Ephrin reverse signaling controls palate fusion via a PI3 kinase-dependent mechanism

Abstract

Secondary palate fusion requires adhesion and epithelial-to-mesenchymal transition (EMT) of the epithelial layers on opposing palatal shelves. This EMT requires transforming growth factor β3 (TGFβ3), and its failure results in cleft palate. Ephrins, and their receptors, the Ephs, are responsible for migration, adhesion, and midline closure events throughout development. Ephrins can also act as signal-transducing receptors in these processes, with the Ephs serving as ligands (termed "reverse" signaling). We found that activation of ephrin reverse signaling in chicken palates induced fusion in the absence of TGFβ3, and that PI3K inhibition abrogated this effect. Further, blockage of reverse signaling inhibited TGFβ3-induced fusion in the chicken and natural fusion in the mouse. Thus, ephrin reverse signaling is necessary and sufficient to induce palate fusion independent of TGFβ3. These data describe both a novel role for ephrins in palate morphogenesis, and a previously unknown mechanism of ephrin signaling.

Figure 1. EphB2 and ephrin-B2 expression in fusing palate epithelium

Day 14.5 embryos from mice harboring: (A) the EB2/LacZ chimeric allele or (B) the LacZ knock-in to the EphB2 locus were sectioned coronally and stained with X-gal. Counter stain is Nuclear Fast Red. βgal expression was found in the palate epithelium, suggesting a role in adhesion and/or fusion. Note the breakup and dispersion of EB2/βgal during EMT.

Figure 2. Eph and ephrin effects on palate fusion

Palatal shelves were dissected from eight day old chicken embryos and cultured in contact on a support for 72 h in the presence of specific treatments, as indicated. Tissues were fixed, paraffin processed, and sectioned. Shown are representative H&E stained sections from each treatment. Note the darkened epithelial layer that disappears as fusion proceeds. H&E stained sections from anterior to posterior were scored on for fusion on a scale of 1 to 5 at anterior, middle, and posterior points and these scores averaged to yield the mean fusion score (MFS) shown. Values are ±SEM, with n=7 to 9 for each group across three separate experiments. Magnification is 100×.

Figure 3. Ephrin dependence of mouse palate fusion

Embryonic day 14.5 mouse palates were cultured in the presence of unclustered EphA4/Fc soluble recombinant protein or IgG Fc control protein as described in the text. Tissues were fixed, paraffin processed, and sectioned. Shown are representative H&E stained sections from each treatment. H&E stained sections from anterior to posterior were scored on for fusion on a scale of 1 to 5 and these scores averaged to yield the mean fusion score (MFS) shown. Values are ±SEM for n=14 palates over four independent experiments. Magnification is 200×.

Figure 4. Effect of PI3K inhibition on Eph-induced palate fusion

Chicken palates were grown with the treatments indicated under the conditions described in the text. Samples grown in TGFβ3 or EphB2 alone fused almost completely. Addition of the PI3K inhibitor LY294002 abrogated fusion with either TGFβ3 or clustered EphB2/Fc. Shown are H&E stained examples of each group with n=16 to 19 for each group from 3 independent experiments. Mean fusion score (MFS) for each is shown ± SEM. Magnification is 100×.

Figure 5. Model of ephrin and TGFβ3 signal transduction in palate fusion

Ephrin and TGFβR signals intersect at a point upstream of PI3K, which is required for fusion. Other possible pathways from eprhin-Bs that do not go through PI3K are not diagrammed. Known possible effectors or ephrin-Bs in reverse signaling are described in the text. Signals from Eph RTKs that induce partial fusion are unknown.