MED-1,2 and the C. elegans Endomesoderm

The EMS cell, formed at the 4-cell stage, is a mesendoderm progenitor: EMS divides to produce an anterior cell, MS, and a posterior cell, E. The E cell generates the entire gut (endoderm) while MS generates mesodermal cells (including muscle and part of the feeding organ, the pharynx). We have shown that both cells are specified by the genes med-1,2.

How do the same genes specify two very different fates? The MS and E cells are made different from each other at the 4-cell stage.

The neighbor of EMS, called P2, polarizes EMS so that the daughter cell closest to P2 becomes the E cell. The P2-EMS signal uses components of the conserved Wnt signaling pathway, which is important for many developmental processes (including cancer in humans). The endpoint of Wnt signaling in the MS/E decision is the differential activity of a regulator called POP-1.  In the MS cell, POP-1 represses the endoderm fate by directly repressing the E-specifying genes end-1 and end-3. In the E cell, this activity of POP-1 is blocked.

GFP::POP-1 localization in vivo

We have shown that the asymmetric activity of POP-1 results from the movement of POP-1 from the nucleus to the cytoplasm. This results in high nuclear levels of POP-1 in the anterior cell, MS, and low nuclear levels in E.  POP-1 movie

Using an in vivo assay that allows visualization of protein-DNA interactions, we found that GFP::POP-1 directly binds to the end-1,3 genes in living embryos.



localization of tbx-35 mRNA in situ

Our recent work has shown that in MS, MED-1,2 activate the regulator tbx-35. Embryos lacking tbx-35 arrest without most of the tissues normally made by MS. Shown at left is the localization of tbx-35 mRNA in an early embryo, using an in situ hybridization protocol developed in our laboratory.

In the absence of POP-1, tbx-35 continues to be activated in the MS cell, suggesting that an additional nuclear Wnt coregulator specifies MS-E differences.


We have found that while the MED-1,2 regulators structurally related to the broad GATA class of transcription factors, they bind a non-canonical target site. As med-like genes are found only in Caenorhabditis species, we hypothesize that this subclass of regulators evolved very recently. We have an ongoing collaboration with the laboratory of Joel Mackay, University of Sydney, Australia to study the solution structure of the MED-1 DNA-binding domain. (See MED-1 Structure page.)

More recently we have found that ceh-51, a target of TBX-35 and at least one other factor, provides MS-specifying functions in tbx-35 mutant embryos. As a result of this, double mutant ceh-51; tbx-35 embryos have a strong MS defect that is similar to the MS defect in med-1,2 mutants. Part of this work was performed in collaboration with Paul Sternberg (Caltech).

We are currently trying to further elucidate the genes that function in MS specification downstream of TBX-35 and CEH-51. Other current projects include examination of expression patterns of other embryonic genes, and evaluation of differences in the endomesoderm gene network in other nematode species.

P2-EMS induction (Goldstein, 1992)   POP-1 (Lin et al., 1995)   Wnt signaling (Rocheleau et al., 1997)
Wnt signaling (Thorpe et al., 1997)  MED-1,2 (Maduro et al., 2001)  POP-1 in vivo (Maduro et al., 2002)
MED-1 binding site (Broitman-Maduro et al., 2005)
 evolution of MED-type factors (Coroian et al., 2006)
tbx-35 specifies MS (Broitman-Maduro et al., 2006)  ceh-51 paperceh-51 and tbx-35 (Broitman-Maduro et al. 2009)