The P-Rex1/Rac signaling pathway as a point of convergence for HER/ErbB receptor and GPCR responses

ABSTRACT

Guanine nucleotide Exchange Factors (GEFs) are responsible for mediating GDP/GTP exchange for specific small G proteins, such as Rac. There has been substantial evidence for the involvement of Rac-GEFs in the control of cancer cell migration and metastatic progression. We have previously established that the Rac-GEF P-Rex1 is a mediator of actin cytoskeleton rearrangements and cell motility in breast cancer cells downstream of HER/ErbB receptors and the G-Protein Coupled Receptor (GPCR) CXCR4. P-Rex1 is highly expressed in luminal A and B breast cancer compared to normal mammary tissue, whereas expression is very low in basal breast cancer, and its expression correlates with the appearance of metastasis in patients. Here, we discuss the involvement of P-Rex1 as an effector of oncogenic/metastatic receptors in breast cancer and underscore its relevance in the convergence of receptor-triggered motile signals. In addition, we provide an overview of our recent findings describing a cross-talk between HER/ErbB receptors and CXCR4, and how this impacts on the activation of P-Rex1/Rac1 signaling, as well as highlight challenges that lie ahead. We propose a model in which P-Rex1 acts as a crucial node for the integration of upstream inputs from HER/ErbB receptors and CXCR4 in luminal breast cancer cells.

P-Rex1: A Rac-GEF implicated in cancer progression

Rac GTPases are small G proteins that regulate essential cellular functions, including the maintenance of cell morphology, polarity, adhesion, and migration. This is achieved mostly through regulation of the actin cytoskeleton structure, with Rac being activated by a variety of stimuli downstream of different receptors including tyrosine-kinases and G-Protein Coupled Receptors (GPCRs).^1-3 As with other small G proteins, Rac activation is regulated as part of a cycle between a GDP-bound inactive form and its GTP-bound active form. The control of this GTPase cycle is tightly regulated by 3 families of proteins. Firstly, the Guanine-nucleotide Exchange Factors (GEFs), which catalyze nucleotide exchange to facilitate GTP binding. In contrast, the GTPase Activating Proteins (GAPs) act as negative regulators, by stimulating the intrinsic GTPase activity of Rac to hydrolyze GTP, and thus lead to its inactivation. The third family is the GDP Dissociation Inhibitors (GDIs), which sequester the inactive GDP-bound form in the cytoplasm, preventing Rac translocation and activation.^4,5

Dysregulation of normal Rac signaling has been known to have a role in many different diseases. For example, hyperactivation of the Rac pathway in cancer results in enhanced migration and invasiveness.^6,7 Such alterations in Rac signaling can involve up-regulation of Rac itself,⁸ expression of the active spliced variant Rac1b^8,9 or, very rarely, mutations in RAC genes.^10,11 However, the most common mechanism that accounts for Rac hyperactivation in breast cancer is the aberrant action of its direct regulators, such as overexpression and/or hyperactivation of Rac-GEFs. Relevant examples in breast cancer include the overexpression of Vav3, which together with Vav2, is implicated in breast tumor growth and lung-specific metastasis,^12,13 and the upregulation of P-Rex1 which was shown to have a prominent role in disease progression.^14,15 Down-regulation of Rac-GAPs in breast cancer has also been shown.^16,17

P-Rex1 was first identified in neutrophils as an exchange factor specific for Rac, where it was shown to regulate ROS production, motility, and polarity in response to activation by PIP₃ and Gβγ.¹⁸ This GEF is a large multidomain protein containing an N-terminal DH-PH domain tandem that is responsible for mediating its GEF activity toward Rac (Fig. 1). It also contains a pair of DEP and PDZ domains, as well as a C-terminal domain with homology to inositol polyphosphate 4-phosphatase (IP4P). The DH-PH tandem is known to be required for the interaction of P-Rex1 with PIP₃ and Gβγ proteins, which have been shown to synergistically regulate its catalytic activity and localization.^18,19 Recent studies have successfully crystalized this DH-PH tandem of P-Rex1 in complex with Rac,^20,21 thus increasing our understanding of the structural mechanism behind these interactions and GEF activity. Post-translational modifications, specifically phosphorylations, as well as protein-protein interactions (such as the recently described association with norbin)²² have been shown to regulate P-Rex1 activity. Of particular note is the phosphorylation by PKA, which inhibits PIP₃- and Gβγ-stimulated P-Rex1 guanine nucleotide exchange activity on Rac.²³ PKA phosphorylates P-Rex1 in the first DEP domain, which was recently shown to mediate an autoinhibitory intramolecular interaction with the DH-PH tandem.²⁴ In contrast, the dephosphorylation of at least one site within the IP4P-like domain by PP1α positively regulates P-Rex1 activity.²⁵ The phosphorylation of P-Rex1 at other sites has also been implicated in the regulation of its GEF activity.^15,26,27 The regulation and function of P-Rex family members has been described in great detail in a recent review.²⁸

Figure 1.

Regulation of P-Rex1 activity. (A) Domain structure of P-Rex1 showing the DH-PH tandem, PDZ domains, DEP domains, and IP4P-like domain. (B) Proposed cycle of activation/deactivation of P-Rex1. In response to receptor activation, inactive P-Rex1 in the cytoplasm is recruited to the plasma membrane by direct interactions with PIP₃, Gβγ subunits and norbin. The interactions with norbin and PIP₃ occur via the PH domain, while Gβγ subunits bind directly to the DH domain. At the plasma membrane P-Rex1 GEF-activity toward Rac is stimulated by PIP₃, Gβγ and norbin, thus resulting in the release of GDP from Rac and the binding of GTP. PKA phosphorylates P-Rex1 at the plasma membrane in Ser436 located in the first DEP domain, which results in intramolecular autoinhibitory interactions between the DH-PH tandem and the first DEP domain.

Since the initial discovery of P-Rex1, this protein has emerged as an important factor in the progression of multiple cancers. For example, P-Rex1 overexpression has been reported in a high proportion of primary metastatic melanomas, as well as some prostate cancers, where it was shown to contribute to the metastatic phenotype.^29,30 Interestingly, another P-Rex family member, P-Rex2, has also been demonstrated as having oncogenic potential, with overexpression identified in a small sample of gastric cancers³¹ as well as mutations found in metastatic melanomas.³² Previous studies by our lab and others have identified P-Rex1 as a mediator of actin reorganization and cell motility in luminal breast cancer cells.^14,15 We showed that overexpression of the P-Rex1 protein occurs in luminal A and B breast cancer, and observed a significant correlation with estrogen receptor expression. On the other hand, P-Rex1 is not expressed in basal/triple negative breast cancer, where other GEFs may control Rac1 activation. P-Rex1 up-regulation in luminal breast cancer involves the demethylation of the PREX1 gene promoter, unlike Vav3, which does not seem to be epigenetically regulated.³³ It became clear that P-Rex1 is an effector of both tyrosine-kinase and G-protein-coupled receptors (GPCRs), and that this Rac-GEF has a primary role as a mediator of motile responses driven by receptor stimulation.¹⁴

P-Rex1 as a node for HER/ErbB receptor tyrosine-kinase and GPCR signals

One of the hallmarks of breast cancer is the hyperactivation of HER/ErbB receptor signaling. The HER/ErbB family of proteins consists of the type I transmembrane growth factor receptors EGFR/HER1/ErbB1, HER2/ErbB2, HER3/ErbB3, and HER4/ErbB4. These receptors contain an extracellular region that binds a ligand (except for the ligandless receptor HER2), a transmembrane domain, and an intracellular catalytic domain. Ligand binding induces a conformational change in the extracellular domain thus promoting HER/ErbB protein homo- or heterodimerization and consequently transphosphorylation of intracellular domains.³⁴ The HER2/HER3 complex is a major oncogenic unit in mammary tumors, with overexpression of HER2 identified in 20–25% of invasive breast cancers.³⁵ This results in enhanced proliferation and survival signaling, and is functionally translated into epithelial-mesenchymal transition, loss of polarity, loss of adhesion, cell invasiveness, genome instability, metabolic changes, and angiogenesis.³⁶ Arguably the most important oncogenic signaling cascades activated by the HER2/HER3 complex are the phosphatidylinositol 3-kinase (PI3K)/Akt, extracellular-signal regulated kinase (ERK), and protein kinase C (PKC) pathways.³⁷ We previously established that activation of HER/ErbB receptors in luminal breast cancer cells leads to a migratory response that is mediated by P-Rex1/Rac1. HER ligands induce the relocalization of P-Rex1 to the plasma membrane, a site where it promotes GTP loading onto Rac1, with the subsequent activation of motility pathways.¹⁴

In addition to the role of HER/ErbB receptors in breast cancer, GPCRs have also been described as essential components of the tumorigenic process. CXCR4, a GPCR that selectively binds the chemokine Stromal cell-Derived Factor-1 (SDF-1/CXCL12), is expressed at high levels in breast cancer when compared to adjacent normal tissue.³⁸ CXCR4 overexpression has been associated with a significant reduction of disease free survival and overall survival in breast cancer patients, as well as increased metastasis to lymph nodes.^39-41 Furthermore, aberrant expression of this chemokine receptor is known to cause changes in invasion, transendothelial migration at the primary tumor site, proliferation, angiogenesis, and the recruitment of immune cells, thus contributing to cancer progression.⁴² CXCR4 controls these relevant functions through the activation of signaling cascades that include PI3K/Akt, NF-κB, Erk, and STAT3 pathways.^43-46 Interestingly, a correlation between CXCR4 and HER2 expression in human breast tumors has been reported, with a significant number of patients exhibiting co-expression of these receptors.⁴⁷ It has also been demonstrated that CXCR4 is required for HER2-mediated invasion in vitro as well as metastasis in vivo, thus establishing a functional link between the HER2 and CXCR4 signaling pathways.⁴⁸

In a previous study,¹⁴ we reported that CXCR4 stimulation with SDF-1 in breast cancer cells leads to Rac1 activation and a motile response, effects that are mediated by P-Rex1. Moreover, we found that CXCR4 was required for the activation of P-Rex1 and Rac1 by the HER/ErbB receptors. Indeed, P-Rex1/Rac1 activation by the HER3 ligand heregulin (HRG) was greatly reduced upon CXCR4 RNAi silencing. A detailed analysis revealed that stimulation of luminal breast cancer cells with HRG caused transactivation of CXCR4, an effect that was not affected by a CXCR4 antagonist and therefore was independent of the CXCR4 ligand SDF-1. HRG treatment led to a significant elevation in Ser324/325- and Ser330-CXCR4 phosphorylation, residues that typically regulate CXCR4 signaling and trafficking, as well as promote arrestin-2 binding to the GPCR. This not only showed that CXCR4 becomes indirectly activated in response to HER3 ligands but also established a role for CXCR4 in the activation of P-Rex1/Rac1 by HER/ErbB receptors. CXCR4 couples to Gα_i, therefore promoting the release of Gβγ subunits upon stimulation. A reasonable hypothesis is that this transactivation represents a major contributor to the activation of the P-Rex1/Rac1-dependent motility, since P-Rex1 is a PIP₃- and Gβγ-dependent Rac-GEF. Not surprisingly, HER3-driven breast cancer cell migration was found to be sensitive to Gα_i inhibition with pertussis toxin.¹⁴ Furthermore, studies by other groups have confirmed the emergence of P-Rex1 as a key regulator in the metastatic response of cancer cells to the stimulation of HER/ErbB and CXCR4 receptors.^15,30

In our recent work,⁴⁹ we demonstrated that, in addition to receptor transactivation, another mechanism of synergism between HER/ErbB receptors and CXCR4 signaling exists that ultimately impacts on P-Rex1/Rac1-driven cell motility. In this study, we observed that sustained exposure to HRG (a scenario that mimics the ligand overexpression that occurs in a significant proportion of breast tumors) sensitizes breast cancer cells through the HER3 receptor to SDF-1-induced Rac1 activation and motility, effects largely mediated by P-Rex1. Knocking down P-Rex1 essentially abolished this sensitizing mechanism. It was found that enhanced Rac1 activation by SDF-1 is due to increased surface expression of CXCR4. It has previously been established that hypoxia stress promotes the expression of CXCR4 through activation of the hypoxia inducible factor-1 α (HIF-1α) transcription factor.^50-52 Like hypoxia, many growth factors and cytokines also up-regulate HIF-1α protein levels under normoxic conditions.^53-55 Remarkably, after continuous stimulation of HER3 receptors with HRG, the levels of HIF-1α protein increased.^49,55 We further demonstrated, through the use of mutagenesis and chromatin immunoprecipitation, that this transcription factor activates the CXCR4 gene via binding to a hypoxia-responsive element (HRE) located in positions -1376 to -1372 in the CXCR4 promoter. The functional consequence was the elevation of CXCR4 surface expression by HRG in luminal breast cancer cells, the augmentation of Rac1 activation via P-Rex1, and the consequent increase in cell motility by SDF-1.⁴⁹ The model we propose is that P-Rex1 is a key integrator of upstream inputs from HER/ErbB receptors and CXCR4 in breast cancer cells. Signals from oncogenic receptors converge on P-Rex1, therefore leading to Rac1 activation, and this occurs by at least 3 different mechanisms (Fig. 2). The first is the direct activation of P-Rex1 by HER/ErbB receptors and CXCR4. The second involves the transactivation of CXCR4 by HER/ErbB receptors in an SDF-1-independent manner, a step that provides the required Gβγ component for P-Rex1 activation. Finally, the P-Rex1/Rac1 motility signaling is enhanced as a consequence of CXCR4 upregulation caused by sustained HER receptor activation. It is likely that other tyrosine-kinase receptors and GPCRs utilize similar mechanisms to signal via P-Rex1 to promote motile and metastatic responses.

Figure 2.

P-Rex1 acts as a convergence point for pro-oncogenic/metastatic pathways. (A) Activation of P-Rex1 by ligand stimulation of HER/ErbB receptors (for example with HRG) and CXCR4 (with SDF-1). PIP₃ and Gβγ subunits generated upon receptor stimulation activate P-Rex1 GEF-activity toward Rac. (B) HER/ErbB receptors transactivate CXCR4 independently of SDF-1, leading to P-Rex1/Rac activation. (C) Sustained stimulation of HER3/ErbB3 stabilizes the expression of the HIF-1α, leading to enhanced transcriptional activation of the CXCR4 gene. This results in up-regulation of CXCR4 surface expression and augmented P-Rex1/Rac1 motility signaling.

Future perspectives

The convergence of multiple tumorigenic signaling inputs on P-Rex1 makes this Rac-GEF an interesting target for anti-cancer therapy. The prospect of a P-Rex1 inhibitor, which would prevent Rac activation and therefore reduce the impact of CXCR4 or HER oncogenic signaling, is particularly thrilling. Despite the obvious difficulties in generating cell permeable P-Rex1 small molecule inhibitors capable of interfering with the P-Rex1-Rac interactions at suitably low concentrations, the recent identification of leading P-Rex1 inhibitor compounds⁵⁶ and the availability of P-Rex1 3-D structures^20,21 are encouraging indicators for near future success. This endeavor would be greatly facilitated by a better understanding of the mechanisms of P-Rex1 regulation, such as phosphorylation modifications and the identification of protein partners, which should provide us with alternative approaches for functionally inhibiting this Rac-GEF.

There are many questions surrounding the role of P-Rex1 in cancer that remain to be answered. For example, the role of P-Rex1 in leukocytes has been well characterized,²⁸ but its role in the interaction between these cells and cancer cells during tumorigenesis is yet to be investigated. Furthermore, most of our current understanding centers on P-Rex1 GEF activity via the DH-PH tandem, while the role of other protein domains or potential GEF-independent functions in cancer are yet to be identified. Novel regulatory mechanisms of GEF activity are also beginning to emerge, with the recent identification of the binding partner norbin,²² as well as insights into how this large multidomain Rac-GEF is regulated by kinases.^24,57 Also, the role of any other Rac-GEFs that may provide redundancy for P-Rex1 has yet to be investigated. Despite the questions that remain unanswered, P-Rex1 has undoubtedly become a prominent pro-oncogenic and pro-metastatic protein in multiple cancers, and further insights into its regulation and function will be of particular interest.

Funding Statement

The work discussed here was supported by grant CA139120 from NIH to M.G.K.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.