Research ArticleNeuroscience

Neuronal and Intestinal Protein Kinase D Isoforms Mediate Na+ (Salt Taste)–Induced Learning

See allHide authors and affiliations

Science Signaling  11 Aug 2009:
Vol. 2, Issue 83, pp. ra42
DOI: 10.1126/scisignal.2000224

Abstract

Ubiquitously expressed protein kinase D (PKD) isoforms are poised to disseminate signals carried by diacylglycerol (DAG). However, the in vivo regulation and functions of PKDs are poorly understood. We show that the Caenorhabditis elegans gene, dkf-2, encodes not just DKF-2A, but also a second previously unknown isoform, DKF-2B. Whereas DKF-2A is present mainly in intestine, we show that DKF-2B is found in neurons. Characterization of dkf-2 null mutants and transgenic animals expressing DKF-2B, DKF-2A, or both isoforms revealed that PKDs couple DAG signals to regulation of sodium ion (Na+)–induced learning. EGL-8 (a phospholipase Cβ4 homolog) and TPA-1 (a protein kinase Cδ homolog) are upstream regulators of DKF-2 isoforms in vivo. Thus, pathways containing EGL-8–TPA-1–DKF-2 enable learning and behavioral plasticity by receiving, transmitting, and cooperatively integrating environmental signals targeted to both neurons and intestine.

Introduction

Many hormones, growth factors, and neurotransmitters elicit diacylglycerol (DAG) synthesis by activating phospholipase C–β (PLC-β), PLC-γ, or PLC-ɛ (1). Upon activation, the PLCs cleave plasma membrane phosphatidylinositol 4,5-bisphosphate (PIP2), to produce DAG and inositol trisphosphate (IP3). IP3, which is hydrophilic, diffuses and binds IP3 receptor channels, which mediate Ca2+ release from the endoplasmic reticulum (2). Membrane-intercalated DAG recruits and activates eight mammalian protein kinase C (PKC) isoforms (3–5). The classical PKCs (cPKCs) α, βI, βII, and γ are activated through the combined action of DAG and Ca2+, whereas DAG alone activates the novel PKCs (nPKCs) δ, ɛ, θ, and η. PKCs regulate ion channel activity, secretion, gene transcription, and other physiological functions by phosphorylating specific Ser-Thr residues in effector proteins. A group of Ser protein kinases collectively called protein kinase D (PKD) also disseminate signals carried by DAG (6, 7). Three mammalian PKDs (PKDs 1 to 3), which are encoded by distinct genes, are expressed ubiquitously. Like PKCs, PKDs have tandem DAG-binding modules (which also bind phorbol esters) and a C-terminal kinase domain (fig. S1).

Hormone-induced increases in DAG recruit PKDs to the plasma membrane (6, 7). Co-recruited PKCs switch on PKD catalytic activity by phosphorylating two serines in the PKD activation loop (A loop). When DAG abundance declines, PKDs translocate from the plasma membrane to the nucleus, cytoplasm, and outer surfaces of various organelles, where increased PKD activity persists for minutes to hours. PKDs and PKCs phosphorylate different substrates (8). Thus, PKDs may expand the temporal and spatial ranges of PKC-mediated signaling pathways.

Hormones that bind Gq/11-coupled receptors activate PKDs in cell lines, suggesting that PKDs regulate many facets of mammalian physiology (7). In various cell culture systems, PKDs control fission and trafficking of Golgi vesicles; modulate antigen-induced signaling in T and B cells; mediate prosurvival responses to oxidative stress; regulate c-Jun N-terminal kinase (JNK)–c-Jun–dependent proliferation; govern adhesion; and elicit nuclear export of class II histone deacetylases (HDACs) (4, 7, 9, 10).

However, the in vivo regulation and physiological functions of PKDs are poorly understood. It is not known whether roles attributed to PKDs from studies on cultured cells represent major or minor PKD functions in vivo. Likewise, we lack knowledge of whether PKDs mediate such functions as survival signaling or Golgi vesicle trafficking ubiquitously or in only a few tissues. Indeed, anticipated functional defects were not observed in PKD-deficient, avian B lymphoma cells (11).

PKD1 mediates induction of pathological cardiac muscle hypertrophy caused by pressure overload or chronic adrenergic signaling (12). Ectopic overexpression of PKD1 promotes endurance in skeletal muscle (13). A current model suggests that these effects depend on PKD-mediated phosphorylation of HDAC5, a repressor of the MEF2 transcription factor (10, 14). However, the normal physiological functions controlled by transcriptional and nontranscriptional PKD effectors remain unknown in muscle (and most other tissues). No studies have identified the physiological processes regulated by PKD2 and PKD3 (and PKD1 in nonmuscle cells) in intact animals.

The magnitude and duration of a PKD-regulated physiological output may reflect integration of signals delivered by combinations of various heterotrimeric guanine nucleotide–binding protein (G protein), PLC, and PKC isoforms. Additional signaling complexity may be introduced by PKD isoforms that function independently or cooperatively. Thus, a major challenge is to determine which PKDs and upstream regulators cooperate to control central biological processes.

The nematode Caenorhabditis elegans provides a powerful system for in vivo analysis of cell signaling pathways (15–17). The development, cell biology, physiology, and many behaviors of the nematode have been characterized in exceptional detail. C. elegans can be studied and analyzed by biochemistry, molecular and cell biology, molecular genetics, and classical genetics. Techniques for gene disruption, epistasis analysis, targeted, cell-specific messenger RNA (mRNA) and protein expression, and quantitative analysis of behavior are well established and enable incisive studies on intact animals. Moreover, C. elegans uses signaling molecules, mechanisms, and pathways that are conserved from nematodes to man (15, 16, 18, 19). Overall, the C. elegans system allows the integrated exploration of functions of specific signaling proteins at the molecular, cellular, and behavioral levels.

We used C. elegans to identify and characterize PKDs and investigate their functions and regulation in vivo. The dkf-2 (for D kinase family) gene encodes a prototypical, 120-kD PKD (20). A 1.5-kb deletion in the gene [dkf-2(pr3) allele] eliminates DKF-2 protein (Fig. 1A), thereby generating a null phenotype (20). DKF-2 depletion does not affect development, feeding, or reproduction. Thus, dkf-2(pr3) null animals and null mutants reconstituted with wild-type (WT) or mutant transgenes can be challenged with stimuli to identify DKF-2 functions and regulators. Using this strategy, we showed that DKF-2 (now renamed DKF-2A) couples DAG to regulation of innate immunity (21).

Fig. 1

The dkf-2 gene encodes DKF-2A and DKF-2B. (A) Western analysis of proteins from WT and dkf-2(pr3) animals shows that partial deletion of the dkf-2 gene extinguishes expression of DKF-2A and DKF-2B. (B) Exons shown as solid rectangles encode DKF-2A and DKF-2B mRNAs. (Introns are not drawn to scale.) DKF-2A protein is encoded by 18 exons (20). Exons shown in blue are unique in DKF-2A mRNA; exons colored red are shared in DKF-2A and DKF-2B mRNAs. Promoter-enhancer DNA that controls synthesis of DKF-2A mRNA was recently characterized (20). Genomic DNA corresponding to nucleotides 2683 to 4897 in intron 6 of the dkf-2A gene (see WormBase sequence T25E12.4) governs transcription of DKF-2B mRNA. DKF-2B mRNA contains a previously unknown 5′ exon (exon 1′, green) and 12 shared exons, numbered 7 to 18. Dual-colored lines show the sizes and exon contributions to PCR-amplified cDNAs obtained using a 5′ SL1 primer and nested 3′ primers derived from exons 8 and 9. (C) The C. elegans genomic DNA sequence that encodes an SL-1 addition site, 5′UTR, exon1′, and intron1 in DKF-2B transcripts is shown. Splice site positions (arrowheads) show how exon1′ is fused to exon 7 in DKF-2B mRNA. The size of the dkf-2A gene is ~43 kb; nucleotide sequence encoding DKF-2B mRNA begins in intron 6 (i.e., previously unknown exon1′) and continues to the 3′ end of the gene (~24 kb). The DNA sequence of the dkf-2 gene is available in WormBase (sequence T25E12.4, www.wormbase.org).

C. elegans senses many environmental stimuli, including ions derived from water-soluble salts. Gradients of sodium salts elicit several types of behavior (22–24). C. elegans is attracted to low concentrations (<200 mM) of NaCl or sodium acetate (salt chemotaxis) but is repelled by high salt concentrations (>200 mM). C. elegans also exhibits an experience-dependent change in behavior (behavioral plasticity). Animals preincubated with 100 mM sodium acetate for 60 min in the absence of food learn to avoid salt and are repelled by a subsequent challenge with 25 mM sodium acetate. In this process, called salt-induced learning, integration of previously obtained sensory information switches the response to an invariant stimulus (low sodium salt) from chemotaxis to avoidance.

Here, we report that the C. elegans dkf-2 gene encodes two PKD isoforms: DKF-2B, which we describe here, accumulates in neurons that constitute chemosensory circuitry; previously characterized DKF-2A is found in intestinal cells. Using molecular genetics in concert with the learning paradigm described above, we discovered that both DKF-2B and DKF-2A are essential for Na+ (salt) taste–induced learning and behavioral plasticity. Na+ avoidance, a learned behavior, is governed by crosstalk between PKD-mediated signaling pathways in nervous tissue and the gut. EGL-8, a PLC-β4 homolog, and TPA-1, a DAG-activated PKCδ homolog, control in vivo activation (and functions) of neuronal DKF-2B and intestinal DKF-2A.

Results

The dkf-2 gene encodes two PKDs, DKF-2A and DKF-2B

We characterized a 120-kD C. elegans PKD isoform that we named DKF-2 (25). DKF-2 and mammalian PKDs have conserved regulatory (C1a, C1b) and kinase domains (fig. S1) that are similar in size and ~75% identical in amino acid sequence (20). A unique N-terminal region (amino acids 1 to 311) distinguishes DKF-2 from other PKDs. Immunoglobulin Gs (IgGs) directed against amino acids 1 to 203 of DKF-2 recognize a 120-kD polypeptide in Western blots of C. elegans proteins and proteins from transfected cells expressing DKF-2 (20). The 120-kD antigen is not detected among proteins derived from C. elegans homozygous for a deletion in the dkf-2 gene [dkf-2(pr3) allele] or nontransfected cells. However, antibodies directed against a C-terminal segment of the PKD (amino acids 1049 to 1070) bound DKF-2 (renamed DKF-2A) and a second protein [Mr (relative molecular mass) ~100,000] in extracts of WT C. elegans (Fig. 1A). A null mutation in the dkf-2 gene abrogated expression of both proteins (Fig. 1A).

To determine the identity of the smaller immunoreactive protein, we performed reverse transcription polymerase chain reaction (RT-PCR) analysis, using mRNA extracted from an asynchronous population of C. elegans as a template. Because most C. elegans mRNAs contain a shared 5′ splice leader sequence (SL-1), we generated complementary DNAs (cDNAs) using an SL-1 primer in conjunction with nested 3′ primers that inversely complement sequences in exons 8 and 9 of the dkf-2 gene [(20) and WormBase archive, sequence T25E12.4]. Sequencing the amplified cDNAs revealed two products. One product matched the 5′ untranslated region (5′UTR), exons 1 to 7, and a portion of exon 8 in DKF-2A cDNA (Fig. 1B). The second contained a previously unknown 5′ nucleotide sequence (Fig. 1, B and C) that preceded sequences identical with exon 7 and part of exon 8 in DKF-2A cDNA. The previously unknown cDNA lacked DKF-2A exons 1 to 6 (Fig. 1B). Amplification and sequencing of the full-length cDNAs, with SL-1 and extreme 3′ primers, showed that dkf-2 encodes two PKD isoforms. The previously unknown transcript encodes a protein composed of 872 amino acids (Mr = 98,000), which we named DKF-2B. Amino acids 1 to 113 in DKF-2B are encoded by an alternative first exon (transcribed with a promoter that lies within intron 6 of the dkf-2A gene) (Fig. 1C). The remaining 12 exons appear in both DKF-2A and -2B mRNAs (Fig. 1B).

The N-terminal sequence (1 to 113) of DKF-2B is unique; the contiguous 759 central and C-terminal amino acids are identical with residues 312 to 1070 of DKF-2A (20). C1a, C1b, PH, and kinase domains are identical in both DKF-2 isoforms. However, divergent N-terminal regions alter the overall electrostatic properties of the kinases: DKF-2A and DKF-2B have predicted pI values of 6.6 and 8.9, respectively. The 5′UTR, open reading frame of alternative exon 1 (exon1′), alternative intron 1, and splice sites linking exon1′ with exon 7 of the dkf-2 gene are provided in Fig. 1C. The dkf-2B promoter-enhancer region precedes exon 1′ and corresponds to nucleotides 2683 to 4897 in intron 6 [5949 base pairs (bp)] of the dkf-2A gene (WormBase archive, sequence T25E12.4). Divergent promoter-enhancer elements direct expression of DKF-2A and DKF-2B in distinct groups of cells in vivo [see (20) and below].

Phorbol ester and DAG induce translocation and activation of DKF-2B

Human embryonic kidney (HEK) 293 cells transfected with a DKF-2B transgene were disrupted and separated into cytosolic and membrane fractions. In the absence of stimuli, ~90% of DKF-2B protein was recovered in the cytosolic fraction (Fig. 2A, upper panel, lanes 1 and 2). Incubation of cells with 1 μM phorbol 12-myristate 13-acetate (PMA) elicited translocation of DKF-2B from cytosolic to membrane fractions (Fig. 2A, upper panel, lanes 3 and 4). Immunofluorescence microscopy confirmed that, under basal conditions, DKF-2B was dispersed in the cytoplasm (Fig. 2B1) and that PMA induced DKF-2B recruitment to the cell periphery and depletion from cytoplasm (Fig. 2B2).

Fig. 2

PMA and bombesin induce translocation of DKF-2B to membranes. (A) HEK293 cells expressing a DKF-2B transgene were treated with GF109203X or vehicle for 1 hour. Subsequently, cells were incubated with PMA or vehicle for 10 min, as indicated. Cytosolic (C) and membrane (M) proteins were isolated and analyzed in a Western blot. The blot shows that PMA elicited PKC-independent translocation of DKF-2B to membranes. The middle and bottom panels show loading controls for membranes (IGF-1 receptor) and cytosol (14-3-3 protein), respectively. (B) HEK293 cells expressing DKF-2B were analyzed by fluorescence microscopy. Cells were incubated with GF109203X (B3) or vehicle (B1 and B2) for 1 hour. Subsequently, cells shown in (B2) and (B3) were treated with PMA for 10 min. (C) HEK293 cells expressing bombesin BB2R and DKF-2B were incubated with bombesin for 5 min (C2 and C3) after preincubation with GF109203X (C3) or vehicle (C1 and C2) for 1 hour. Cells were stained and the intracellular location of DKF-2B was determined. (B) and (C) show that PMA and bombesin induced PKC-independent translocation of DKF-2B to the cell periphery. Experiments were replicated three times and yielded similar results.

PMA elicited phosphorylation of the DKF-2B activation loop (A loop) and a 1800% increase in DKF-2B catalytic activity (Fig. 3A). GF109203X, an inhibitor of all DAG-activated PKCs, suppressed PMA-induced phosphorylation and activation of DKF-2B (Fig. 3A, lanes 2 and 4), but did not affect PMA-mediated translocation of DKF-2B to the membrane (Fig. 2, A, upper panel, lanes 5 to 8, and B3). DKF-2B protein isolated from unstimulated cells was resolved into two polypeptides (apparent Mr values of 98,000 and 100,000) by denaturing electrophoresis (Fig. 3, A and B). The doublet is presumably generated by differential phosphorylation or another posttranslational modification that does not significantly stimulate kinase activity (Fig. 3A, compare phosphotransferase activities for lanes 1 and 2). Apparent Mr values of DKF-2B polypeptides increased by ~2000 when cells were incubated with PMA. Small upshifts in Mr are often caused by protein phosphorylation. GF109203X partially inhibited the PMA-induced increase in DKF-2B molecular weight and suppressed A-loop phosphorylation and kinase activation (Fig. 3A, lanes 2 and 4). Thus, endogenous PKC phosphorylated DKF-2B and stimulated its activity.

Fig. 3

PMA and bombesin activate DKF-2B. (A) Transfected HEK293 cells expressing DKF-2B were treated with PMA or vehicle for 10 min before lysis. Duplicate samples of cells were incubated with the pan-PKC inhibitor GF109203X for 1 hour before PMA addition. DKF-2B was immunoprecipitated and assayed for catalytic activity. An immunoblot shows that similar amounts of DKF-2B were used in each assay (middle panel). PMA-induced increases in apparent Mr of DKF-2B (middle panel) are likely due to phosphorylation. A duplicate Western blot (bottom panel) was probed with IgGs directed against phosphoserines in the DKF-2B A loop. (B) Transfected HEK293 cells expressing both DKF-2B and BB2R were incubated with bombesin or vehicle for 10 min before lysis. Duplicate samples of cells were incubated with GF109203X for 1 hour before bombesin addition. DKF-2B was immunoprecipitated and assayed for kinase activity and A-loop phosphorylation. (A) and (B) show that PMA and the BB2R–Gq–PLC-β–DAG pathway elicit phosphorylation of the DKF-2B A loop and substantial increases in DKF-2B catalytic activity. Inhibition of PKCs by GF109203X blocks DKF-2B phosphorylation and activation. Experiments were repeated three times and similar results were obtained.

Binding of the peptide bombesin with the BB2 bombesin receptor (BB2R) elicits the Gq and PLC-β–dependent activation of C. elegans DKF-1 and DKF-2A (20, 26). Treatment of HEK293 cells harboring BB2R and DKF-2B transgenes with 100 nM bombesin triggered A-loop phosphorylation and increased DKF-2B catalytic activity by 500% (Fig. 3B, lanes 1 and 2). Moreover, bombesin promoted enrichment of DKF-2B at the cell periphery and its concomitant depletion from cytoplasm (Fig. 2C, panels 1 and 2). GF109203X suppressed bombesin-induced DKF-2B phosphorylation and activation (Fig. 3B, lanes 2 and 4), but not translocation of DKF-2B (Fig. 2C3).

The data suggest that BB2R-stimulated accumulation of DAG enables the C1 domain–dependent recruitment of DKF-2B and PKC to plasma membrane; DAG activates PKC, which phosphorylates the DKF-2B A loop, triggering a conformational change that markedly increases PKD catalytic activity.

DKF-2B is differentially expressed in neurons

We microinjected a dkf-2B::DKF-2B–green fluorescent protein (GFP) construct into C. elegans (25). Animals transmitting the transgene in the germ line were selected by monitoring GFP by fluorescence microscopy. GFP fluorescence was detected in ~20 neurons (Fig. 4). In the head, DKF-2B–GFP accumulated in amphid sensory neurons (ASE, AFD, ASG, AWC, and AWA), integrative interneurons (AIM, RIC, AIY, AVF, AVJ, AVK, ADA, and RMG), neurons involved in social feeding and O2 sensing (AUA, URX, PQR, AQR, and BAG), and the neurosecretory motor neuron NSM (Fig. 4, C and E). DKF-2B–GFP also accumulated in PVQ, PVR, and PVW interneurons in the tail (Fig. 4, D and F). The presence of DKF-2B in <15% of C. elegans neurons suggests that it is involved in regulating cell- and circuit-specific functions in the nervous system. DKF-2B–GFP was observed in late embryos, L1 to L4 larvae, and adult animals.

Fig. 4

DKF-2B is expressed in ~20 head and tail neurons. Animals expressing a dkf-2B::DKF-2B–GFP transgene were analyzed by fluorescence microscopy to determine DKF-2B localization. Full-length DKF-2B–GFP is found in a limited number of head and tail neurons in L2 larvae (A1) and adult (B1) animals. (A2) and (B2) are bright-field images of (A1) and (B1), respectively. (C) Head neurons expressing DKF-2B–GFP. (D) Tail neurons expressing DKF-2B–GFP. Names of neurons are indicated. Thirteen neurons that express DKF-2B–GFP are in sharp focus in (C) and (D). Diagrams in (E) (head) and (F) (tail) depict locations of neurons (colored cell bodies) expressing DKF-2B–GFP in the context of the organization of the nervous system and anatomy of C. elegans. Cells that express DKF-2A–GFP in vivo were identified previously (20, 21).

DKF-2A and -2B are essential for salt-induced learning underlying behavioral plasticity

We compared the abilities of WT and DKF-2–deficient animals to detect sodium acetate in chemotaxis assays (see Materials and Methods). WT C. elegans sensed 25 mM sodium acetate and moved to salt-enriched areas of test plates (Fig. 5A). However, the chemotaxis index (CI) values for WT animals declined after 30 min of exposure to sodium salt in the absence of food. Two distinct processes, desensitization (adaptation) and associative learning, could account for this time-dependent decrease in CI. Desensitization occurs when receptors are uncoupled from downstream effectors during prolonged exposure to a stimulus (here, sodium acetate). Associative learning reflects a change in behavior due to coupling of a positive stimulus (sodium acetate) with a negative cue. Because standard chemotaxis assays are performed in the absence of food, starvation serves as a powerful negative cue. Consequently, addition of food to chemotaxis plates should suppress aversive learning, but have no effect on desensitization.

Fig. 5

DKF-2 deficiency abrogates salt-induced learning, but salt detection and osmotic avoidance are normal. (A) WT and DKF-2–deficient [dkf-2(pr3)] animals were assayed for salt chemotaxis, with 25 mM sodium acetate. Error bars represent SEM. *P < 0.001 compared to WT values. (B) Learning assays were performed and the CI was calculated. A positive CI indicates attraction (maximal attraction = 1.0); a negative CI denotes avoidance (maximal avoidance = −1.0). WT and DKF-2–depleted animals were assayed for chemotaxis to 25 mM sodium acetate after preincubation in the presence or absence (control) of 100 mM sodium acetate for 1 hour. Error bars represent SEM. *P < 0.001 compared to WT preincubated with sodium acetate; Student’s t test. (C) WT C. elegans and DKF-2–deficient mutants were assayed for avoidance of 1 M sodium acetate. Together, (A), (B), and (C) show that DKF-2–deficient animals detect sodium salt and avoid high osmolarity normally. However, WT animals learn to avoid sodium acetate after prolonged exposure to salt in the absence of food, but learning is severely impaired in DKF-2–depleted C. elegans. Similar results were obtained in three replications of these experiments.

When assay plates were supplemented with food (Escherichia coli OP50), WT animals rapidly and persistently accumulated in quadrants of test plates that contained 25 mM sodium acetate. The CI value did not decline after 30 min. Instead, a near-maximal CI was maintained for ~3 hours (Fig. 5A). Thus, WT C. elegans learn to avoid sodium acetate during the chemotaxis assay if food is not available.

Animals lacking DKF-2A and -2B migrated to 25 mM sodium acetate, but remained concentrated in salt-enriched areas for 3 hours in the absence or presence of food (Fig. 5A). Thus, DKF-2 deficiency does not compromise detection of sodium salts. Rather, it disrupts salt-induced learning.

The contributions of DKF-2B and DKF-2A to behavioral plasticity were analyzed further with a more specific learning paradigm. DKF-2–deficient [dkf-2(pr3)] and wild-type animals were tested for salt chemotaxis after 1 hour of preincubation in the presence or absence of 100 mM sodium acetate (“learning assays”). Both preincubations and chemotaxis assays were performed in the absence of food. WT C. elegans preincubated without salt (“control group”) were attracted to 25 mM sodium acetate (CI = +0.93), but previous exposure to sodium acetate elicited a strong avoidance response (CI = −0.48) (Fig. 5B). After salt-free preincubation, DKF-2A– and DKF-2B–deficient animals had a high CI, similar to the response of WT nematodes. However, the dkf-2(pr3) null mutants did not learn to avoid salt, as indicated by a CI value of +0.63 (Fig. 5B). Neither WT nor DKF-2A– and DKF-2B–null animals avoided 25 mM sodium acetate when preincubation buffer was supplemented with E. coli OP50 (food) (fig. S2), indicating that the assay measures associative learning.

To determine whether DKF-2A and -2B isoforms affect C. elegans’ abilities to detect and escape from an environment containing a high salt concentration (supraphysiological osmolarity), dkf-2(pr3) animals were exposed to 1 M sodium acetate. Both WT and mutant animals efficiently avoided 1 M sodium acetate (Fig. 5C). Thus, DKF-2 isoforms are not required for salt detection or avoidance of high osmolarity.

DKF-2 isoforms mediate Na+-induced, anion-independent learning

C. elegans left ASE neuron (ASEL) senses Na+, whereas the right ASE neuron (ASER) detects Cl− (27). Consequently, the neuronal circuitry and signaling mechanisms underlying salt-induced learning will vary depending on the contributions of cation and anion. We investigated whether DKF-2 isoforms mediate learning induced by Na+, acetate, or both ions. WT and dkf-2(pr3) animals were attracted to 25 mM tris-acetate (Fig. 6A). Preincubation with 100 mM tris-acetate in the absence of food did not cause subsequent avoidance of 25 mM tris-acetate (learning) by either WT or DKF-2–deficient animals. WT C. elegans showed strong chemotaxis to 25 mM sodium acetate after preincubation with 100 mM tris acetate (Fig. 6B). However, either 50% or 100% substitution of Na+ for Tris+ in preincubation buffer elicited pronounced avoidance responses (CI values of −0.47 and −0.59, respectively) in WT animals (Fig. 6B). In contrast, DKF-2–depleted animals were attracted to 25 mM sodium acetate (CI values of 0.4 to 0.65) irrespective of the salt used during preincubation. WT C. elegans and dkf-2(pr3) null mutants were moderately attracted to 25 mM choline chloride (Fig. 6C). After preincubation with 100 mM choline chloride, WT and DKF-2–deficient animals robustly avoided 25 mM chloride (Fig. 6C, CI ~−0.6). Thus, DKF-2 deficiency does not impair Cl− sensing by ASER or learning per se. Overall, the results indicate that tris cation and acetate anion do not elicit DKF-2A– and DKF-2B–mediated learning. Rather, DKF-2 isoforms are specifically engaged in signaling underlying Na+-induced learning and plasticity.

Fig. 6

DKF-2 isoforms mediate Na+-induced learning. (A) WT and DKF-2–deficient [dkf-2(pr3)] nematodes were preincubated in the presence or absence (control) of 100 mM tris-acetate for 1 hour. Subsequently, animals were assayed for chemotaxis to 25 mM tris-acetate. Error bars represent SEM. Results in (A) show that both WT and DKF-2–deficient C. elegans detect tris acetate, but neither tris nor acetate induces aversive learning. (B) WT and DKF-2–depleted animals were preincubated in the presence or absence (control) of sodium acetate, tris-acetate, or an equimolar mixture of sodium acetate and tris-acetate, as indicated. Animals were subsequently assayed for chemotaxis to 25 mM sodium acetate. *P < 0.001 compared to WT animals preincubated with 100 mM sodium acetate. †P < 0.001 compared to WT animals preincubated with 50 mM sodium acetate plus 50 mM tris-acetate; Bonferroni t test. The results in (B) reveal that DKF-2 deficiency severely disrupts Na+-induced learning and plasticity. (C) WT and dkf-2 null C. elegans were preincubated in the presence or absence (control) of 100 mM choline chloride. Subsequently, animals were assayed for chemotaxis to 25 mM choline chloride. The data in (C) show that animals lacking DKF-2 isoforms are not generally compromised in sensing or learning; they detect and learn to avoid Cl− in a normal manner. All experiments were repeated three times and similar results were obtained for each replication.

Both neuronal DKF-2B and intestinal DKF-2A are required for Na+-induced learning

dkf-2(pr3) null C. elegans lacks both DKF-2A and DKF-2B (Fig. 7A). DKF-2A is expressed robustly in intestinal cells and modestly in two head neurons (20); DKF-2B is expressed in ~20 neurons that lack DKF-2A (Fig. 4). To determine whether one or both DKF-2 isoforms are crucial for Na+-induced learning, we created transgenic animals (in a dkf-2 null background) that produced DKF-2B–GFP, DKF-2A–GFP, or both DKF-2–GFP isoforms under control of endogenous promoters (Fig. 7A). These GFP-tagged kinases were detected as 150- and 130-kD fusion proteins in Western blots (Fig. 7A). Similar amounts of DKF-2A–GFP and DKF-2B–GFP were detected in transgenic nematodes.

Fig. 7

Both DKF-2B and DKF-2A are essential for Na+-induced learning. (A) A Western immunoblot documents expression of DKF-2A–GFP (dkf-2A::DKF-2A–GFP transgene), DKF-2B–GFP (dkf-2B::DKF-2B–GFP transgene), or both DKF-2-GFP isoforms in a dkf-2(pr3) null background. (B) The Na+-induced learning paradigm was applied. The indicated C. elegans strains were assayed for chemotaxis to 25 mM sodium acetate after preincubation with 100 mM sodium acetate. Error bars represent SEM. *P < 0.001 compared with WT animals preincubated with 100 mM sodium acetate; Dunnett’s t test. Results in (B) show that both DKF-2 isoforms are required for Na+-induced learning. (C) WT C. elegans and transgenic animals expressing DKF-2A–GFP and DKF-2B–GFP in neurons (dkf-2 null background) were assayed for Na+-induced learning as described in (B). The dkf-2A(ΔGATA) promoter drives expression of DKF-2A–GFP in two head neurons (20); the dkf-2B promoter enables synthesis of DKF-2B–GFP in distinct neurons. Intestinal cells of transgenic animals lack both DKF-2 isoforms. *P < 0.001 compared with WT animals preincubated with 100 mM sodium acetate, Student’s t test. The data in (C) show that, in the absence of intestinal DKF-2A, synthesis of DKF-2B and DKF-2A in neurons does not restore Na+-dependent learning. All experiments were replicated three times and yielded similar results.

DKF-2B–GFP alone did not rescue the Na+-induced learning defect of dkf-2(pr3) null animals (Fig. 7B). When DKF-2A–GFP was the only PKD present, Na+ chemotaxis (sensing) was partially inhibited, as indicated by a decline in CI from +0.87 to +0.22. Preincubation-induced Na+ avoidance was not restored (Fig. 7B). In contrast, animals learned to avoid 25 mM Na+ when DKF-2A and DKF-2B were expressed simultaneously (Fig. 7B). Thus, both DKF-2A and DKF-2B are required for Na+-induced learning and behavioral plasticity.

Because DKF-2A is detected in two head neurons, as well as in intestine, it was possible that Na+-dependent learning was exclusively controlled by DKF-2 isoforms in neurons. We used a mutated dkf-2A promoter, which directs synthesis of DKF-2A–GFP exclusively in head neurons (20), in combination with a WT dkf-2B::DKF-2B–GFP transgene to produce animals that synthesized DKF-2A and -2B isoforms in appropriate neurons, but not in intestinal cells. These animals were defective for Na+-induced learning (CI = +0.55 versus −0.55 for WT C. elegans) and had a phenotype indistinguishable from that of dkf-2(pr3) null C. elegans (Fig. 7C). Thus, expression of DKF-2 isoforms in both neurons and intestinal cells is required for Na+-induced learning and behavioral plasticity.

To rigorously confirm the preceding conclusions, we generated transgenic animals (in a dkf-2 null background) that express DKF-2 isoforms under the control of classical neuron and intestine specific gene promoters. The inx-16 gene, which encodes a gap junction protein, is exclusively and constitutively expressed in larval and adult intestine (28). Na+-dependent learning was not restored when inx-16 promoter and enhancer DNA directed expression of DKF-2A–GFP in intestinal cells of dkf-2(pr3) null animals (fig. S3). The rgef-1 (F25B3.3) gene encodes a Ras-Rap guanosine 5′-triphosphate (GTP) exchange factor (WormBase). The rgef-1 promoter is active in all neurons, but is silent in nonneuronal cells (29, 30). Expression of an rgef-1::DKF-2B–GFP transgene in neurons failed to rescue the learning defect in dkf-2 null animals (fig. S3). However, simultaneous expression of inx-16::DKF-2A–GFP and rgef-1::DKF-2B–GFP transgenes in intestinal cells and neurons, respectively, restored robust Na+-dependent learning and behavioral plasticity in dkf-2(pr3) animals. These observations replicate the results obtained when DKF-2A–GFP and DKF-2B–GFP were expressed under regulation of endogenous dkf-2A and dkf-2B promoters (Fig. 7B). Thus, cooperation between neuronal DKF-2B and intestinal DKF-2A is essential for progression of a Na+-induced learning process.

EGL-8 and TPA-1 control DKF-2A and -2B activation and functions in vivo

If WT DKF-2 isoforms are not activated by upstream signaling proteins, then C. elegans will exhibit the same learning defect as dkf-2(pr3) animals. Consequently, we used loss of Na+-induced learning as an assay, in combination with molecular genetics, to identify in vivo PKD regulators. C. elegans expresses two PLC-β homologs. Animals homozygous for severe loss-of-function mutations in plc-2 (PLC-β2 homolog) and egl-8 (PLC-β4 homolog) genes were tested in the learning assay. Disruption of the plc-2 gene had no effect on Na+-induced learning (Fig. 8A). In contrast, animals with defective EGL-8 protein (md1977 or n488 alleles of egl-8) did not learn to avoid 25 mM Na+ after preincubation with 100 mM sodium acetate (Fig. 8A). Defects in Na+-induced learning were qualitatively and quantitatively similar in egl-8(n488) and dkf-2(pr3) single mutants and an egl-8(n488);dkf-2(pr3) double mutant (Fig. 8A). These results place EGL-8 and DKF-2A or 2B in the same pathway and indicate that DAG production (or an increase in free cytoplasmic Ca2+ or both) is essential for salt taste–induced learning.

Fig. 8

EGL-8, DAG, and TPA-1 regulate DKF-2 isoforms and Na+-induced learning. (A) C. elegans strains that have severe loss-of-function mutations in plc-2, egl-8, or dkf-2 genes (ok1761, md1971, n488, and pr3 alleles) were assayed for Na+-induced learning. Results in (A) show that EGL-8 is required for Na+-dependent, aversive learning. EGL-8 and DKF-2 isoforms function in the same pathway. (B) C. elegans strains that have single or double-null or strong loss-of-function mutations in pkc-1, pkc-2, tpa-1, or dkf-2 genes (ok563, ok328, km530, and pr3 alleles, respectively) were tested for Na+-dependent learning. Results in (B) show that TPA-1, but not other PKCs, is essential for Na+-induced aversive learning and exerts its actions in the same pathway as that of DKF-2 isoforms. (C) Chemotaxis indices were determined for C. elegans strains that have single or double loss-of-function mutations in either dgk-1 or egl-8 genes (ok1462 and n488 alleles, respectively) or both genes to assess Na+-induced learning and plasticity. Data in (C) indicate that DAG, a signaling molecule generated by EGL-8, partially restores Na+-induced learning when PLC activity is diminished. Error bars represent SEM. *P < 0.001 compared with WT animals preincubated with 100 mM sodium acetate; ‡P < 0.001 compared with egl-8(n488) animals preincubated with 100 mM sodium acetate. Dunnett’s t test was used in (A) and (B); Bonferroni’s t test was applied in (C). (D) tpa-1(k530), egl-8(n488), and WT strains of C. elegans were incubated with 100 mM sodium acetate for 1 hour in the absence of food. Duplicate batches of animals were used as untreated controls. Nematodes were disrupted and endogenous DKF-2A and DKF-2B were immunoprecipitated and analyzed in a Western blot (Materials and Methods). The blot in the upper panel shows that similar amounts of DKF-2A and -2B proteins were isolated in each precipitation. The blot in the lower panel shows that Na+ and starvation-induced A-loop phosphorylation (and hence activation) of DKF-2A and -2B (lane 2) was suppressed in EGL-8 or TPA-1–deficient animals (lanes 4 and 6). All experiments in the figure were repeated three times and similar results were obtained.

DKF-2B and 2A isoforms are activated by PKC-mediated phosphorylation of A-loop serines [Fig. 4, A and B, and (20)]. C. elegans has three genes that encode DAG-regulated PKCs. PKC-2 is homologous with the mammalian Ca2+- and DAG-activated PKCs α, βI, βII, and γ; it is expressed in neurons, intestine, and other tissues (31). PKC-1 corresponds to vertebrate PKCs ɛ and η; it is expressed in sensory and other neurons (32). TPA-1 is analogous to PKCs δ and θ; it accumulates in neurons, intestine, and several other cell types (33). Both TPA-1 and PKC-1 are controlled by DAG alone. We asked whether one or multiple PKCs regulate DKF-2B in the nervous system in vivo.

pkc-1(ok563) null, pkc-2(ok328) null, and TPA-1 loss-of-function [tpa-1(k530)] mutants are viable, fertile, and develop normally. Single and double mutants lacking PKC-1 or PKC-2 or both isoforms avoided 25 mM Na+ after preincubation with 100 mM sodium acetate (CI values less than −0.4) (Fig. 8B). Thus, Ca2+- and DAG-regulated PKC-2 and its targets, as well as DAG-controlled PKC-1 and its effector proteins, are not implicated in signaling pathways that control salt-induced plasticity. TPA-1 deficiency had no effect on Na+ detection. However, TPA-1 depletion markedly impaired Na+-dependent learning (CI ~+0.4), yielding the same phenotype as DKF-2A and -2B deficiency (Fig. 8B). Phenotypes of dkf-2(pr3) null animals and double mutants lacking TPA-1 and both DKF-2 isoforms were similar. These results indicate that TPA-1 regulates DKF-2B [and DKF-2A, as previously shown (21)] in vivo. Because TPA-1 is activated by DAG alone, these observations indicate that DKF-2B is part of a PLC-regulated, DAG-controlled signaling pathway. DKF-2B links signals carried by DAG and an nPKC to a central physiological function, experience-dependent learning.

Results from the preceding genetic and functional analyses indicate that deleterious mutations in the tpa-1 and egl-8 genes should suppress in vivo activation of DKF-2 isoforms. To explicitly test this proposition, we incubated tpa-1(k530), egl-8(n488), and WT strains of C. elegans with 100 mM sodium acetate for 1 hour in the absence of food. Duplicate groups of animals were used as untreated controls. After detergent-soluble proteins were isolated from the nematodes, DKF-2A and -2B were immunoprecipitated with IgGs directed against a shared C-terminal epitope. Western analysis of precipitated proteins revealed that loss of TPA-1 or EGL-8 activity did not substantially alter the abundance of DKF-2A and DKF-2B polypeptides (Fig. 8D, upper panel). IgGs directed against the diphosphorylated activation loop of DKF-2 isoforms detected a low degree of DKF-2B phosphorylation (indicating low kinase activity) in untreated WT animals (Fig. 8D, lower panel, lane 1). Moreover, DKF-2A phosphorylation and a substantial increase in DKF-2B phosphorylation were evident when WT animals were simultaneously exposed to sodium acetate and starved (Fig. 8D, lower panel, lane 2). In contrast, DKF-2 isoforms were not phosphorylated in control or starved, sodium acetate–treated animals lacking TPA-1 or EGL-8 (Fig. 8D, lower panel, lanes 3 to 6). Thus, loss of Na+-induced learning in TPA-1– or EGL-8–depleted animals (Fig. 8, A and B) was linked to lack of activation loop phosphorylation and, hence, inactivation of DKF-2A and DKF-2B (Fig. 8D).

DAG kinase gene disruption restores learning in EGL-8–deficient animals

Diacylglycerol kinases (DGKs) catalyze the synthesis of phosphatidic acid from DAG. Consequently, loss-of-function mutations in dgk genes increase intracellular DAG content. DGK-1, a human DGKθ homolog, is expressed in neurons. A partially deleted dgk-1(ok1462) allele encodes a DGK-1 polypeptide that lacks ~40% of the kinase domain, indicating it is incapable of phosphorylating DAG. Defective DGK-1 (which promotes DAG accumulation) did not affect learning (Fig. 8C). However, loss of Na+-dependent learning in PLC-deficient (DAG-depleted) C. elegans [egl-8(n488)] was suppressed in a dgk-1(ok1462), egl-8(n488) double mutant (Fig. 8C). A CI value of −0.15 showed that double mutants avoided 25 mM sodium acetate after preincubation with 100 mM Na+; egl-8(n488) animals were attracted to 25 mM sodium acetate (CI = +0.40) under the same conditions. Thus, DAG-controlled, TPA-1–DKF-2–mediated signal transduction governs Na+-induced learning and behavioral plasticity.

WT transgenes rescue learning defects in tpa-1 and egl-8 mutants

TPA-1– and EGL-8–deficient animals were reconstituted with WT transgenes controlled by endogenous promoters (tpa-1::TPA-1 and egl-8::EGL-8, respectively). Expression of the WT PKCδ and PLC-β4 homologs restored robust, Na+-induced, aversive learning in the transgenic strains (fig. S4, A and B). The results confirm that EGL-8 and TPA-1 specifically regulate activation and functions of DKF-2 isoforms in vivo.

Discussion

Two distinct promoters govern dkf-2 gene transcription in different cell types. The respective mRNAs encode DKF-2A in intestinal cells and DKF-2B in neurons. Activated PKDs create distinct branches in DAG-PKC–controlled signaling networks because they translocate from the plasma membrane to various intracellular locations and phosphorylate effectors that are not PKC substrates (6–8). Consequently, expression of DKF-2B in ~20 pairs of sensory neurons and interneurons expands and diversifies the physiological impact of DAG-PKC–mediated signaling in C. elegans.

DKF-2B and DKF-2A have identical C1 and kinase domains. Like DKF-2A, DKF-2B is recruited to membranes by PMA or DAG. Co-recruited PKC switches on DKF-2B catalytic activity by phosphorylating A-loop serines 727 and 731. Thus, the regulatory and catalytic properties of the DKF-2 isoforms are similar. Unique N-terminal regions confer markedly different predicted isoelectric points on DKF-2A (pI = 6.6) and DKF-2B (pI = 8.9). Dissimilar electrostatic properties could facilitate interactions with distinct binding partners or differentially alter the stabilities of DKF-2A and DKF-2B polypeptides.

To ensure survival and optimize reproduction, C. elegans must detect beneficial environmental stimuli (food, essential ions, and the like), but avoid toxins, bacterial pathogens, and starvation. C. elegans’ behavior toward attractive chemosensory signals (such as Na+) can be modified by associated cues (such as starvation) and previous experience. Behavioral plasticity, based on associative learning, is a conserved neurophysiological process that enables organisms ranging from nematodes to mammals to adapt to a dynamic environment. We found that DAG, a PKC δ-θ homolog (TPA-1), and two PKDs (DKF-2A and -2B) are essential for salt taste (Na+)–induced behavioral plasticity in C. elegans. These studies directly link PKDs to regulation of a critical nervous system function and yield insights into the molecular basis for learning.

Unlike WT C. elegans, DKF-2–deficient animals do not switch from attraction to 25 mM Na+ to aversion after preconditioning with 100 mM sodium acetate (minus food). Reconstitution of dkf-2(pr3) null animals with a dkf-2A::DKF-2A–GFP transgene did not restore Na+-induced learning as would be anticipated if it were governed solely by neurons expressing DKF-2B. Unexpectedly, neither a dkf-2B::DKF-2B–GFP transgene nor expression of both DKF-2B–GFP and DKF-2A–GFP in appropriate neurons rescued the learning defect caused by dkf-2 gene disruption. However, transgenic animals expressing DKF-2A–GFP and DKF-2B–GFP in intestine and neurons, respectively, were indistinguishable from WT C. elegans in Na+-dependent learning assays. Thus, this learning program is switched on only when signals carried by DAG are disseminated by PKD isoforms (DKF-2A and DKF-2B) in discrete populations of cells. The Na+-induced behavioral change is evidently triggered by a binary stimulus detector system that is embedded in nervous tissue and the gut. Thus, PKD-regulated behavioral output reflects integration of signaling information acquired by both neurons and intestinal cells. Signals could be integrated via mono- or bidirectional crosstalk. Increased intestinal DKF-2A activity promotes accumulation of mRNAs encoding numerous secreted, hormone-like peptides (21). Diffusible peptides could bind to and activate neuronal receptors, thereby coupling intestinal DKF-2A to regulation of neuronal physiology. DKF-2B activation could also trigger release of neuropeptides that carry signals from neurons to receptors on intestinal cells.

Na+ avoidance is a form of associative learning generated by two cues: sodium salt and starvation (24). The primary Na+ detecting neuron, ASEL, has ciliated endings that contact the external milieu. ASEL and its downstream synaptic partners (interneurons) express DKF-2B. Thus, our observations suggest that DKF-2B is essential for synaptic transmission and neuropeptide secretion underlying aversive learning. We speculate that DKF-2A contributes to behavioral plasticity by transducing a starvation signal in intestine. Activation of intestinal DKF-2A would promote secretion of gut hormones that mediate crosstalk with the nervous system as described above. Signals emanating from similar, DAG-PKD–controlled pathways in neurons and intestinal epithelial cells are integrated to generate a learned alteration in behavior, Na+ avoidance (fig. S5).

Expression of intestinal DKF-2A–GFP in dkf-2(pr3) animals substantially reduced Na+ detection (Fig. 7B and fig. S3). Animals lacking either DKF-2A or both DKF-2 isoforms detected Na+ normally (Fig. 7B and fig. S3). These results suggest that DKF-2A activity elicits release of a diffusible signaling molecule from intestine that either reduces the sensitivity of the ASEL neuron to environmental Na+ or diminishes signal strength in the chemosensory circuitry after Na+ is encountered. Although the underlying molecular mechanism is unknown, the observations indicate that a PLC- and DAG-activated intestinal PKD generates endocrine regulatory signals that are routed to a specific chemosensory circuit in the nervous system. This provides further support for the idea that signaling mediated by DKF-2 isoforms coordinates and integrates physiological activities of the gut and nervous system.

Expression of DKF-2B is not essential for maximal sensitivity to environmental Na+ ions (Fig. 7B and fig. S3). However, coexpression of DKF-2B and DKF-2A in WT and reconstituted dkf-2 null animals enables highly efficient Na+ sensing (CI greater than +0.9). Thus, DKF-2B evidently generates a counterregulatory signal that suppresses the inhibitory action of DKF-2A. The signaling molecule could act locally (on neurons) or be targeted to intestinal cells through diffusion. Identification of the molecules and mechanisms underlying this mode of bidirectional neuron–intestinal cell communication is an important objective for future investigations.

The effects of DKF-2A and DKF-2B on Na+ detection per se appear to be separate and distinct from their effects on Na+-dependent learning and plasticity. Animals expressing DKF-2B alone or lacking both DKF-2 isoforms sense Na+ normally, but are highly impaired in learning. dkf-2(pr3) null C. elegans, as well as transgenic animals expressing either DKF-2A–GFP or DKF-2B–GFP (dkf-2 null background), can detect increased Na+ content in the external environment (albeit with quantitatively different CI values). However, none of these animals learns to avoid Na+. The reduced Na+-sensing ability of transgenic animals expressing only DKF-2A is not due to a general defect in learning. These animals detect and learn to avoid Cl− ions normally (fig. S6), yielding essentially the same CI values as that of WT and dkf-2(pr3) null nematodes (Fig. 6C). Overall, decreased Na+ detection by animals expressing DKF-2A–GFP alone does not affect results, interpretations, or conclusions regarding the roles of DKF-2A and DKF-2B in learning and plasticity.

The ASER and ASEL amphid sensory neurons contact the external environment and detect K+, Cl− and Na+ ions (34). Ion sensing is coupled to behavior through the activation of downstream signaling pathways in ASE neurons and neurons controlled by ASE-derived neurotransmitters or neuropeptides (22, 23, 35, 36). ASE neurotransmitters propagate transsynaptic signals to several interneurons (22, 35, 37), which synapse with a group of motor neurons that control the animal’s orientation and movement toward increasing concentrations of ions. To explain learning, Jansen and colleagues (22) propose that the same interneurons can receive inhibitory signals (neuoropeptides, neurotransmitters) emanating from sensory neurons that control avoidance of noxious stimuli. They postulate that, on prolonged exposure to salt, ASE secretes distinct neuropeptides that switch on avoidance signaling. Thus, the interneurons process and integrate opposing attraction and avoidance signals to generate dynamic, plastic behavior. Signaling initiated by guanosine 3′,5′-monophosphate (cGMP)–gated calcium channels, Gqα-GTP, and Gβγ subunits has been implicated in salt-induced learning (22, 23). However, little is known about the critical downstream regulatory proteins, signaling modules and mechanisms that link cGMP or heterotrimeric G proteins to associative learning.

DKF-2B is robustly expressed in ASE neurons. However, body cavity neurons and numerous integrative interneurons also contain DKF-2B. Thus, DKF-2B could potentially contribute to Na+-induced learning by controlling synaptic output of ASE; mediating signal reception and transmission in body cavity neurons; regulating signal processing and integration in higher-level interneurons that translate the differential between attractive and aversive inputs into a modulated behavioral response or some combination of these. DKF-2B is also a potential effector for DAG generated when Gq stimulates EGL-8.

Salt-taste learning is also linked to the insulin–phosphatidylinositol 3-kinase (PI3K) signaling pathway in ASER (35). Mutations in daf-2 (which encodes an insulin receptor), age-1 (which encodes PI3K), and akt-1 (which encodes the serine-threonine kinase Akt) abolish salt-induced learning. Neurotransmitter released by ASER creates a negative feedback loop by triggering insulin (INS-1) secretion from adjacent AIA interneurons. INS-1 activates the DAF-2 pathway, which down-regulates synaptic output from ASER (35). AIA forms direct reciprocal synapses with ASER, but not ASEL (38).

The study implicating the insulin signaling pathway in salt-induced learning used NaCl as a stimulus (35). Because ASER senses Cl−, but not Na+, the insulin pathway evidently mediates avoidance of the anion. In contrast, DKF-2A and -2B are indispensable for Na+-induced learning; acetate anion used in our studies had no effect on learning. A lack of direct reciprocal synaptic connections between AIA and ASEL further suggests that the negative feedback model for ASER (Cl− avoidance) may not apply to Na+-induced behavioral plasticity.

The cellular basis for contribution of DKF-2 isoforms to learning differs markedly from the ASER-insulin single-cell model. Expression of DKF-2B and -2A in neurons is necessary, but not sufficient for Na+-induced learning. An indispensable input into Na+-induced plasticity is generated by coactivation of DKF-2A in intestinal epithelial cells.

Our data indicate that the PLC-β4 homolog EGL-8 produces DAG, which controls activation of DKF-2 isoforms and Na+-induced plasticity. In contrast, the PLC-β2 homolog PLC-2 is not an upstream regulator of PKDs. The phenotypes of egl-8 and dkf-2 null single mutants and egl-8, dkf-2(pr3) double mutants are similar, suggesting PLCs γ and ɛ are probably not involved in learning underlying Na+ avoidance.

TPA-1, a DAG-activated PKC δ-θ homolog, controls the in vivo activation (and functions) of DKF-2B in neurons and DKF-2A in intestine. C. elegans PKCs 1 and 2 are expressed in many neurons (31, 32). pkc-2 gene transcription is controlled by three promoters and transcripts are modified by differential splicing, yielding mRNAs encoding six PKCs homologous with mammalian Ca2+- and DAG-activated PKCs (31). Alternative splicing generates two DAG-activated PKC ɛ-η homologs, PKCs 1A and 1B (32). Despite their diversity and potential multifunctionality, PKCs 1 and 2 and their substrate effectors are not engaged in Na+-dependent learning.

Most mammalian DAG-controlled PKCs can phosphorylate and activate PKDs in cultured cells (6, 7), suggesting that widely expressed PKCs could serve as redundant PKD activators in signaling cascades. The discovery that C. elegans PKDs are specifically regulated by TPA-1 in vivo provides a distinct counterpoint to the concept of redundant regulation. Studies using transfection and overexpression, small interfering RNA and dominant-negative kinase strategies in homogeneous cell culture models have provided invaluable insights into PKC-mediated regulation of PKD. However, extrapolation of these findings to in vivo systems has inherent limitations. The unique relationship between conserved TPA-1 and DKF-2 isoforms in C. elegans suggests that reassessing the roles of individual PKC isoforms in PKD activation will be a priority when in vivo models for PKD regulation are established in mice.

DAG-activated PKCs have been implicated in synaptic plasticity underlying long-term depression (LTD), long-term potentiation (LTP), and various aspects of behavioral learning in animals with complex nervous systems (39–42). For example, Aplysia PKCɛ mediates reversal of serotonin-induced synaptic depression, whereas a Ca2+ and DAG-regulated PKC promotes increased neurotransmitter release at nondepressed synapses. Mammalian PKCγ is required for the induction of LTP in hippocampal neurons by high-frequency electrical stimulation. PKCα is involved in the development of LTD at parallel fiber–Purkinje cell synapses, where it stimulates endocytosis of type 2 metabotropic glutamate receptors. However, the identities, exact roles, and regulation of the PKC isoforms that affect specific facets of synaptic plasticity are incompletely understood. Here, we link Na+-induced behavioral plasticity in C. elegans to activation of a single PKC isoform, TPA-1, in neurons and nonneuronal cells. Inclusion of downstream PKDs and their effectors in the PKC signaling pathway extends our understanding of the molecular basis of plasticity. The mechanisms that enable the selective activation of DKF-2A and -2B by TPA-1 and insulation of DKF-2 isoforms from other DAG-activated PKCs remain to be determined. Activation of DKF-2A and DKF-2B by a single PKC increases signaling specificity by bypassing the overlapping substrate specificity of PKC isoforms. PKDs phosphorylate and regulate proteins that are not substrates for other DAG-activated PKCs. Identification of TPA-1 as the upstream regulator of DKF-2A and DKF-2B in vivo challenges the generalization that PKCδ is not involved in plasticity (39). Results and interpretations elaborated above suggest that exploring the roles of mammalian PKDs in LTP, LTD, and other aspects of learning may advance understanding of plasticity in the central nervous system.

C. elegans uses a signaling pathway that includes EGL-8, DAG, TPA-1, and a DKF-2 isoform to transmit and integrate environmental signals targeted to neurons and intestine. However, functions of this signaling pathway are not limited to either plasticity or physiological outputs dependent on interactions between different cell types. The intestinal EGL-8–DAG–TPA-1–DKF-2A pathway promotes induction of mRNAs that mediate innate immunity in animals lacking DKF-2B (21). Major challenges and goals for future studies will be discovery and characterization of critical DKF-2A and -2B effectors and elucidation of mechanisms by which DKF-2 effectors control distinct and shared physiological functions.

Materials and Methods

C. elegans strains and culture

C. elegans strains were cultivated at 20°C on agar plates seeded with E. coli OP50. The C. elegans Genetics Center (St. Paul, MN) provided plc-2(ok1761), egl-8(md1971), egl-8(n488), dgk-1(ok1462), and tpa-1(k530) mutants. pkc-1(ok563), pkc-2(ok328), and dkf-2(pr3) null mutants were obtained through collaboration with the C. elegans Knockout Consortium (Vancouver, BC, Canada). Standard protocols were used to generate pkc-1(ok563);pkc-2(ok328) double null mutants and animals carrying chromosomally integrated copies dkf-2A::DKF-2A–GFP and dkf-2B::DKF-2B–GFP transgenes. Mutants were backcrossed six times into WT background.

Antibodies

Polyclonal IgGs directed against epitopes in an N-terminal segment of DKF-2A (amino acids 1 to 203) and recombinant GFP were produced in rabbits and purified as previously described (20). Peptides corresponding to diphosphorylated DKF-2A–2B A loop (GEKpSFRRpSVVG), nonphosphorylated A loop (GEKSFRRSVVG), and the C terminus of DKF-2A–2B (HESDDVRWQAYEKEHNVTPVYV) (43), were synthesized at Proteintech Group (Chicago, IL). Cys was added to the N termini of peptides to enable covalent coupling with maleimide-activated keyhole limpet hemocyanin carrier protein (for antigen injection) or SulfoLink agarose beads (Pierce Biotechnology) to generate affinity chromatography resins. IgGs directed against C-terminal and diphosphorylated A-loop peptides were produced in rabbits and purified as previously reported (32, 44). In the latter case, antiserum was cycled over a column of nonphosphorylated peptide three times. Subsequently, the antiserum was applied to a phosphorylated A-loop peptide–agarose column for affinity purification of desired IgGs. IgGs were stored in phosphate-buffered saline–50% glycerol at −20°C. IgGs against 14-3-3 and IGF1-Rβ were purchased from Santa Cruz Biotechnology.

Western immunoblot analysis

Proteins were obtained from C. elegans, cytosolic and membrane fractions of transfected cells, or transfected cells lysed with buffer containing 1% Triton X-100 as previously described (20, 21). Proteins were size-fractionated by denaturing electrophoresis and transferred to a polyvinylidene difluoride membrane as previously reported (20, 45, 46). Each lane of the Western blot received 30 μg of protein. Blots were probed with affinity-purified IgGs that bind the C termini of both DKF-2A and DKF-2B or the diphosphorylated A-loop peptide that is identical in DKF-2A and DKF-2B (see above). Antigen-antibody complexes were visualized by using peroxidase-coupled secondary antibodies and an enhanced chemiluminescence procedure (46). Chemiluminescence signals were captured on x-ray film. All Western blot assays were repeated three times and similar results were obtained.

Protein kinase assay

DKF-2B– or DKF-2B–GFP was immunoprecipitated from detergent extracts of transfected cells as previously described (20). Catalytic activity was quantified by measuring incorporation of 32P radioactivity from [γ-32P]adenosine 5′-triphosphate into Syntide-2 as previously reported (20, 21). Kinase assays were replicated three times and similar results were obtained.

RT-PCR

Total RNA was extracted from young adult WT animals with TRIZOL Reagent (Invitrogen). RNA was purified on an RNeasy column (Qiagen). Superscript III reverse transcriptase (Invitrogen) catalyzed first-strand cDNA synthesis, with purified RNA as a template and oligothymidilate as primer. cDNA encoding full-length DKF-2B open reading frame was synthesized by PCR, using KOD DNA polymerase (Novagen) and primers corresponding to sequences in the 5′ (Fig. 1C) and 3′ (20) UTRs of DKF-2B transcripts. cDNA was verified by sequencing and cloned into the pCDNA 3.1 expression vector (20).

Transgenes and transgenic animals

We fused genomic DNA (2.2 kb) flanking the 5′ end of the dkf-2B structural gene to cDNA encoding full-length DKF-2B in the C. elegans expression vector pPD 95.79 (26, 47). We eliminated the translation stop codon in the DKF-2B ORF to enable in-frame fusion with vector DNA encoding GFP. Authentic promoter and enhancer elements ensure normal temporal, developmental, and cell-specific expression of DKF-2B–GFP. To determine if the GFP tag altered regulatory or catalytic properties of the kinase or both, we transfected cultured cells with DKF-2B or DKF-2B–GFP transgenes and assayed kinase activity. Like WT DKF-2B, the DKF-2B–GFP fusion kinase showed low basal phosphotransferase activity (fig. S7). Incubation of cells with PMA promoted large, parallel increases in the catalytic activities of DKF-2B and DKF-2B–GFP (fig. S7). GF109203X blocked PMA-induced activation of both kinases. Thus, DKF-2B–GFP is appropriately regulated and functional.

Genomic promoter-enhancer DNA (830 bp) that precedes the 5′ end of the inx-16 gene was provided by E. Jorgensen (University of Utah). Promoter-enhancer DNA (2.7 kb) that flanks the 5′ end of the rgef-1 structural gene was obtained from L. Chen in the Rubin laboratory. inx-16 and rgef-1 promoter DNAs were fused to cDNAs encoding full-length DKF-2A–GFP and DKF-2B–GFP, respectively, in the C. elegans expression vector pPD 95.79, as previously described (26, 47). A transgene in which egl-8 promoter DNA (2.5 kb) drives expression of full-length EGL-8 cDNA (48) was provided by J. Kaplan (Harvard Medical School). Promoter-enhancer DNA (2.6 kb of DNA immediately adjacent to the 5′ end of the tpa-1 structural gene) and full-length TPA-1 cDNA were cloned by PCR, with genomic DNA and a cDNA library, respectively, as templates. Amplified DNAs were cloned into the expression vector pPD 95.77 to create a tpa-1::TPA-1 transgene, as previously described (26, 47) Transgenic animals were generated as previously reported (26, 49).

Immunofluorescence microscopy

Transfected cells expressing DKF-2B–GFP were fixed, permeabilized, and incubated successively with IgGs against DKF-2 and fluorescein isothiocyanate–tagged secondary antibodies as previously described (50). Fluorescence signals corresponding to antigen-antibody complexes were recorded with a Zeiss Axio Imager Z1 microscope.

C. elegans expressing a dkf-2B::DKF-2B–GFP transgene was analyzed by fluorescence microscopy to determine DKF-2B localization. For low-magnification microscopy, GFP-derived fluorescence was recorded from fixed animals with a Zeiss Axio Imager Z1 microscope and images were processed with Zeiss AxioVision software. For higher-resolution microscopy, live animals were anesthetized with 1-phenoxy-2-propanol and fluorescence signals were captured with an Olympus AX70 microscope equipped with an Olympus Optronics camera and Magnafire 2.1 software. Image processing and analysis were performed with Adobe photoshop CS2. All neurons expressing DKF-2B–GFP were identified in multiple lines of transgenic animals.

Chemotaxis assay

Chemotaxis to water-soluble compounds was assayed in Petri plates divided into quadrants (22, 23). Pairs of opposite quadrants were filled with buffered agar [2% agar, 10 mM MOPS (pH 7.2), 1 mM CaCl2, 1 mM MgSO4] containing or lacking dissolved attractant (25 mM sodium acetate). Adjacent quadrants were connected with a thin layer of agar. After young adult nematodes were washed three times with CTX buffer [5 mM KH2PO4/K2HPO4 (pH 6.6), 1 mM CaCl2, 1 mM MgSO4], 160 to 200 animals were placed at the intersection of the four quadrants. In the standard assay, the distribution of animals in the quadrants was determined after 10 min. (An extended time course for chemotaxis is provided in Fig. 5A.) The CI was calculated as follows: CI = (A − C)/(A + C), where A is the number of animals in quadrants containing sodium acetate, and C is the number of nematodes in quadrants lacking dissolved attractant.

Learning assay

Young adult animals were washed and incubated with CTX buffer alone (control) or CTX buffer containing 100 mM sodium acetate (or other salts, see Fig. 6) for 1 hour (22). Subsequently, animals were collected by sedimentation and tested in the chemotaxis assay described above to determine CI values. Assays were repeated three times and yielded similar results.

High-osmolarity avoidance assay

Opposite quadrants of petri plates were filled with buffered agar lacking or containing 1 M sodium acetate (22). Animals were placed at the intersection of the quadrants and CI values were determined as described above. CI values below zero provide an avoidance index (−1.0 corresponds to maximal aversion).

Other methods

ExPASY ProtParam software, provided by the Swiss Institute of Bioinformatics (www.expasy.ch/tools/protparam.html), was used to predict pI values for DKF-2A and -2B. Detailed procedures for cell culture, expression of DKF-2 isoforms in transfected cells, preparation of cytosol and total membrane proteins, biochemical assay of DKF-2A–2B translocation in intact cells, immunoprecipitation, in vitro kinase assays, Western immunoblot assays, immunofluorescence microscopy, electrophoresis of proteins, DNA sequencing, protein quantification, construction of transgenes and transgenic animals, and creation of animals expressing an integrated dkf-2A::DKF-2A–GFP transgene are as previously described (20, 21, 26, 45, 49–51)

Acknowledgments

We thank J. Sze and D. Hall for helpful discussions and advice. This work was supported by NIH grant GM080615 to C. S. Rubin.

Supplementary Materials

www.sciencesignaling.org/cgi/content/full/2/83/ra42/DC1

Fig. S1. Domain organization of PKDs and PKCɛ.

Fig. S2. Food suppresses Na+-induced learning.

Fig. S3. Both neuronal DKF-2B and intestinal DKF-2A are essential for Na+-induced, aversive learning.

Fig. S4. WT transgenes rescue learning defects in TPA-1– and EGL-8–deficient animals.

Fig. S5. PLC-DAG-PKC-PKD signaling modules in neurons and intestinal cells cooperatively mediate Na+-induced, associative learning.

Fig. S6. Animals differentially expressing DKF-2A are not generally compromised in sensing or learning.

Fig. S7. DKF-2B–GFP and WT DKF-2B have similar properties.

References and Notes

  1. 1.↵
    1. S. G. Rhee
    , Regulation of phosphoinositide-specific phospholipase C. Annu. Rev. Biochem. 70, 281–312 (2001).
    OpenUrlCrossRefPubMedWeb of Science
  2. 2.↵
    1. M. J. Berridge,
    2. R. F. Irvine
    , Inositol phosphates and cell signalling. Nature 341, 197–205 (1989).
    OpenUrlCrossRefPubMedWeb of Science
  3. 3.↵
    1. A. C. Newton
    , Regulation of the ABC kinases by phosphorylation: Protein kinase C as a paradigm. Biochem. J. 370, 361–371 (2003).
    OpenUrlCrossRefPubMedWeb of Science
  4. 4.↵
    1. M. Spitaler,
    2. D. A. Cantrell
    , Protein kinase C and beyond. Nat. Immunol. 5, 785–790 (2004).
    OpenUrlCrossRefPubMedWeb of Science
  5. 5.↵
    1. H. Mellor,
    2. P. J. Parker
    , The extended protein kinase C superfamily. Biochem. J. 332, 281–292 (1998).
    OpenUrlAbstract/FREE Full Text
  6. 6.↵
    1. Q. J. Wang
    , PKD at the crossroads of DAG and PKC signaling. Trends Pharmacol. Sci. 27, 317–323 (2006).
    OpenUrlCrossRefPubMedWeb of Science
  7. 7.↵
    1. E. Rozengurt,
    2. O. Rey,
    3. R. T. Waldron
    , Protein kinase D signaling. J. Biol. Chem. 280, 13205–13208 (2005).
    OpenUrlFREE Full Text
  8. 8.↵
    1. K. Nishikawa,
    2. A. Toker,
    3. F. J. Johannes,
    4. Z. Songyang,
    5. L. C. Cantley
    , Determination of the specific substrate sequence motifs of protein kinase C isozymes. J. Biol. Chem. 272, 952–960 (1997).
    OpenUrlAbstract/FREE Full Text
  9. 9.↵
    1. R. B. Vega,
    2. B. C. Harrison,
    3. E. Meadows,
    4. C. R. Roberts,
    5. P. J. Papst,
    6. E. N. Olson,
    7. T. A. McKinsey
    , Protein kinases C and D mediate agonist-dependent cardiac hypertrophy through nuclear export of histone deacetylase 5. Mol. Cell. Biol. 24, 8374–8385 (2004).
    OpenUrlAbstract/FREE Full Text
  10. 10.↵
    1. T. A. McKinsey
    , Derepression of pathological cardiac genes by members of the CaM kinase superfamily. Cardiovasc. Res. 73, 667–677 (2007).
    OpenUrlAbstract/FREE Full Text
  11. 11.↵
    1. P. Liu,
    2. A. M. Scharenberg,
    3. D. A. Cantrell,
    4. S. A. Matthews
    , Protein kinase D enzymes are dispensable for proliferation, survival and antigen receptor-regulated NFκB activity in vertebrate B-cells. FEBS Lett. 581, 1377–1382 (2007).
    OpenUrlCrossRefPubMedWeb of Science
  12. 12.↵
    1. J. Fielitz,
    2. M. S. Kim,
    3. J. M. Shelton,
    4. X. Qi,
    5. J. A. Hill,
    6. J. A. Richardson,
    7. R. Bassel-Duby,
    8. E. N. Olson
    , Requirement of protein kinase D1 for pathological cardiac remodeling. Proc. Natl. Acad. Sci. U.S.A. 105, 3059–3063 (2008).
    OpenUrlAbstract/FREE Full Text
  13. 13.↵
    1. M. S. Kim,
    2. J. Fielitz,
    3. J. McAnally,
    4. J. M. Shelton,
    5. D. D. Lemon,
    6. T. A. McKinsey,
    7. J. A. Richardson,
    8. R. Bassel-Duby,
    9. E. N. Olson
    , Protein kinase D1 stimulates MEF2 activity in skeletal muscle and enhances muscle performance. Mol. Cell. Biol. 28, 3600–3609 (2008).
    OpenUrlAbstract/FREE Full Text
  14. 14.↵
    1. B. C. Harrison,
    2. M. S. Kim,
    3. E. van Rooij,
    4. C. F. Plato,
    5. P.J. Papst,
    6. R. B. Vega,
    7. J. A. McAnally,
    8. J. A. Richardson,
    9. R. Bassel-Duby,
    10. E. N. Olson,
    11. T. A. McKinsey
    , Regulation of cardiac stress signaling by protein kinase d1. Mol. Cell. Biol. 26, 3875–3888 (2006).
    OpenUrlAbstract/FREE Full Text
  15. 15.↵
    1. D. M. Tobin,
    2. C. I. Bargmann
    , Invertebrate nociception: Behaviors, neurons and molecules. J. Neurobiol. 61, 161–174 (2004).
    OpenUrlCrossRefPubMedWeb of Science
  16. 16.↵
    1. G. I. Patterson,
    2. R. W. Padgett
    , TGFβ-related pathways. Roles in Caenorhabditis elegans development. Trends Genet. 16, 27–33 (2000).
    OpenUrlCrossRefPubMedWeb of Science
  17. 17.↵
    1. H. Schulenburg,
    2. C. L. Kurz,
    3. J. J. Ewbank
    , Evolution of the innate immune system: The worm perspective. Immunol. Rev. 198, 36–58 (2004).
    OpenUrlCrossRefPubMedWeb of Science
  18. 18.↵
    1. P. Huang,
    2. M. J. Stern
    , FGF signaling in flies and worms: More and more relevant to vertebrate biology. Cytokine Growth Factor Rev. 16, 151–158 (2005).
    OpenUrlCrossRefPubMedWeb of Science
  19. 19.↵
    1. N. Moghal,
    2. P. W. Sternberg
    , Multiple positive and negative regulators of signaling by the EGF-receptor. Curr. Opin. Cell Biol. 11, 190–196 (1999).
    OpenUrlCrossRefPubMedWeb of Science
  20. 20.↵
    1. H. Feng,
    2. M. Ren,
    3. L. Chen,
    4. C. S. Rubin
    , Properties, regulation, and in vivo functions of a novel protein kinase D: Caenorhabditis elegans DKF-2 links diacylglycerol second messenger to the regulation of stress responses and life span. J. Biol. Chem. 282, 31273–31288 (2007).
    OpenUrlAbstract/FREE Full Text
  21. 21.↵
    1. M. Ren,
    2. H. Feng,
    3. Y. Fu,
    4. M. Land,
    5. C. S. Rubin
    , Protein kinase D is an essential regulator of C. elegans innate immunity. Immunity 30, 521–532 (2009).
    OpenUrlCrossRefPubMedWeb of Science
  22. 22.↵
    1. R. K. Hukema,
    2. S. Rademakers,
    3. M. P. Dekkers,
    4. J. Burghoorn,
    5. G. Jansen
    , Antagonistic sensory cues generate gustatory plasticity in Caenorhabditis elegans. EMBO J. 25, 312–322 (2006).
    OpenUrlCrossRefPubMedWeb of Science
  23. 23.↵
    1. G. Jansen,
    2. D. Weinkove,
    3. R. H. Plasterk
    , The G-protein γ subunit gpc-1 of the nematode C. elegans is involved in taste adaptation. EMBO J. 21, 986–994 (2002).
    OpenUrlCrossRefPubMedWeb of Science
  24. 24.↵
    1. S. Saeki,
    2. M. Yamamoto,
    3. Y. Iino
    , Plasticity of chemotaxis revealed by paired presentation of a chemoattractant and starvation in the nematode Caenorhabditis elegans. J. Exp. Biol. 204, 1757–1764 (2001).
    OpenUrlAbstract
  25. ↵ C. elegans protein, mRNA, and cDNA are named with capital letters (DKF-2); the corresponding gene is identified with lowercase italics (dkf-2). In transgene constructs (e.g., dkf-2B::DKF-2B), promoter-enhancer DNA is identified in lowercase italics; DNA encoding protein is designated with capital letters. A double colon separates promoter from protein coding DNA.
  26. 26.↵
    1. H. Feng,
    2. M. Ren,
    3. S. L. Wu,
    4. D. H. Hall,
    5. C. S. Rubin
    , Characterization of a novel protein kinase D: Caenorhabditis elegans DKF-1 is activated by translocation-phosphorylation and regulates movement and growth in vivo. J. Biol. Chem. 281, 17801–17814 (2006).
    OpenUrlAbstract/FREE Full Text
  27. 27.↵
    1. J. T. Pierce-Shimomura,
    2. S. Faumont,
    3. M. R. Gaston,
    4. B. J. Pearson,
    5. S. R. Lockery
    , The homeobox gene lim-6 is required for distinct chemosensory representations in C. elegans. Nature 410, 694–698 (2001).
    OpenUrlCrossRefPubMedWeb of Science
  28. 28.↵
    1. M. A. Peters,
    2. T. Teramoto,
    3. J. Q. White,
    4. K. Iwasaki,
    5. E. M. Jorgensen
    , A calcium wave mediated by gap junctions coordinates a rhythmic behavior in C. elegans. Curr. Biol. 17, 1601–1608 (2007).
    OpenUrlCrossRefPubMedWeb of Science
  29. 29.↵
    1. Z. Altun-Gultekin,
    2. Y. Andachi,
    3. E. L. Tsalik,
    4. D. Pilgrim,
    5. Y. Kohara,
    6. O. Hobert
    , A regulatory cascade of three homeobox genes, ceh-10, ttx-3 and ceh-23, controls cell fate specification of a defined interneuron class in C. elegans. Development 128, 1951–1969 (2001).
    OpenUrlAbstract/FREE Full Text
  30. 30.↵
    1. H. R. Brignull,
    2. F. E. Moore,
    3. S. J. Tang,
    4. R. I. Morimoto
    , Polyglutamine proteins at the pathogenic threshold display neuron-specific aggregation in a pan-neuronal Caenorhabditis elegans model. J. Neurosci. 26, 7597–7606 (2006).
    OpenUrlAbstract/FREE Full Text
  31. 31.↵
    1. A. Islas-Trejo,
    2. M. Land,
    3. I. Tcherepanova,
    4. J. H. Freedman,
    5. C. S. Rubin
    , Structure and expression of the Caenorhabditis elegans protein kinase C2 gene. Origins and regulated expression of a family of Ca2+-activated protein kinase C isoforms. J. Biol. Chem. 272, 6629–6640 (1997).
    OpenUrlAbstract/FREE Full Text
  32. 32.↵
    1. M. Land,
    2. A. Islas-Trejo,
    3. J. H. Freedman,
    4. C. S. Rubin
    , Structure and expression of a novel, neuronal protein kinase C (PKC1B) from Caenorhabditis elegans. PKC1B is expressed selectively in neurons that receive, transmit, and process environmental signals. J. Biol. Chem. 269, 9234–9244 (1994).
    OpenUrlAbstract/FREE Full Text
  33. 33.↵
    1. Y. Tabuse,
    2. T. Sano,
    3. K. Nishiwaki,
    4. J. Miwa
    , Molecular evidence for the direct involvement of a protein kinase C in developmental and behavioural susceptibility to tumour-promoting phorbol esters in Caenorhabditis elegans. Biochem. J. 312, 69–74 (1995).
    OpenUrlAbstract/FREE Full Text
  34. 34.↵
    1. C. I. Bargmann,
    2. H. R. Horvitz
    , Chemosensory neurons with overlapping functions direct chemotaxis to multiple chemicals in C. elegans. Neuron 7, 729–742 (1991).
    OpenUrlCrossRefPubMedWeb of Science
  35. 35.↵
    1. M. Tomioka,
    2. T. Adachi,
    3. H. Suzuki,
    4. H. Kunitomo,
    5. W. R. Schafer,
    6. Y. Iino
    , The insulin/PI 3-kinase pathway regulates salt chemotaxis learning in Caenorhabditis elegans. Neuron 51, 613–625 (2006).
    OpenUrlCrossRefPubMedWeb of Science
  36. 36.↵
    1. M. Matsuki,
    2. H. Kunitomo,
    3. Y. Iino
    , Goα regulates olfactory adaptation by antagonizing Gqα-DAG signaling in Caenorhabditis elegans. Proc. Natl. Acad. Sci. U.S.A. 103, 1112–1117 (2006).
    OpenUrlAbstract/FREE Full Text
  37. 37.↵
    1. P. Sengupta
    , Generation and modulation of chemosensory behaviors in C. elegans. Pflugers Arch. 454, 721–734 (2007).
    OpenUrlCrossRefPubMedWeb of Science
  38. 38.↵
    1. J. G. White,
    2. E. Southgate,
    3. J. N. Thomson,
    4. S. Brenner
    , The structure of the nervous system of the nematode Caenorhabditis elegans. Phil. Trans. R. Soc. Lond. B Biol. Sci. 314, 1–340 (1986).
    OpenUrlAbstract/FREE Full Text
  39. 39.↵
    1. W. S. Sossin
    , Isoform specificity of protein kinase Cs in synaptic plasticity. Learn. Mem. 14, 236–246 (2007).
    OpenUrlAbstract/FREE Full Text
  40. 40.
    1. M. Leitges,
    2. J. Kovac,
    3. M. Plomann,
    4. D. J. Linden
    , A unique PDZ ligand in PKCα confers induction of cerebellar long-term synaptic depression. Neuron 44, 585–594 (2004).
    OpenUrlCrossRefPubMedWeb of Science
  41. 41.
    1. D. J. Linden
    , Neuroscience. From molecules to memory in the cerebellum. Science 301, 1682–1685 (2003).
    OpenUrlAbstract/FREE Full Text
  42. 42.↵
    1. N. Saito,
    2. Y. Shirai
    , Protein kinase Cγ (PKCγ): Function of neuron specific isotype. J. Biochem. 132, 683–687 (2002).
    OpenUrlAbstract/FREE Full Text
  43. ↵ Abbreviations for amino acids are as follows: A, Ala; D, Asp; E, Glu; F, Phe; G, Gly; H, His; K, Lys; L, Leu; N, Asn; P, Pro; Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; and Y, Tyr.
  44. 44.↵
    1. S. L. Wu,
    2. J. Staudinger,
    3. E. N. Olson,
    4. C. S. Rubin
    , Structure, expression, and properties of an atypical protein kinase C (PKC3) from Caenorhabditis elegans. PKC3 is required for the normal progression of embryogenesis and viability of the organism. J. Biol. Chem. 273, 1130–1143 (1998).
    OpenUrlAbstract/FREE Full Text
  45. 45.↵
    1. A. H. Hirsch,
    2. S. B. Glantz,
    3. Y. Li,
    4. Y. You,
    5. C. S. Rubin
    , Cloning and expression of an intron-less gene for AKAP 75, an anchor protein for the regulatory subunit of cAMP-dependent protein kinase IIβ. J. Biol. Chem. 267, 2131–2134 (1992).
    OpenUrlAbstract/FREE Full Text
  46. 46.↵
    1. C. Ndubuka,
    2. Y. Li,
    3. C. S. Rubin
    , Expression of a kinase anchor protein 75 depletes type II cAMP-dependent protein kinases from the cytoplasm and sequesters the kinases in a particulate pool. J. Biol. Chem. 268, 7621–7624 (1993).
    OpenUrlAbstract/FREE Full Text
  47. 47.↵
    1. A. Fire,
    2. S. W. Harrison,
    3. D. Dixon
    , A modular set of lacZ fusion vectors for studying gene expression in Caenorhabditis elegans. Gene 93, 189–198 (1990).
    OpenUrlCrossRefPubMedWeb of Science
  48. 48.↵
    1. M. R. Lackner,
    2. S. J. Nurrish,
    3. J. M. Kaplan
    , Facilitation of synaptic transmission by EGL-30 Gqα and EGL-8 PLCβ: DAG binding to UNC-13 is required to stimulate acetylcholine release. Neuron 24, 335–346 (1999).
    OpenUrlCrossRefPubMedWeb of Science
  49. 49.↵
    1. J. H. Freedman,
    2. L. W. Slice,
    3. D. Dixon,
    4. A. Fire,
    5. C. S. Rubin
    , The novel metallothionein genes of Caenorhabditis elegans. Structural organization and inducible, cell-specific expression. J. Biol. Chem. 268, 2554–2564 (1993).
    OpenUrlAbstract/FREE Full Text
  50. 50.↵
    1. Y. Li,
    2. C. Ndubuka,
    3. C. S. Rubin
    , A kinase anchor protein 75 targets regulatory (RII) subunits of cAMP-dependent protein kinase II to the cortical actin cytoskeleton in non-neuronal cells. J. Biol. Chem. 271, 16862–16869 (1996).
    OpenUrlAbstract/FREE Full Text
  51. 51.↵
    1. H. Feng,
    2. M. Ren,
    3. C. S. Rubin
    , Conserved domains subserve novel mechanisms and functions in DKF-1, a Caenorhabditis elegans protein kinase D. J. Biol. Chem. 281, 17815–17826 (2006).
    OpenUrlAbstract/FREE Full Text
View Abstract

Article Tools

  • Email
  • Download Powerpoint
  • Print

  • Alerts
  • Citation tools

  • Neuronal and Intestinal Protein Kinase D Isoforms Mediate Na+ (Salt Taste)–Induced Learning

    By Ya Fu, Min Ren, Hui Feng, Lu Chen, Zeynep F. Altun, Charles S. Rubin

    Science Signaling : ra42

    Protein kinase D signaling in the intestine and nervous system enables worms to learn to avoid salt.

    CiteULike logo Connotea logo del.icio.us logo Digg logo Facebook logo Google logo Mendeley logo Reddit logo Twitter logo

Related Content

Similar Articles in:

Citing Articles in:

Navigate This Article