Genome-wide screens in mice uncover regulators of oncogenic growth

We performed a series of unbiased in vivo screens of ~16,000 mouse genes, designed to identify candidates based on their potential to elicit a quantitative effect on epidermal growth during embryogenesis and oncogenic hyperplasia. By conducting our studies in the dynamic environment of the physiological system, we uncover many previously overlooked candidates in addition to the well-known growth regulators.

HFSP Long-Term Fellows Slobodan Beronja and Naoki Oshimori and colleagues
authored on Mon, 28 October 2013

Identification of mechanisms that regulate tissue growth is paramount to understanding human development, and to designing strategies to counter diseases such as cancer. Current functional studies of tissue growth include genome-wide cellular RNAi screening approaches, but as the in vitro paradigm incompletely models the complexity of the physiological environment, these have been largely focused on identifying cell intrinsic regulators of growth.

We have recently developed a method to conduct RNAi-mediated gene function studies in the mouse (ref. 1). Using ultrasound-guided microinjection of lentivirus into the amniotic space of mid-gestation embryos we achieved efficient, stable, and specific transduction of the mouse epidermis that, when coupled with lentiviral RNAi, enabled rapid assessment of gene function in skin morphogenesis. Given the efficacy of the RNAi approach in mouse skin in single gene or pathway studies (ref. 2, 3), we have now extended the scale of our analyses by several orders of magnitude and conducted the first genome-wide RNAi-mediated screens in mice.

Figure: From gastrulation to birth: developmental series of mouse embryos from gestational days 9 to 18.

We focused our study on epidermal growth over the nine days (gastrulation to birth) of normal development, when a single epidermal progenitor gives rise to a clone of ~40 cells, and during oncogenic (HrasG12V-induced) hyperplasia, where the overall growth rate is ~3-fold faster. We tested 16,000 genes and uncovered many expected and unexpected candidates. Importantly, by eliminating genes that were identified as important under both normal and oncogenic conditions, consistent with their general role as housekeeping/viability genes, we were left with ~250 candidates for oncogene-specific growth regulators. This is an especially attractive set as they represent genes that could be targeted in cancer without eliciting any ill effect on normal tissue homeostasis.

Among the top HrasG12V-dependent screen hits, we showed that β-catenin, an effector of the Wnt-signaling pathway, and Mllt6, a gene previously identified as a fusion partner of MLL in leukemia, are essential for embryonic hyperproliferation, where they promote mitogenic effects of Wnt-signaling. Importantly both of the candidates were also critical for initiation and maintenance of HrasG12V-dependent epidermal tumors in mice and in human squamous cell carcinoma xenograft models. Together, these results validate our experimental approach and strongly argue that developmental growth in mice is a sound model for uncovering genes that are essential in human tumorigenesis.


RNAi screens in mice identify physiological regulators of oncogenic growth. Beronja S, Janki P, Heller E, Lien WH, Keyes B, Oshimori N, and Fuchs E. Nature. 2013 Sep 12; 501(7466):185-90. doi: 10.1038/nature12464.

Other references

(Ref. 1) S. Beronja et al., Nature Medicine, 2010 16: 821-7.
(Ref. 2) S. Williams et al., Nature, 2011 470: 353-8.
(Ref. 3) N. Oshimori & E. Fuchs. Cell Stem Cell, 2012 10: 63-75.

Pubmed link

Nature news and Views commentary

Faculty 1000

Beronja Lab website