Pebble (PBL) in Drosophila melanogaster

Ect2 is a guanine nucleotide exchange factor (GEF) of the Rho GTPases (Tatsumoto et al. 1999). It was isolated as a cDNA clone with capability of converting mouse fibroblasts to malignantly transformed cells (Miki et al. 1991, 1993). Transforming Ect2 variants lacked the N-terminal regulatory domain, and efficiently induced focus formation, anchorage-independent cell growth, and cell invasiveness in mouse fibroblasts. Injection of athymic nude mice with Ect2 transformants resulted in tumor formation. While full-length Ect2 is localized only in the nucleus, its oncogenic variants are detected in the cytoplasm as well. In M phase, Ect2 distributes in the entire cells after nuclear membrane breakdown, and is concentrated on mitotic spindles and then central spindles. In the last stage of M phase, Ect2 translocates to the midbody, which is formed between two emerging daughter cells. Perturbation of Ect2 function in M phase resulted in the inhibition of cytokinesis (cellular division) without major effects on mitosis (Tatsumoto et al. 1999). These findings established that Ect2 is a critical regulator of cytokinesis. Subsequent studies on Ect2 as well as the Rho negative regulator MgcRacGAP clarified the molecular mechanisms of cytokinesis (Yuce et al. 2005; Kamijo et al. 2006); chromosome separation initiates the formation of central spindles where MgcRacGAP and MKLP1 (mitosis-specific motor protein) form a tight complex. Recruitment of Ect2 to this complex activates Rho at the cell equator leading to cytokinesis.

Structural Properties

The full-length Ect2 contains several structural motifs (Fig. 1). The C-terminal half of the molecule contains the Dbl (diffuse B-lymphoma oncogene)-homology (DH) and Pleckstrin-homology (PH) domains, which are motifs of the guanine nucleotide factor (GEF) of the Rho GTPases. Most of Rho GEFs contain each of the domains, and usually the DH domain is located at the N-terminus of the PH domain. While the Rho GEF motif is the combination of the DH and PH domains, there are a number of proteins containing the PH domain alone, most of which function as a phosphatidylinositol binding site as well as a membrane anchoring site. The very C-terminal region of Ect2 is a serine- and threonine-rich domain and functions as a regulator of cell transformation. Transforming Ect2, which was identified as cDNA clone capable of transforming NIH 3 T3 fibroblasts, lacks the N-terminal half and thus contains only the C-terminal half of the full-length Ect2 (Miki et al. 1993).

Open image in new window

The N-terminal half of Ect2 contains the BRCA1 terminal repeat (BRCT) domains (Tatsumoto et al. 1999). Identification of the fission yeast cut5 gene revealed a new homology region between Cut5 and the N-terminal half of Ect2. This region consists of two BRCT domains. Cut5 contains four repeats, whereas Ect2 has two. Fission yeast cut5 mutants show a very peculiar phenotype: no coordination between mitosis and cytokinesis, which results in the induction of cytokinesis during mitosis. The BRCT domains of Ect2 have at least two functions. First, they associate with the catalytic domain of Ect2, which renders the molecule inactive. Both the BRCT domains are considered essential to this function, as the expression of both the repeats, but not the either single repeat blocks cytokinesis as a dominant-negative mutant of Ect2. The central S domain may function as a joint of the N- and C-terminal halves, and phosphorylation of the S domain is known to affect the configuration (Hara et al. 2006). Since the BRCT sequence is known to function as a phosphospecific protein-binding motif, the Ect2 BRCT domains may function after the C-terminal domain is phosphorylated. A second function of the Ect2 BRCT domains is to interact with the Rho regulator MgcRacGAP (Yuce et al. 2005). Presumably, the conversion of the closed Ect2 structure to the open form stimulates the binding to MgcRacGAP. Phosphorylation of MgcRacGAP by Plk1 also stimulates Ect2-MgcRacGAP association (Wolfe et al. 2009).

Biochemical Functions

Ect2 functions as an activator of the Rho family of small GTPases, which are represented by RhoA, Rac1, and Cdc42 (Miki et al. 2002). Rho GTPases function as molecular switches in cells and alternate between two states: an active GTP-bound state and an inactive GDP-bound state (Fig. 2). Ect2 converts inactive Rho to active Rho by replacing bound GDP by GTP. The GTP-bound form of Rho is specifically recognized by target/effector proteins to transmit the signals further downstream. As the biological functions of RhoA, Rac1, and Cdc42 are regulation of the formation of actin stress fibers, membrane ruffling, and filopodia, respectively, most of the Rho-mediated signals involve actin remodeling. Among the known Rho targets, Rho kinase (ROCK), Citron kinase, and mDia are involved in Ect2-mediated signaling. Inactivation of a Rho GTPase is carried out through the enhancement of its intrinsic GTPase activity by a cognate GTPase activating protein (GAP). A Rho GAP, designated MgcRacGAP, is known to function together with Ect2 to regulate cytokinesis and chromosome separation (Minoshima et al. 2003).

Open image in new window

Inhibition of Ect2 function in cultured cells strongly inhibits cytokinesis without visible effects on mitosis, generating binucleate cells. Therefore, Ect2 is a critical regulator of cytokinesis (Miki 2005). In Drosophila, a known gene (pebble), whose mutation results in cytokinesis defects, was shown to encode the fly ortholog of the human ECT2 (Prokopenko et al. 2000). Ect2 is localized in the nucleus in interphase. Upon nuclear membrane breakdown in M phase, Ect2 is found in the entire cells and then concentrated at the central spindle (Fig. 3). The Rho negative regulator MgcRacGAP and the plus-end directed motor protein MKLP1 form a stable complex and associate with the central spindle. As described below, a set of complex phosphorylation/dephosphorylation events on both Ect2 and MgcRacGAP appear to induce the association of Ect2 to MgcRacGAP on the central spindle (Yuce et al. 2005; Hara et al. 2006). The Ect2 protein that has been recruited to the central spindle then activates Rho GTPases by guanine nucleotide exchange, and activated Rho translocates to the cell equator, where it activates the Rho target Citron kinase. Citron kinase then phosphorylates the regulatory subunit of myosin heavy chain kinase. These biochemical events induce contractile ring formation and contraction (Kamijo et al. 2006).

Open image in new window

Although Ect2 knockdown mainly inhibits cytokinesis, it also impairs the attachment of mitotic spindles to the kinetochores, leading to a delay of prometaphase and abnormal chromosomal separation (Oceguera-Yanez et al. 2005). MgcRacGAP is involved in this function as well. While RhoA is a critical molecular switch in cytokinesis regulation, Cdc42 mainly functions in chromosome alignment in this mitotic role of Ect2.

Ect2 might play a role in epithelial cell polarity. The polarity regulator Par6 was identified as an Ect2-interacting protein (Liu et al. 2004). Ect2 also associates with the other polarity complex components, consisting of Par3, atypical protein kinase (PKCiota or PKCzeta), and Cdc42 (see Fig. 4). Coexpression of Ect2 and Par6 efficiently activates Cdc42, and overexpression of Ect2 activates aPKC activity. In epithelial cells, Ect2 localization was detected at the cell–cell junction as well as the nucleus. Thus, Ect2 appears to involve epithelial cell polarity, although the other Rho GEF Tiam-1 plays a major role.

Open image in new window

The central domain of Ect2, designated the S domain, functions as a joint of the catalytic C-terminal and the regulatory N-terminal halves. It contains several phosphorylation sites, which may function as switches for configuration changes (Miki et al. 2009). Phosphorylation at T341 of Ect2 by Cdk1 in M phase affects configuration of the molecule (open and closed forms) and may control the accessibility of Ect2-binding proteins (Hara et al. 2006). Phosphorylation of T412 by Cdk1 also induces binding of another M phase kinase Plk1, leading to phosphorylation of Ect2 by Plk1 (Niiya et al. 2006). It is known that this phosphorylation is required for the exchange activity of Ect2. Plk1 appears to have multiple functions in M phase. Its N-terminal half contains the regulatory Polo-box domain (PBD), which can associate with phosphorylated peptides. Cdk1 functions as a “priming kinase” in this case: Cdk1 phosphorylation stimulates Plk1 binding through the PBD domain, and the kinase domain of Plk1 subsequently phosphorylates the substrate that the Plk1 has bound. Through this mechanism, Plk1 is recruited to various cellular compartments during M phase. Plk1 phosphorylation of Ect2 and MgcRacGAP is critical for cytokinesis regulation. The N-terminal domain of Ect2 mainly consists of two BRCT repeats, which are capable of binding to phosphorylated proteins. This domain functions to localize Ect2 to the central spindle. On the other hand, Ect2 is in a closed conformation in its inactive state through the interaction of the N- and C-terminal halves. The phosphorylation status of T431 regulates the configuration. Inhibition of Plk1 by a chemical inhibitor prevents Ect2 association to the central spindle as well as binding to MgcRacGAP (Wolfe et al. 2009). Thus, Plk1 phosphorylates MgcRacGAP, which may help Ect2 to associate with MgcRacGAP. On the other hand, Ect2 is also phosphorylated by Plk1, which is required for its activity. Complex phosphorylation events might regulate cytokinesis through these protein kinases.

Although full-length Ect2 is a nuclear protein, its transforming variants display various degrees of cytoplasmic localization (Fig. 5). Subsequent analysis revealed that main nuclear localization signals located at the S domain are lost in these transforming variants (Saito et al. 2004; Justilien and Fields 2009). Whereas the possible nuclear functions of Ect2 should be clarified, Ect2 might be sequestered in the nucleus not to activate Rho GTPases in interphase. Transforming variants of Ect2 that are partially located in the cytoplasm can activate Rho GTPases leading to malignant transformation. Detailed analyses indicate that deletion of the main nuclear localization signals induces only a fraction of transforming activity of Ect2. Other factors, including phosphorylation status and configuration changes, are also important for the transforming activity of Ect2. Nonetheless, nuclear localization of Ect2 is a critical factor of Ect2 transforming activity.

Open image in new window

Despite the cytoplasmic mislocalization of transforming variants of Ect2, deletion or rearrangement of the ECT2 gene in human cancers has not been reported. Recent analyses of huge number of human cancers using microarray analyses revealed that Ect2 is a prognostic marker of several cancers including breast cancer. It is also reported that Ect2 expression inversely correlates to prognosis of glioblastomas (Sano et al. 2006). As described above, Ect2 associates with the tertiary complex consisting of the polarity determinant protein Par6, an atypical protein kinase C, and a scaffold protein Par3 (Liu et al. 2004, 2006) (Fig. 4). The gene for an aPKC, PKCiota, is located at the proximity of the ECT2 gene in human chromosome 3. Both the loci are co-amplified in several cancers including non-small cell lung carcinomas (NSLCs). Interestingly, Ect2 is mislocalized to the cytoplasm in these cancer cell lines, reminiscent of the phenotype of transforming Ect2. When either Par6 or PKCiota is knocked down in these cells, however, Ect2 localization returns to the nucleus. Thus, upregulation of the polarity complex appears to retain Ect2 in the cytoplasm in these cancer cell lines. It is well established that the Rho family GTPases, Cdc42 and Rac1, are involved in cell polarity. Ect2 knockdown in these cells inhibits tumorigenicity and invasiveness, indicating that Ect2 activates the Rac1 pathway through the polarity complex in these cells leading to malignant transformation.

Space limitations preclude comprehensive referencing. See also references therein.

Ect2 was originally isolated as a cDNA clone with an ability to morphologically transform mouse fibroblasts. The catalytic domain (C-terminal half) catalyzes guanine nucleotide exchange on Rho GTPases, whereas the regulatory domain (N-terminal half) contains two BRCT domains, which regulate the catalytic activity and also function to interact with other signaling molecules such as MgcRacGAP. Ect2 is a critical regulator of cytokinesis. Ect2 also interacts with the polarity determinants containing Par3, Par6, aPKC, and Rac1. In some cancer cells, amplification of a chromosomal region containing the genes for Ect2 and aPKC results in overproduction of these proteins. Cytoplasmic localization of Ect2/Par complex appears to cause malignant transformation of the cells.