The very C-terminus of PRK1/PKN is essential for its activation by RhoA and downstream signaling

Abstract

PRK1 is a lipid- and Rho GTPase-activated serine/threonine protein kinase implicated in the regulation of receptor trafficking, cytoskeletal dynamics and tumorigenesis. Although Rho binding has been mapped to the HR1 region in the regulatory domain of PRK1, the mechanism involved in the control of PRK1 activation following Rho binding is poorly understood. We now provide the first evidence that the very C-terminus beyond the hydrophobic motif in PRK1 is essential for the activation of this kinase by RhoA. Deletion of the HR1 region did not completely abolish the binding of PRK1-ΔHR1 to GTPγS-RhoA nor the activation of this mutant by GTPγS-RhoA in vitro. In contrast, removing of the last six amino acid residues from the C-terminus of PRK1 or truncating of a single C-terminal residue from PRK1-ΔHR1 completely abrogated the activation of these mutants by RhoA both in vitro and in vivo. The critical dependence of the very C-terminus of PRK1 on the signaling downstream of RhoA was further demonstrated by the failure of the PRK1 mutant lacking its six C-terminal residues to augment lisophosphatidic acid-elicited neurite retraction in neuronal cells. Thus, we show that the HR1 region is necessary but not sufficient in eliciting a full activation of PRK1 upon binding of RhoA. Instead, such activation is controlled by the very C-terminus of PRK1. Our results also suggest that the very C-terminus of PRK1, which is the least conserved among members of the protein kinase C superfamily, is a potential drug target for pharmacological intervention of RhoA-mediated signaling pathways.

Introduction

Protein kinase C-related kinase 1 (PRK1, also known as PKN or PKNα) is a lipid-activated serine/threonine protein kinase and a member of the protein kinase C (PKC) superfamily [1], [2]. PRK1 was originally identified in the liver as a protease-activated serine/threonine kinase via conventional protein chemistry [3], [4], [5]. The cDNAs encoding PRK1 were subsequently cloned by independent groups using homology-based PCR cloning and low-stringency screening of cDNA library [1], [2], [6].

PRK1 consists of a single polypeptide chain of approximately 116 kDa with a regulatory domain at the N-terminal half of the protein and a catalytic domain in the C-terminal half of the kinase. The catalytic domain of PRK1 has 50% homology to that of the PKC, while its regulatory domain is less conserved as it lacks a C1 domain and has a C2-related domain that does not bind Ca²⁺. In the N-terminal, there is a region (residues 37–281) that consists of three repeats of approximately 80 amino acid residues with each containing a leucine zipper-like motif [7]. Due to its high homology with the corresponding region in its sister enzyme PRK2 [8], this region is known as homology region-1 (HR1), encompassing HR1a, HR1b and HR1c (Fig. 1A). Interestingly, the HR1 domain is not found in PKCs [1]. Therefore, PRK1 is not activated by classical PKC activators such as diacylglycerol, phorbol ester or Ca²⁺. Instead, it is activated by phospholipids and unsaturated fatty acids, notably by cardiolipin [1], [2], [5].

PRK1 is found in all eukaryotic organisms and its amino acid sequence is highly conserved during evolution [9]. For example, the Xenopus PRK1 possesses 75% amino acid sequence homology to that of human and rat PRK1 [10]. Although PRK1 is a late addition to the PKC superfamily, it might represent the precursor of the mammalian PKC during evolution. In the yeast, the PKC1p from Saccharomyces cerevisiae and the Pck1P and Pck2P from Saccharomyces pombe resemble PRKs structurally as they possess two copies of the HR1 domain at the N-terminus but lack the functional C1 or C2 domains. Furthermore, yeast PKCs are biochemically distinct from mammalian PKCs in that they are activated by Rho GTPase but not by diacylglycerol or calcium [11]. In fact, human PRK1 was one of the first RhoA effectors discovered in 1996 by the seminal work at two laboratories showing that PRK1 binds to and is activated by GTP-bound RhoA in mammalian cells [12], [13]. The importance of PRK1 in development is underscored by the finding that a loss-of-function mutation in the Drosophila Prk1 gene results in a dorsal closure defect during embryogenesis that is lethal to the developing fly embryo [14]. In mammalian cells, PRK1 has been shown to act downstream of Rho GTPase to regulate the intracellular trafficking of the epidermal growth factor receptor in HeLa carcinoma cells [15]; to modulate stress-induced gene expression in NIH3T3 and HEK293T cells [16] and to mediate the formation of cortical actin in mouse embryonic fibroblasts [17].

Elegant biochemical analyses revealed that a segment in the regulatory domain spanning amino acid residues 33–111 constitutes the minimal region in PRK1 that is required for the interaction with RhoA [18]. Further studies showed that HR1a and HR1b subdomains can each bind with RhoA in a GTP-dependent manner [25]. Detailed molecular interactions between PRK1 and RhoA were mapped by solving the crystal structure of PRK1 residues 13–98 (HR1a) complexed to guanosine 5′-3-O-thio-triphosphate (GTPγS)-RhoA at 2.2 Å resolution [19]. It was found that the HR1a domain forms an antiparallel coiled-coil (ACC) and binds to two distinct surface sites of RhoA, one within the switch II and the other with the switch I region of RhoA via hydrophobic and electrostatic interactions. The ACC-like structure was also found in HR1b of PRK1 [20]. The equilibrium dissociation constant between RhoA and the PRK1 HR1a or HR1b domain is estimated to be in the range of 150–300 nM and 180 μM, respectively [20], [21]. However, it is not clear whether the HR1 region of PRK1 is necessary and sufficient for the interaction with RhoA and for transducing signals from RhoA to downstream cellular machinery. Given the recently identified roles of PRK1 in transducing signals of RhoA to regulate cytoskeletal dynamics in neuronal cells and in initiating transcriptional superactivation of the androgen receptor during the progression of prostate cancer [22], [23], we carried out studies to elucidate molecular mechanisms underlying the control of the activation of PRK1 by RhoA. In particular, we wished to determine whether the HR1 region is the sole determinant for the activation of PRK1 by RhoA both in vitro and in cells. By employing a strategy that combines mutagenesis with biochemical analysis, we find unexpectedly that the very C-terminus of PRK1 is critical for the full activation of PRK1 by RhoA and subsequent downstream signaling in mammalian cells.