. Dr. Beronja: “This allowed us to approach this study in an unbiased and comprehensive way”. Surprisingly, from all the potential candidates, the introduction of the oncogenic mutant PIKCAH1047R, found in most epithelial cancers, actually inhibited the expansion of cells (and not promoting it as expected). To confirm this result, the authors injected specific lentiviral constructs in reporter mice in which cells expressing the oncogenes (infected cells) express a green fluorescent protein (GFP) while non-infected cells express a red fluorescent protein (RFP). They observed that cells expressing PIKCAH1047R formed smaller clones than the control, validating the in vivo screen.
To understand the mechanism underlying this unexpected result, the authors turned to a mouse model expressing a physiological level of PI3KCAH1047R. They observed an increased proliferation of skin progenitor cells without expansion of these clones, although there was no increase in apoptosis or senescence. Thus, it was clear that another mechanism was governing this tumor suppressive effect of the oncogenic mutation. The authors developed an elegant in vivo EdU/BrdU pulse chase assay, “a method that enabled [them] to, for the first time, assess cell fate choice in vivo and precisely quantify rates of progenitor cell renewal and differentiation”, says Dr. Beronja. They observed that a greater proportion of cells expressing PI3KCAH1047R differentiated quickly after dividing, thus decreasing the renewal rate of the progenitors that would have otherwise expanded as a tumor clone. RNA sequencing confirmed the establishment of a transcriptional program governing differentiation in progenitor cells expressing PI3KCAH1047R. As Dr. Beronja highlights, they “discovered stem cell differentiation as the primary block to oncogene-driven clonal expansion in skin epithelium”.
Using a two-photon intra-vital microscopy technique, they demonstrated that the oncogenic mutation of PI3KCA led to an increased proportion of asymmetrical divisions (differentiation of one of the two daughter cells) and symmetrical differentiation (differentiation of the two daughter cells) despite a reduced symmetrical renewal of skin progenitors. This imbalance is at the origin of the phenotype observed.
The researchers performed a second in vivo genetic screen to identify drivers of differentiation or proliferation. Using ultra-sound guided injection, 1062 shRNAs were introduced in utero, followed by an EdU pulse to identify proliferating cells. They found that the shRNA against Sh3rf1, a gene coding for a protein downstream of the PI3K/AKT signaling, was enriched in differentiated cells and in Edu-negative (non-proliferative) cells. They also demonstrated that active PI3K signaling leads to the phosphorylation of SH3RF1 by AKT, which prevents SH3RF1 from activating the JNK signaling pathway that inhibits the differentiation of skin progenitor cells. In other words, oncogenic activation of the PI3K signaling relieves the brake on the differentiation program. Dr. Beronja explains how their work provides “a detailed cellular and molecular mechanism for how skin may tolerate and therefore accumulate oncogenic lesions over time without showing any functional or morphological defects”.
“The study was focused on the most commonly mutated oncogene in epithelial cancer, and further highlighted the importance and feasibility of using physiological models of tissue growth where the full complement of cellular behaviors and molecular mechanisms can be investigated. It also suggests a completely new and unanticipated hypothesis that stem cell differentiation is the dominant tumor suppressive force in skin. In effect, this suggests that stem cell fate choice rather than the trinity of proliferation/apoptosis/senescence is the critical driver of cancer”, says Dr. Beronja. This provocative thought is currently under investigation in his laboratory.
This work was supported by the National Institute of Health, the Cell & Molecular Biology Training Grant and a Thomsen Family Fellowship.
Fred Hutch faculty member Dr Slobodan Beronja contributed to this research.
Ying Z, Sandoval M, Beronja S. 2018. Oncogenic activation of PI3K induces progenitor cell differentiation to suppress epidermal growth. Nature Cell Biology. Nov;20(11):1256-1266.