Differential recruitment of CD44 isoforms by ErbB ligands reveals an involvement of CD44 in breast cancer.

Members of the CD44 family of transmembrane glycoproteins control cell signaling pathways from numerous cell surface receptors, including receptor tyrosine kinases (RTKs). The decisive factor (ligand, RTKs or both) that controls the recruitment of specific CD44 isoforms is still unknown. We investigated this question by using the EGFR signaling pathway, in which one receptor can be activated by a broad range of ligands. By means of siRNA-mediated downregulation of CD44 expression and blocking experiments, we identified CD44v6 as a co-receptor for EGF- and ER-induced ErbB1 activation and for NRG1-induced ErbB3 and ErbB4 activation. In contrast, TGFα is independent of all CD44 isoforms, even though it addresses the same receptor pairs as EGF. Moreover, the heparin-sulfated CD44v3 isoform is required for HB-EGF-induced EGFR signaling. These data suggest that specific CD44 isoforms are recruited in a ligand-dependent manner as co-receptors in the EGFR signaling pathways and that the specificity is determined by the ligand and not by the receptors themselves. The in vivo relevance of this interplay between CD44 isoforms and EGFR ligands is underlined by the decreased metastatic spreading of mammary carcinomas in mice treated with a CD44v6-specific peptide. Most importantly, we found a clear correlation between the presence of CD44v6/ErbB1 complexes in breast cancer patients and lymph node metastases.

Subfunctionalization and neofunctionalization of vertebrate Lef/Tcf transcription factors.

Invertebrates express a multitude of Wnt ligands and all Wnt/ß-catenin signaling pathways converge to only one nuclear Lef/Tcf. In vertebrates, however, four distinct Lef/Tcfs, i.e. Tcf-1, Lef, Tcf-3, and Tcf-4 fulfill this function. At present, it is largely unknown to what extent the various Lef/Tcfs are functionally similar or diversified in vertebrates. In particular, it is not known which domains are responsible for the Tcf subtype specific functions. We investigated the conserved and non-conserved functions of the various Tcfs by using Xenopus laevis as a model organism and testing Tcfs from Hydra magnipapillata, Caenorhabditis elegans and Drosophila melanogaster. In order to identify domains relevant for the individual properties we created series of chimeric constructs consisting of parts of XTcf-3, XTcf-1 and HyTcf. Rescue experiments in Xenopus morphants revealed that the three invertebrate Tcfs tested compensated the loss of distinct Xenopus Tcfs: Drosophila Tcf (Pangolin) can substitute for the loss of XTcf-1, XTcf-3 and XTcf-4. By comparison, Caenorhabditis Tcf (Pop-1) and Hydra Tcf (HyTcf) can substitute for the loss of only XTcf-3 and XTcf-4, respectively. The domain, which is responsible for subtype specific functions is the regulatory CRD domain. A phylogenetic analysis separates Tcf-1/Lef-1 from the sister group Tcf-3/4 in the vertebrate lineage. We propose that the vertebrate specific diversification of Tcfs in vertebrates resulted in subfunctionalization of a Tcf that already united most of the Lef/Tcf functions.