Nucleotide Release Sequences in the Protein Kinase SRPK1 Accelerate Substrate Phosphorylation

Protein phosphorylation represents an essential posttranslational modification that regulates signaling processes in the cell. There are now over 500 known protein kinases in the human genome and it is estimated that up to 30% of the proteins in the cell may be phosphorylated by this large enzyme family at one or more sites (1, 2). The protein kinases consist of a conserved kinase domain comprised of a small, ATP/ binding lobe and a larger substrate-binding lobe. Sequences flanking the kinase domain can serve to either positively or negatively regulate the activity of the kinase domain. Some protein kinases contain autoinhibitor sequences outside the kinase domain that bind in the active site prohibiting substrate association in a classic competitive manner (3). Activation is oftentimes achieved by phosphorylation within the kinase domain at the activation loop and/or by subunit binding that optimally arranges active-site residues (4, 5). These activation modes generally involve large increases in the phosphoryl transfer rate and sometimes a shift in mechanism from slow transfer to rate-limiting ADP release (6). Indeed, many active protein kinases have highly efficient phosphoryl transfer steps and slow ADP release steps that limit substrate turnover (7-14). Given the general role of rate-limiting ADP release for activated protein kinases (15) an exchange factor capable of enhancing or repressing this step would directly affect protein phosphorylation rates. Other nucleoside triphosphate-utilizing enzymes are known to incorporate such mechanisms. For example, Rho GTPases are activated by a guanine exchange factor that accelerates GDP release (16, 17). Based on the kinetic data, factors that enhance ADP release in protein kinases have the potential of accelerating protein phosphorylation rates and impacting cell signaling pathways. The splicing of early mRNA transcripts is catalyzed by a large macromolecular complex known as the spliceosome. In addition to several core RNA molecules, the spliceosome contains numerous auxiliary proteins important for assembly and splice/ site selection. SR proteins are essential splicing factors that establish 5′ and 3′ splice sites on precursor mRNA. They are composed of one or two RNA recognition motifs (RRMs)1 that bind specific mRNA regions establishing intro-exon boundaries. All SR proteins contain a C-terminal RS domain rich in long Arg-Ser repeats. Phosphorylation of these repeats by the serine protein kinase SRPK1 is essential for nuclear translocation of the SR protein and interaction with the spliceosome (18-20). SRPK1 is considered constitutively active since it does not require prior phosphorylation in its kinase domain for SR protein phosphorylation. SRPK1 and its family members are unique protein kinases whose kinase domains are bifurcated by a spacer insert domain [SID] (Fig.1A). The SID is as large as the kinase domain, interacts with chaperones and maintains a cytoplasmic pool of the kinase for SR protein phosphorylation (21). SRPK1 also contains a long N-terminal extension whose physiological function is poorly understood (Fig.1A). This region outside the kinase domain may be important for physiological control as it has been shown that phosphorylation of the N-terminus enhances activity (22) whereas binding of the nuclear scaffold protein SAFB to the N-terminus represses activity (23). Figure 1 Diffusion Limits of ATP and SRSF1 in the Active Site of Wild-Type & Mutant Forms of SRPK1 Much of our knowledge regarding how SRPKs recognize and phosphorylate SR proteins has been informed by the structural and kinetic analyses of SRPK1 and one of its substrate targets, SRSF1 (aka ASF/SF2) (24). Kinetic studies have shown that SRPK1 binds tightly to the RS domain near the C-terminal end of a long Arg-Ser dipeptide repeat known as the RS1 segment and then moves in an N-terminal direction modifying approximately 10-12 serines (25). X-ray structural studies of a truncated form of SRPK1 with SRSF1 reveal that this lengthy polyphosphorylation reaction is facilitated by an electronegative docking groove in the large lobe of the kinase domain (26). The docking groove initially binds the N-terminal portion of RS1 feeding the C-terminal end into the active site for phosphorylation initiation. The N-terminal Arg-Ser repeats then sequentially translate from the docking groove to the active site, a process that leads to unfolding of the final β strand in RRM2 which then occupies the docking groove (27). Rapid quench flow studies indicate that the rate of serine phosphorylation and translocation of the Arg-Ser stretch through the docking groove and active site are fast whereas the release of the product ADP is rate-limiting at each step (11). Although SRSF1 binds with very high affinity at the start of the reaction (Kd =20-50 nM) (28, 29), the addition of the first eight phosphates reduces binding affinity by up to two orders of magnitude (11). These changes along with phosphorylation-dependent decreases in ADP release rate help to terminate the multi-site reaction and release the processed SR protein from SRPK1. Although the present X-ray model of SRPK1 has provided keen insights into the mechanism of SR protein binding and phosphorylation, the N-terminus and most of the SID (about ½ of the kinase) have been omitted to aid in crystallization. The SID is highly charged and predicted to lack regular secondary structure. Recent hydrogen-deuterium [H-D] exchange data indicate that most of the SID with the exception of two short segments near the SID/kinase boundaries is unstable and likely to be unfolded (30). Such a large, unstructured region in the SID could provide attachment points for the chaperones, thus, pinning the kinase in the cytoplasm. A small segment of the N- and C-terminal ends of the SID, included in the expression construct to aid in crystallization, adopts helical conformations possibly explaining the observed stability of these regions in the H-D exchange experiments. While not included in the X-ray structure, the N-terminus appears to possess stable regions based on H-D exchange studies suggesting that it may adopt some secondary structure and/or pack onto the kinase domain (30). Prior studies have shown that while the kinase domain of SRPK1 is active and capable of phosphorylating SR proteins, the rate of this process is increased by the presence of the SID and N-terminus (30). Given the enormous size of these noncataytic regions (Fig.1A), we wished to determine how they enhance SR protein phosphorylation and whether a smaller region or regions within the SID and N-terminus could be responsible for the observed catalytic enhancements. Using a combination of rapid quench flow, viscosometric and deletion analyses we showed that a small segment of the SID predicted to adopt a helical conformation (Sα1) and a portion of the N-terminus function cooperatively to increase the rate of ADP release and consequently increase the phosphorylation of the SR protein SRSF1. Overall, the data imply that these noncatalytic regions in SRPK1 constitute a nucleotide release factor that is likely to interact with the N-terminal lobe of the kinase domain. This is the first such observation of a structural segment outside the critical kinase domain that up-regulates protein phosphorylation through modulation of the nucleotide pocket and ATP/ADP exchange rates.