Wnts are secreted from cells, albeit rarely in a soluble form. The Porcupine protein may be involved in secretion or ER transport, as Wingless is retained in the ER in porcupine mutant Drosophila embryos (Kadowaki 1996 . van den Heuvel 1993).

In the extracellular space, several secreted proteins have been shown to bind directly to Wnts, to modulate Wnt activity. The secreted Frps resemble the ligand binding domains of the Frizzled receptor. In Xenopus, the secreted Cerberus has a Wnt binding domain. WIFs form another group of secreted Wnt binding factors Hsieh 1999.. Dickkopf in Xenopus antagonizes Wnt action, but it has not been shown to bind to Wnt directly. The interactions between Wnts and the Frizzled receptors may be regulated by Glycosaminoglycans, for which Wnt proteins have a high affinity. (Bradley 1990). In Drosophila, Dally has been proposed the be one of those Glycosaminoglycan modified proteins and to act as a co-receptor for Wingless ( Lin 1999.; Tsuda1999.)

In cells not exposed to the Wnt signal, Arm/b-catenin levels are kept low through interactions with the protein kinase zw3 (GSK-3b), APC and Axin (Behrens Itoh ., Hamada, 1999.) GBP is another GSK binding protein which down-regulates the pathway (Yost et al), but has only been found in vertebrates. Axin is dephosphorylated by the Wnt signal, which leads to loss of binding of b-catenin (Willert, 1999)

Arm/b-catenin is degraded, after phosphorylation by GSK-3, through the ubiquitin pathway (Aberle 1997.), involving interactions with Slimb/b-TrCP (Jiang 1998, , Spencer 1999, Marikawa 1998,; reviewed in Maniatis 1999 ), The DIX domain in Axin is similar to a domain in Dishevelled, and may promote interacions between these two domains (Hsu 1999). Axin also binds to the phosphatase PP2A. (Hsu 1999), while the B56 subunit of PP2A interacts with APC (Seeling 1999)

The wg/Wnt signal leads, through its receptor (members of the Frizzled (Fz) family) and Dishevelled (Dsh) to inactivation of Zw3/GSK. While the mechanism of action of Dsh is not known, it interacts with a number of other molecules, including Casein Kinase 1 (CK1; Peters 1999, Sakanaka, 1999) Casein Kinase 2 (CK2; Willert 1997) and GBP/Frat1 (Li 1999). CK1 appears to be an essential step in Wnt signaling, both in Xenopus and in C. elegans (Peters 1999). It should be noted that CK1 and CK2 are unrelated kinases. Dsh can also bind to the Phosphatase PP2C (PP2C) which is able to dephosphorylate Axin (Strovel, 1999). In Drosophila, the naked cuticle gene (naked) acts as an inducible inhibitor of Wingless signaling (Zeng et al, 2000)

A direct interaction between Dsh, GBP/Frat1 and Zw3/GSK may be the keystep in the inactivation of Zw3/GSK (Li L1999). As a consequence, Arm is released from the Axin complex (Willert, 1999,) increases in concentration and can bind to TCF1/Lef-1/pangolin. This complex can activate nuclear target genes. Loss of APC in mammalian cells can also lead to a critical loss over Arm control, leading to cell transformation, most likely by activation of other target genes. Arm/b-catenin can also bind to the cytoplasmic tail of Cadherin. The Drosophila Teashirt (Tsh) protein is able to bind to the COOH-terminus of Armadillo, to modulate wingless signaling (Gallet 1999, Gallet 1998)

In the nucleus, in the absence of the Wnt signal,TCF acts as a repressor of Wnt/Wg target genes (Brannon 1997, Bienz 1998 . Riese 1997; also in C. elegans Lin 1998). TCF can form a complex with Groucho (Roose 1998) and CBP (Waltzer 1998.) repressing gene transcription when Wnt signaling is inactive (Cavallo 1998) CtBP acts as another co-repressor binding to TCF (Brannon, 1999). The repressing effect of Groucho is mediated by interactions with Histone Deacetylases (HDAC, Chen 1999).

Tcf is antagonized by phosphorylation, and the protein kinase Lit-1 (in C. elegans; the Drosophila homolog of Lit1 is Nemo and a vertebrate homolog is called NLK) is implicated in direct phosphorylation (Rocheleau .; Ishitani ; Meneghini .). The kinase activity of Lit1/NLK/Nemo is stimulated by another kinase, TAB1/TAK1 (or MOM-4 in the worm). (Rocheleau .; Ishitani ,; Meneghini .). Rocheleau et al. have proposed that the Lit1/NLK/Nemo kinase is activated by Wrm-1 (the C. elegans Armadillo) but it is not clear whether this also the case in other species.

Arm/b-catenin can convert TCF into a transcriptional activator of the same genes that are repressed by TCF alone (reviewed in Nusse, 1999). Arm/b-catenin activity in the nucleus may also be regulated by interactions with other members of the HMG-box family (to which TCF belongs) including XSox17 (Zorn et al, 1999). Specificity of activation of target genes can be achieved by interaction with other factors, for example the Smad4 protein which mediates TGF-beta signaling (Nishita et al, 2000).

The Drosophila segment polarity gene lines is a stage specific modulator of wingless signaling, acting in the nucleus (Hatini, 2000)

There are now many target genes of this pathway, listed in a separate table.

See the Homepage of Randall Moon's lab for information on alterntive pathways, involving Calcium release and PKC.