BIOLOGICAL OVERVIEW

Activation of kinases is central to signal transduction during neural activation. Familiarity with two cellular kinases is essential for understanding this process: Protein kinase A (PKA), is activated by elevated levels of cyclic AMP generated by adenylyl cyclase (Rutabaga in Drosophila); a second, Protein kinase C (PKC), is activated by a lipid dependent mechanism. Activation of both kinases takes place in the two main routes of signal transduction during neural activation, and both kinases have been implicated in the cellular processes involved in learning.

Protein kinase C operates at one of the most unusual crossroads of cellular activation, one dependent on hydrolysis of membrane lipids. The membrane lipid component phosphatidylinosityl 4,5-bisphosphate (PIP2), generated by the addition of phosphate residues to the sugar (inositol) moiety of the membrane lipid phosphatidylinositol, is hydrolyzed by the enzyme phospholipase C-beta to give the lipid diacylglycerol (DAG) and the sugar phosphate known as inositol 1,4,5-trisphospate (IP3). While diacylglycerol activates protein kinase C, IP3 releases Ca2+ from the endoplasmic reticulum. Each of these enzymatic processes are sensitive to the level of Ca2+ in the cell. Initiating the cellular activation cascade is the activation of phospholipase C-beta by heterotrimeric G proteins linked to G protein coupled receptors. The activation of PKC by the membrane lipid DAG is mimicked by phorbol esters whose structure resembles the structure of DAG.

Drosophila has three PKC enzymes, one eye specific [inactivation no afterpotential C or dPKC53E(ey)] and two others whose expressions are confined to the head. The isoform called dPKC53E(brain) is closely linked to dPKC53E(eye). Among the Drosophila genes, dPKC53E(brain) (Rosenthal, 1987) is most similar to the classical genes (PKCalpha, beta and gamma), while dPKC98F has greater amino acid identity with, and a domain arrangement characteristic of, PKCdelta (also termed nPKC) (Schaeffer, 1989).

The concept of memory as a unified process involving a linear series of biochemical and cellular activation steps and a single anatomical localization is no longer tenable. There are multiple kinds of memory, residing in multiple locations and generated by multiple biochemical and cellular processes (Tully, 1994). Studies in insects, in particular the honeybee (Hammer, 1995) and Drosophila, have been one of the main stimuli for reaching this conclusion. For a discussion of the multiple memory systems in mammals, see CaM Kinase II: Evolutionary homologs: CaM kinase II and learning and behavior. Studies in humans reveal that selective effects on a single component of memory have been correlated with damage to particular brain regions, such that short- and long-term memory can function independently of one another. In some cases, an affected individual cannot repeat a string of three items immediately after presentation but may nonetheless perform normally in tests for 24 hour memory. In this situation, short-term memory is not a prerequisite for long-term memory (Squire, 1992 and references).

Mutants for genes involved in memory in Drosophila (linotte, latheo, dunce, rutabaga, amnesiac, Protein kinase A, dCREB2, and radish) produce specific memory defects that are classified by mutant phenotype. linotte and latheo produce a defect in the acquisition phase (learning) of memory formation while subsequent memory formation is normal. Dunce and Rutabaga, both involved in the cyclic AMP cascade, are involved in short term memory that decays within the course of hours. Mutants for amnesiac and PKA are defective in medium term memory, which declines over the course of a day or two. dCREB2 and radish are involved in memory of longer duration (long term memory and anesthesia resistent memory). These are memory components that last for at least a day (Greenspan, 1995 and references).

Since mutants in PKC are not available in Drosophila, a biochemical genetic trick is performed to study the role of PKC in neural growth and fly memory. The inactive form of PKC results from the presence of an autoinhibitory domain that occupies the catalytic site. Flies were made homozygous for a transgene that expresses a protein fragment consisting of a specific pseudosubstrate inhibitor of PKC. The transgene pseudosubstrate inhibitor is derived from PKC from the fly eye. Controlling the PKC inhibitor by means of heat shock circumvents any possibility of developmental effects such as those that may occur in knockout mutants. This strategy also has the advantage of avoiding the non-specific effects of drugs. Following heat shock, flies express the pseudosubstrate inhibitor at high levels. This system was used to study the role of PKC in the growth of cultured, differentiating neuroblasts. Cell cultures expressing a low concentration of the pseudosubstrate inhibitor show an alteration in neurite growth. Such cultures exhibit a significant decrease in the number of neural processes as compared to control cultures.

Drosophila larvae form two types of neuronal processes. Type I processes have large buttons with few branches, and type II processes are thinner and beaded, with more branches. Both type I and type II processes are diminished in vitro in cells from flies expressing the pseudosubstrate inhibitor. In in vitro neuronal cell clusters the proportional amounts of neuroblasts, ganglion mother cells (GMCs), and mature neurons are little affected. It is concluded that endogenously inhibited protein kinase C in transgenic Drosophila embryonic neuroblasts down regulates the outgrowth of type I and II processes of cultured mature neurons (Broughton, 1996a). The effect of PKC on development of neural processes in cultured cells is likely to be due to modification of cytoskeletal targets by PKC. Another enzyme, CaM kinase II, evokes an inhibition of neural arborization and outgrowth in Drosophila cultured neurons. CaM kinase II is transported along processes after neuronal contact has been made, at a time when functional synapses might be forming (Broughton, 1996b)

A cytoskeletal target for PKC has been identified in mammals. Dynamin is a microtubule-binding protein with a microtubule activating GTPase activity. The gene encoding Dynamin is mutated in shibire, a Drosophila mutant defective in endocytosis in nerve terminals and other cells. Mammalian dynamin is phosphorylated by Protein kinase C and is dephosphorylated upon neural excitation. The GTPase activity of dynamin is enhanced 12-fold when phosphorylated by PKC. The dynamins are therefore a group of nerve terminal phosphorproteins whose GTPase is regulated by phosphorylation in parallel with synaptic vesicle recycling. The regulation of dynamin GTPase could serve as a trigger for the rapid endocytosis of synaptic vesicles after exocytosis (Robinson, 1993).

As PKC modifies neural process formation in Drosophila, it might also act to modify components of learning and memory. A test for such a role has been made in flies expressing the PKC pseudosubstrate. The effects of heat-inducible PKC inhibitor on courtship conditioning (an associative conditioning paradigm) were examined. Normally, a male's courtship activity is suppressed after exposure to a mated female. (For a more detailed description of this study see Calcium/calmodulin dependent protein kinase II, Biological overview). Flies mutant for PKC show an alteration in performance, that is, they fail to show immediate suppression of courtship, a trait exhibited by normal flies. Mutant flies nevertheless develop a normal memory of training episodes. In other words, they remember and forget no differently from controls (Kane, 1997).

The behavior of transgenic flies shows that disrupting PKC function dissociates immediate performance from ultimate retention and decay rate of conditioned memory. In so doing, it demonstrates that immediate performance, at least for this form of learning, occurs in parallel with memory formation, not as a precursor. These results suggest a mechanism in which a common stimulus, the mated female, triggers separate events: a PKC-dependent process that feeds back immediately on the animal's behavior and one or more separate processes that lead to memory formation. The existence of such a mechanism supports the idea that distinct biochemical mechanisms are needed, each with its own characteristic time course, to fill all of the gaps from the organism's first experience of an effective conditioning stimulus until the consolidation of its most durable memory. No single, unitary mechanism appears to be able to be able to span all of these intervals (Kane, 1997).

A similar result has been obtained in the nudibranch mollusk Hermissenda (Crow, 1993). Blockade of PKC in Hermissenda selectively affects the in vitro expression of short-term conductance changes associated with learning, but has no effect on those that are long-term. Similarly, in a physiological preparation of Aplysia, long-term synaptic facilitation can be set up in the absence of short-term facilitation after pharmacological blockade of one type of serotonin (Emptage, 1993) thought to activate PKC (Sugita, 1992). Moreover, the time course for the development of long-term facilitation, which is diagnostic of the normal duration of short-term facilitation, resembles the time course seen in the Drosophila study for the PKC-mediated effect on performance (Kane, 1997).

Given the pleiotropy of PKC (the fact that it affects many pathways) it is not obvious how to account for the apparent specificity of the behavioral defect in flies. One would expect that interference with PKC, implicated in neural process formation, would interfere with long term memory, thought to involve alterations in neural connectivity. However PKC is involved in early learning processes and not long term memory. The failure of PKC-inhibited flies to manifest expression of immediate conditioning can be interpreted as a transient failure of retrieval in the earliest stages of memory (Kane, 1997). These flies are truely LD, not "learning disabled" but those that may be said to "learn differently."

GENE STRUCTURE

Adult flies contain three transcripts of dPKC53E(brain) (Rosenthal, 1987).

The dPKC98F gene encodes a major 5.5 kb transcript that is expressed throughout development. The level of this transcript is greatly reduced in embryos and correlates with the increased expression of two additional transcripts of 4.3 and 4.5 kb (Schaeffer, 1989).

Exons - 14 for dPKC53E(brain) (Rosenthal, 1987).

PROTEIN STRUCTURE

Amino Acids - 639 for dPKC53E(brain), 634 for dPKC98F and 631 for dPKC53E(eye)

Structural Domains

The three Drosophila PKC proteins share structural features with their mammalian counterparts. These include: a consensus ATP binding site located near the middle of the molecule, a pair of cysteine-containing "zinc finger" motifs near the amino terminus of PKC, and an eight amino acid pseudosubstrate domain located toward the N-terminus. This sequence, which resembles the PKC phosphorylation site on target proteins, has been implicated in autoregulation of the enzyme. The two 53E sequences are more homologous to one another than to 98F. dPKC53E(brain) (Rosenthal, 1987) among Drosophila genes is most similar to the classical genes (PKCalpha, beta and gamma). The 98F sequence contains a 634 amino acid open reading frame with homology to mammalian PKCs. However, the 98F sequence diverges from "classical" mammalian PKC genes in a region near the amino terminus. The more recently identified mammalian PKC-related genes (PKCdelta, epsilon and eta) also diverge in this domain. The 98F gene has greater amino acid identity with, and a domain arrangement characteristic of PKCdelta (also termed nPKC). The third PKC protein, dPKC53E(eye) is a photoreceptor cell-specific PKC, and lies on the second chromosome within 50 kb of dPKC53E(brain) (Schaeffer, 1989).

The protein kinase C (PKC) binding site used by PKC activators such as phorbol esters and diacylglycerols (DAGs) has been characterized by means of molecular modeling and site-directed mutagenesis studies. Based upon a NMR-determined solution structure of the second cysteine rich domain of PKC alpha, molecular modeling has been used to study the structures of the complexes formed between the PKC receptor and a number of PKC ligands, phorbol esters, and DAGs. Site-directed mutagenesis studies have identified a number of residues important to the binding of phorbol esters to PKC. Analysis of the molecular modeling and mutagenesis results allows the development of a binding model for PKC ligands for which the precise binding nature is defined. The calculated hydrogen bond energies between the protein and various ligands in this binding model are consistent with their measured binding affinities. The binding site for phorbol esters and DAGs is located in a highly conserved, hydrophobic loop region formed by residues 6-12 and 20-27. For the binding elements in phorbol esters, the oxygen atom at C20 contributes most to the overall binding energy; the oxygen at C3 also plays a significant role. The oxygen atom at C12 is not directly involved in the interaction between phorbol esters and PKC. These results also suggest that the oxygens at C9 and C13 are involved in PKC binding, while the oxygen at C4 is of minimal significance. These results are consistent with known structure-activity relationships in the phorbol ester family of compounds. Comparisons with the X-ray structure show that although the X-ray data support the results for oxygens at C3, C12, and C20 of phorbol esters, they suggest different roles for oxygens at C4, C9, and C13 (Wang, 1996).

Three-dimensional models of the five functional modules in human protein kinase C alpha (PKC alpha) have been generated on the basis of known related structures. The catalytic region at the C-terminus of the sequence and the N-terminal auto-inhibitory pseudo-substrate have been modeled using the crystal structure complex of cAMP-dependent protein kinase (cAPK) and PKI peptide. While the N-terminal helix of the catalytic region of PKC alpha is predicted to be in a different location compared with cAPK, the C-terminal extension is modeled like that in the cAPK. The predicted permissive phosphorylation site of PKC alpha, Thr 497, is found to be entirely consistent with the mutagenesis studies. Basic Lys and Arg residues in the pseudo-substrate make several specific interactions with acidic residues in the catalytic region and may interact with the permissive phosphorylation site. Models of the two zinc-binding modules of PKC alpha are based on nuclear magnetic resonance and crystal structures of such modules in other PKC isoforms while the calcium phospholipid binding module (C2) is based on the crystal structure of a repeating unit in synaptotagmin I. Phorbol ester binding regions in zinc-binding modules and the calcium binding region in the C2 domain are similar to those in the basis structures. A hypothetical model of the relative positions of all five modules has the putative lipid binding ends of the C2 and the two zinc-binding domains pointing in the same direction (Srinivasan, 1996).

The structure of the second activator-binding domain of PKC delta has been determined in complex with phorbol 13-acetate, which binds in a groove between two pulled-apart beta strands at the tip of the domain. The C3, C4, and C20 phorbol oxygens form hydrogen bonds with main-chain groups whose orientation is controlled by a set of highly conserved residues. Phorbol binding caps the groove and forms a contiguous hydrophobic surface covering one-third of the domain, explaining how the activator promotes insertion of PKC into membranes (Zhang, 1995).

Binding of phorbol esters to protein kinase C (PKC) has been regarded as dependent on phospholipids, with phosphatidylserine being the most effective for reconstituting binding. By using a purified single cysteine-rich region from PKC delta expressed in Escherichia coli it can be shown that specific binding of [3H]phorbol 12,13-dibutyrate to the receptor still takes place in the absence of the phospholipid cofactor. However, phorbol ester binds to the cysteine-rich region with 80-fold lower affinity in the absence of phosphatidylserine, as in its presence. Similar results are observed with the intact recombinant PKC delta isolated from insect cells. When different phorbol derivatives are examined, distinct structure-activity relations for the cysteine-rich region are found in the presence and absence of phospholipid (Kazanietz, 1995a).

Peptides B and C (two subdomains of the regulatory domain of PKC-gamma) bind phorbol ester with good affinity in the presence of phosphatidylserine. Like PKC itself, these peptides also recognize other PKC activators, including dioctanoylglycerol and teleocidin B-4, and exhibit an ability to differentiate phorbol ester from its C-4 epimer. Peptide B becomes conformationally ordered only in the presence of phospholipid, suggesting that the regulatory domain of PKC itself might be organized for activation only when associated with the lipid bilayer, where its activator (diacylglycerol) is encountered (Wender, 1995).

Based on marked differences in the enzymatic properties of diacylglycerols compared with phorbol ester-activated protein kinase C (PKC), it was proposed that activation induced by these compounds may not be equivalent. In the present study, direct evidence is provided showing that phorbol esters and diacylglycerols bind simultaneously to PKC alpha. Using a novel binding assay employing the fluorescent phorbol ester, sapintoxin-D (SAPD), evidence for two sites of high and low affinity was obtained. Thus, both binding and activation dose-response curves for SAPD are double sigmoidal, which is also observed for dose-dependent activation by the commonly used phorbol ester, 4beta-12-O-tetradecanoylphorbol-13-acetate (TPA). TPA removes high affinity SAPD binding and also competes for the low affinity site. By contrast with TPA, low affinity binding of SAPD is inhibited by sn-1,2-dioleoylglycerol (DAG), while binding to the high affinity site is markedly enhanced. Again contrasting with both TPA and DAG, the potent PKC activator, bryostatin-I (B-I), inhibits SAPD binding to its high affinity site, while low affinity binding is unaffected. Based on these findings, a model for PKC activation is proposed in which binding of one activator to the low affinity site allosterically promotes binding of a second activator to the high affinity site, resulting in an enhanced level of activity. Overall, the results provide direct evidence that PKCalpha contains two distinct binding sites, with affinities that differ for each activator in the order: DAG > phorbol ester > B-I and B-I > phorbol ester > DAG, respectively (Slater, 1996).

The C2 domain is a Ca(2+)-binding motif of approximately 130 residues in length originally identified in the Ca(2+)-dependent isoforms of protein kinase C. Single and multiple copies of C2 domains have been identified in a growing number of eukaryotic signaling proteins that interact with cellular membranes and mediate a broad array of critical intracellular processes, including membrane trafficking, the generation of lipid-second messengers, activation of GTPases, and the control of protein phosphorylation. As a group, C2 domains display the remarkable property of binding a variety of different ligands and substrates, including Ca2+, phospholipids, inositol polyphosphates, and intracellular proteins. Expanding this functional diversity is the fact that not all proteins containing C2 domains are regulated by Ca2+, suggesting that some C2 domains may play a purely structural role or may have lost the ability to bind Ca2+. This review summarizes the information currently available regarding the structure and function of the C2 domain and provides a novel sequence alignment of 65 C2 domain primary structures. The alignment predicts that C2 domains form two distinct topological folds, illustrated by the recent crystal structures of C2 domains from synaptotagmin 1 and phosphoinositide-specific phospholipase C-delta 1. The alignment highlights residues that may be critical to the C2 domain fold or required for Ca2+ binding and regulation (Nalefski, 1996).

Certain isoforms of protein kinase C (PKC) require both Ca2+ and phospholipid for optimum activity. However, little is known about the nature of the interaction between PKC and Ca2+. The present study demonstrates that the isolated regulatory domain of PKC beta 1, when synthesized as a fusion protein in Escherichia coli, binds 45Ca2+ with high affinity, but only in the presence of phosphatidylserine or 12-O-tetradecanoyl-phorbol-13-acetate. This binding is highly selective for Ca2+ since it is preferentially inhibited by excess non-radioactive Ca2+ when compared with the cations Mg2+, Mn2+, Na+, or K+. It appears, therefore, that the binding of Ca2+ to PKC requires a complex tertiary structure in the regulatory domain (Luo, 1993a).

Activation of certain isoforms of protein kinase C (cPKCs) requires Ca2+ and is associated with a conserved C2 domain that is not present in Ca(2+)-independent isoforms (nPKCs). The site(s) of Ca2+ binding and the role of the C2 domain have not been previously identified. Ca2+ is bound mainly to the C1 domain of PKC beta 1, but the C2 domain confers specificity for Ca2+ binding when compared with Mg2+ and Mn2+. It is proposed that in cPKCs there is selective binding of Ca2+ to a pocket formed by the C1 and C2 domains. This induces a change in conformation that activates the enzyme. In nPKCs, the cation binding pocket is less specific for Ca2+ because it lacks the C2 domain. Therefore, divalent cations like Mg2+ can bind to it, thereby abrogating the requirement of Ca2+ for enzyme activation (Luo, 1993b).

date revised: 28 Apr 97 

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.