Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 Feb 8.
Published in final edited form as: FEBS Lett. 2018 Feb 8;592(4):537–546. doi: 10.1002/1873-3468.12990

Modulation of GSK3β autoinhibition by Thr-7 and Thr-8

Yixin Tong 1,2, Sohyun Park 1, Di Wu 1, Thurl E Harris 1, Christopher A Moskaluk 3, David L Brautigan 4, Zheng Fu 1,*
PMCID: PMC5829016  NIHMSID: NIHMS937093  PMID: 29377106

Abstract

Glycogen synthase kinase 3β (GSK-3β) is a pivotal signaling node that regulates a myriad of cellular functions and is deregulated in many pathological conditions, making it an attractive therapeutic target. Inhibitory Ser-9 phosphorylation of GSK3β by AKT is an important mechanism for negative regulation of GSK3β activity upon insulin stimulation. Here, we report that Thr-7 and Thr-8 residues located in the AKT/PKB substrate consensus sequence on GSK3β are essential for insulin-stimulated Ser-9 phosphorylation in vivo and for GSK3β inactivation. Intestinal cell kinase (ICK) phosphorylates GSK3β Thr-7 in vitro and in vivo. Thr-8 phosphorylation partially inhibits GSK3β, but Thr-7 phosphorylation promotes GSK3β activity and blocks phospho-Ser-9-dependent GSK3β autoinhibition. Our findings uncover novel mechanistic and signaling inputs involved in the autoinhibition of GSK3β.

Keywords: GSK3β, phosphorylation, autoinhibition, ICK, AKT, insulin

Introduction

Glycogen synthase kinase 3β (GSK3β) is a multi-functional protein kinase that plays an important role in amazingly diverse intracellular signaling actions by targeting a broad range of substrates, including metabolic enzymes, transcriptional factors, and structural and signaling proteins [1, 2]. Major mechanisms that regulate GSK3β activity include phosphorylation, protein complex formation, and intracellular localization [1–3]. Phosphorylation of GSK3β at serine 9 [4] and tyrosine 216 [5] is the most widely studied mechanism of “Yin-Yang” regulation. Under basal/resting conditions, GSK3β is constitutively active in cells. Tyr-216 phosphorylation at the activation loop is critical for GSK3β activity by inducing a conformational change that allows substrates to bind GSK3β [6]. Either the autokinase activity [7, 8] or a distinct tyrosine kinase [9, 10] was reported to account for Tyr-216 phosphorylation. Growth factors induce phosphorylation of GSK3β at Ser-9, which triggers an autoinhibitory effect on GSK3β activity. Several members of the AGC (protein kinase A/G/C) family of protein kinases, including Akt [11–13], p90Rsk [4, 14–16], p70 S6K [4, 17], PKA [18, 19], and certain isoforms of PKC [20, 21], have been found to phosphorylate GSK3β Ser-9 under diverse upstream stimuli and conditions.

Although GSK3β does not have a strict substrate consensus motif, it is one of a few unique protein kinases that strongly prefer prior phosphorylation of their substrates by a priming kinase at a proximal serine/threonine C-terminal to the target residue [22]. Structural analysis has revealed that GSK3β relies on a phosphorylated serine or threonine residue at the P+4 position in the substrate for the alignment of its β- and α-helical domains into a catalytically active conformation [23–25]. A recent crystal structure of GSK3 bound to its phosphorylated N-terminus showed that the pS9 auto-inhibitory peptide binds to GSK3 as a pseudo-substrate by occupying the primed substrate binding pocket and stabilizes a catalytically competent conformation [26]. In the inhibitory peptide-binding pocket, the phosphate of phosphorylated Ser-9 in the primed P+4 position forms hydrogen bonds with Arg-96, Arg-180 and Lys-205, and Thr-7 at the P+2 position packs against Phe-67 [26]. Mutation of Phe-67 to alanine severely impairs GSK3β catalytic activity, suggesting that the interaction between Phe-67 and Thr-7 is important for stabilizing a catalytically active conformation [26]. Highly conserved Thr-7 and Thr-8 residues are adjacent to Ser-9 within the AKT/PKB substrate consensus sequence R-P-R-T7-T8-S9-F on the N-terminus of GSK3β. Although Thr-7 and Thr-8 are not stringently required for AKT substrate specificity, they are positioned in the AKT-GSK3β binding interface for potential contacts with residues (e.g. Glu-279) in the AKT activation segment [27]. These structural information together raised the question as to whether Thr-7 and Thr-8 play any significant role in AKT phosphorylation of Ser-9 and GSK3β autoinhibition.

Materials and methods

Reagents and plasmids

Human insulin was from Lilly USA (Indianapolis, IN, USA). Calyculin A was from Calbiochem (distributed through MilliporeSigma, Burlington, MA, USA). All synthetic polypeptides were from ABI Scientific (Sterling, VA, USA). The MISSION TRC ICK shRNA Target Set and the control non-targeting shRNA vector were obtained from Sigma (St. Louis, MO, USA), as described in [28]. GST-ICK WT (wild type) and K33M (kinase dead) in pEBG-GST were described previously in [29, 30]. HA-GSK3β WT, S9A, and K85A in pCDNA3 (Addgene plasmids #14753, #14754, and #14755) were gifts from Dr. Jim Woodgett [14] through Addgene (Cambridge, MA, USA). HA-GSK3β T7A and T8A mutants were generated using the QuikChange Site-Directed Mutagenesis Kit from Stratagene (San Diego, CA, USA) and confirmed by DNA sequencing at DNA Sciences Core, University of Virginia.

Cell Lines, transfection, and infection

HEK293T and RIE-Ras12V (gift from Dr. John Barnard, The Ohio State University) cells were maintained at 37°C and 5% CO2 in DMEM supplemented with 10% FBS. RIE-1 rat intestinal epithelial cells were obtained from Ken Brown (Cambridge, UK). RIE-Ras12V cells were established by stable transfection of the parental RIE cells with pSV2-H-Ras(12V) containing human sequences encoding the constitutively active H-Ras(12V) protein, as described in [31]. HEK293T cells were transfected using either a calcium phosphate protocol as described in [30, 32] or the lipofectamine 2000 reagent following the manufacturer’s instruction (Invitrogen, Carlsbad, CA, USA). Lentiviral shRNA particles were generated in HEK293T cells as described in [28, 33] and used to infect RIE-Ras12V cells for RNA interference.

Antibodies, Western blot, and peptide competition assay

The polyclonal GSK3β phospho-Thr-7 antibody was generated in rabbits against keyhole limpet hemocyanin-coupled phospho-GSK3β peptide RPR[pT7]TSFAES (aa 4–13) at GenScript (Piscataway, NJ, USA). Phosphopeptide-specific antibodies were affinity-purified through a positive selection over phosphopeptide antigens followed by negative selections over non-phosphopeptide antigens and phospho-T7-deficient GSK3β-T7A protein antigens. ICK antibody was described in [28]. All commercial antibodies for signaling pathways were from Cell Signaling Technology, Inc. (Danvers, MA, USA). Methods for preparation of cell extracts and Western blotting were detailed in [33, 34]. Blocking peptide competition assay was used to validate the specific reactivity of phosphosite-specific antibodies (Fig. S1). Primary antibodies (1–2 μg/ml) were pre-incubated with blocking peptides (200-fold molar excess) at 4°C for 16–18 hours. The antibody-peptide mixture was cleared of any insoluble proteins by centrifugation at 15K rpm for 15 minutes prior to the use of pre-incubated antibodies for Western blot.

In vitro kinase assay

To assess GSK3β Thr-7 phosphorylation by ICK and GSK3β Ser-9 phosphorylation by AKT, HA-GSK3β kinase-dead (KD) proteins were expressed in HEK293T cells, affinity-purified from cell extracts by anti-HA agarose beads. HA-GSK3β substrates (0.5–1 μg) were then incubated with active His-ICK1-291 [28, 32] or His-Akt1/PKBα (Millipore, Dundee, UK) proteins (50–100 ng) and 100 μM [γ-32P]-ATP (PerkinElmer, Waltham, MA, USA) in kinase buffer [50 mM HEPES (pH 7.5), 10 mM MgCl2, 2 mM DTT, and complete protease and phosphatase inhibitor cocktails (Roche, Basel, Switzerland)] at 30°C for 20 min. After the reaction was terminated by addition of 2X SDS sample buffer, the reaction samples were processed and analyzed for radioactivity by autoradiograph as described in [28, 32]. For non-radioactive assay, the above reaction mixtures were incubated with unlabeled ATP and the reactivity of GSK3β against phospho-T7 and phospho-S9 antibodies was analyzed by Western blot.

To measure GSK3β activity against primed substrates, HA-GSK3β (WT, T7A, T8A, and S9A) proteins were expressed in HEK293T cells. After insulin (100 nM) stimulation for 60 min, recombinant GSK3β proteins were affinity-purified and then incubated with 60 μM Phospho-Glycogen Synthase Peptide-2 (p-GSP2) (Millipore, Burlington, MA, USA) [4] and 100 μM [γ-32P]-ATP (PerkinElmer, Waltham, MA, USA) in kinase buffer at 30°C for 20 min. The reaction mixtures were spotted onto phosphocellulose paper (p81, Whatman), washed in 0.75% phosphoric acid, rinsed in acetone, and counted for radioactivity in a liquid scintillation counter.

Peptide inhibition assay

Peptides were prepared by solid-phase synthesis, HPLC purified, and verified for composition and purity (>95%) by mass-spectrometry (ABI Scientific Inc., Sterling, VA, USA). Peptides were derived from the N-terminal residues 3–12 of GSK3β [24]: phospho-S9 [GRPRTT(pS)FAE], phospho-T7 [GRPR(pT)TSFAE], phospho-T8 [GRPRT(pT)SFAE], phospho-T7/S9 [GRPR(pT)T(pS)FAE], and non-phospho-control [GRPRTTSFAE]. Phospho-glycogen synthase peptide-2 (p-GSP2) (Millipore, Burlington, MA, USA) was used as the prime-phosphorylated substrate for GSK3β [4]. Purified recombinant GSK3β (BioLabs, Ipswich, MA, USA), 60 μM p-GSP2 substrates, and 1 mM GSK3β N-terminal peptides in kinase buffer supplemented with complete protease and phosphatase inhibitors (Roche) and 100 μM [γ-32P]-ATP (PerkinElmer) were incubated at 30°C for 20 min. The reaction mixtures were assessed for radioactivity on phosphocellulose paper as described above.

Statistical analysis

Quantified experimental data were analyzed by the Student t-test. Data were reported as mean ± standard deviation (SD). P-values less than 0.05 were considered as significant.

Results

Highly conserved Thr-7 and Thr-8 residues are essential for insulin-stimulated Ser-9 phosphorylation in vivo and autoinhibition of GSK3β

Peptide and protein library screening has defined the substrate specificity for AKT/PKB as RXRXXS/TΦ, with strong preference for arginine residues at both -3 and -5 positions and hydrophobic amino acids (Φ) in the +1 position [35]. The structure of AKT/PKB in complex with a GSK3β peptide containing residues 3–12 of GSK3β provided the structural basis for this substrate selectivity [27]. In the N-terminal segment of GSK3β, like Ser-9 both Thr-7 and Thr-8 are highly conserved across different species (Fig. 1A). Thr-8 at the -1 position is solvent exposed; Thr-7 at the -2 position forms a hydrogen bond with Glu-279 in the activation segment of AKT [27]. Despite these insights from biochemical and structural analyses of the N-terminal peptide of GSK3β, it remains to be determined whether Thr-7 and Thr-8 play an important role in the full-length GSK3β when it interacts with AKT in response to insulin.

Figure 1.

Figure 1

Open in a new tab

Alanine mutations of highly conserved Thr-7 and Thr-8 on the N-terminus of GSK3β impairs Ser-9 phosphorylation in vivo and GSK3β autoinhibition by insulin. (A) In the N-terminus of human GSK3β, highly conserved Thr-7 and Thr-8 residues are located adjacent to the inhibitory phosphorylation site Ser-9 within the AKT substrate consensus sequence R-P-R-T7-T8-S9-F. (B) HA-GSK3β wild type (WT) and T-to-A mutant (T7A, T8A, S9A) proteins were expressed in HEK293T cells, purified by anti-HA agarose beads, and incubated in vitro with cold ATP and His-Akt1 at 30°C for various time points as indicated. Total and phospho-S9 specific signals of HA-GSK3β were detected on Western blot. Phospho-S9 signals were quantified and normalized against total GSK3β signals. (C) HA-GSK3β (WT, T7A, T8A) proteins were expressed in HEK293T cells under normal fed condition. Whole cell extracts were Western blotted for total and phospho-S9 signals of endogenous and HA-tagged GSK3β. Shown are data of duplicates. (D) HA-GSK3β (WT, T7A) proteins were expressed in HEK293T cells under normal fed condition. After anti-HA immunoprecipitation, total and phospho-S9 signals of HA-GSK3β as well as endogenous AKT co-immunoprecipitated with HA-GSK3β were analyzed by Western blot. Shown are data of triplicates. (E) HA-GSK3β WT or mutant (T7A, T8A) proteins were expressed in HEK293T cells. After overnight starvation in 0.1% low-serum medium, cells were treated with 100 nM insulin for various time ponits as indicated. Phospho-S9 specific and total GSK3β signals in whole cell extracts (WCE) or in anti-HA immunoprecipitates were analyzed by Western blot. Shown are representative data from three independent experiments. (F) HA-GSK3β WT and mutant (T7A, T8A, S9A) proteins were expressed in HEK293T cells. After starvation and insulin stimulation (100 nM, 60 min), cells were lysed and HA-GSK3β proteins were affinity-purified on anti-HA agarose beads. HA-GSK3β WT proteins purified from unstimulated cells served as the control. In vitro kinase assay was conducted to assess GSK3β activity toward p-GS peptide substrates. Shown are GSK3β activities relative to the control, mean ± SD, n=3 independent experiments, *P<0.01.

To examine the role of Thr-7 and Thr-8 in AKT phosphorylation of Ser-9, we conducted both in vitro and in vivo assays for Ser-9 phosphorylation on HA-tagged GSK3β proteins containing the T7A or T8A point mutation. In vitro, alanine mutations replacing either Thr-7 or Thr-8 did not prevent Ser-9 phosphorylation of HA-GSK3β proteins by AKT (Fig. 1B). Compared with wild type HA-GSK3β, AKT phosphorylates HA-GSK3β containing the T7A mutation equally well, but phosphorylates HA-GSK3β containing the T8A mutation to a lesser extent (Fig. 1B). In contrast, mutating either Thr-7 or Thr-8 to alanine significantly attenuated Ser-9 phosphorylation of HA-GSK3β in cells at the basal level (Fig. 1C) and during starvation followed by insulin stimulation (Fig. 1E). Impaired Ser-9 phosphorylation on HA-GSK3β (T7A) mutant proteins did not correlate with a significant reduction in the amount of AKT associated with HA-GSK3β (T7A), suggesting that Thr-7 mutation does not significantly alter the binding affinity between AKT and GSK3β (Fig. 1D). Using a peptide competition assay, we showed that p-Ser9 recognition was displaced to a comparable extent by the blocking polypeptide containing p-Ser9 and by the same polypeptide containing Thr7 mutated to Ala (Fig. S1). This observation eliminated the concern that an adjacent mutation has any significant impact on epitope recognition by the phospho-Ser9-specific antibody used in Fig. 1.

We then assessed the effects of T7A and T8A mutations on GSK3β activity upon insulin stimulation. HA-tagged GSK3β wild-type (WT) and mutant proteins purified from insulin-stimulated HEK293T cells were assessed for phosphotransferase activity against prime phosphorylated glycogen synthase (p-GS) peptide substrates. HA-GSK3β WT proteins purified from unstimulated HEK293T cells served as the baseline control. Consistent with prior studies, HA-GSK3β WT proteins containing elevated p-S9 signals after insulin stimulation exhibited a significant reduction in GSK3β activity. This inhibitory effect on GSK3β was abolished by not only the S9A mutation but also the T7A or the T8A mutation (Fig. 1F). We concluded from these results that conserved Thr-7 and Thr-8 residues are essential for insulin-stimulated Ser-9 phosphorylation in vivo and autoinhibitory effect on GSK3β activity toward its primed substrates.

The effect of T7A or T8A mutation on GSK3β Ser-9 phosphorylation cannot be abolished by inhibition of PPP family phosphatases

T7A or T8A mutation resulted in a decrease in Ser-9 phosphorylation of GSK3β in cells (Fig. 1). A significant question was whether dephosphorylation of p-Ser-9 was enhanced when Thr-7 or Thr-8 was mutated to Ala. To test this possibility, after insulin stimulation we treated cells with Calyculin A, a potent, cell permeable inhibitor of PPP (phosphoprotein phosphatases) family including PP1, PP2A, PP4, PP5, PP6. As expected, Calyculin A treatment markedly enhanced insulin-induced phosphorylation of mTOR, S6K, GSK3, and ERK (Fig. 2A). Inhibition of phosphatase activities was not sufficient to prevent decreased phosphorylation of Ser-9 on GSK3β caused by T7A or T8A mutation (Fig. 2B). This result suggests that phosphatase actions are not the main driver of the effect of the alanine substitution for Thr-7 or Thr-8 on Ser-9 phosphorylation of GSK3β.

Figure 2.

Figure 2

Open in a new tab

The inhibitory effect of the T7A or T8A mutation on GSK3β Ser-9 phosphorylation cannot be abolished by inhibition of phosphatase activities. (A) HEK293T cells were starved in low-serum medium overnight, and then treated with 100 nM insulin for 30 min followed by 50 nM Calyculin A for 30 min prior to lysis. Whole cell extracts were blotted for both total and phosphorylation signals of mTOR, S6K, GSK3, and ERK. (B) HA-GSK3β WT or mutant (T7A, T8A) proteins were expressed in HEK293T cells. After overnight starvation in 0.1% low-serum medium, cells were treated with 100 nM insulin for 30 min followed by 50 nM Calyculin A for 30 min prior to lysis. HA-GSK3β proteins were isolated from whole cell lysate by anti-HA immunoprecipitation. Phospho-S9 and total GSK3β signals in anti-HA immunoprecipitates were analyzed by Western blot. Note that Calyculin A is a potent, cell permeable inhibitor of PPP family phosphatases including PP1, PP2A, PP4, PP5, and PP6.

ICK can specifically phosphorylate GSK3β Thr-7 in vitro and in vivo

Thr-7 in the N-terminus of GSK3β fits in the substrate consensus motif R-P-X-[S/T]-P/A/S/T for intestinal cell kinase (ICK) as determined in [32]. Thr-7 residue and the ICK substrate consensus sequence R-P-R-T7-T found in human GSK3β are well conserved across different species (Fig. 1A). ICK is a ubiquitously expressed serine/threonine protein kinase in the CMGC (CDK/MAPK/GSK3/CLK) group of the human kinome [29, 30]. We therefore tested whether ICK is an upstream kinase for GSK3β-T7. GSK3β phospho-T7 and phospho-S9 antibodies were utilized to monitor specific phosphorylation of GSK3β Thr-7 and Ser-9 by Western blot. The phosphosite specificity of these two antibodies was validated by a peptide competition assay (Fig. S1). Synthetic peptides containing p-S9 or p-T7 specifically competed with antibodies recognizing p-S9 or p-T7, respectively (Fig. S1 A, D). Furthermore, phosphorylation of adjacent sites did not seem to significantly perturb recognition by GSK-3β phosphosite-specific antibodies (Fig. S1). The phospho-peptide with both p-T7 and p-S9 sites appeared to compete with each phosphosite-specific antibody to p-S9 or p-T7 equally well in comparison to polypeptides containing only one site phosphorylated (Fig. S1 B, D).

Using HA-tagged catalytically-inactive GSK3β proteins as the substrate, we showed that purified His-AKT and His-ICK1-291 proteins [28, 32] can phosphorylate GSK3β in vitro (Fig. 3A). By Western blot, we further showed the strong reactivity of HA-GSK3β to phospho-S9 and phospho-T7 specific antibodies after co-incubating HA-GSK3β with His-AKT and His-ICK1-291 respectively in vitro (Fig. 3B). This result confirmed the phosphosite specificity of AKT on Ser-9 and ICK on Thr-7. We further validated the specificity of ICK on GSK3β Thr-7 by showing that only T7A, but not T8A or S9A mutation, can abolish phospho-T7 signals from ICK (Fig. 3C). To test if ICK phosphorylates GSK3β-T7 in vivo, we employed both over-expression and knock-down approaches. Co-expression of HA-GSK3β with GST-ICK wild-type (WT), as compared with the kinase-dead (KD) mutant of ICK, in HEK293T cells resulted in a significant increase in the phospho-T7 signals which was completely abolished by the T7A mutation (Fig. 3D). RIE-Ras12V cells (H-Ras12V oncogene transformed rat intestinal epithelial cells) displaying abundant phospho-GSK3β T7 signals were treated with validated ICK-specific lentiviral shRNA [28, 33]. Phospho-T7 signals were clearly down-regulated as a result of reduced ICK expression (Fig. 3E). We conclude from these results that ICK is an upstream kinase for GSK3β Thr-7 both in vitro and in vivo.

Figure 3.

Figure 3

Open in a new tab

ICK phosphorylates GSK3β Thr-7 in vitro and in vivo. (A) HA-GSK3βKD (kinase dead mutant) was expressed in HEK293T cells and purified by anti-HA agarose beads. Purified HA-GSK3βKD substrates were incubated in vitro with 32γP-ATP and His-AKT or His-ICK1-291. Shown here is the 32P autoradiograph. (B) A non-radioactive kinase assay was conducted in vitro using wild type HA-GSK3β proteins as the substrate. Phospho-T7/S9 signals on HA-GSK3β were detected by phosphosite-specific antibodies on Western blot. (C) HA-GSK3β (WT, T7A, T8A, S9A) proteins were expressed in HEK293T cells, purified by anti-HA agarose beads, and incubated in vitro with cold ATP and His-ICK1-291. Total and phosphosite specific signals of HA-GSK3β were detected on Western blot. (D) Wild type (WT) or kinase dead (KD) GST-ICK proteins were co-expressed with WT or T7A mutant HA-GSK3β proteins in HEK293T cells. Total and phospho-T7 signals on purified HA-GSK3β proteins were detected on Western blot. (E) RIE-Ras12V cells were infected with lentivirus expressing either an validated ICK-specific shRNA (shICK) or the control shRNA (shCTL). Phospho-GSK3β T7 specific and total GSK3β and ICK signals were shown on Western blot.

Thr-8 phosphorylation partially inhibits GSK3β; but Thr-7 phosphorylation increases GSK3β activity and blocks phospho-S9-dependent inhibition of GSK3β

Structure of GSK3β suggests an intramolecular inhibitory mechanism in which the phosphorylated N-terminus of GSK3β acts as a competitive pseudosubstrate, occupying the positively charged binding pocket for the primed substrates [23–26]. Given the proximity of Thr-7 and Thr-8 to Ser-9 on the N-terminus of GSK3β, we postulated that phospho-T7 or phospho-T8 may play a similar autoinhibitory role as phospho-S9 in the regulation of GSK3β activity toward its primed substrates. To test this hypothesis, we examined whether phospho-T7 or phosphor-T8 can substitute for phospho-S9 to inhibit GSK3β activity in vitro using the classic phosphopeptide inhibition assay. In consistent with previous reports [24, 25], phospho-S9 peptides derived from the N-terminus of GSK3β effectively inhibited GSK3β activity as compared with unphosphorylated peptides (Fig. 4). In comparison, phospho-T8 peptides only partially inhibited GSK3β activity, and surprisingly, phospho-T7 peptides exhibited a stimulatory rather than inhibitory effect on GSK3β activity (Fig. 4). We also made an intriguing observation that phospho-T7/S9 peptides did not produce any significant inhibition of GSK3β (Fig. 4). These results indicate that Thr-8 phosphorylation can partially trigger autoinhibition of GSK3β, whereas Thr-7 phosphorylation disrupts phospho-S9-dependent autoinhibition of GSK3β.

Figure 4.

Figure 4

Open in a new tab

Thr-8 phosphorylation triggers a partial inhibition of GSK3β; In contrast, Thr-7 phosphorylation enhances GSK3β activity and blocks phospho-S9-mediated inhibition of GSK3β. GSK3β N-terminal peptides that are either unphosphorylated (CTL) or Ser-9 phosphorylated (pS9) or Thr-7 phosphorylated (pT7) or Thr-8 phosphorylated (pT8) or Thr-7/Ser-9 double-phosphorylated (pT7/S9) were incubated with active GSK3β kinase and primed phospho-GS peptide substrates for in vitro kinase assay. Radioactivities relative to the control were shown, mean ± SD, n=3 independent experiments, *P<0.01, **P<0.001, NS = not significant.

Discussion

Our findings here provided new insights into the phosphorylation-dependent autoregulation mechanism of GSK3β. It has been well-established that in response to insulin stimulation, AKT inhibits GSK3β activity via phosphorylation of Ser-9 [11]. Based on structural and biochemical evidence, the current model predicts that the phosphorylated N-terminal segment of GSK3β forms a loop long enough to bring phosphorylated Ser-9 in direct contact with the basic residues (Arg-96, Arg-180, and Lys-205) in the substrate binding grove, thus blocking access of the substrate to the substrate binding cleft (Fig. 5A) [23–26]. Given the proximity of Thr-7 and Thr-8 to Ser-9, it is conceivable that phosphorylated Thr-7 and Thr-8 residues may also participate in interactions with basic residues R96/R180/K205 in the substrate binding pocket to block substrate access to the active catalytic sites. However, our data underscores the selectivity of phospho-S9 and to a lesser extent of phospho-T8 in mediating the autoinhibitory effect of GSK3β (Fig. 5A, C). We also made a novel observation that Thr-7 phosphorylation exerts a stimulatory rather than inhibitory effect on GSK3β, and phospho-Thr7 can prevent phospho-Ser9-induced autoinhibition of GSK3β (Fig. 5D). This surprising result is indeed consistent with the prediction from a crystal structure of GSK3 that the packing of Thr-7 in the inhibitory peptide against Phe-67 in the glycine-rich loop of the N-terminal lobe (aa 62–70) is important for stabilizing a catalytically competent conformation [26]. A possible explanation for the stimulatory effect of phospho-T7 is that by forming electrostatic interactions and hydrogen bonds with Phe-67 and Phe-93, the phosphate group on phosphorylated Thr-7 further stabilizes the active conformation of GSK3β, and meanwhile alters the conformation of the binding interface between phospho-S9 and basic residues R96/R180/K205 in the substrate binding pocket, thus unblocking substrate access to the catalytic site. This intriguing hypothesis warrants further investigation.

Figure 5.

Figure 5

Open in a new tab

New insights into the autoinhibition model of GSK3β. (A) It has been well-established that insulin induces AKT phosphorylation of GSK3β-S9, which triggers conformational changes that allow the N-terminus of GSK3β to block the substrates from accessing its binding pocket, resulting in autoinhibition. (B) Shown in this study, the alanine mutation of highly conserved Thr-7 or Thr-8 residue within the AKT substrate consensus sequence R-P-R-T-T-S-F at the N-terminus of GSK3β impairs in vivo Ser-9 phosphorylation and GSK3β autoinhibition by insulin. (C) Shown in this study, Thr-8 phosphorylation triggers a partial inhibition of GSK3β as compared with Ser-9 phosphorylation. The upstream stimuli or the pathophysiological conditions that induce Thr-8 phosphorylation are unknown. (D) Shown in this study, intestinal cell kinase (ICK) specifically phosphorylates GSK3β-T7. Thr-7 phosphorylation blocks phospho-S9-mediated inhibition of GSK3β. The mechanisms underlying ICK activation and the crosstalk between pT7-pS9 in the regulation of GSK3β activity in vivo are unknown.

Our data here demonstrated that although Thr-7 and Thr-8 residues are not strigently required for AKT phosphorylation of Ser-9 in vitro, they are essential for basal and insulin-stimulated Ser-9 phosphorylation in vivo and phospho-S9 induced GSK3β autoinhibition (Fig. 5B). One possible explanation for different in vitro and in vivo results is that the in vitro system using only purified AKT and GSK3β proteins cannot recapitulate the complexity of multi-protein interactions that may be involved in AKT phosphorylation of GSK3β-S9 in vivo. We postulate that the alanine mutation at Thr-7 or Thr-8 disrupts hydrogen bonds that are critical for the formation of an optimal interface between AKT and GSK3β N-terminus, and thus hinders AKT phosphorylation of Ser-9 and GSK3β inactivation. Another possible explanation for the effect of T7A or T8A mutation on Ser-9 phosphorylation is through enhanced phosphatase reaction and turnover rate of phospho-S9. Our data from Calyculin inhibition experiment provided a strong argument against phosphatases as the main contributor to the observed effect of T7A or T8A mutation on Ser-9 phosphorylation. However, due to the nonquantitative nature of the experiment, we cannot conclude that the turover rate of phospho-S9 is unaffected by Thr-7/Thr-8 mutation.

We hereby provided compelling evidence that ICK is an upstream kinase for GSK3β Thr-7. In cells, overexpression of ICK led to a concurrent increase in GSK3β Thr-7 and Ser-9 phosphorylation, and conversely downregulation of ICK resulted in a similar decrease in both GSK3β phospho-Thr7 and phospho-Ser9 signals (Fig. S2). These results raised the hypothesis that pre-phosphorylation of Thr-7 by ICK may promote Ser-9 phosphorylation. Many significant questions remain to be addressed. For example, does pre-phosphorylation with ICK enhance the ability of GSK-3β to act as an AKT substrate? Furthermore, is the ability of ICK to promote Ser-9 phosphorylation in cells dependent on Thr-7 or independent of potential regulation of AKT? In the structure of AKT complexed with GSK3β3–12 peptide, GSK3β Thr-7 is predicted to form a hydrogen bond with AKT Glu-279 [27]. Thus, it is possible that Thr-7 phosphorylation may promote favorable conformational changes in the binding interface between GSK3β and AKT that facilitate Ser-9 phosphorylation by AKT. The exploration of the mechanisms underlying pThr7-pSer9 crosstalk is of great significance and merits further experimentation.

In our current model for GSK3β autoinhibition (Fig. 5), a significant question that remains to be addressed is that under what physiological stimuli/conditions Thr-7 and Thr-8 are phosphorylated in vivo. A post-translational modification scan of critical signaling proteins by immunoaffinity enrichment coupled with mass spectrometry revealed three phospho-residues (Thr-7, Thr-8, and Ser-9) in the N-terminus of GSK3β, R-P-R-T7-T8-S9-F-A-E-S [36]. While phospho-Ser-9 site has been confirmed by site-specific methods, phospho-Thr-7/Thr-8 sites were assigned using only proteomic discovery-mode mass spectrometry, as summarized in PhosphoSitePlus®. Our mass spectrometry data suggest that unlike Ser-9, both Thr-7 and Thr-8 are candidate phosphosites of very low stoichiometry under insulin stimulation (Sherman N.E. and Fu Z., unpublished data). This is consistent with prior observations from us and others that unlike MAP kinases, ICK is not catalytically activated by serum or growth factors, such as insulin and EGF [28, 29]. Another possible explanation is that phospho-T7/T8 sites undergo much faster turnovers than phospho-S9 site after insulin stimulation. Elevated expression of ICK and its upstream activating kinase CCRK/CDK20 (cell cycle related kinase/cyclin-dependent kinase 20) has been reported in various types of cancers [34, 37–39]. Whether Thr-7 phosphorylation is significantly up-regulated in cancer and contributes to tumor development and progression by promoting GSK3β activity is an intriguing question and currently under investigation.

In conclusion, within the AKT substrate consensus motif RPRT7T8S9F in the highly conserved N-terminus of GSK3β, Thr-7 and Thr-8 are essential for insulin-stimulated Ser-9 phosphorylation in vivo and GSK3β autoinhibition. Intestinal cell kinase (ICK) can specifically phosphorylate GSK3β Thr-7 in vitro and in vivo. While Thr-8 phosphorylation can only partially mimic Ser-9 phosphorylation in terms of autoinhibition of GSK3β on primed substrates, Thr-7 phosphorylation increases rather than reduces GSK3β activity and blocks phospho-S9-mediated autoinhibition of GSK3β. These new findings advanced the current model of GSK3β regulation by uncovering novel signaling inputs that are critical for GSK3β autoinhibition.

Supplementary Material

Supp FigS1-2

Acknowledgments

This work was supported by the National Institute of Health grant CA195273 to Z.F. We are indebted to our colleagues Thomas W. Sturgill, Michael J. Weber, and Daniel Gioeli for insightful discussions and critical comments. We thank our colleague Nicholas E. Sherman at W. M. Keck Biomedical Mass Spectrometry for expert evaluation of MS data and our undergraduate trainees Eric Wang, Emily L. Breeding, Yoon Seon Oh, So Hee Son, and Hun Ki Hong for technical assistance.

Abbreviations

GSK3

glycogen synthase kinase 3

ICK

intestinal cell kinase

AKT/PKB

protein kinase B

WT

wild type

KD

kinase dead

Footnotes

Author Contributions

Z.F. conceived the project; Z.F., Y.T., T.E.H., C.A.M., D.L.B. designed experiments; Y.T., S.H.P., D.W., Z.F. performed experiments; Z.F., Y.T., T.E.H., C.A.M., D.L.B. analyzed and interpreted data; Z.F. wrote the manuscript; D.L.B. revised the manuscript; All authors have reviewed the manuscript and approved for submission.

References

  • 1.Grimes CA, Jope RS. The multifaceted roles of glycogen synthase kinase 3beta in cellular signaling. Prog Neurobiol. 2001;65:391–426. doi: 10.1016/s0301-0082(01)00011-9. [DOI] [PubMed] [Google Scholar]
  • 2.Kaidanovich-Beilin O, Woodgett JR. GSK-3: Functional Insights from Cell Biology and Animal Models. Front Mol Neurosci. 2011;4:40. doi: 10.3389/fnmol.2011.00040. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Doble BW, Woodgett JR. GSK-3: tricks of the trade for a multi-tasking kinase. J Cell Sci. 2003;116:1175–1186. doi: 10.1242/jcs.00384. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Sutherland C, Leighton IA, Cohen P. Inactivation of glycogen synthase kinase-3 beta by phosphorylation: new kinase connections in insulin and growth-factor signalling. Biochem J. 1993;296(Pt 1):15–19. doi: 10.1042/bj2960015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Hughes K, Nikolakaki E, Plyte SE, Totty NF, Woodgett JR. Modulation of the glycogen synthase kinase-3 family by tyrosine phosphorylation. EMBO J. 1993;12:803–808. doi: 10.1002/j.1460-2075.1993.tb05715.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Bax B, Carter PS, Lewis C, Guy AR, Bridges A, Tanner R, Pettman G, Mannix C, Culbert AA, Brown MJ, et al. The structure of phosphorylated GSK-3beta complexed with a peptide, FRATtide, that inhibits beta-catenin phosphorylation. Structure. 2001;9:1143–1152. doi: 10.1016/s0969-2126(01)00679-7. [DOI] [PubMed] [Google Scholar]
  • 7.Lochhead PA, Kinstrie R, Sibbet G, Rawjee T, Morrice N, Cleghon V. A chaperone-dependent GSK3beta transitional intermediate mediates activation-loop autophosphorylation. Mol Cell. 2006;24:627–633. doi: 10.1016/j.molcel.2006.10.009. [DOI] [PubMed] [Google Scholar]
  • 8.Cole A, Frame S, Cohen P. Further evidence that the tyrosine phosphorylation of glycogen synthase kinase-3 (GSK3) in mammalian cells is an autophosphorylation event. Biochem J. 2004;377:249–255. doi: 10.1042/BJ20031259. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Kim L, Liu J, Kimmel AR. The novel tyrosine kinase ZAK1 activates GSK3 to direct cell fate specification. Cell. 1999;99:399–408. doi: 10.1016/s0092-8674(00)81526-3. [DOI] [PubMed] [Google Scholar]
  • 10.Lesort M, Jope RS, Johnson GV. Insulin transiently increases tau phosphorylation: involvement of glycogen synthase kinase-3beta and Fyn tyrosine kinase. J Neurochem. 1999;72:576–584. doi: 10.1046/j.1471-4159.1999.0720576.x. [DOI] [PubMed] [Google Scholar]
  • 11.Cross DA, Alessi DR, Cohen P, Andjelkovich M, Hemmings BA. Inhibition of glycogen synthase kinase-3 by insulin mediated by protein kinase B. Nature. 1995;378:785–789. doi: 10.1038/378785a0. [DOI] [PubMed] [Google Scholar]
  • 12.Shaw M, Cohen P, Alessi DR. Further evidence that the inhibition of glycogen synthase kinase-3beta by IGF-1 is mediated by PDK1/PKB-induced phosphorylation of Ser-9 and not by dephosphorylation of Tyr-216. FEBS Lett. 1997;416:307–311. doi: 10.1016/s0014-5793(97)01235-0. [DOI] [PubMed] [Google Scholar]
  • 13.van Weeren PC, de Bruyn KM, de Vries-Smits AM, van Lint J, Burgering BM. Essential role for protein kinase B (PKB) in insulin-induced glycogen synthase kinase 3 inactivation. Characterization of dominant-negative mutant of PKB. J Biol Chem. 1998;273:13150–13156. doi: 10.1074/jbc.273.21.13150. [DOI] [PubMed] [Google Scholar]
  • 14.Stambolic V, Woodgett JR. Mitogen inactivation of glycogen synthase kinase-3 beta in intact cells via serine 9 phosphorylation. Biochem J. 1994;303(Pt 3):701–704. doi: 10.1042/bj3030701. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Saito Y, Vandenheede JR, Cohen P. The mechanism by which epidermal growth factor inhibits glycogen synthase kinase 3 in A431 cells. Biochem J. 1994;303(Pt 1):27–31. doi: 10.1042/bj3030027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Eldar-Finkelman H, Seger R, Vandenheede JR, Krebs EG. Inactivation of glycogen synthase kinase-3 by epidermal growth factor is mediated by mitogen-activated protein kinase/p90 ribosomal protein S6 kinase signaling pathway in NIH/3T3 cells. J Biol Chem. 1995;270:987–990. doi: 10.1074/jbc.270.3.987. [DOI] [PubMed] [Google Scholar]
  • 17.Zhang HH, Lipovsky AI, Dibble CC, Sahin M, Manning BD. S6K1 regulates GSK3 under conditions of mTOR-dependent feedback inhibition of Akt. Mol Cell. 2006;24:185–197. doi: 10.1016/j.molcel.2006.09.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Fang X, Yu SX, Lu Y, Bast RC, Jr, Woodgett JR, Mills GB. Phosphorylation and inactivation of glycogen synthase kinase 3 by protein kinase A. Proc Natl Acad Sci U S A. 2000;97:11960–11965. doi: 10.1073/pnas.220413597. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Li M, Wang X, Meintzer MK, Laessig T, Birnbaum MJ, Heidenreich KA. Cyclic AMP promotes neuronal survival by phosphorylation of glycogen synthase kinase 3beta. Mol Cell Biol. 2000;20:9356–9363. doi: 10.1128/mcb.20.24.9356-9363.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Goode N, Hughes K, Woodgett JR, Parker PJ. Differential regulation of glycogen synthase kinase-3 beta by protein kinase C isotypes. J Biol Chem. 1992;267:16878–16882. [PubMed] [Google Scholar]
  • 21.Cook D, Fry MJ, Hughes K, Sumathipala R, Woodgett JR, Dale TC. Wingless inactivates glycogen synthase kinase-3 via an intracellular signalling pathway which involves a protein kinase C. EMBO J. 1996;15:4526–4536. [PMC free article] [PubMed] [Google Scholar]
  • 22.Thomas GM, Frame S, Goedert M, Nathke I, Polakis P, Cohen P. A GSK3-binding peptide from FRAT1 selectively inhibits the GSK3-catalysed phosphorylation of axin and beta-catenin. FEBS Lett. 1999;458:247–251. doi: 10.1016/s0014-5793(99)01161-8. [DOI] [PubMed] [Google Scholar]
  • 23.ter Haar E, Coll JT, Austen DA, Hsiao HM, Swenson L, Jain J. Structure of GSK3beta reveals a primed phosphorylation mechanism. Nat Struct Biol. 2001;8:593–596. doi: 10.1038/89624. [DOI] [PubMed] [Google Scholar]
  • 24.Dajani R, Fraser E, Roe SM, Young N, Good V, Dale TC, Pearl LH. Crystal structure of glycogen synthase kinase 3 beta: structural basis for phosphate-primed substrate specificity and autoinhibition. Cell. 2001;105:721–732. doi: 10.1016/s0092-8674(01)00374-9. [DOI] [PubMed] [Google Scholar]
  • 25.Frame S, Cohen P, Biondi RM. A common phosphate binding site explains the unique substrate specificity of GSK3 and its inactivation by phosphorylation. Mol Cell. 2001;7:1321–1327. doi: 10.1016/s1097-2765(01)00253-2. [DOI] [PubMed] [Google Scholar]
  • 26.Stamos JL, Chu ML, Enos MD, Shah N, Weis WI. Structural basis of GSK-3 inhibition by N-terminal phosphorylation and by the Wnt receptor LRP6. Elife. 2014;3:e01998. doi: 10.7554/eLife.01998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Yang J, Cron P, Good VM, Thompson V, Hemmings BA, Barford D. Crystal structure of an activated Akt/protein kinase B ternary complex with GSK3-peptide and AMP-PNP. Nat Struct Biol. 2002;9:940–944. doi: 10.1038/nsb870. [DOI] [PubMed] [Google Scholar]
  • 28.Fu Z, Kim J, Vidrich A, Sturgill TW, Cohn SM. Intestinal cell kinase, a MAP kinase-related kinase, regulates proliferation and G1 cell cycle progression of intestinal epithelial cells. Am J Physiol Gastrointest Liver Physiol. 2009;297:G632–640. doi: 10.1152/ajpgi.00066.2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Togawa K, Yan YX, Inomoto T, Slaugenhaupt S, Rustgi AK. Intestinal cell kinase (ICK) localizes to the crypt region and requires a dual phosphorylation site found in map kinases. J Cell Physiol. 2000;183:129–139. doi: 10.1002/(SICI)1097-4652(200004)183:1<129::AID-JCP15>3.0.CO;2-S. [DOI] [PubMed] [Google Scholar]
  • 30.Fu Z, Schroeder MJ, Shabanowitz J, Kaldis P, Togawa K, Rustgi AK, Hunt DF, Sturgill TW. Activation of a nuclear Cdc2-related kinase within a mitogen-activated protein kinase-like TDY motif by autophosphorylation and cyclin-dependent protein kinase-activating kinase. Mol Cell Biol. 2005;25:6047–6064. doi: 10.1128/MCB.25.14.6047-6064.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Gangarosa LM, Sizemore N, Graves-Deal R, Oldham SM, Der CJ, Coffey RJ. A raf-independent epidermal growth factor receptor autocrine loop is necessary for Ras transformation of rat intestinal epithelial cells. J Biol Chem. 1997;272:18926–18931. doi: 10.1074/jbc.272.30.18926. [DOI] [PubMed] [Google Scholar]
  • 32.Fu Z, Larson KA, Chitta RK, Parker SA, Turk BE, Lawrence MW, Kaldis P, Galaktionov K, Cohn SM, Shabanowitz J, et al. Identification of yin-yang regulators and a phosphorylation consensus for male germ cell-associated kinase (MAK)-related kinase. Mol Cell Biol. 2006;26:8639–8654. doi: 10.1128/MCB.00816-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Bolick DT, Chen T, LAOA, Tong Y, Wu D, Joyner LT, 2nd, Oria RB, Guerrant RL, Fu Z. Intestinal cell kinase is a novel participant in intestinal cell signaling responses to protein malnutrition. PLoS One. 2014;9:e106902. doi: 10.1371/journal.pone.0106902. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Chen T, Wu D, Moskaluk CA, Fu Z. Distinct expression patterns of ICK/MAK/MOK protein kinases in the intestine implicate functional diversity. PLoS One. 2013;8:e79359. doi: 10.1371/journal.pone.0079359. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Obata T, Yaffe MB, Leparc GG, Piro ET, Maegawa H, Kashiwagi A, Kikkawa R, Cantley LC. Peptide and protein library screening defines optimal substrate motifs for AKT/PKB. J Biol Chem. 2000;275:36108–36115. doi: 10.1074/jbc.M005497200. [DOI] [PubMed] [Google Scholar]
  • 36.Stokes MP, Farnsworth CL, Moritz A, Silva JC, Jia X, Lee KA, Guo A, Polakiewicz RD, Comb MJ. PTMScan direct: identification and quantification of peptides from critical signaling proteins by immunoaffinity enrichment coupled with LC-MS/MS. Mol Cell Proteomics. 2012;11:187–201. doi: 10.1074/mcp.M111.015883. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Wu GQ, Xie D, Yang GF, Liao YJ, Mai SJ, Deng HX, Sze J, Guan XY, Zeng YX, Lin MC, et al. Cell cycle-related kinase supports ovarian carcinoma cell proliferation via regulation of cyclin D1 and is a predictor of outcome in patients with ovarian carcinoma. Int J Cancer. 2009;125:2631–2642. doi: 10.1002/ijc.24630. [DOI] [PubMed] [Google Scholar]
  • 38.An X, Ng SS, Xie D, Zeng YX, Sze J, Wang J, Chen YC, Chow BK, Lu G, Poon WS, et al. Functional characterisation of cell cycle-related kinase (CCRK) in colorectal cancer carcinogenesis. Eur J Cancer. 2010;46:1752–1761. doi: 10.1016/j.ejca.2010.04.007. [DOI] [PubMed] [Google Scholar]
  • 39.Feng H, Yu Z, Tian Y, Lee YY, Li MS, Go MY, Cheung YS, Lai PB, Chan AM, To KF, et al. A CCRK-EZH2 epigenetic circuitry drives hepatocarcinogenesis and associates with tumor recurrence and poor survival of patients. J Hepatol. 2015;62:1100–1111. doi: 10.1016/j.jhep.2014.11.040. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supp FigS1-2
NIHMS937093-supplement-Supp_FigS1-2.pptx (169.8KB, pptx)

RESOURCES