Our Research

Epithelial renewal in the Drosophila intestine


To understand how cell-cell interactions, diet, the microbiota, and infection regulate the proliferation of intestinal stem cells, facilitating tissue self-renewal in healthy animals and, in tumor models, a loss of growth control.


Model explaining epithelial renewal in the Drosophila midgut.

Figure 1: Model explaining epithelial renewal in the Drosophila midgut. Cell lineage depicted with black arrows, cell signaling interactions depicted with red.

Like mammals, insects have intestinal stem cells (ISCs) that drive dynamic self-renewal of the gut epithelium. The simplicity of the Drosophila midgut and the superior genetic tools available make this an attractive model for studies of stem cell biology and gut homeostasis. Cellular and molecular similarities to the human gut suggest relevance to diseases like colorectal cancer and inflammatory bowel disease. In exploring how ISC division is controlled, we discovered that damage to the intestinal epithelium results in a burst of stem cell division that facilitates rapid regeneration. To understand this regenerative response we have analyzed the functions of most of the major signaling pathways in each of the five intestinal cell types. We discovered that gut epithelial stress or damage induces JNK and p38-Map Kinase activity and suppresses Hippo signaling in epithelial enterocytes (ECs), and that this triggers the production of secreted cytokines (upd2, upd3) and EGFR ligands (Vn, Spi, Krn) that stimulate JAK/STAT and Ras/MAPK activity in the intestinal stem cells, activating them to grow and divide (Fig 1).


Our elucidation of a feedback system explaining gut homeostasis (Fig 1) highlights new areas of research that can enhance our understanding of tissue biology and human disease. These include: how damage is sensed by the gut epithelium, how stress signaling activates growth factor and cytokine signaling and inflammation, how growth factor signaling activates stem cells, and how transformed stem cells develop into tumors and subvert the mechanisms that mediate normal gut homeostasis. Our understanding of the intestinal stem cell niche, how stem cell pools are controlled, and how differentiation is coordinated with ISC division is also incomplete. Classic questions in cell biology, such as how growth factor signaling activates cell growth and the cell cycle machinery, also remain insufficiently answered, and ripe for study in this system. Projects addressing each of these questions are currently underway in the lab. 

Ongoing Research

  1. How is gut epithelial damage sensed? 
    • When gut enterocytes (EC) are damaged or lost, the surrounding epithelium responds by producing IL-6- Leptin-like cytokines (Upd2, Upd2), which trigger a signaling cascade that activates ISCs to grow, divide, and produce new ECs. Our understanding of this damage response is so rudimentary that we don’t even know if the intial signals are chemical or mechanical. To identify the genes and pathways the gut uses to sense damage and trigger Upd3 expression, we are testing candidate pathways (Jnk, p38, Hpo) and processes (Ca++, ROS signaling), and performing a genome-scale RNAi screen. Genes found to be essential for damage sensing will be characterized and assigned to regulatory pathways. The results should yield general insight into how epithelial integrity is maintained, and how dysfuntion in damage sensing systems contributes to losses in epithelial homeostasis that lead to diseases like chronic inflammation and cancer.
  2. How does EGFR signaling activate ISC proliferation?
    • Drosophila migdut with ISC-derived tumors.

      Figure 2: Drosophilamigdut with ISC-derived tumors (green cells) generated by co-expressing NotchRNAi andRasV12 in stem cells. DNA is blue and expression of a reporter gene for a stress responsive cytokine (Upd3-LacZ) is red.

      Signaling via the epidermal growth factor receptor (EGFR), Ras, and the mitogen activated protein kinase (MapK) is required and sufficient for ISC activation downstream of gut epithelial damage. Although intensively studied and central to an overwhelming number of cancers, how Ras/Mapk signaling actually drives cell growth and division is surprisingly poorly understood. We recently identified the transcriptional repressor Capicua (Cic) and two of its targets, the ETS transcription factors Pnt and Ets21C, as critical Ras-dependent effectors of ISC activation. Each of these genes is associated with human cancer, most likely as effectors of the Ras oncogene. Our current studies combine target gene identification with functional tests to determine the mechanisms by which these transcription factors activate ISC growth and division. We are also testing whether there are Cic- and ETS-independent functions of MapK signaling, and if there are, we will undertake to identify the critical targets.
  3. What niche and host factors support the growth of ISC-derived tumors?
    • We developed a tumor model in which intestinal stem cells expressing a differentiation inhibitor (NotchRNAi) and an EGFR signaling activator (RasV12G, RafAct, Pnt, Ers21C, or CicRNAi) generate rapidly growing tumors (Fig 2) that can be serially transplanted in flies. After transplantation these aggressive tumors grow rapidly in many parts of the fly, recruit trachea, and kill the host. A gene replacement strategy is being used to determine the functions required to allow niche-independent stem cell tumor growth. Using genetically modified hosts, are also testing the requirement from the host animal, checking processes such as cell adhesion, epithelial tension, oxygen supply, blood cells, and nutrition, in niche-independent tumor growth. An RNAseq screen will also be used to identify other potentially relevant host processes. We hope to determine the full catalog of factors and conditions required for transformed ISCs to grow outside the stem cell niche.
  4. Can ISC proliferation and differentiation be recapitulated in vitro? 
    • The lack of long-term primary cell or organ culture remains a major limitation to Drosophila as a model system for research. We are attempting to develop primary culture of Drosophila intestinal stem cells (ISCs) by testing combinations of genetic manipulations and culture conditions. A protocol for culturing ISCs will allow many new approaches for studying stem cell biology. These include live analysis of cell proliferation, lineage asymmetry and differentiation, cell-based RNAi screening, and molecular techniques like RNAseq, ChIP, and protein mass spectrometry that work best with large numbers of synchronous homogeneous cells.

Growth-dependent control of the Cell Cycle


To understand how the initiation of DNA replication (G1/S transition) is coordinated with rates of cell growth and controlled by growth factor signaling.


In growing animal cells, rates of cell cycle progression are usually tightly coupled to rates of cell growth (mass accumulation), such that cells divide each time they double their mass. The growth of animal cells is limited by nutrition via TOR signaling, and also regulated by signaling through receptor tyrosine kinases, cytokine receptors, and G-protein coupled-receptors. Genes in all of these pathways are highly associated with cancer development, and can be targeted for cancer therapy. A central question in cell biology is how growth factor signaling via these conduits controls the cell cycle. Many studies, both classic ones using human cells and from our lab using the Drosophila system, indicate that growth-coupled cell cycle control is executed principally at G1àS rather than G2àM transitions. Studies with human cells have engendered a model, depicted in many textbooks, wherein growth factor signaling triggers phosphorylation cascades that lead to the transcriptional activation of genes encoding G1 Cyclin-dependent Kinase (Cdk) complexes, which in turn trigger DNA replication (Fig 3A). However our studies in Drosophila, as well as many studies from mice and yeast, do not support central aspects of this model. Instead, we suggest an alternative mechanism wherein translational regulation of limiting cell cycle regulators determines rates of G1àS progression (Fig 3B).

Textbook model (A) and our revised model (B) for growth factor-regulated cell cycle entry.

Figure 3: Textbook model (A) and our revised model (B) for growth factor-regulated cell cycle entry. A is modified from Weinberg’s text “The Biology of Cancer” (2014). Similar diagrams may be found in Alberts et al (2014) 

We have been researching this issue using endocycling cells of the Drosophila gut and salivary glands. Endocycles consist of DNA Synthesis (S) and Gap phases (G) without intervening Mitoses (M). In most endoreplicating cells nutrition-dependent Insulin/TOR signaling is the principle driver of cell growth and DNA replication, but other inputs such as EGFR/MAPK signaling can also have important growth promoting functions. In 2011 we reported that the fly’s endocycles employ a novel bi-phasic oscillator in which the transcription factor E2F1 promotes cyclin E (cycE)transcription, CycE/Cdk2 then triggers S-phase, and S-phase in turn causes the destruction of E2F1 by activating a DNA-replication specific ubiquitin ligase, CRL4cdt2 (Fig 4). Strikingly, factors that alter protein synthetic rates such as nutrition, TOR or Ras activity regulate E2F protein levels and thereby make endocycle progression growth-dependent. Our tests indicated that E2F protein levels in this context are controlled translationally, suggesting that translational efficiency provides the “missing link” between cell growth and G1àS progression in Drosophila. Recent studies in our lab indicate that in another endoreplicating cell type, the midgut Enteroblast (EB), EGFR/Ras/MapK signaling drives endocycling, also via translational control of E2F1. Notably, the mechanism we propose (Fig 4) need not be relevant only to endocycles, but potentially controls G0 and/or G1 length in proliferating diploid cells.


Our ongoing studies have two goals. First, we are testing our translational model for coupling cell growth to G1àS progression, using reporter mRNAs derived from E2F1 and alterations in signaling and nutrition to modulate translation. Our intention is to confirm the translational model, define the mRNA sequences that mediate translational efficiency, and determine how they do this. Second, we would like to know which elements of the endocycle mechanism we’ve elucidated in Drosophila (Fig 4) are conserved in mammalian cell cycles and in G1àS control in general. To investigate the question of conservation we are assaying the periodicity of tagged human E2F proteins, and their dependence on CRL4Cdt2, in cultured human cells.

Ongoing Research:

Model for growth-dependence of G1S transitions.

Figure 2: Drosophila migdut with ISC-derived tumors (green cells) generated by co-expressing NotchRNAi and RasV12 in stem cells.DNA is blue and expression of a reporter gene for a stress responsive cytokine (Upd3-LacZ) is red.

  1. Translational control of E2F1.
    • To further validate our model for G1àS control, we are characterizing the serum- Ras-, and TOR-dependent translational control of a set of E2F 5’UTR reporters. If these tests show that E2F 5’ UTRs mediate E2F translation, we will create further 5’ UTR deletion constructs to fine-map the mRNA sequences responsible (these will be translation suppression sequences). Then, CRISPR technology will be used to mutate these regulatory sites in the E2F locus in cells and flies, and their cell cycle progression phenotypes will be assayed. We expect such mutants to have shortened G1 periods, and to show some degree of growth factor- and nutrition-independence for G1àS progression. Using the E2F 5’ UTR elements defined above, we will screen for upstream translational regulators.
  2. How does EGFR/MAPK signaling drive cell growth?
    • EGFR/RAS/MAPK signaling drives the growth of many types of animal cells, most notably cancer cells transformed by activated mutant forms of RAS, RAF, NF2, ErbB2, and EGFR. It’s amazing that we don’t understand how this signaling pathway drives cell growth, but it clearly does. In Drosophila, the most profound growth effects we’ve discovered are in the adult intestinal stem cells (ISCs), Enteroblasts (Fig 1, Fig 2) and their progenitors, the larval “Adult Midgut Progenitor” cells (AMPs). We are taking a multi-faceted approach to determine how EGFR signaling makes these cells grow. Epistasis tests are being used to test candidate downstream effectors such as RSK, AKT, and TOR components, and transcriptome and proteome profiling is being performed to identify potential targets.
  3. Translational control of mitotic cycles via E2F1.
    • Although most of our work has focused on endocycling cells, we believe that the cell cycle ingrowing mitotic cells is also highly subject to translational control. To test this theory, we are evaluating Drosophila E2F1 as a growth sensor in two mitotic cell types: wing progenitors and intestinal stem cells. In these experiments we alter oncogenic growth signaling (Myc, PI3K, Ras, Tor, Hpo) and measure how this affects levels of E2F1 protein, mRNA, and rates of cell division. If the results support translational control, the specific mechanism will be sought using E2F1-UTR mutants and trans-acting factors identified as above. These experiments will provide examples of how a growth sensor regulates proliferation in stem and progenitor cells, further validating the mechanism’s relevance to development and human disease.
  4. Do human cells use growth-sensing cell cycle regulators?
    • Answering this question could revise a central, cancer-relevant paradigm in cell biology and present new strategies for cellular growth control. In this project we are using ribosome profiling of normal human epithelial cells (RPE-1) to identify growth factor-dependent, translationally regulated genes that potentially regulate the cell cycle. Candidate genes so defined will be functionally tested to determine whether they actually regulate cell proliferation. Using ribosome density maps, we will address specific mechanisms of translational control and how these interface with growth and growth factor signaling.