AUTHORS DISPUTED THIS EVENT ALTHOUGH ORIGINALLY PROPOSING...

AUTHORS DISPUTED THIS EVENT ALTHOUGH ORIGINALLY PROPOSING IT BECAUSE THEY SAY (Raptis, Adan, Daniel) THAT THE INCREASE IN RAC1/CDC42 PROTEIN LEVEL UPON CADHERIN ENGAGEMENT MAKES IT DIFFICULT TO DISTINGUISH ANY TRUE ACTIVATION DOWNSTREAM OF CADHERIN ENGAGEMENT FROM INCREASED ACTIVATION THAT IS THE CONSEQUENCE OF INCREASED PROTEIN STABILITY, WHILE ACTIVATION CAN BE THROUGH OTHER MEANS. CDH1 (E-cadherin) engagement in confluent cell cultures triggers a dramatic surge in total RAC1 and CDC42 activity, assessed through the amount of GTP bound RAC1 (Nakagawa et al. 2001: canine kidney cell line MDCKII; Noren et al. 2001: MDCKII cells, human embryonic kidney cell line HEK293; Kovacs et al. 2002: Chinese hamster ovary cell line CHO, engineered to express human CDH1; Hoshino et al. 2004: mouse EL cell line derived from mouse L fibroblasts engineered to express ectopic CDH1; Mertens et al. 2005: primary mouse keratinocytes; Jin et al. 2005: MDCK cells; Wang et al. 2006: L cells; Kraemer et al. 2007: CHO cells expressing human CDH1; Kusano et al. 2008: human colon carcinoma cell line Caco2: RAC1 activation upon CDH1-mediated cell-cell adhesion is inhibited by overexpression of KEAP1, a component of an E3 ubiquitin ligase complex; Deplazes et al. 2009: human metastatic breast cancer cell line MDA-MB-435S expressing recombinant human CDH1) and CDC42 (Kim et al. 2000: human breast cancer cell line MCF-7; Noren et al. 2001: MDCKII cells, human embryonic kidney cell line HEK293; Kraemer et al. 2007: CHO cells expressing human CDH1). Increase in RAC1 and CDC42 activity is not seen in the confluent mouse fibroblast cell line NIH 3T3 (Noren et al. 2001). The activity of RHOA, on the other hand, decreases in the confluent MDCKII and HEK293 cultures, while it increases in the confluent NIH 3T3 culture (Noren et al. 2001). RAC1 is preferentially activated upon CDH1 engagement compared with CDC42 (Kraemer et al. 2007).

Multiple studies described below report that the activation of RAC1 and CDC42 promotes CDH1-mediated cell-cell adhesiveness in confluent cells and inhibits it in sub-confluent cells. RAC1 and CDC42 were reported to promote CDH1-mediated cell-cell adhesion by inhibiting IQGAP1 (Kuroda et al. 1998, Fukata et al. 1999). IQGAP1, an effector of activated CDC42 and RAC1, in the absence of GTP-bound RAC1 and CDC42 localizes and interacts with CDH1 and CTNNB1 (beta-catenin) at sites of cell-cell contact in EL cells, inducing the dissociation of CTNNA1 (alpha-catenin) from the CDH1:CTNNB1 complex, and decreasing CDH1-mediated cell-cell adhesion (Kuroda et al. 1998, Fukata et al. 1999). GTP-bound RAC1 and CDC42 inhibit the binding of the IQGAP1 to CTNNB1 in vitro, thus preventing the IQGAP1-mediated dissociation of CTNNA1 from the CDH1:CTNNB1 complex (Fukata et al. 1999). In a study using MDCKII cells, however, it was reported that CDH1-mediated activation of RAC1 leads to increased CDH1-mediated cell-cell adhesion through IQGAP1, creating a positive feedback loop (Noritake et al. 2004). The Ca2+ binding protein Calmodulin (CALM1) was reported to compete with IQGAP1 for binding to the CDH1:catenin complex, thus increasing homophilic CDH1 adhesions (Li et al. 1999). In a study using pancreatic carcinoma cell line PANC-1, it was reported that IQGAP1 co-localization with the CDH1:catenin complex results in increased cell adhesion, and that active RAC1 sequestered IQGAP1 from the CDH1:catenin complex, leading to decreased adhesion (Hage et al. 2009), which contradicts previously published findings described above.

In human epidermal keratinocyte cell line SCC12F, the establishment of CDH1-based cell-cell adhesions leads to distribution of wild-type, dominant-negative, and constitutively active GFP-labeled recombinant RAC1 proteins to sites of these cell-cell adhesions, where RAC1 proteins co-localize with CDH1 and catenins (Akhtar et al. 2000). Co-localization of RAC1 with CDH1 at cell-to-cell contacts and activation of RAC1 at these sites was confirmed in confluent human keratinocytes of Sf, Kf and AEK strains (Erasmus et al. 2009). CDH1-mediated activation of RAC1 in keratinocytes requires EGFR and the RAC1 GEF DOCK1 (also known as DOCK180) (Erasmus et al. 2015). Contrary to RAC1, the establishment of CDH1-mediated cell-to-cell adhesions in confluent human keratinocytes results in downregulation of CDC42 activity (Erasmus et al. 2009). Expression of the dominant-negative RAC1 did not lead to disruption of CDH1-mediated cell-cell adhesions (Akhtar et al. 2000), as would be expected from previously published findings (Kuroda et al. 1998). Expression of a constitutively active RAC1 in sub-confluent cultures of SCC12F cells, however, disrupts CDH1-mediated cell-cell adhesions through promoting clathrin-independent CDH1 endocytosis, which allows for cell spreading (Akhtar and Hotchin 2001). In sub-confluent MDCK cells, RAC1-mediated internalization of CDH1 was shown to depend on EZR (ezrin) (Pujuguet et al. 2003). PAK1 is involved in disruption of CDH1-mediated adhesions downstream of RAC1 and possibly RAC3 in human keratinocytes (Lozano et al. 2008), In subconfluent normal human Sf keratinocytes, RAC1 activation was reported to lead to PAK1-mediated internalization of CDH1 through macropinocytosis, where GTP-bound RAC1 activates PAK1, leading to PAK1-mediated phosphorylation of a RAB regulator GDI2 (also known as RabGDIβ), which then associates with RAB5 and RAB11 and promotes their retrieval from membranes (Erasmus et al. 2021). In human breast carcinoma cell line MCF-7, treatment with EGF induces RAC1-mediated pinocytosis of CDH1 (Bryant et al. 2007). In confluent cultures of SCC12F cells, constitutively active RAC1 does not disrupt CDH1-mediated cell-cell adhesions (Akhtar and Hotchin 2001), and in confluent MDCK cells active RAC1 does not promote CDH1 internalization (Pujuguet et al. 2003). In MDCKII cells, EGFP-tagged recombinant RAC1 was also shown to co-localize with CDH1 at sites of cell-cell contact and to translocate to the cytosol upon disruption of CDH1-mediated cell-cell adhesion (Nakagawa et al. 2001). In human melanoma cells lines grown at 30-80% confluence, increased expression level of CDH1 negatively correlates with the level of GTP-bound CDC42 but has no effect on level of GTP-bound RAC1, and neither RAC1 nor CDC42 activity affect the protein levels of CDH1 (Cerdido et al. 2024).

In mouse sarcoma cell line S180, RAC1 and CDC42 are activated when CDH1-mediated adhesions are formed between cells grown in suspension and RAC1 and CDC42-mediated actin cytoskeleton remodeling strengthens the adhesions (Chu et al. 2004).

While the recruitment of RAC1 to CDH1 cell-cell adhesions is not PI3K dependent, full activation of RAC1 at CDH1 cell-cell adhesions sites appears to require PI3K activity (Nakagawa et al. 2001, Kovacs et al. 2002). PI3K and RAC1 co-localize with CDH1 at sites of homophilic CDH1 cell-cell contacts (Kovacs et al. 2002; Perez et al. 2007 :MDCK cells were used). Using FRET imaging, it was shown that the establishment of CDH1 cell-cell adhesions leads to an increase in GTP-bound RAC1 but not CDC42 in a PI3K-dependent manner (Hoshino et al. 2004).

The establishment of nectin trans-dimers and nectin-mediated activation of CDC42 was shown to precede the formation of homophilic CDH1 trans-dimers and RAC1 activation in EL cells and MDCKII cells (Hoshino et al. 2004), although the initial observation suggested that CDC42 and RAC1 activity were not essential for the association of NECTIN1 and CDH1 during the formation of adherens junctions (Honda et al. 2003). Nectin-mediated activation of CDC42 in L and MDCK cells was shown to involve a CDC42 guanine nucleotide exchange factor (GEF) FARP2 (also known as FERM or FRG), while the downstream activation of RAC1 requires phosphorylation of VAV2, a RAC1 GEF, by SRC (Kawakatsu et al. 2005). Nectins and CDH1 were also reported to cooperate to rapidly induce and then suppress RAC1 activity during initial adhesion in MDCK cells (Kitt and Nelson 2011).

In primary mouse keratinocytes, CDH1-based cell-cell adhesion activates RAC1 through RAC1 GEF TIAM1, and RAC1 activation is required for the maturation of tight junctions (Mertens et al. 2005). In CHO cells expressing human CDH1 and in mouse mammary epithelial cell line NMuMG, both TIAM1 and TIAM2 are recruited to homophilic CDH1 dimers, but are not critical for activation of RAC1 (Kraemer et al. 2007). Activation of RAC1 by CDH1-mediated cell-cell engagement was reported to activate DYRK1B (MIRK) serine/threonine kinase (Jin et al. 2005). TIAM1 can also translocate to the nucleus, where, in complex with TRIM28, it binds to the CDH1 gene locus and the protocadherin gene cluster, promoting repressive H3K9me3 methylation at these loci and silencing of transcription of CDH1 and protocadherins (Ginn et al. 2023).

In mouse embryonic stem cells (mESCs), artificial CDH1 engagement leads to activation of RAC1 through paxillin (PXN), but the cell spreading is required in addition to homophilic CDH1 engagement (Mattias et al. 2014). In mESCs, RAC1 acts as a negative regulator of non-junctional cadherin complexes that, under a low mechanical tension, promotes CTNNB1 dissociation and fragmentation of CDH1 complexes (Liu et al. 2025).

At medium cell densities, CDH1 engagement stimulates proliferation of the rat kidney cell line NRK-52E and human mammary epithelial cell line MCF-10A in a RAC1 activation-dependent manner, but is not needed for inhibition of proliferation in these cell lines at high cell densities (Liu et al. 2006). The establishment of cell polarity in NRK-52E cells, however, requires CDH1 engagement-mediated activation of CDC42 but not RAC1 or RHOA (Desai et al. 2009).

CDH1-mediated activation of RAC1 may be CTNND1-dependent (Liu et al. 2006, Deplazes et al. 2009). A CDH1 loss-of-function mutant with an in-frame deletion of exon 8 isolated from gastric cancer does not activate RAC1 and is also defective in recruiting CTNND1 to sites of cell-to-cell contact (Deplazes et al. 2009).

Contrary to CDH1, homophilic CDH11 engagement, in L cells engineered to express CDH11, leads to decrease in RAC1 activity level, as measured by the amount of GTP-bound RAC1 (Kiener et al. 2006). Also, a counterbalance between CDH11 expression and RAC1 activity was reported in the spreading dynamics of mouse mesenchymal stem cells (Jiang et al. 2021). However, during the embryonic development of the mouse hindbrain, in particular inferior olivary and facial motor nuclei, CDH11 promotes RAC1 activation through direct interaction with the RAC1 GEF TRIO (Backer et al. 2007). In Xenopus, CDH11 was also reported to interact with TRIO in migrating cranial neural crest cells, leading to activation of RAC1, CDC42, and RHOA (Kashef et al. 2009). In migrating human breast cancer cell lines BT-549, Hs578T, MCF-7, MDA-MB-231, and T47D, CDH11 is required for the recruitment of TRIO to the plasma membrane and activation of RAC1 (Li et al. 2011). CDH11-mediated RAC1 activation through TRIO is implicated in aortic valve calcification in humans and pig and mouse disease models (Vaidya et al. 2022). In mouse Balb/c3T3 and 10T½ fibroblast cell lines, CDH11 engagement increases the activity as well as protein levels of RAC1 and CDC42 GTPases (Geletu et al. 2013), while CAV1-mediated downregulation of CDH11 in mouse Balb/c3T3 and NIH 3T3 fibroblasts leads to downregulation of RAC1 activity (Geletu et al. 2018). Activation of RAC1 by CDH11 in mouse Balb/c3T3 cells was reported to be SRC-dependent (Adan et al. 2022). RAC1 activity is downregulated in Cdh11-/- mouse bone marrow derived macrophages and dendritic cells, consistent with CDH11-mediated activation of RAC1 (Chavula et al. 2024).