Dibutyryl cyclic AMP-induced phosphorylation of specific proteins in adenohypophysial cells

The objective of this work was to identify the natural substrates of cyclic AMP-dependent protein kinase in pituitary cells. Studies were performed using 2 systems: intact pituitary cells stimulated with dibutyryl cyclic AMP (DBC) after preincubation with [gamma-32P]. Phosphorylation of proteins was analyzed by two-dimensional gel electrophoresis, followed by autoradiography. In intact cells, the only clear and reproducible effect of DBC stimulation is increased phosphorylation of 3 proteins (termed A, B, and C), each with a molecular weight of about 20 000 dalton. The time-course and dose-dependence of phosphorylation of A, B and C are generally similar to that for DBC-induced hormone secretion, which is consistent with a role for these proteins in the secretory mechanism. When [gamma-32P]ATP is added to cell extracts, proteins A, B, and C are not measurably phosphorylated, either in the absence or presence of cyclic AMP. This observation suggests that proteins A, B and C may not be directly phosphorylated by cyclic AMP-dependent protein kinase, but may be phosphorylated indirectly by a second kinase. On the other hand, growth hormone and prolactin are readily phosphorylated in cell extracts by cyclic AMP-dependent protein kinase (although they are not phosphorylated in vivo). This finding makes clear the need for caution in interpreting results from broke cell systems.     

Activation induced by luteinizing hormone of type II protein kinase dependent on cyclic adenosine monophosphate and phosphorylation of soluble proteins in porcine granulosa cells

The present experiments were designed to study whether exogenous LH could elicit acute cyclic AMP-mediated activation of cyclic AMP-dependent protein kinase and phosphorylation of cellular protein in intact porcine granulosa cells. Incubation of porcine granulosa cells (from 3 to 5 mm diameter follicles) with 2 microgram luteinizing hormone/ml (LH) caused a significant rise of cellular cyclic AMP content within 2 min of the addition of LH. The increase was dose-dependent and occurred between doses of 0.2 and 2.0 microgram LH/ml. Luteinizing hormone also caused a time- and dose-dependent dissociation of the type II cyclic AMP-dependent protein kinase isozyme in porcine granulosa cells. Luteinizing hormone (0.05--2 microgram/ml) significantly dissociated the cyclic AMP-dependent protein kinase between 2 and 30 min after stimulation. The protein kinase dissociation was a specific effect of LH and was not elicited by either adrenocorticotrophic hormone or prolactin. During the period of LH-induced protein kinase activation, several soluble granulosa cell proteins, ranging in molecular weights from about 43 000 to 99 000, became phosphorylated in a time-dpeendent and hormone-specific manner. The results suggest that cyclic AMP-mediated activation of granulosa cell type II cyclic AMP-dependent protein kinase may be a prerequisite in the short-term molecular action of LH leading to LH-specific phosphorylation of several soluble granulosa cell proteins of an as yet unidentified function.     

Phosphorylation of myosin light chain kinase from vascular smooth muscle by cAMP- and cGMP-dependent protein kinases

The enzyme, myosin light chain kinase, has been purified to homogeneity from bovine aortic vascular smooth muscle. Approximately 10 mg of enzyme could be obtained from 1 kg of fresh aortas with an overall yield of 26% of the original activity. The vascular myosin light chain kinase has a molecular weight of 160 000 by sodium dodecyl sulfate polyacrylamide gel electrophoresis. Antiserum raised to the aortic myosin light chain kinase in rabbits strongly inhibited phosphotransferase activity. In addition, the antiserum was used to identify myosin kinase in a crude homogenate of vascular smooth muscle by radioimmunoblotting. A single species of the enzyme (Mr = 160 000) was identified. The bovine aortic myosin kinase could be phosphorylated by both cyclic AMP- and GMP-dependent protein kinases. Approximately 2 mols PO4/mole of enzyme could be incorporated by the cyclic AMP-dependent protein kinase in the absence of calmodulin. If Ca2+ and calmodulin were included in the reaction mixture, phosphate incorporation by the cyclic AMP-dependent protein kinase was reduced to 1 mol and phosphorylation by cyclic GMP-dependent protein kinase was completely inhibited. These results were confirmed by tryptic peptide mapping. Two distinct phosphopeptides were identified: site-1 and site-2. Both could be phosphorylated by the cyclic AMP-dependent protein kinase but only site-1 was phosphorylated by the cyclic GMP-dependent enzyme. In the presence of Ca2+ and calmodulin, phosphorylation by cAMP-dependent protein kinase was restricted to site-1. The effect of phosphorylation on myosin light chain kinase activity was determined. Only phosphorylation by cyclic AMP-dependent protein kinase was found to alter the requirement of myosin kinase for calmodulin. The K0.5 (i.e. the concentration of calmodulin required for half-maximal enzyme activation) for calmodulin was 5 nM for the unphosphorylated myosin kinase. With 2 mol PO4/mol myosin kinase incorporated, the K0.5 for calmodulin was increased to 82 nM. When only 1 mol PO4/mol myosin kinase was incorporated, no effect on calmodulin requirement was observed. Moreover, single site phosphorylation had no effect on other activity parameters, including Km for ATP and for light chains. Our studies suggest that cyclic AMP-dependent protein kinase may play an important role in the regulation of vascular myosin kinase activity. Moreover, our results indicate that cyclic GMP-dependent protein kinase does not affect calmodulin-activation of myosin kinase or several other activity parameters.     

LH-dependent steroid production and protein phosphorylation in culture of rat tumour Leydig cells

The possible cause of the decreased responsiveness of cultured Leydig cells to stimulation with LH was investigated with rat tumour Leydig cells cultured at 32 degrees C for 2 days. Pregnenolone production and phosphorylation of proteins were studied in combination with the activity of cyclic AMP-dependent protein kinase. Pregnenolone production in cultured tumour Leydig cells decreased from 45 +/- 16 on day 0 to 14 +/- 7 on day 2 (ng/60 min/mg protein) under basal conditions and from 226 +/- 108 on day 0 to 30 +/- 24 on day 2 (ng/60 min/ng protein) under LH-stimulated conditions (means +/- SD, n = 4). This decrease may be accounted for by decreased capacity of cholesterol side-chain cleavage which decreased (in the presence of 25-hydroxycholesterol) from 1.68 +/- 0.26 on day 0 to 0.74 +/- 0.48 on day 2 (microgram/60 min/mg protein; mean +/- SD, n = 4) and decreased activity of cyclic AMP-dependent protein kinase (as apparent from a direct assay of protein kinase activity and the extent of phosphorylation of LH-dependent phosphoproteins).     

Alterations in type I and type II protein kinases in the thyroid and adenohypophysis in response to a dietary goitrogen

The goitrogen, propylthiouracil (PTU), when administered in the diet of white leghorn chickens resulted in a marked increase in weight of the thyroid gland within 7 days which was maximal (160% of control) within 14 days. This system was used to determine the alterations in cyclic AMP-dependent protein kinase activity ratios and to assess the amounts of type I and type II cyclic AMP-dependent protein kinases present in the thyroid as well as any changes related to hypertrophy and hyperplasia of the thyroid. There was a marked elevation in the cyclic AMP-dependent protein kinase activity ratio ($\text{?cAMP}\text{+cAMP}$) 7 days after initiation of treatment as well as an increase in the total amount of supernatant cyclic AMP-dependent protein kinase activity (+cAMP) which doubled within 14 days of dietary PTU. Both type I and type II cyclic AMP-dependent protein kinase activities were detectable in the thyroid and adenohypophysis. Type I comprised 10% and type II 90% of the total cyclic AMP-dependent protein kinase activity in both tissues. The total amount of type I protein kinase activity increased significantly by 7 days. By 14 days, both type I and II protein kinases were increased twofold. The amount of type I returned to a control value by 21 days whereas type II remained elevated. Adenohypophysial type II protein kinase activity decreased to 70% of control by 14 days. Cyclic AMP is known to exert its effect on trophic responses through the activation of cyclic AMP-dependent protein kinase. These data implicate both activation of the enzymes as well as increased amounts of types I and II protein kinase activity in the thyroid in response to goitrogen-induced hypertrophy and hyperplasia.     

Cyclic AMP-dependent endogenous phosphorylation of a microtubule-associated protein.

Microtubules prepared from chick brain homogenates by successive cycles of assembly-disassembly were found to contain two high-molecular-weight proteins, designated microtubule-associated protein1 and microtubule-associated protein2. Microtubule-associated protein2 (apparent molecular weight 300,000 by sodium dodecyl sulfate-polyacrylamide gel electrophoresis) was the preferred substrate for an endogenous cyclic AMP-dependent protein kinase which appeared to be an integral component of the microtubules. The initial rate of phosphorylation of microtubule-associated protein2 was enhanced 4- to 6-fold by cyclic AMP, with half-maximal stimulation occurring at 2 times 10-7 M cyclic AMP. Under optimal conditions, a total of 1.0 and 1.9 mol of phosphate was incorporated per mole of microtubule-associated protein2, in the absence and presence of cyclic AMP, respectively. Cyclic AMP also stimulated the phosphorylation of tubulin, but the rate of phosphate incorporation per mol of tubulin was only 0.15% that of microtubule-associated protein2. The data raise the possibility that the cyclic AMP-dependent phosphorylation of microtubule-associated protein 2 may play a role in microtubule assembly or function.     

Inactivation of rabbit muscle phosphorylase phosphatase by cyclic AMP-dependent kinas.

Partially purified rabbit skeletal muscle phosphorylase phosphatase (EC 3.1.3.17; phosphoprotein phosphohydrolase) was inactivated when it was incubated with exogenous cyclic AMP-dependent protein kinase (EC 2.7.1.37; ATP:protein phosphotransferase), cyclic AMP, and ATP-Mg. Subsequent separation of the phosphatase by acrylamide gel electrophoresis or sucrose density centrifugation resulted in reactivation of the enzyme. The phosphatase decreased in molecular weight from approximately 70,000 to 52,000, and a phosphorylated inhibitor with molecular weight of 26,000 was found. Reactivation of phosphatase also occurred when it was incubated with MnCl2 or trypsin. The inhibitor was effective at less than 10(-8) M and was relatively heat stable. Its activity was destroyed by tryptic digestion and by dephosphorylation by a Mn-stimulated phosphatase. These observations support the possibility that phosphorylase phosphatase activity is controlled by cyclic AMP-dependent protein kinase and a Mn-stimulated phosphatase by a reaction involving phosphorylation and dephosphorylation of a protein phosphatase inhibitor.     

Fluorescent photoaffinity labeling: Adenosine 3′,5′-cyclic monophosphate receptor sites

An approach to the study of protein receptor sites in protein mixtures or supramolecular assemblies by using fluorescence spectroscopy is described. This approach, fluorescent photoaffinity labeling, combines the merits of photoaffinity labeling to attain site-directed reactivity with the probing power of fluorescent ligands. A fluorescent photoaffinity label for cyclic AMP receptor sites of cyclic AMP-dependent protein kinases was synthesized in both unlabeled and radioactive forms. The probe, 8-azido-1,N(6)-ethenoadenosine 3',5'-cyclic monophosphate, mimics cyclic AMP in its ability to stimulate the phosphotransferase activity of the protein kinases and strongly competes with cyclic AMP for its binding sites in all preparations so far tested. Photolysis, after equilibration of protein kinase and 8-azido-1,N(6)-ethenoadenosine 3',5'-cyclic monophosphate in the dark, effects binding of the intermediate nitrene irreversibly and specifically to the cyclic AMP sites with the development of fluorescence. Excess reagent and low molecular weight photolytic products are removable by dialysis. Studies of a crude beef heart preparation containing cyclic AMP-dependent protein kinase suggest that the cyclic AMP binding sites are hydrophobic in nature and strongly immobilize the adenine moiety of the cyclic nucleotide.     

Retinal horizontal cell gap junctional conductance is modulated by dopamine through a cyclic AMP-dependent protein kinase.

The action of many neuromodulators is mediated by intracellular second messengers such as cyclic AMP. In the retina, exogenously applied dopamine alters the conductance of gap junctions between cultured horizontal cells and this effect is mediated by cyclic AMP. However, it is not known how cyclic AMP modulates horizontal cell gap junction function. Here I report that cyclic AMP works by way of a cyclic AMP-dependent protein kinase. Cyclic AMP-dependent protein kinase injected into coupled horizontal cells from white bass (Roccus chrysops) rapidly and reversibly uncoupled the cells, mimicking the actions of dopamine. The threshold for the effect was between 0.06 and 0.03 microM. Injection of Walsh inhibitor of protein kinase [Walsh, D. A., Ashby, C. D., Gonzalez, C., Calkins, D., Fischer, E. H. & Krebs, E. G. (1971) J. Biol. Chem. 246, 1977-1985] blocked the effect of dopamine. Thus, the action of dopamine is to raise intracellular levels of cyclic AMP, which then activates a cyclic AMP-dependent protein kinase. Although not tested, it is likely that the cyclic AMP-dependent protein kinase phosphorylates a protein, possibly a gap junction protein, to alter conductance.     

Divergent actions of protein kinase modulator in regulating mammalian cyclic GMP-dependent and cyclic AMP-dependent protein kinases.

Protein kinase modulator can either augment or depress phosphorylation of substrate proteins catalyzed by cyclic GMP-dependent and cyclic AMP-dependent protein kinases prepared from mammalian and arthropod tissues. Alteration by the modulator of the phosphorylating activity of the protein kinases is considered to be due to modification of protein substrate specificity subsequent to interaction of the modulator with the catalytic subunits of the enzymes. It is likely that the physiologic role of the modulator is to monitor the opposing effects of cyclic GMP and cyclic AMP by regulating the activity of respective protein kinases. Thus the modulator may function as a biologic "fine tuner" providing on additional mechanism by which the signals imparted to cells by physiologic stimuli can be correctly expressed.     

Fructose-bisphosphatase as a substrate of cyclic AMP-dependent protein kinase

We have tested rat liver fructose-bisphosphatase (D-fructose-1,6-bisphosphate 1-phosphohydrolase, EC 3.1.3.11) and three other gluconeogenic fructose-bisphosphatases as substrates for the catalytic subunit of cyclic AMP-dependent protein kinase. In contrast to the rat liver enzyme, homogeneous preparations of mouse liver, rabbit liver, and pig kidney fructose-bisphosphatase could not be phosphorylated by the kinase. Comparative sodium dodecyl sulfate/polyacrylamide gel electrophoresis of the four above fructose-bisphosphatases revealed that the subunit molecular weight of the isolated rat liver enzyme (ca. 40,000-42,000) was greater than that of mouse liver, rabbit liver, and pig kidney fructose-bisphosphatases (ca. 36,000-37,000). Treatment of 32P-labeled rat liver fructose-bisphosphatase with trypsin resulted in the conversion of the rat liver enzyme to an active species with a subunit molecular weight identical to that of the three other enzymes, with complete loss of the 32P-labeled site. Identical trypsin treatment of pig kidney fructose-bisphosphatase caused no change in the molecular weight of the enzyme. The results suggest that the purified mouse liver, rabbit liver, and pig kidney fructose-bisphosphatases are not substrates for the cyclic AMP-dependent protein kinase in vitro because they lack the phosphorylation-site peptide.     

Actin nascent chains are substrates for cyclic AMP-dependent phosphorylation in vivo.

Two-dimensional gel electrophoresis of extracts of S49 mouse lymphoma cells labeled with [35S]methionine in the presence of inducers or analogs of cyclic AMP reveals a protein that both affinity purification and peptide mapping show to be a form of nonmuscle actin. This actin species also exhibits cyclic AMP-dependent labeling with [32P]phosphate, and, after acid hydrolysis, 32P label is found associated with phosphoserine. Phosphorylated actin does not appear when cells prelabeled with [35S]methionine are treated with an inducer of cyclic AMP in the presence of emetine, an inhibitor of protein synthesis; this suggests that only the nascent form of actin is a substrate for cyclic AMP-dependent phosphorylation. As well as differing slightly in isoelectric points, beta and gamma actins are found to yield different partial proteolytic cleavage products with staphylococcal protease. This microheterogeneity in the major cellular actin component is repeated in both the metabolically labile delta/epsilon actin and phosphorylated actin, suggesting that these three forms of actin derive from the same two gene products.     

Follicle-stimulating hormone-dependent phosphorylation of vimentin in cultures of rat Sertoli cells.

Endogenous protein phosphorylation was investigated in cultured rat Sertoli cells after treatment with follicle-stimulating hormone (FSH) and pharmacological agents that activate cAMP-dependent protein kinases. In intact Sertoli cells, both phosphorylation and dephosphorylation of proteins occurred in response to treatment with these agents. Studies using cell-free preparations suggest that four phosphoproteins phosphorylated by cAMP or the catalytic subunit of cAMP-dependent protein kinase were also phosphorylated in a FSH-dependent manner in intact cells. These data suggest that FSH-dependent phosphorylation in Sertoli cells occurs through activation of a cAMP-dependent protein kinase. A FSH-dependent phosphoprotein with a molecular weight of 58,000 was identified as the intermediate filament protein vimentin, based on its migration in two-dimensional gels and its peptide map. The cellular distribution of vimentin was monitored by immunofluorescence in Sertoli cells after treatment with FSH. Results of this study support a role for intermediate filaments in FSH-dependent events in Sertoli cells.     

Regulation of protein synthesis in reticulocyte lysates: phosphorylation of methionyl-tRNAf binding factor by protein kinase activity of translational inhibitor isolated from hemedeficient lysates.

A previous study demonstrated that the translational inhibitor from lysates of heme-deficient rabbit reticulocytes is associated with a protein kinase activity. Chromatography of this inhibitor preparation on phosphocellulose yields two distinct protein kinase activities, PC1 and PC2. PC1, which consitutes about 90% of the activity in the unresolved preparation, does not inhibit protein synthesis in lysates, but actively phosporylates calf thymus histone II in a 3':5'-cyclic AMP-denpendent reaction. PC2 contains the translational inhibitor, phosphorylates histone poorly, and is not cyclic AMP-dependent. While [gamma-32P]ATP as the phosphate donor, the two kinase fractions were analyzed with the putative substrates, salt-washed 40S ribosomal subunits, and the initiation factor that mediates the binding of Met-tRNAf to the 40S subunit. PC1 is inactive with the initiation factor, but phosphorylates 40S subunits at a single major site that migrates as a 31,000-dalton band in sodium dodecyl sulfate-acrylamide gels; phosphorylation requires cyclic AMP. Similar phosphorylation of the reticulocyte 40S site (31,000 daltons) can be demonstrated with other cyclic AMP-dependent kinases from reticulocytes, rat liver, and bovine heart muscle. PC2 phosphorylates the small subunit (38,000 daltons) but not the large subunit(s) of the initiation factor; the reaction does not require cyclic AMP. PC2 does not phosphorylate 40S subunits. In the presence of 40S subunits, the initiation factor appears to be rapidly bound in a manner that effectively blocks phosphorylation of the initiation factor by PC2; under the same conditions phosphorylation of the 40S subunit by PC1 is not affected. The initiation factor has been shown to reverse the inhibitions of protein chain initiation induced in lysates by heme deficiency, double-stranded RNA, oxidized glutathione, or the purified translational inhibitor. The observation that the Met-tRNAf binding factor is phosphorylated by PC2 supports the hypothesis that this initiation factor is a target for the action of the translational inhibitor activated in heme deficiency.     

Interconversion of Cyclic Nucleotide-Activated and Cyclic Nucleotide-Independent Forms of a Protein Kinase from Beef Heart

A protein kinase activated by cyclic nucleotides was purified from beef heart. Upon exposure to adenosine 3':5'-cyclic monophosphate (cyclic AMP) during gel filtration on Sephadex G-200, the protein kinase dissociated into a cyclic nucleotide-independent protein kinase and a cyclic nucleotide-binding protein. A similar or identical cyclic nucleotide-independent protein kinase could be obtained in highly purified form by clution from a DEAE-cellulose column with 10(-6) M cyclic AMP; the cyclic AMP-binding protein was apparently retained by the resin. The addition of cyclic nucleotide-binding protein to cyclic nucleotide-independent protein kinase resulted in the reappearance of cyclic nucleotide-dependent protein kinase, which could be isolated by filtration on Sephadex G-200 in the absence of cyclic AMP. These results confirm and extend previous suggestions that cyclic nucleotides activate protein kinases by dissociating them from inhibitory, cyclic nucleotide-binding proteins.     

Protein kinase C in pig thyroid cells: activation, translocation and endogenous substrate phosphorylating activity in response to phorbol esters

The phorbol ester, TPA, induced the intracellular redistribution of protein kinase C in intact thyroid cells; it caused within 5 min of incubation a 90% decrease of the cytosolic protein kinase C and an increase of the membrane-associated enzyme activity which appeared to be fully activated by TPA. TSH at concentrations which gave the maximal stimulation on various parameters of iodine metabolism induced the translocation of only 10-15% of protein kinase C from the cytosol to the membrane fraction. TPA induced a 2-fold increase in the incorporation of [32P]phosphate into cellular proteins and selectively activated the phosphorylation of two molecular species: a 180,000 Da protein and to a lesser extent a 170,000 Da protein in dispersed pig thyroid cells prelabeled with [32P]orthophosphate. The effect of TPA was maximum after 5 min of incubation and was concentration-dependent between 1 nM and 1 microM. The two phosphorylated substrates were only found in the cytosolic fraction. The TPA-induced phosphorylation of the 180,000 Da protein was observed in thyroid cells in suspension, in thyroid cell monolayers and follicle-like reassociated cells. In these three experimental situations, the 180,000 Da protein was not phosphorylated in response to TSH. Incubation of thyroid cell cytosolic fraction in the presence of [32P]ATP with calcium and phospholipid led to the phosphorylation of few proteins among which a 180,000 Da component. These proteins were not phosphorylated in the cytosol of TPA-treated cells, a finding in agreement with the translocation of protein kinase C. These results indicate that (1) the activation-translocation of thyroid protein kinase C induced by TPA is associated with the phosphorylation of selective substrates, and (2) TSH, even at high concentration, failed to exert the same action as TPA on protein kinase C in pig thyroid cells.     

Effects of phospholamban phosphorylation catalyzed by adenosine 3':5'-monophosphate- and calmodulin-dependent protein kinases on calcium transport ATPase of cardiac sarcoplasmic reticulum

To elucidate the role of 22000-dalton protein phospholamban, a putative regulator of Ca2+-dependent ATPase of cardiac sarcoplasmic reticulum, we examined the relationship between cyclic AMP- and calmodulin-dependent phosphorylation of phospholamban and their effects on ATPase activity and calcium transport of cardiac sarcoplasmic reticulum. Cardiac microsomes were incubated with [gamma-32P]ATP or unlabeled ATP, catalytic subunit of cyclic AMP-dependent protein kinase and/or exogenous calmodulin, and subsequently assayed for ATPase activity and calcium uptake by cardiac sarcoplasmic reticulum. Cyclic AMP-dependent phosphorylation of phospholamban was independent of Ca2+, whereas calmodulin-dependent phosphorylation of phospholamban was dependent on Ca2+ within a range between 0.2 and 50 microM. Cyclic AMP- and calmodulin-dependent phosphorylation of phospholamban occurred independently; when both kinases were operative, the amounts of phosphorylation were additive. Under these conditions, the phosphoproteins formed by cyclic AMP- and calmodulin-dependent protein kinases electrophoretically migrated as 11000-dalton components when sodium dodecyl sulfate-solubilized phosphoproteins were boiled prior to polyacrylamide gel electrophoresis. The ATPase activity was stimulated by either cyclic AMP- or calmodulin-dependent phosphorylation of phospholamban at Ca2+ concentrations up to 2 microM. The extents of stimulation of ATPase activity were additive when both types of phosphorylation were functional. Calcium uptake was similarly augmented by cyclic AMP- and/or calmodulin-dependent phosphorylation of phospholamban. These results indicate that Ca2+-dependent ATPase and calcium transport of cardiac sarcoplasmic reticulum are regulated by phospholamban phosphorylation catalyzed by cyclic AMP- and calmodulin-dependent protein kinases, thus suggesting a dual role of phospholamban in active calcium transport.     

Octopamine- and Serotonin-Stimulated Phosphorylation of Specific Protein in the Abdominal Ganglion of Aplysia california

Phosphorylation of a protein (or proteins) of molecular weight 120,000 in the Aplysia abdominal ganglion, as measured by incorporation of [(32)P] or [(33)P]sodium phosphate in vitro followed by separation of the phosphoproteins on sodium dodecyl sulfate-polyacrylamide gels, was specifically stimulated by incubation in the presence of the putative neurotransmitters octopamine or serotonin. The stimulatory effect of octopamine and serotonin was inhibited by phentolamine and methysergide, respectively, and was mimicked by incubation in the presence of dibutyryl cyclic AMP. Label-chase experiments indicated that the difference between control and octopamine-treated ganglia persists for several hours after removal of the drug from the incubation medium. This result suggests that neurotransmitters may produce relatively long-lasting changes in a phosphoprotein in the ganglion, perhaps in postsynaptic cells.     

Characterization of a N-bromoacetyl-L-thyroxine affinity-labeled 55-kilodalton protein as protein disulfide isomerase in cultured glial cells

In glial cells, thyroid hormone regulates the polymerization state of the actin cytoskeleton by a mechanism that does not require protein synthesis or the nuclear T3 receptor. Using the affinity label N-bromoacetyl-L-T4, we identified a thyroid hormone-binding protein of 55 kDa (glial-p55) in cultured glial cells that is unique from the type II iodothyronine 5'-deiodinase and which exhibits a T4-dependent shift from a membrane-associated pool to the F-actin cytoskeleton. In a number of other cell types, a 55-kDa thyroid hormone-binding protein has been identified and shown to be a subunit of the enzyme protein disulfide isomerase (PDI). In this study we have characterized glial-p55 and compared it with purified rat hepatic PDI. Glial-p55 appears to be identical to PDI by peptide fragmentation analysis, using multiple methods. The hydrodynamic properties of glial-p55, determined by molecular sieve chromatography and sucrose density centrifugation, showed that this protein has a native molecular mass of 115 kDa, which is in close agreement with that reported for PDI. Like PDI, glial-p55 is an acidic protein with a pI of 4.8 or less. Thus, the 55-kDa thyroid hormone-binding protein in glial cells appears to be PDI.