Structure and dimerization of the kinase domain from yeast Snf1, a member of the Snf1/AMPK protein family

The Snf1/AMPK kinases are intracellular energy sensors, and the AMPK pathway has been implicated in a variety of metabolic human disorders. Here we report the crystal structure of the kinase domain from yeast Snf1, revealing a bilobe kinase fold with greatest homology to cyclin-dependant kinase-2. Unexpectedly, the crystal structure also reveals a novel homodimer that we show also forms in solution, as demonstrated by equilibrium sedimentation, and in yeast cells, as shown by coimmunoprecipitation of differentially tagged intact Snf1. A mapping of sequence conservation suggests that dimer formation is a conserved feature of the Snf1/AMPK kinases. The conformation of the conserved alphaC helix, and the burial of the activation segment and substrate binding site within the dimer, suggests that it represents an inactive form of the kinase. Taken together, these studies suggest another layer of kinase regulation within the Snf1/AMPK family, and an avenue for development of AMPK-specific activating compounds.     

Crystal structure of the protein kinase domain of yeast AMP-activated protein kinase Snf1

AMP-activated protein kinase (AMPK) is a master metabolic regulator, and is an important target for drug development against diabetes, obesity, and other diseases. AMPK is a hetero-trimeric enzyme, with a catalytic (alpha) subunit, and two regulatory (beta and gamma) subunits. Here we report the crystal structure at 2.2A resolution of the protein kinase domain (KD) of the catalytic subunit of yeast AMPK (commonly known as SNF1). The Snf1-KD structure shares strong similarity to other protein kinases, with a small N-terminal lobe and a large C-terminal lobe. Two negative surface patches in the structure may be important for the recognition of the substrates of this kinase.     

Pak1 protein kinase regulates activation and nuclear localization of Snf1-Gal83 protein kinase

Three kinases, Pak1, Tos3, and Elm1, activate Snf1 protein kinase in Saccharomyces cerevisiae. This cascade is conserved in mammals, where LKB1 activates AMP-activated protein kinase. We address the specificity of the activating kinases for the three forms of Snf1 protein kinase containing the beta-subunit isoforms Gal83, Sip1, and Sip2. Pak1 is the most important kinase for activating Snf1-Gal83 in response to glucose limitation, but Elm1 also has a significant role; moreover, both Pak1 and Elm1 affect Snf1-Sip2. These findings exclude the possibility of a one-to-one correspondence between the activating kinases and the Snf1 complexes. We further identify a second, unexpected role for Pak1 in regulating Snf1-Gal83: the catalytic activity of Pak1 is required for the nuclear enrichment of Snf1-Gal83 in response to carbon stress. The nuclear enrichment of Snf1 fused to green fluorescent protein (GFP) depends on both Gal83 and Pak1 and is abolished by a mutation of the activation loop threonine; in contrast, the nuclear enrichment of Gal83-GFP occurs in a snf1Delta mutant and depends on Pak1 only when Snf1 is present. Snf1-Gal83 is the only form of the kinase that localizes to the nucleus. These findings, that Pak1 both activates Snf1-Gal83 and controls its nuclear localization, implicate Pak1 in regulating nuclear Snf1 protein kinase activity.     

Characterization of protein kinase C theta activation loop autophosphorylation and the kinase domain catalytic mechanism

Protein kinase C theta (PKCtheta), a member of the Ca(2+)-independent novel subfamily of PKCs, is required for T-cell receptor (TCR) signaling and IL2 production. PKCtheta-deficient mice have impaired Th2 responses in a murine ova-induced asthma model, while Th1 responses are normal. As an essential component of the TCR signaling complex, PKCtheta is a unique T-cell therapeutic target in the specific treatment of T-cell-mediated diseases. We report here the PKCtheta autophosphorylation characteristics and elucidation of the catalytic mechanism of the PKCtheta kinase domain using steady-state kinetics. Key phosphorylated residues of the active PKCtheta kinase domain expressed in Escherichia coli were characterized, and mutational analysis of the kinase domain was performed to establish the autophosphorylation and kinase activity relationships. Initial velocity, product inhibition, and dead-end inhibition studies provided assignments of the kinetic mechanism of PCKtheta(362)(-)(706) as ordered, wherein ATP binds kinase first and ADP is released last. Effects of solvent viscosity and ATPgammaS on PKCtheta catalysis demonstrated product release is partially rate limiting. Our studies provide important mechanistic insights into kinase activity and phosphorylation-mediated regulation of the novel PKC isoform, PKCtheta. These results should aid the design and discovery of PKCtheta antagonists as therapeutics for modulating T-cell-mediated immune and respiratory diseases.     

beta-subunits of Snf1 kinase are required for kinase function and substrate definition

The Snf1 kinase and its mammalian homolog, the AMP-activated protein kinase, are heterotrimeric enzymes composed of a catalytic alpha-subunit, a regulatory gamma-subunit and a beta-subunit that mediates heterotrimer formation. Saccharomyces cerevisiae encodes three beta-subunit genes, SIP1, SIP2 and GAL83. Earlier studies suggested that these subunits may not be required for Snf1 kinase function. We show here that complete and precise deletion of all three beta-subunit genes inactivates the Snf1 kinase. The sip1Delta sip2Delta gal83Delta strain is unable to derepress invertase, grows poorly on alternative carbon sources and fails to direct the phosphorylation of the Mig1 and Sip4 proteins in vivo. The SIP1 sip2Delta gal83Delta strain manifests a subset of Snf phenotypes (Raf(+), Gly(-)) observed in the snf1Delta 10 strain (Raf(-), Gly(-)), suggesting that individual beta-subunits direct the Snf1 kinase to a subset of its targets in vivo. Indeed, deletion of individual beta-subunit genes causes distinct differences in the induction and phosphorylation of Sip4, strongly suggesting that the beta-subunits play an important role in substrate definition.     

Roles of the glycogen-binding domain and Snf4 in glucose inhibition of SNF1 protein kinase

The SNF1/AMP-activated protein kinase (AMPK) family is required for adaptation to metabolic stress and energy homeostasis. The gamma subunit of AMPK binds AMP and ATP, and mutations that affect binding cause human disease. We have here addressed the role of the Snf4 (gamma) subunit in regulating SNF1 protein kinase in response to glucose availability in Saccharomyces cerevisiae. Previous studies of mutant cells lacking Snf4 suggested that Snf4 counteracts autoinhibition by the C-terminal sequence of the Snf1 catalytic subunit but is dispensable for glucose regulation, and AMP does not activate SNF1 in vitro. We first introduced substitutions at sites that, in AMPK, contribute to nucleotide binding and regulation. Mutations at several sites relieved glucose inhibition of SNF1, as judged by catalytic activity, phosphorylation of the activation-loop Thr-210, and growth assays, although analogs of the severe human mutations R531G/Q had little effect. We further showed that alterations of Snf4 residues that interact with the glycogen-binding domain (GBD) of the beta subunit strongly relieved glucose inhibition. Finally, substitutions in the GBD of the Gal83 beta subunit that are predicted to disrupt interactions with Snf4 and also complete deletion of the GBD similarly relieved glucose inhibition of SNF1. Analysis of mutant cells lacking glycogen synthase showed that regulation of SNF1 is normal in the absence of glycogen. These findings reveal novel roles for Snf4 and the GBD in regulation of SNF1.     

Conformational changes in the activation loop of the insulin receptor's kinase domain.

Low catalytic efficiency of basal-state protein kinases often depends on activation loop residues blocking substrate access to the catalytic cleft. Using the recombinant soluble form of the insulin receptor's kinase domain (IRKD) in its unphosphorylated state, activation loop conformation was analyzed by limited proteolysis. The rate of activation loop cleavage by trypsin is slow in the apo-IRKD. Bound Mg-adenine nucleoside di- and triphosphates increased the cleavage rate with half-maximal effects observed at 0.4-0.9 mM nucleotide. Adenosine monophosphate at concentrations up to 10 mM was not bound appreciably by the IRKD and had virtually no impact on activation loop cleavage. Amino-terminal and carboxy-terminal core-flanking regions of the IRKD had no statistically significant impact on the ligand-dependent or -independent activation loop cleavages. Furthermore, the core-flanking regions did not change the inherent conformational stability of the active site or the global stability of the IRKD, as determined by guanidinium chloride-induced denaturation. These measurements indicate that the intrasterically inhibitory conformation encompasses > or =90% of the ligand-free basal state kinase. However, normal intracellular concentrations of Mg-adenine nucleotides, which are in the millimolar range, would favor a basal-state conformation of the activation loop that is more accessible.     

Multiple activation loop conformations and their regulatory properties in the insulin receptor's kinase domain

Low catalytic efficiency of protein kinases often results from intrasteric inhibition caused by the activation loop blocking the active site. In the insulin receptor's kinase domain, Asp-1161 and Tyr-1162 in the peptide substrate-like sequence of the unphosphorylated activation loop can interact with four invariant residues in the active site: Lys-1085, Asp-1132, Arg-1136, and Gln-1208. Contributions of these six residues to intrasteric inhibition were tested by mutagenesis, and the unphosphorylated kinase domains were characterized. The mutations Q1208S, K1085N, and Y1162F each relieved intrasteric inhibition, increasing catalytic efficiency but without changing the rate-limiting step of the reaction. The mutants R1136Q and D1132N were virtually inactive. Steric accessibility of the active site was ranked by relative changes in iodide quenching of intrinsic fluorescence, and A-loop conformation was ranked by limited tryptic cleavage. Together these ranked the openness of the active site cleft as R1136Q approximately D1132N > or = D1161A > Y1162F approximately K1085N > Q1208S > or = wild-type. These findings demonstrate the importance of specific invariant residues for intrasteric inhibition and show that diverse activation loop conformations can produce similar steady-state kinetic properties. This suggests a broader range of regulatory properties for the activation loop than expected from a simple off-versus-on switch for kinase activation.     

Yeast Pak1 kinase associates with and activates Snf1

Members of the Snf1/AMP-activated protein kinase family are activated under conditions of nutrient stress by a distinct upstream kinase. Here we present evidence that the yeast Pak1 kinase functions as a Snf1-activating kinase. Pak1 associates with the Snf1 kinase in vivo, and the association is greatly enhanced under glucose-limiting conditions when Snf1 is active. Snf1 kinase complexes isolated from pak1Delta mutant strains show reduced specific activity in vitro, and affinity-purified Pak1 kinase is able to activate the Snf1-dependent phosphorylation of Mig1 in vitro. Purified Pak1 kinase promotes the phosphorylation of the Snf1 polypeptide on threonine 210 within the activation loop in vitro, and an increased dosage of the PAK1 gene causes increased Snf1 threonine 210 phosphorylation in vivo. Deletion of the PAK1 gene does not produce a Snf phenotype, suggesting that one or more additional protein kinases is able to activate Snf1 in vivo. However, deletion of the PAK1 gene suppresses many of the phenotypes associated with the deletion of the REG1 gene, providing genetic evidence that Pak1 activates Snf1 in vivo. The closest mammalian homologue of yeast Pak1 kinase, calcium-calmodulin-dependent protein kinase kinase beta, may play a similar role in mammalian nutrient stress signaling.     

TAK1 mitogen-activated protein kinase kinase kinase is activated by autophosphorylation within its activation loop

TAK1, a member of the mitogen-activated kinase kinase kinase family, is activated in vivo by various cytokines, including interleukin-1 (IL-1), or when ectopically expressed together with the TAK1-binding protein TAB1. However, this molecular mechanism of activation is not yet understood. We show here that endogenous TAK1 is constitutively associated with TAB1 and phosphorylated following IL-1 stimulation. Furthermore, TAK1 is constitutively phosphorylated when ectopically overexpressed with TAB1. In both cases, dephosphorylation of TAK1 renders it inactive, but it can be reactivated by preincubation with ATP. A mutant of TAK1 that lacks kinase activity is not phosphorylated either following IL-1 treatment or when coexpressed with TAB1, indicating that TAK1 phosphorylation is due to autophosphorylation. Furthermore, mutation to alanine of a conserved serine residue (Ser-192) in the activation loop between kinase domains VII and VIII abolishes both phosphorylation and activation of TAK1. These results suggest that IL-1 and ectopic expression of TAB1 both activate TAK1 via autophosphorylation of Ser-192.     

Rate-limiting domain and loop motions in arginine kinase

Arginine kinase catalyzes the reversible transfer of a phosphoryl group between ATP and arginine. It is the arthropod homologue of creatine kinase, buffering cellular ATP levels. Crystal structures of arginine kinase, in substrate-free and substrate-bound forms, have revealed large conformational changes associated with the catalytic cycle. Recent nuclear magnetic resonance identified movements of the N-terminal domain and a loop comprising residues I182--G209 with conformational exchange rates in the substrate-free enzyme similar to the turnover rate. Here, to understand whether these motions might be rate-limiting, we determined activation barriers for both the intrinsic dynamics and enzyme turnover using measurements over a temperature range of 15-30 °C. (15)N transverse relaxation dispersion yields activation barriers of 46 ± 8 and 34 ± 12 kJ/mol for the N-terminal domain and I182--G209 loop, respectively. An activation barrier of 34 ± 13 kJ/mol was obtained for enzyme turnover from steady-state kinetics. The similarity between the activation barriers is indeed consistent with turnover being limited by backbone conformational dynamics and pinpoints the locations of potentially rate-limiting motions.     

Ubp8 and SAGA regulate Snf1 AMP kinase activity

Posttranslational modifications of histone proteins play important roles in the modulation of gene expression. The Saccharomyces cerevisiae (yeast) 2-MDa SAGA (Spt-Ada-Gcn5) complex, a well-studied multisubunit histone modifier, regulates gene expression through Gcn5-mediated histone acetylation and Ubp8-mediated histone deubiquitination. Using a proteomics approach, we determined that the SAGA complex also deubiquitinates nonhistone proteins, including Snf1, an AMP-activated kinase. Ubp8-mediated deubiquitination of Snf1 affects the stability and phosphorylation state of Snf1, thereby affecting Snf1 kinase activity. Others have reported that Gal83 is phosphorylated by Snf1, and we found that deletion of UBP8 causes decreased phosphorylation of Gal83, which is consistent with the effects of Ubp8 loss on Snf1 kinase functions. Overall, our data indicate that SAGA modulates the posttranslational modifications of Snf1 in order to fine-tune gene expression levels.     

Subunits of the Snf1 kinase heterotrimer show interdependence for association and activity

The Snf1 kinase and its mammalian orthologue, the AMP-activated protein kinase (AMPK), function as heterotrimers composed of a catalytic alpha-subunit and two non-catalytic subunits, beta and gamma. The beta-subunit is thought to hold the complex together and control subcellular localization whereas the gamma-subunit plays a regulatory role by binding to and blocking the function of an auto-inhibitory domain (AID) present in the alpha-subunit. In addition, catalytic activity requires phosphorylation by a distinct upstream kinase. In yeast, any one of three Snf1-activating kinases, Sak1, Tos3, or Elm1, can fulfill this role. We have previously shown that Sak1 is the only Snf1-activating kinase that forms a stable complex with Snf1. Here we show that the formation of the Sak1.Snf1 complex requires the beta- and gamma-subunits in vivo. However, formation of the Sak1.Snf1 complex is not necessary for glucose-regulated phosphorylation of the Snf1 activation loop. Snf1 kinase purified from cells lacking the beta-subunits do not contain any gamma-subunit, indicating that the Snf1 kinase does not form a stable alphagamma dimer in vivo. In vitro kinase assays using purified full-length and truncated Snf1 proteins demonstrate that the kinase domain, which lacks the AID, is significantly more active than the full-length Snf1 protein. Addition of purified beta- and gamma-subunits could stimulate the kinase activity of the full-length alpha-subunit but only when all three subunits were present, suggesting an interdependence of all three subunits for assembly of a functional complex.     

Cyclic AMP-dependent protein kinase regulates the subcellular localization of Snf1-Sip1 protein kinase

The Snf1/AMP-activated protein kinase family has diverse roles in cellular responses to metabolic stress. In Saccharomyces cerevisiae, Snf1 protein kinase has three isoforms of the beta subunit that confer versatility on the kinase and that exhibit distinct patterns of subcellular localization. The Sip1 beta subunit resides in the cytosol in glucose-grown cells and relocalizes to the vacuolar membrane in response to carbon stress. We show that translation of Sip1 initiates at the second ATG of the open reading frame, yielding a potential site for N myristoylation, and that mutation of the critical glycine abolishes relocalization. We further show that the cyclic AMP-dependent protein kinase (protein kinase A [PKA]) pathway maintains the cytoplasmic localization of Sip1 in glucose-grown cells. The Snf1 catalytic subunit also exhibits aberrant localization to the vacuolar membrane in PKA-deficient cells, indicating that PKA regulates the localization of Snf1-Sip1 protein kinase. These findings establish a novel mechanism of regulation of Snf1 protein kinase by the PKA pathway.     

Subunit phosphorylation and activation of phosphorylase kinase in perfused rat hearts

The potential correlations between phosphorylase kinase subunit phosphorylation and activation have been examined using 32P-perfused rat hearts exposed to a variety of hormonal stimuli. Phosphate incorporation was measured after isolation of the enzyme by immunoprecipitation from heart extracts. Time courses of catecholamine or glucagon treatment produced a rapid rise in both the activity and the beta subunit phosphorylation of the enzyme, and a slightly slower increase in alpha' subunit phosphorylation. For short durations of catecholamine stimulation, the ratio of phosphate in the alpha' versus beta subunit was dependent upon hormone dose. After removal of hormone, both inactivation and alpha' subunit dephosphorylation were fairly slow, while the beta subunit was dephosphorylated more rapidly. For all of the above conditions, activation correlated with both alpha' and beta subunit phosphorylation. The maximum level of phosphate incorporation observed in response to hormonal stimulation is estimated to be approximately 1.3-1.7 mol of [32P]phosphate/mol of (alpha' beta gamma delta)4, divided about equally between the alpha' and beta subunits. When hearts were treated with hormone either in the absence of added calcium or in the presence of a calcium channel blocker, the time courses of subunit phosphorylation and activation were similar to those seen with standard perfusion conditions, suggesting that if any Ca2+-dependent autophosphorylation of phosphorylase kinase were occurring it does not make a major contribution to the observed hormonal responses. The complicated relationships observed here between phosphorylase kinase subunit phosphorylation and activation for the most part provide physiological affirmation of the patterns observed in vitro, but they also show some possible differences of potential interest.     

Protein kinase C phosphorylates protein kinase D activation loop Ser744 and Ser748 and releases autoinhibition by the pleckstrin homology domain

Persistent activation of protein kinase D (PKD) via protein kinase C (PKC)-mediated signal transduction is accompanied by phosphorylation at Ser(744) and Ser(748) located in the catalytic domain activation loop, but whether PKC isoforms directly phosphorylate these residues, induce PKD autophosphorylation, or recruit intermediate upstream kinase(s) is unclear. Here, we explore the mechanism whereby PKC activates PKD in response to cellular stimuli. We first assessed in vitro PKC-PKD transphosphorylation and PKD activation. A PKD738-753 activation loop peptide was well phosphorylated by immunoprecipitated PKC isoforms, consistent with similarities between the loop and their known substrate specificities. A similar peptide with glutamic acid replacing Ser(748) was preferentially phosphorylated by PKCepsilon, suggesting that PKD containing phosphate at Ser(748) is rapidly targeted by this isoform at Ser(744). When incubated in the presence of phosphatidylserine, phorbol 12,13-dibutyrate and ATP, intact PKD slowly autophosphorylated in the activation loop but only at Ser(748). In contrast, addition of purified PKCepsilon to the incubation mixture induced rapid Ser(744) and Ser(748) phosphorylation, concomitant with persistent 2-3-fold increases in PKD activity, measured using reimmunoprecipitated PKD to phosphorylate an exogenous peptide, syntide-2. We also further examined pleckstrin homology domain-mediated PKD regulation to determine its relationship with activation loop phosphorylation. The high constitutive activity of the pleckstrin homology (PH) domain deletion mutant PKD-deltaPH was not abrogated by mutation of Ser(744) and Ser(748) to alanines, suggesting that one function of activation loop phosphorylation in the PKD activation mechanism is to relieve autoinhibition by the PH domain. These studies provide evidence of a direct PKCepsilon-PKD phosphorylation cascade and provide additional insight into the activation mechanism.     

Phosphoregulation of mixed-lineage kinase 1 activity by multiple phosphorylation in the activation loop

Mixed-lineage kinase 1 (MLK1) is a mitogen-activated protein kinase kinase kinase capable of activating the c-Jun NH(2)-terminal kinase (JNK) pathway. Full-length MLK1 has 1104 amino acids and a domain structure identical to MLK2 and MLK3. Immunoblot and mass spectrometry show that MLK1 is threonine (and possibly serine) phosphorylated in or near the activation loop. A kinase-dead mutant is not, consistent with autophosphorylation. Mutation to alanine of any of the four serine or threonine residues in the activation loop reduces both the activity of the recombinant kinase domain and JNK pathway activation driven by full-length MLK1 expressed in mammalian cells. Furthermore, the gel mobility of the mutant MLK1s is closer to that of the kinase-dead than wild type, consistent with reduced phosphorylation. Thr312 is the key residue: MLK1[T312A] retains only basal activity (about 1-2% of wild type), and its gel mobility is indistinguishable from kinase-dead. Thr312 does not suffice, however; phosphorylation of multiple sites is necessary for full activation of MLK1. An activation mechanism consistent with these data involves phosphorylation of multiple sites in the activation loop, with phosphorylation of Thr312 required for full phosphorylation. This mechanism is broadly similar to that previously reported for MLK3 [Leung, I. W., and Lassam, N. (2001) J. Biol. Chem. 276, 1961-1967], but the key residue differs.