Differential down-regulation of protein kinase C isozymes

Types I, II, and III protein kinase C have been shown to be products of, respectively, gamma, beta, and alpha genes of this enzyme family (Huang, F. L., Yoshida, Y., Nakabayashi, H., Knopf, J. L., Young, W. S., III, and Huang, K.-P. (1987) Biochem. Biophys. Res. Commun. 149, 946-952). Incubation of the highly purified rat brain protein kinase C isozymes with trypsin (kinase/trypsin (w/w) = 100) under identical conditions results in a preferential degradation of types I and II enzymes, whereas the type III enzyme was relatively resistant to tryptic proteolysis. Degradation of the type III enzyme by trypsin could be facilitated with the addition of Ca2+, phosphatidylserine, and dioleoylglycerol; none of these components alone was effective. Limited proteolysis of the three protein kinase C isozymes generated distinctive fragments for each isozyme, indicating that each isozyme has different trypsin-sensitive sites. Tryptic digestion of the type III protein kinase C was used as a model to determine the effects of various modulators on protein kinase C degradation. While Ca2+ and phosphatidylserine together were sufficient to convert the type III protein kinase C from a trypsin-insensitive to a -sensitive form, addition of dioleoylglycerol greatly reduced the Ca2+ requirement for such a conversion. Among the various phospholipids tested, in the presence of either dioleoylglycerol or phorbol ester, phosphatidylserine, cardiolipin, and phosphatidic acid were the most effective, and phosphatidylcholine and phosphatidylethanolamine were the least effective in supporting the digestion of type III protein kinase. Other acidic phospholipids, such as lysophosphatidylserine and phosphatidylinositol, were also effective in supporting the degradation in the presence of phorbol ester but not in the presence of dioleoylglycerol. The relevance of these proteolytic reactions to physiological responses was assessed with phorbol ester on rat basophilic leukemia RBL-2H3 cells, which contained both types II and III protein kinase C. Immunoblot analysis with the isozyme-specific antibodies revealed that phorbol ester induced a faster degradation of type II than that of type III isozyme in these cells. The results demonstrate that the various protein kinase C isozymes have different susceptibilities to proteolysis in vitro, when tested with trypsin, as well as to endogenous proteases in intact cells.     

Tissue distribution and developmental expression of protein kinase C isozymes

Protein kinase C is a ubiquitous enzyme found in a variety of mammalian tissues and is especially highly enriched in brain and lymphoid organs. Based on biochemical and immunological analyses, we have identified three types of protein kinase C isozyme (designated types I-III) from rat brain. Monospecific antibodies against each of the protein kinase C isozymes were prepared for the determination of tissue distribution, subcellular localization, and developmental changes of these enzymes. The various protein kinase C isozymes were found to be distinctively distributed in different tissues: the type I enzyme in brain; the type II enzyme in brain, pituitary and pineal glands, spleen, thymus, retina, lung, and intestine; and the type III enzyme in brain, pineal gland, retina, and spleen. The rat brain enzymes were differentially distributed in different subcellular fractions. The type I enzyme appeared to be most lipophilic and was recovered mostly in the particulate fractions (80-90%) regardless of the EGTA- or Ca2+-containing buffer used in the homogenization. Significant amounts (30-40%) of the type II and III enzymes were recovered in the cytosolic fraction with EGTA-containing buffer. The expressions of different protein kinase C isozymes appear to be differently controlled during development. In rat brain, both type II and III enzymes were found to increase progressively from 3 days before birth up to 2-3 weeks of age and remained constant thereafter. However, the expression of the type I enzyme displayed a different developmental pattern; it was very low within 1 week, and an abrupt increase was observed between 2 and 3 weeks of age. In thymus, the type II enzyme was found to be maximal shortly after birth; whereas the same kinase in spleen was very low within 2 weeks of age, and a significant increase was observed between 2 and 3 weeks. These results demonstrate that protein kinase C isozymes are distinctively distributed in different tissues and subcellular locales and that their expressions are controlled differently during development.     

Biochemical characterization of rat brain protein kinase C isozymes

Biochemical characteristics of three rat brain protein kinase C isozymes, types I, II, and III, were compared with respect to their protein kinase and phorbol ester-binding activities. All three isozymes appeared to be alike in their phorbol ester-binding activities as evidenced by their similar Kd for phorbol 12,13-dibutyrate and requirements for Ca2+ and phospholipids. However, differences with respect to the effector-mediated stimulation of protein kinase activity were detectable among these isozymes. The type I enzyme could be stimulated by cardiolipin to a greater extent than those of the type II and III enzymes. In the presence of cardiolipin, the concentrations of dioleoylglycerol or phorbol 12,13-dibutyrate required for half-maximal activation (A1/2) of the type I enzyme were nearly an order of magnitude lower than those for the type II and III enzymes. In the presence of phosphatidylserine, differences in the A1/2 of dioleoylglycerol and phorbol 12,13-dibutyrate for the three isozymes of protein kinase C were less significant than those measured in the presence of cardiolipin. Nevertheless, the A1/2 of these two activators for the type I enzyme were lower than those for the type II and III enzymes. At high levels of phosphatidylserine (greater than 15 mol %), binding of phorbol 12,13-dibutyrate to the type I enzyme evoked a corresponding stimulation of the kinase activity, whereas binding of this phorbol ester to the type II and III enzymes produced a lesser degree of kinase stimulation. For all three isozymes, the concentrations of phosphatidylserine required for half-maximum [3H]phorbol 12,13-dibutyrate binding were almost an order of magnitude less than those for kinase stimulation. Consequently, neither isozyme exhibited a significant kinase activity at lower levels of phosphatidylserine (less than 5 mol %) and phorbol 12,13-dibutyrate (50 nM), a condition sufficient to promote near maximal phorbol ester binding. In addition to their different responses to the various activators, the three protein kinase C isozymes also have different Km values for protein substrates. The type I enzyme appeared to have lower Km values for histone IIIS, myelin basic protein, poly(lysine, serine) (3:1) polymer, and protamine than those for the type II and III enzymes. These results documented that the three protein kinase C isozymes were distinguishable in their biochemical properties. In particular, the type I enzyme, which is a brain-specific isozyme, is distinct from the type II and III enzymes, both have a widespread distribution among different tissues.     

Differential localization of protein kinase C isozymes in retinal neurons

We report the immunohistochemical localization of protein kinase C isozymes (types I, II, and III) in the rabbit retina using the monospecific monoclonal antibodies MC-1a, MC-2a, and MC-3a. Using immunoblot analysis of partially purified protein kinase C preparations of rabbit retina, types II and III isozymes alone were detected. The activity of type III was the stronger. By light microscopic immunohistochemical analysis, retinal neurons were negative for type I and positive for type II and type III isozymes. Type II was more diffusely distributed through the retinal layers, but was distinctive in ganglion cells, bipolar cells, and outer segments. The immunoreactivity was stronger for type III isozyme, and it was observed in mop (rod) bipolar cells and amacrine cells. By using immunoelectron microscopy, the cytoplasm of the cell body, the axon, and dendrites of the mop bipolar cells were strongly immunoreactive for type III. The so-called rod bipolar cells were for the first time seen to form synapses with rod photoreceptor cells. These differential localizations of respective isozymes in retinal neurons suggest that each isozyme has a different site of function in each neuron.     

Isozymic forms of protein kinase C in regenerating rat liver

The expression of multiple forms of protein kinase C (PK-C) was studied in regenerating rat liver using hydroxyapatite column chromatography. Two forms of the enzyme were found in the cytosolic as well as membrane fraction of livers from partially hepatectomized rats. The kinetic variation in the activation of these two liver isozymes by fatty acids, phosphatidylserine and diacylglycerol was similar to that reported for the PK-C subspecies from rat brain, designated types II and III. Intracellular redistribution of PK-C caused by phorbol 12-myristate 13-acetate (PMA) was concentration-dependent and was due to translocation of isozyme III, because type II was insensitive to 5 x 10(-8) M PMA. The activity ratio of the two isozymes in either the particulate or cytosolic fraction was the same at 22 h as compared to 4 h after partial hepatectomy.     

Type I protein kinase C isozyme in the visual-information-processing pathway of monkey brain

Previously using PKC isozyme-specific antibodies for immunoblot analysis, we demonstrated the heterogeneous distribution of PKC isozymes in various regions of monkey and rat brains and that type I PKC was most abundant in cerebellum, hippocampus, amygdala, and cerebral cortex (Huang et al.: J Biol Chem 262:15714-15720, 1987). Using these antibodies, we have also demonstrated that type I, II, and III PKC are products of PKC genes gamma, beta, and alpha, respectively (Huang et al.: Biochem Biophys Res Commun 149:946-952, 1987). By immunocytochemical analysis, type I PKC-specific antibody showed strong reactivity in various types of neuron in hippocampal formation, amygdala, cerebellum, and neocortex. In hippocampal formation, granule cells of dentate gyrus and pyramidal cells of hippocampus were heavily stained. By immunoblot analysis, relative levels of PKC isozymes in several areas of monkey cerebral cortex involved in the visual information processing and storage were determined. Both type II and III PKCs appeared to be evenly distributed and at moderate levels, type I PKC formed a gradient of increasing concentration rostral along the cerebral cortex of occipital to temporal and then to the limbic areas. Neurobehavioral studies have demonstrated that the neocortical and limbic areas of the anterior and medial temporal regions participate more directly than the striate, prestriate, and posterior temporal regions in the storage of visual representations and that both hippocampus and amygdala are important in the memory formation. As type I PKC is present at high levels in hippocampus, amygdala, and anterior temporal lobe, we predict that the type I protein kinase C may participate in the plastic changes important for mnemonic function.     

Phosphorylation of the EGF receptor from A431 epidermoid carcinoma cells by three distinct types of protein kinase C

Three distinct types of protein kinase C obtained from rat brain cytosol phosphorylated the EGF receptor of A431 epidermoid carcinoma cells at different rates. This receptor was phosphorylated most rapidly by type III protein kinase C, but slowly by type I enzyme. Type II enzyme showed intermediate activity. Chromatographic analysis indicated that A431 cells possessed only one of the three types found in rat brain, which apparently corresponded to type III enzyme. This type of protein kinase C, that is encoded by the alpha-sequence or a closely related sequence, appeared to be expressed commonly in many tissues and organs. The result implies that type III enzyme may play roles in growth promotion.     

Endogenous substrates of rat heart protein kinase C type I, II, and III isozymic forms in cardiac sarcolemma.

The endogenous substrate proteins of rat cardiac protein kinase C type I, II, and III isozymic forms were studied in rat cardiac sarcolemma. The 19-, 21-, 29-, 35-, and 95-kDa proteins were phosphorylated by both types II and III, but not type I. The extent of phosphorylation by individual protein kinase C isozymic forms was additive and equal to the extent of phosphorylation observed when a mixture of isozymic forms was employed. The extent of phosphorylation of the 21-kDa protein by type III was much higher than that by type II. These results suggest that the protein kinase C isozymes have preferences for specific endogenous substrate proteins. The phosphorylation of these endogenous substrate proteins by protein kinase C isozymes probably plays a role in cardiac cell functions.     

Phosphorylation of aortic plasma membranes by protein kinase C

A calcium-sensitive, phospholipid-dependent protein kinase (protein kinase C) and its three isozymes were purified from rat heart cytosolic fractions utilizing a rapid purification method. The purified protein kinase C enzyme showed a single polypeptide band of 80 KDa on SDS-polyacrylamide gel electrophoresis, and was totally dependent on the presence of Ca²⁺ and phospholipid for activity. Diacylglycerol was also found to stimulate enzymatic activity. Autophosphorylation of the purified PKC showed an 80 KDa polypeptide. The identity of the purified protein was also verified with monoclonal antibodies specific for PKC. Further fractionation of the purified PKC on a hydroxylapatite column yielded three distinct peaks of enzyme activity, corresponding to type I, II and III based on similar chromatographic behaviour as the rat brain enzyme. All three forms were entirely Ca^2– and phosphatidylserine dependent. Type II was found to be the most abundant. Type I was found to be highly unstable. PKC activity studies demonstrate that types II and III isozymic forms are different with respect to their sensitivity to Ca²⁺.Abbreviations PKC Protein Kinase C - SDS Sodium Dodecyl Sulfate - PAGE Polyacrylamide Gel Electrophoresis - K_m Michaelis constant - NBT Nitro-Blue Tetrazolium - BCIP 5-Bromo-4-Chloro-3-Indolyl Phosphate     

Differential sensitivity of neural and non-neural protein kinase isozymes to cyclic AMP

Cyclic AMP-dependent protein kinase, which plays a major role in metabolic and genetic regulation, consists of two classes of isozymes denoted as type I and type II. The type II isozyme, moreover, consists of two subclasses denoted as neural and non-neural based upon immunochemical differences between the enzyme isolated from bovine brain and heart, respectively. Whereas the catalytic (C) subunits of these three isozymes are quite similar, all three isozymes differ with respect to their regulatory (R) subunits. In the present report, we have compared the sensitivities to cyclic AMP of the type I and type II isozymes in several tissues from a single species (rat). The sensitivities of the three isozymes to cyclic AMP were type I much greater than non-neural type II greater than neural type II. We suggest that the differences in sensitivity to cyclic AMP of isozymes present in the same cell provides the cell with a dynamic range of responses to the widely varying alterations in cellular cyclic AMP levels produced by regulatory first messengers.     

Immunohistochemical evidence for the overexpression of protein kinase C in proliferative diseases of human thyroid.

We report immunohistochemical evidence for the overexpression of protein kinase C in various proliferative diseases of human thyroid. Immunohistochemical characterization of various surgically removed thyroid tissues, viz., cancer tissues: papillary carcinoma and follicular carcinoma; adenoma tissues: tubular, trabecular and colloid adenomas; adenomatous goiter; and normal thyroid was done using the monospecific monoclonal antibodies MC-1a, MC-2a and MC-3a, each of which is specific for types I, II and III isozymes of protein kinase C, respectively. For protein kinase C type II, a remarkable difference in staining intensity was noted between the cancerous and normal tissues. The cytoplasm of papillary and follicular carcinoma cells stained more intensely than that of normal thyroid cells. In the benign tumor and adenomatous goiter tissues, stronger staining was noted in the papilliform-proliferating portion and cubic epithelial cells. In the normal thyroid tissues, epithelial cells of greater height were more strongly stained than simple squamous epithelial cells. These results indicated that protein kinase C type II isozyme is expressed in larger amounts in cancerous and proliferative tissues of the human thyroid.     

Characterization and regulation of heart adenosine 3':5'-monophosphate-dependent protein kinase isozymes.

There is broad species variation in the type of cAMP-dependent protein kinase isozyme present in supernatant fractions of heart homogenates as determined by DEAE-cellulose chromatography, Isozyme I, which elutes at less than 0.1 M NaCl, is predominant in mouse and rat hearts; while isozyme II, which elutes at greater than 0.1 M NaCl, is the predominant type in beef and guinea pig. Human and rabbit hearts contain about equal amounts of the two types. The type I heart kinases are more easily dissociated into free regulatory and catalytic subunits by incubation with histone than are the type II kinases, and the separated regulatory and catalytic subunits of isozyme II of rat heart reassociate more rapidly than the subunits of isozyme I under the conditions used. The data from several experiments using rat heart indicate that the basal activity ratio of the protein kinase in crude extracts (approximately 0.15) is due mainly to basal endogenous cAMP and that cAMP elevation accounts entirely for the epinephrine effect on the enzyme. Addition of epinephrine and 1-methyl-3-isobutylxanthine to the perfusate causes a rapid (1 min) increase in cAMP, active supernatant protein kinase, and active phosphorylase in perfused hearts of both rat (mainly isozyme I) and guinea pig (mainly isozyme II). The elevation percentage in cAMP is about the same in the two species, but the increase in active protein kinase is greater in rat heart. If hearts from either animal are perfused continually (10 min) with epinephrine (0.8 muM) and 1-methyl-3-isobutylxanthine (10 muM), the cAMP level, active protein kinase, and active phosphorylase remain elevated. Likewise, all parameters return rapidly to the basal levels when epinephrine and 1-methyl-3-isobutylxanthin are removed. Most of the epinephrine effect on the rat heart supernatant kinase is retained at 0 degrees if cAMP is removed by Sephadex G-25 chromatography, although this procedure completely reverses the epinephrine effect in the guinea pig heart. The epinephrine effect on the rabbit heart kinase (approximately equal amounts of isozymes I and II) is partially reversed by Sephadex G-25. These species differences can be accounted for by differences in association-dissociation behavior of the isozymes in vitro. The data suggest that epinephrine causes activation of both isozymes. The activity present in the particulate fraction comprises nearly half of the total cAMP-dependent protein kinase activity in homogenates of rabbit heart. Triton X-100 extracts of low speed particulate fractions from hearts of each species tested, including rat heart, contain predominantly or entirely the type II isozyme, suggesting differences in intracellular distribution of the isozymes. The binding of the protein kinase to the particulate fraction is apparently due to the properties of the regulatory subunit component. Differences in topographical distribution of the isozymes could provide for differences in either physiological regulation or substrate specificity.     

Localization of the type III isozyme of hexokinase at the nuclear periphery.

The distribution of the type III isozyme of hexokinase (ATP:D-hexose 6-phosphotransferase, EC 2.7.1.1) in rat kidney, liver, spleen, lung, and brain was determined immunohistochemically, using a monoclonal antibody generated against the enzyme purified from rat Novikoff hepatoma.In all tissues, specific cell types exhibited intense staining at the nuclear periphery, as confirmed by analysis using confocal microscopy. Isolated nuclei from kidney or liver were devoid of detectable type III hexokinase, although the enzyme was found in the "soluble" fraction from kidney or liver homogenates; these results suggest that the type III isozyme is associated in a labile manner with the external surface of the nucleus, with this association being disrupted by conventional homogenization and nuclear isolation procedures. The nuclear localization of the type III isozyme contrasts with previously demonstrated association of the type I and II isozymes with mitochondria. The physiological significance of a nuclear localization for the type III isozyme remains unclear. However, it was noted that many of the cells exhibiting prominent nuclear staining for type III hexokinase are endothelial or epithelial cells, suggesting a possible relationship between nuclear type III hexokinase and transport functions which are prominent in such cells.     

A novel brain-specific 25 kDa protein (p25) is phosphorylated by a Ser/Thr-Pro kinase (TPK II) from tau protein kinase fractions

A novel brain-specific 25 kDa protein (p25) was purified from a bovine brain extract. The protein was phosphorylated by Ser/Thr-Pro kinase (TPK II) in tau protein kinase fractions at the Ser residues of Ser-Pro sequences. Using immunoblot analysis, the protein was found only in brain extracts, and was most abundant in the brain regions such as cerebrum and hippocampus, but less abundant in cerebellum, medulla oblongata and olfactory bulb. The protein was detected in rat, bovine and human brain extracts, indicating that this protein specifically exists in mammalian brain tissues.     

Immunochemical evidence that three protein kinase C isozymes increase in abundance during HL-60 differentiation induced by dimethyl sulfoxide and retinoic acid

Activity of the Ca2+/phospholipid-dependent protein kinase C has been shown to increase during differentiation of the human promyelocytic leukemia cell line HL-60 by dimethyl sulfoxide and retinoic acid (Zylber-Katz, E., and Glazer, R. I. (1985) Cancer Res. 45, 5159-5164). Antipeptide antibodies were prepared that specifically recognize the alpha, beta, and gamma isozymes of protein kinase C in rat brain cytosol and HL-60 cell extracts. The three isozymes do not share a common tissue distribution pattern. The gamma enzyme is abundant in brain but a relatively minor component in HL-60 cells; the opposite is true for the alpha enzyme. All three isozymes increase at least 2-fold in abundance in HL-60 cells exposed to 1.2% dimethyl sulfoxide for 48 h. The increase in abundance of the alpha and beta isoforms reaches 7- and 5-fold, respectively, by 96 h without further increase in the abundance of the gamma isozyme. Similarly, all three isozymes increase at least 1.5-fold in abundance after 48 h and 3-fold after 96 h with 1 microM retinoic acid. No further increase in the abundance of any of the isozymes is seen between 96 and 144 h of incubation with retinoic acid. The increase in protein kinase C activity is not limited to the cytosolic forms of the enzyme; a parallel increase in membrane-associated protein kinase C is also observed during differentiation. Approximately 10% of total protein kinase C activity is membrane-associated in both control and differentiating cells. These studies provide the first immunochemical evidence that all three protein kinase C isozymes increase during HL-60 cell differentiation, and they suggest that the increase in the isozyme levels may be coordinately regulated.     

Effect of diabetes on calcium/calmodulin dependent protein kinase-II from rat brain.

Changes in the protein levels and activity of Ca2+/Calmodulin dependent protein kinase II (CaM kinase II) level were studied in cytosolic and particulate fractions from cerebral hemisphere, cerebellum, brain stem, thalamus and hypothalamus regions of rat brain after 4 and 12 weeks of induction of diabetes. Streptozotocin induced diabetes, resulted in pronounced increase of CaM kinase II activity as determined by the kinase activity assay. The total amount of enzyme protein (alpha-subunit specific) also showed increase as revealed by western blotting. Parallel studies were also made in age matched control rats and insulin treated diabetic rats. The increase in CaM kinase II activity was more pronounced in the 12 weeks diabetic group. Insulin treatment of diabetic rats, resulted in recovery of enzyme activity near to control values from majority of the brain regions studied. The expression of alpha-subunit specific CaM kinase II correlates with the enzyme activity in the diabetic rat brain.     

Purification and characterization of creatine kinase isozymes from the nurse shark Ginglymostoma cirratum

Creatine kinase from nurse shark brain and muscle has been purified to apparent homogeneity. In contrast to creatine kinases from most other vertebrate species, the muscle isozyme and the brain isozyme from nurse shark migrate closely in electrophoresis and, unusually, the muscle isozyme is anodal to the brain isozyme. The isoelectric points are 5.3 and 6.2 for the muscle and brain isozymes, respectively. The purified brain preparation also contains a second active protein with pI 6.0. The amino acid content of the muscle isozyme is compared with other isozymes of creatine kinase using the Metzger Difference Index as an estimation of compositional relatedness. All comparisons show a high degree of compositional similarity including arginine kinase from lobster muscle. The muscle isozyme is marginally more resistant to temperature inactivation than the brain isozyme; the muscle protein does not exhibit unusual stability towards high concentrations of urea. Kinetic analysis of the muscle isozyme reveals Michaelis constants of 1.6 mM MgATP, 12 mM creatine, 1.2 mM MgADP and 50 mM creatine phosphate. Dissociation constants for the same substrate from the binary and ternary enzyme-substrate complex do not differ significantly, indicating limited cooperatively in substrate binding. Enzyme activity is inhibited by small planar anions, most severely by nitrate. Shark muscle creatine kinase hybridizes in vitro with rabbit muscle or monkey brain creatine kinase; shark brain isozyme hybridizes with monkey brain or rabbit brain creatine kinase. Shark muscle and shark brain isozymes, under a wide range of conditions, failed to produce a detectable hybrid.