Biochemical purification uncovers mammalian sterile 3 (MST3) as a new protein kinase for multifunctional protein kinases AMPK and SIK3

The AMP-activated protein kinase (AMPK) and AMPK-related kinase salt-inducible kinase 3 (SIK3) regulate many important biological processes ranging from metabolism to sleep. Liver kinase B1 is known to phosphorylate and activate both AMPK and SIK3, but the existence of other upstream kinases was unclear. In this study, we detected liver kinase B1–independent AMPK-related kinase phosphorylation activities in human embryonic kidney cells as well as in mouse brains. Biochemical purification of this phosphorylation activity uncovered mammalian sterile 20–like kinase 3 (MST3). We demonstrate that MST3 from human embryonic kidney cells could phosphorylate AMPK and SIK3 in vivo. In addition, recombinant MST3 expressed in and purified from Escherichia coli could directly phosphorylate AMPK and SIK3 in vitro. Moreover, four other members of the MST kinase family could also phosphorylate AMPK or SIK3. Our results have revealed new kinases able to phosphorylate and activate AMPK and SIK3.     

Thrombin activates AMP-activated protein kinase in endothelial cells via a pathway involving Ca2+/calmodulin-dependent protein kinase kinase beta

AMP-activated protein kinase (AMPK) is a sensor of cellular energy state in response to metabolic stress and other regulatory signals. AMPK is controlled by upstream kinases which have recently been identified as LKB1 or Ca2+/calmodulin-dependent protein kinase kinase beta (CaMKKbeta). Our study of human endothelial cells shows that AMPK is activated by thrombin through a Ca2+-dependent mechanism involving the thrombin receptor protease-activated receptor 1 and Gq-protein-mediated phospholipase C activation. Inhibition of CaMKK with STO-609 or downregulation of CaMKKbeta using RNA interference decreased thrombin-induced AMPK activation significantly, indicating that CaMKKbeta was the responsible AMPK kinase. In contrast, downregulation of LKB1 did not affect thrombin-induced AMPK activation but abolished phosphorylation of AMPK with 5-aminoimidazole-4-carboxamide ribonucleoside. Thrombin stimulation led to phosphorylation of acetyl coenzyme A carboxylase (ACC) and endothelial nitric oxide synthase (eNOS), two downstream targets of AMPK. Inhibition or downregulation of CaMKKbeta or AMPK abolished phosphorylation of ACC in response to thrombin but had no effect on eNOS phosphorylation, indicating that thrombin-stimulated phosphorylation of eNOS is not mediated by AMPK. Our results underline the role of Ca2+ as a regulator of AMPK activation in response to a physiologic stimulation. We also demonstrate that endothelial cells possess two pathways to activate AMPK, one Ca2+/CaMKKbeta dependent and one AMP/LKB1 dependent.     

Regulation of brain-type creatine kinase by AMP-activated protein kinase: Interaction,phosphorylation and ER localization

AMP-activated protein kinase (AMPK) and cytosolic brain-type creatine kinase (BCK) cooperate under energy stress to compensate for loss of adenosine triphosphate (ATP) by either stimulating ATP-generating and inhibiting ATP-consuming pathways, or by direct ATP regeneration from phosphocreatine, respectively. Here we report on AMPK-dependent phosphorylation of BCK from different species identified by in vitro screening for AMPK substrates in mouse brain. Mass spectrometry, protein sequencing, and site-directed mutagenesis identified Ser6 as a relevant residue with one site phosphorylated per BCK dimer. Yeast two-hybrid analysis revealed interaction of active AMPK specifically with non-phosphorylated BCK. Pharmacological activation of AMPK mimicking energy stress led to BCK phosphorylation in astrocytes and fibroblasts, as evidenced with a highly specific phospho-Ser6 antibody. BCK phosphorylation at Ser6 did not affect its enzymatic activity, but led to the appearance of the phosphorylated enzyme at the endoplasmic reticulum (ER), close to the ER calcium pump, a location known for muscle-type cytosolic creatine kinase (CK) to support Ca²⁺-pumping.     

Resveratrol ameliorates high glucose-induced protein synthesis in glomerular epithelial cells

High glucose-induced protein synthesis in the glomerular epithelial cell (GEC) is partly dependent on reduction in phosphorylation of AMP-activated protein kinase (AMPK). We evaluated the effect of resveratrol, a phytophenol known to stimulate AMPK, on protein synthesis. Resveratrol completely inhibited high glucose stimulation of protein synthesis and synthesis of fibronectin, an important matrix protein, at 3 days. Resveratrol dose-dependently increased AMPK phosphorylation and abolished high glucose-induced reduction in its phosphorylation. We examined the effect of resveratrol on critical steps in mRNA translation, a critical event in protein synthesis. Resveratrol inhibited high glucose-induced changes in association of eIF4E with eIF4G, phosphorylation of eIF4E, eEF2, eEF2 kinase and, p70S6 kinase, indicating that it affects important events in both initiation and elongation phases of mRNA translation. Upstream regulators of AMPK in high glucose-treated GEC were explored. High glucose augmented acetylation of LKB1, the upstream kinase for AMPK, and inhibited its activity. Resveratrol prevented acetylation of LKB1 and restored its activity in high glucose-treated cells; this action did not appear to depend on SIRT1, a class III histone deacetylase. Our data show that resveratrol ameliorates protein synthesis by regulating the LKB1–AMPK axis.     

Ligand-regulated peptide aptamers that inhibit the 5'-AMP-activated protein kinase

In an effort to extend the peptide aptamer approach, we have developed a scaffold protein that allows small molecule ligand control over the presentation of a peptide aptamer. This scaffold, a fusion of three protein domains, FKBP12, FRB, and GST, presents a peptide linker region for target protein binding only in the absence of the small molecule Rapamycin or other non-immunosuppressive Rapamycin derivatives. Here we describe the characterization of ligand-regulated peptide aptamers that interact with and inhibit the 5'-AMP-activated protein kinase (AMPK). AMPK, a central regulator of cellular energy homeostasis, responds to high cellular AMP/ATP ratios by promoting energy producing pathways and inhibiting energy consuming biosynthetic pathways. We have characterized 15 LiRPs of similar, poly-basic sequence and have determined that they interact with the substrate peptide binding region of both AMPK alpha1 and alpha2. These proteins, some of which serve as poor substrates of AMPK, inhibit the kinase as pseudosubstrates in a Rapamycin-regulated fashion in vitro, an effect that is largely competitive with substrate peptide and mediated by an increase in the kinase's apparent K(m) for substrate peptide. This pseudosubstrate inhibition of AMPK by LiRP proteins reduced the AMP stimulation of AMPK in vitro and caused the inhibited state of the kinase to kinetically resemble the basal, unstimulated state of AMPK, providing potential insight into the molecular mechanisms of AMP stimulation of AMPK.     

Identification of a Nuclear Export Signal in the Catalytic Subunit of AMP-activated Protein Kinase

The metabolic regulator AMP-activated protein kinase (AMPK) maintains cellular homeostasis through regulation of proteins involved in energy-producing and -consuming pathways. Although AMPK phosphorylation targets include cytoplasmic and nuclear proteins, the precise mechanisms that regulate AMPK localization, and thus its access to these substrates, are unclear. We identify highly conserved carboxy-terminal hydrophobic amino acids that function as a leptomycin B–sensitive, CRM1-dependent nuclear export sequence (NES) in the AMPK catalytic subunit (AMPKα). When this sequence is modified AMPKα shows increased nuclear localization via a Ran-dependent import pathway. Cytoplasmic localization can be restored by substituting well-defined snurportin-1 or protein kinase A inhibitor (PKIA) CRM1-binding NESs into AMPKα. We demonstrate a functional requirement in vivo for the AMPKα carboxy-terminal NES, as transgenic Drosophila expressing AMPKα lacking this NES fail to rescue lethality of AMPKα null mutant flies and show decreased activation loop phosphorylation under heat-shock stress. Sequestered to the nucleus, this truncated protein shows highly reduced phosphorylation at the key Thr172 activation residue, suggesting that AMPK activation predominantly occurs in the cytoplasm under unstressed conditions. Thus, modulation of CRM1-mediated export of AMPKα via its C-terminal NES provides an additional mechanism for cells to use in the regulation of AMPK activity and localization.     

AMP-activated protein kinase mediates VEGF-stimulated endothelial NO production

Vascular endothelial growth factor (VEGF) is an important regulator of endothelial cell function. VEGF stimulates NO production, proposed to be a result of phosphorylation and activation of endothelial NO synthase (eNOS) at Ser1177. Phosphorylation of eNOS at this site also occurs after activation of AMP-activated protein kinase (AMPK) in cultured endothelial cells. We therefore determined whether AMPK mediates VEGF-stimulated NO synthesis in endothelial cells. VEGF caused a rapid, dose-dependent stimulation of AMPK activity, with a concomitant increase in phosphorylation of eNOS at Ser1177. Infection of endothelial cells with an adenovirus expressing a dominant negative mutant AMPK partially inhibited both VEGF-stimulated eNOS Ser1177 phosphorylation and NO production. VEGF-stimulated AMPK activity was completely inhibited by the Ca(2+)/calmodulin-dependent protein kinase kinase inhibitor, STO-609. Stimulation of AMPK via Ca(2+)/calmodulin-dependent protein kinase kinase represents a novel signalling mechanism utilised by VEGF in endothelial cells that contributes to eNOS phosphorylation and NO production.     

Nature of the intrinsic protein kinases involved in phosphorylation of non-histone proteins in intact prostatic nuclei: further identification of androgen-sensitive protein kinase reactions

Nuclei isolated from rat ventral prostate contain a number of messenger-dependent and -independent protein kinases. Studies were undertaken to determine the relative contribution of these protein kinases in phosphorylation of non-histone proteins (NHPs) in isolated nuclei. The data suggest that messenger-dependent protein kinases such as those dependent on cAMP or Ca²⁺/calmodulin or Ca^2–/phospholipid may be present in very small amounts in intact isolated nuclei, and thus appear not to be significantly involved in phosphorylation of endogenous NHPs. Messenger-independent nuclear associated protein kinases PK-N1 and PK-N2 are known to catalyze the phosphorylation of NHPs in vitro (Goueli SA, et al., Eur J Biochem 113: 45–51, 1980). Of these, the intrinsic heparin-sensitive PK-N2 as compared with heparin-insensitive PK-N1 appeared to be the predominant protein kinase engaged in phosphorylation of NHPs in intact nuclei. About 78–88% of NHP phosphorylation in intact nuclei was inhibited by heparin suggesting that the remaining 12–22% phosphorylation of NHPs was catalyzed via the heparin-insensitive protein kinase(s). Further, the data provide additional evidence that heparin-sensitive PK-N2 is the one that is most responsive to androgenic status in the animal.Abbreviations NHP Non-Histone Protein - PMSF Phenylmethylsulfonyl Fluoride - DTT Dithiothreitol - SDS Sodium Dodecyl Sulfate     

Linker engineering for fusion protein construction: Improvement and characterization of a GLP-1 fusion protein

Protein engineering has been successfully applied in protein drug discovery. Using this technology, we previously have constructed a fusion protein by linking the globular domain of adiponectin to the C-terminus of a glucagon-like peptide-1 (GLP-1) analog. Herein, to further improve its bioactivity, we reconstructed this fusion protein by introducing linker peptides of different length and flexibility. The reconstructed fusion proteins were overexpressed in Escherichia coli and purified using nickel affinity chromatography. Their agonist activity towards receptors of GLP-1 and adiponectin were assessed in vitro by using luciferase assay and AMP-activated protein kinase (AMPK) immunoblotting, respectively. The effects of the selected fusion protein on glucose and lipid metabolism were evaluated in mice. The fusion protein reconstructed using a linker peptide of AMGPSSGAPGGGGS showed high potency in activating GLP-1 receptor and triggering AMPK phosphorylation via activating the adiponectin receptor. Remarkably, the optimized fusion protein was highly effective in lowering blood glucose and lipids in mice. Collectively, these findings demonstrate that the bioactivity of this GLP-1 fusion protein can be significantly promoted by linker engineering, and indicate that the optimized GLP-1 fusion protein is a promising lead structure for anti-diabetic drug discovery.     

A low-temperature-responsive gene from barley encodes a protein with single-stranded nucleic acid-binding activity which is phosphorylated in vitro

A low-temperature-responsive gene, blt 801, isolated from a winter barley (Hordeum vulgare L.) cDNA library prepared from leaf meristematic tissue, was sequenced. The deduced amino acid sequence predicts a glycine-rich RNA-binding protein (GR-RNP) which was homology to stress-responsive GR-RNPs from several other plant species. BLT 801 is a two-domain protein, the amino-terminal domain comprises a consensus RNA-binding domain similar to that found in many eukaryotic genes and the carboxy-terminal domain is extremely glycine-rich (68.5% glycine). Blt 801 mRNA also accumulates in response to the phytohormone abscisic acid. The protein encoded by blt 801 has been produced as a recombinant fusion protein using a bacterial expression vector. The fusion protein, a chimaera of glutathione S-transferase and BLT 801, has been used in studies to determine nucleic acid binding and other characteristics. Binding studies with single-stranded nucleic acids show that BLT 801 has affinity for homoribopolymers G, A and U but not C, it also binds to single-stranded DNA and selects RNA molecules containing open loop structures enriched in adenine but low in cytosine. BLT 801 has a consensus motif for phosphorylation by cAMP protein kinase (PKA) at the junction between the two domains which can be phosphorylated by PKA in vitro and which, by analogy to animal studies, may have significance for controlling enzyme function.     

An activating mutation in the γ1 subunit of the AMP-activated protein kinase

The AMP-activated protein kinase (AMPK) is a heterotrimeric protein composed of a catalytic subunit and two regulatory subunits, β and γ. The γ subunit is essential for enzyme activity by virtue of its binding to the C-terminus of the subunit and appears to play some role in the determination of AMP sensitivity. We demonstrate that a γ1R70Q mutation causes a marked increase in AMPK activity and renders it largely AMP-independent. This activation is associated with increased phosphorylation of the subunit activation loop T172. These in vitro characteristics of AMPK are also reflected in increased intracellular phosphorylation of one of its major substrates, acetyl-CoA carboxylase. These data illustrate the importance of the γ1 subunit in the regulation of AMPK and its modulation by AMP.     

A two-dimensional screen for AMPK substrates identifies tumor suppressor fumarate hydratase as a preferential AMPKα2 substrate

AMP-activated protein kinase (AMPK) is emerging as a central cellular signaling hub involved in energy homeostasis and proliferation. The kinase is considered as a suitable target for pharmacological intervention in several energy-related pathologies like diabetes type II and cancer, although its signaling network is still incompletely understood. Here we apply an original two-dimensional in vitro screening approach for AMPK substrates that combines biophysical interaction based on surface plasmon resonance with in vitro phosphorylation. By enriching for proteins that interact with a specific AMPK isoform, we aimed to identify substrates that are also preferentially phosphorylated by this specific AMPK isoform. Application of this screen to full-length AMPK α2β2γ1 and soluble rat liver proteins identified the tumor suppressor fumarate hydratase (FH). FH was confirmed to interact with and to be preferentially phosphorylated by the AMPKα2 isoform by using yeast-two-hybrid and in vitro phosphorylation assays. AMPK-mediated phosphorylation of FH significantly increased enzyme activity in vitro and in vivo, suggesting that it is a bona fide AMPK substrate. In vivo, AMPKα2 is supposed to target the cytosolic/nuclear pools of FH, whose tumor suppressor function relies on DNA damage repair and inhibition of HIF-1α-signaling.     

Protein kinase substrate recognition studied using the recombinant catalytic domain of AMP-activated protein kinase and a model substrate

We have expressed a truncated form of the alpha1 kinase domain of AMP-activated protein kinase (AMPK) in Escherichia coli as a glutathione-S-transferase fusion (GST-KD). A T172D mutant version did not require prior phosphorylation and was utilized for most subsequent studies. We have also created a recombinant substrate (GST-ACC) by expressing 34 residues around the major phosphorylation site (Ser79) on rat acetyl-CoA carboxylase-1/alpha (ACC1) as a GST fusion. This was an excellent substrate that was phosphorylated with similar kinetic parameters to ACC1 by both native AMPK and the bacterially expressed kinase domain. We also constructed a structural model for the binding of the ACC1 sequence to the kinase domain, based on crystal structures for related protein kinases. The model was tested by making a total of 25 mutants of GST-ACC and seven mutants of GST-KD, and measuring kinetic parameters with different combinations. The results reveal that AMPK and ACC1 interact over a much wider region than previously realized (>20 residues). The features of the interaction can be summarised as follows: (i) an amphipathic helix from P-16 to P-5 on the substrate binds in a hydrophobic groove on the large lobe of the kinase; (ii) basic residues at P-6 and P-4 bind to two acidic patches (D215/D216/D217 and E103/D100/E143, respectively), on the large lobe; (iii) a histidine at P+3 interacts with D56 on the small lobe; (iv) the side-chain of P+4 leucine could not be precisely positioned, but a new finding was that asparagine or glutamine could replace a hydrophobic residue at this position. These interactions position the serine residue to be phosphorylated in close proximity to the gamma-phosphate group of ATP. Although based on modelling rather than a determined structure, this represents one of the most detailed studies of the interaction between a kinase and its substrate achieved to date.     

AMP-activated protein kinase kinase: detection with recombinant AMPK alpha1 subunit

The AMP-activated protein kinase (AMPK) is a heterotrimeric serine/threonine protein kinase important for the responses to metabolic stress. It consists of a catalytic alpha subunit and two non-catalytic subunits, beta and gamma, and is regulated both by the allosteric action of AMP and by phosphorylation of the alpha and beta subunits catalyzed by AMPKK(s) and autophosphorylation. The Thr172 site on the alpha subunit has been previously characterized as an activating phosphorylation site. Using bacterially expressed AMPK alpha1 subunit proteins, we have explored the role of Thr172-directed AMPKKs in alpha subunit regulation. Recombinant alpha1 subunit proteins, representing the N-terminus, have been expressed as maltose binding protein (MBP) 6x His fusion proteins and purified to homogeneity by Ni(2+) chromatography. Both wild-type alpha1(1-312) and alpha1(1-312)T172D are inactive when expressed in bacteria, but the former can be fully phosphorylated (1 mol/mol) on Thr172 and activated by a surrogate AMPKK, CaMKKbeta. The corresponding AMPKalpha1(1-392), an alpha construct containing its autoinhibitory sequence, can be similarly phosphorylated, but it remains inactive. In an insulinoma cell line, either low glucose or 5-aminoimidazole-4-carboxamide ribonucleoside (AICAR) treatment leads to activation and T172 phosphorylation of endogenous AMPK. Under the same conditions of cell incubation, we have identified an AMPKK activity that both phosphorylates and activates the recombinant alpha1(1-312), but this Thr172-directed AMPKK activity is unaltered by low glucose or AICAR, indicating that it is constitutively active.     

The connexin43 gap junction protein is phosphorylated by protein kinase A and protein kinase C: In vivo and in vitro studies

There is general agreement that the connexin43 gap junction protein is a substrate for phosphorylation by protein kinase C but there is no similar consensus regarding the action of protein kinase A. Our previous studies demonstrated that channels formed by connexin43 were reversibly gated in response to microinjected protein kinase A and protein kinase C, but we did not determine whether these effects involved direct action on the connexin43 protein. Using a combination of in vivo metabolic labeling and in vitro phosphorylation of recombinant protein and synthetic peptides, we now find that connexin43 is a relatively poor substrate for purified protein kinase A compared to protein kinase C, but that phosphorylation can be accelerated by 8-Br-cAMP (8-bromoadenosine 3,5-cyclic monophosphate) which also enhances connexin43 synthesis but at a much slower rate than phosphorylation. Phosphorylation of a critical amino acid, Ser364, by protein kinase A, appears to be necessary for subsequent multiple phosphorylations by protein kinase C. However, protein kinase C can phosphorylate connexin43 at a reduced level in the absence of prior phosphorylation. The results suggest that the correct regulation of channels formed by connexin43 may require sequential phosphorylations of this protein by protein kinase A and protein kinase C.     

Inhibition of Kir2.1 (KCNJ2) by the AMP-activated protein kinase

The inward rectifier K⁺ channel Kir2.1 participates in the maintenance of the cell membrane potential in a variety of cells including neurons and cardiac myocytes. Mutations of KCNJ2 encoding Kir2.1 underlie the Andersen–Tawil syndrome, a rare disorder clinically characterized by periodic paralysis, cardiac arrhythmia and skeletal abnormalities. The maintenance of the cardiac cell membrane potential is decreased in ischaemia, which is known to stimulate the AMP-activated serine/threonine protein kinase (AMPK). This energy-sensing kinase stimulates energy production and limits energy utilization. The present study explored whether AMPK regulates Kir2.1. To this end, cRNA encoding Kir2.1 was injected into Xenopus oocytes with and without additional injection of wild type AMPK (AMPKα1 + AMPKβ1 + AMPKγ1), of the constitutively active ^γR70QAMPK (α1β1γ1(R70Q)), of the kinase dead mutant ^αK45RAMPK (α1(K45R)β1γ1), or of the ubiquitin ligase Nedd4-2. Kir2.1 activity was determined in two-electrode voltage-clamp experiments. Moreover, Kir2.1 protein abundance in the cell membrane was determined by immunostaining and subsequent confocal imaging. As a result, wild type and constitutively active AMPK significantly reduced Kir2.1-mediated currents and Kir2.1 protein abundance in the cell membrane. Expression of wild type Nedd4-2 or of Nedd4-2^S795A lacking an AMPK phosphorylation consensus sequence downregulated Kir2.1 currents. The effect of wild type Nedd4-2 but not of Nedd4-2^S795A was significantly augmented by additional coexpression of AMPK. In conclusion, AMPK is a potent regulator of Kir2.1. AMPK is at least partially effective through phosphorylation of the ubiquitin ligase Nedd4-2.     

Chemically reprogramming the phospho‐transfer reaction to crosslink protein kinases to their substrates

The proteomic mapping of enzyme–substrate interactions is challenged by their transient nature. A method to capture interacting protein kinases in complexes with a single substrate of interest would provide a new tool for mapping kinase signaling networks. Here, we describe a nucleotide‐based substrate analog capable of reprogramming the wild‐type phosphoryl‐transfer reaction to produce a kinase‐acrylamide‐based thioether crosslink to mutant substrates with a cysteine nucleophile substituted at the native phosphorylation site. A previously reported ATP‐based methacrylate crosslinker (ATP‐MA) was capable of mediating kinase crosslinking to short peptides but not protein substrates. Exploration of structural variants of ATP‐MA to enable crosslinking of protein substrates to kinases led to the discovery that an ADP‐based methacrylate (ADP‐MA) crosslinker was superior to the ATP scaffold at crosslinking in vitro. The improved efficiency of ADP‐MA over ATP‐MA is due to reduced inhibition of the second step of the kinase–substrate crosslinking reaction by the product of the first step of the reaction. The new probe, ADP‐MA, demonstrated enhanced in vitro crosslinking between the Src tyrosine kinase and its substrate Cortactin in a phosphorylation site‐specific manner. The kinase–substrate crosslinking reaction can be carried out in a complex mammalian cell lysate setting, although the low abundance of endogenous kinases remains a significant challenge for efficient capture.     

Protein phosphorylation in rat cardiac microsomes: Effects of inhibitors of protein kinase A and of phosphatases

The phosphorylation of rat cardiac microsomal proteins was investigated with special attention to the effects of okadaic acid (an inhibitor of protein phosphatases), inhibitor 2 of protein phosphatase 1 and inhibitor of cyclic AMP-dependent protein kinase (protein kinase A). The results showed that okadaic acid (5 µM) modestly but reproducibly augmented the protein kinase A-catalyzed phospholamban (PLN) phosphorylation, although exerted little effect on the calcium/calmodulin kinase-catalyzed PLN phosphorylation. Microsomes contained three other substrates (M_r 23, 19 and 17 kDa) that were phosphorylated by protein kinase A but not by calcium/calmodulin kinase. The protein kinase A-catalyzed phosphorylation of these three substrates was markedly (2-3 fold) increased by 5 µM okadaic acid. Calmodulin was found to antagonize the action of okadaic acid on such phosphorylation. Protein kinase A inhibitor was found to decrease the protein kinase A-catalyzed phosphorylation of microsomal polyp eptides. Unexpectedly, inhibitor 2 was also found to markedly decrease protein kinase A-catalyzed phosphorylation of phospholamban as well these other microsomal substrates. These results are consistent with the views that protein phosphatase 1 is capable of dephosphorylating membrane-associated phospholamban when it is phosphorylated by protein kinase A, but not by calcium/calmodulin kinase, and that under certain conditions, calcium/calmodulin-stimulated protein phosphatase (protein phosphatase 2B) is also able to dephosphorylate PLN phosphorylated by protein kinase A. Additionally, the observations show that protein phosphatase 1 is extremely active against the three protein kinase A substrates (M_r 23, 19 and 17 kDa) that were present in the isolated microsomes and whose state of phosphorylation was particularly affected in the presence of dimethylsulfoxide. Protein phosphatase 2B is also capable of dephosphorylating these three substrates. (Mol Cell Biochem 175: 109–115, 1997