AJP - Heart Fuel your research with LabChart
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Am J Physiol Heart Circ Physiol 289: H1941-H1950, 2005. First published June 17, 2005; doi:10.1152/ajpheart.01111.2004
0363-6135/05 $8.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
289/5/H1941    most recent
01111.2004v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (4)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Robia, S. L.
Right arrow Articles by Walker, J. W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Robia, S. L.
Right arrow Articles by Walker, J. W.

Novel determinant of PKC-{epsilon} anchoring at cardiac Z-lines

Seth L. Robia, Misuk Kang, and Jeffery W. Walker

Department of Physiology, University of Wisconsin, Madison, Wisconsin

Submitted 2 November 2004 ; accepted in final form 7 June 2005


   ABSTRACT
TOP
 ABSTRACT
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
The Z-line represents a critical link between the transverse tubule network and cytoskeleton of cardiac cells with a role in anchoring structural proteins, ion channels, and signaling molecules. Protein kinase C-{epsilon} (PKC-{epsilon}) regulates cardiac excitability, cardioprotection, and growth, possibly as a consequence of translocation to the Z-line/T tubule region. To investigate the mechanism of PKC-{epsilon} translocation, fragments of its NH2-terminal 144-amino acid variable domain, {epsilon}V1, were fused with green fluorescent protein and evaluated by quantitative Fourier image analysis of decorated myocytes. Deletion of 23 amino acids from the NH2-terminus of {epsilon}V1, including an EAVSLKPT motif important for binding to a receptor for activated C kinase (RACK2), reduced but did not abolish Z-line binding. Further deletions of up to 84 amino acids from the NH2-terminus of {epsilon}V1 also did not prevent Z-line decoration. However, deletions of residues 85–144 from the COOH-terminus strongly reduced Z-line binding. COOH-terminal deletions caused 2.5-fold greater loss of binding energy ({Delta}{Delta}G) than did NH2-terminal deletions. Synthetic peptides derived from these regions modulated {epsilon}V1 binding and cardiac myocyte function, but also revealed considerable heterogeneity within populations of adult cardiac myocytes. The COOH-terminal subdomain important for Z-line anchoring maps to a surface in the {epsilon}V1 crystal structure that complements the eight-amino acid RACK2 binding site and two previously identified membrane docking motifs. PKC-{epsilon} anchoring at the cardiac Z-line/T tubule appears to rely on multiple points of contact probably involving protein-lipid and protein-protein interactions.

protein kinase C-{epsilon}

THE PROTEIN KINASE C (PKC) family of serine-threonine kinases is taxonomically divided into three isozyme categories based on differential regulation by various cofactors (28, 30). In the heart, PKC mediates signaling from several G protein-coupled receptors, including {alpha}1-adrenergic, endothelin, and ANG II receptors (6). PKC substrate specificity has been determined in vitro to be quite general, both in terms of the wide variety of protein substrates phosphorylated and the low degree of isoform selectivity of a particular substrate (28, 29). This belies the apparent need for a diversity of isoforms and suggests that substrate specificity in vivo arises from other recognition mechanisms. In this regard, much attention has been focused upon "translocation," the physical redistribution of PKC to specific locations within the cell in response to second messenger activators such as Ca2+ and/or diacylglycerol (21). Despite dynamic reversibility, translocation and anchoring significantly reduce the apparent diffusion coefficient of PKC (42), and may help confer specificity by constraining PKC to act on substrates within defined regions of the cell. The molecular mechanisms underlying this process have not been fully elucidated.

This study focused on PKC-{epsilon}, one of the calcium-insensitive novel isoforms, which has been specifically implicated in L-type channel regulation (18), regulation of thin filament Ca2+ sensitivity (11, 41), and ischemic preconditioning (12, 23, 34, 35). PKC-{epsilon} has been shown to bind through its first variable domain ({epsilon}V1) to the coatomer protein {beta}'COP [also designated receptor for activated C kinase (RACK2)] when activated by phosphatidylserine and diacylglycerol (4). PKC-{epsilon}-RACK2 binding, and thus kinase translocation, were proposed to be mediated by a discrete 8-residue RACK-binding sequence, EAVSLKPT, near the NH2-terminus of this {epsilon}V1 domain (7).

The crystal structure of the {epsilon}V1 domain was recently solved by Ochoa et al. (32), who proposed a different mode of {epsilon}V1-mediated anchoring of PKC-{epsilon} to membranes. In their model, key hydrophobic and electrostatic interactions occur between surface loops 1 and 3 of {epsilon}V1 and the phospholipid membrane (3, 24, 32). Mg2+ is also hypothesized to play a critical role in coordinating charged residues at the binding interface. This model is more analogous to Ca2+-dependent membrane binding of the homologous C-2 domain of classical PKCs (30) and contrasts with the RACK model in its supposition of a distributed binding interface, rather than a discrete eight-amino acid binding site.

We have previously described decoration of saponin-permeabilized cardiac myocytes with PKC-{epsilon} green fluorescent protein (GFP) fusion constructs as an assay for investigating the localization and kinetics of PKC-{epsilon} anchoring (36). This method provides the ability to simultaneously measure dynamic anchoring properties while observing the spatial distribution of the anchoring sites, all under conditions of defined solution composition, pH, ionic strength, and probe concentration. Using this assay, we have now evaluated the relative binding efficacy of a series of deletion mutants of the {epsilon}V1 regulatory domain of PKC-{epsilon} to determine contributions of specific sequences to anchoring. The constructs were designed to explicitly test the role of previously identified motifs in anchoring of PKC-{epsilon} in cardiac myocytes and therefore contained various combinations of the eight-amino acid RACK binding site and membrane docking loops 1 and 3. The data reveal a new segment of PKC-{epsilon} important for Z-line anchoring in cardiac myocytes. Complementary experiments with synthetic peptides were generally consistent with the deletion analysis and provided insight into the physiological role of PKC-{epsilon} translocation in myocytes. Overall, the data suggest important refinements to existing models of PKC-{epsilon} anchoring and translocation, which are discussed in the context of the crystal structure of the {epsilon}V1 regulatory domain.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Molecular biology and protein expression. Amplification of the desired fragments of {epsilon}V1 was accomplished with PCR. Oligonucleotides flanking the desired regions of the {epsilon}V1 domain were designed with integral restriction sites for fragment subcloning. Forward-oriented oligonucleotides were engineered with 5' BamHI sites, and reverse oligonucleotides were engineered with 5' XhoI sites. The oligonucleotide sequences for forward and reverse primers used for PCR cloning were as follows: for fragment 1–144: 5'-ggggatccgatggtagtgttcaatggcctg-3' and 5'-ggctcgagtctcttcattgtctttaggtgc-3'; for fragment 24–144: 5'-ggggatccgtcgctgcgccatgcggtggga-3' and 5'-ggctcgagtctcttcattgtctttaggtgc-3'; for fragment 1–84: 5'-ggggatccgatggtagtgttcaatggcctg-3' and 5'-ggctcgactaaagacggccagctcgatctt-3'; for fragment 85–144: 5'-ggggatccgcacgatgcccccatcggctac-3' and 5'-ggctcgagtctcttcattgtctttaggtgc-3'; for fragment 94–144: 5'-ggggatccgttcgtggccaactgcaccatc-3' and 5'-ggctcgagtctcttcattgtctttaggtgc-3'; for fragment 14–93: 5'-ggggatccggaggccgtgagcttgaagccc-3' and 5'-ggctcgagtgtcgtcgtagccgatgggggc-3'. Custom synthetic DNA was purchased from Operon (Alameda, CA). The gene encoding enhanced green fluorescent protein (EGFP) was excised from phosphorylated EGFP (Clontech) and inserted in the histidine-tag vector pTrcHisB (Invitrogen) out of frame to prevent EGFP expression. To engineer {epsilon}V1 fragment protein constructs with EGFP fused to their COOH-termini, cDNA sequences encoding {epsilon}V1 truncation mutants were inserted in this vector at the 5'-end of the EGFP start codon. This restored the proper reading frame and founded fluorescent green colonies, which facilitated screening. Histidine-tagged GFP-fusion constructs were expressed in DH5{alpha} cells and purified with a Ni2+-nitriloacetic acid (NTA) column (Qiagen, Valencia, CA) as described (36). Concentrations of purified GFP-fusion proteins were determined by fluorimetric comparison with standard preparations of GFP. In addition, purified fusion proteins were subjected to 12% PAGE, stained with Coomassie blue, and quantified by visual comparison with BSA standards.

Identities and purities of fusion protein bands were further confirmed by staining with InVision His-tag protein stain and by running partially denaturing gels that permitted GFP fluorescence of gel bands to be monitored. Preparations were further purified or discarded if quantification by fluorimetry and PAGE gave mismatched concentration estimates.

Preparation of unlabeled {epsilon}V1. {epsilon}V1 was subcloned into pGEX-4T-1 (Amersham Biosciences, Piscataway, NJ) and expressed as a GST fusion protein in Escherichia coli. The protein was purified with glutathione-Sepharose 4B and cleaved from the support with thrombin, according to the manufacturer's instructions. Purity and concentration were estimated by SDS-PAGE.

Alexa Fluor-568 succinimidyl ester labeling of {epsilon}V1. Unlabeled {epsilon}V1 was dialyzed for 3 h against PBS, pH 8.1, to remove Tris and dithiothreitol before labeling. Lysine-reactive Alexa Fluor-568 succinimidyl ester (Molecular Probes, Eugene, OR) was dissolved in dimethyl sulfoxide at 10 mg/ml and added to dialyzed {epsilon}V1 solution at a dye-to-protein weight ratio of 1:7.5. The labeling reaction proceeded for 2 h at room temperature. Unincorporated dye was removed using a Sephadex G50 spin column (Amersham Biosciences). Spectrophotometric analysis of product indicated a protein-to-dye molar ratio of 25:1.

Electron microscopy. Adult rat ventricular myocytes were prepared as described (15), pelleted at 800 g for 45 s, and fixed with 4% paraformaldehyde and 0.5% glutaraldehyde. Additional sample preparation and imaging were performed per standard protocols at the University of Wisconsin Electron Microscopy Facility.

Myocyte decoration. Adult rat ventricular myocytes were permeabilized with 100 µg/ml saponin in relaxing solution (in mM: 100 KCl, 1 MgCl2, 2 EGTA, 4.5 ATP, 10 imidazole, and 1 dithiothreitol, pH 7.0) as described (36). In some cases, cells were permeabilized in fresh 0.5% Triton X-100 (Pierce Biotechnology, Rockford, IL). Permeabilized myocytes were washed extensively in Relaxing Solution and blocked with 2% BSA to prevent nonspecific binding. {epsilon}V1 fragment GFP protein constructs were applied at a final concentration of 200–800 nM, and Alexa-568-labeled {epsilon}V1 was applied at 90 nM. Confocal microscopy was performed using 488 nm illumination and 522 nm acquisition for GFP-fusion proteins and 568 nm illumination and 605 nm acquisition for Alexa-568-labeled {epsilon}V1. The intensity of striations was quantified by two-dimensional fast-Fourier transform (FFT) analysis using the public domain software Image J or a custom script that runs in Origin (Microcal Software, Northampton, MA). First-order peaks in the transformed power spectrum real image were manually circumscribed, and the peak volume was integrated for each image. Integrated volumes were normalized to cell sectional area. Results were similar to those obtained with lineout peak height analysis (36). Striation pattern intensities for fragments of {epsilon}V1 were normalized to the intensity obtained for an equal concentration of full-length {epsilon}V1-GFP-(1–144) on the same experimental day.

Reversibility of binding was tested by dilution of decorated myocytes with buffer and observing the loss of myocyte fluorescence, as well as by measuring the recovery of fluorescence after regional photobleaching (36). For peptide-blocking experiments, permeabilized myocytes were preincubated with 100 µM EAVSLKPT or 4 µM HDAPIGYD for 15 min at room temperature and then decorated with 200 nM {epsilon}V1-GFP.

Myocyte twitch measurements. Adult rat ventricular myocytes were evaluated within 6 h after enzymatic isolation as described (19). Briefly, myocytes were subjected to electrical field stimulation at 0.5 Hz, 40 volts, and 21–23°C in 1 mM Ca2+ Ringer solution (in mM: 125 NaCl, 5 KCl, 2 NaH2PO4, 5 sodium pyruvate, 1.2 MgSO4, 11 glucose, 0.5 CaCl2, and 25 HEPES, pH 7.4) after settling on the glass floor of a custom perfusion chamber. Twitches were digitized at 30 Hz using a Video Edge detector (Crescent Electronics, Sandy, UT). Peptides were dissolved at a final concentration of 500 nM in 1 mM Ca2+ Ringer and perfused in the stimulation chamber at 0.4–0.5 ml/min. Peptides with an NH2-terminal cysteine (side chain protecting group: 3-nitro 2-pyridinesulfenyl) were prepared on an ABI 432a solid-phase synthesizer with 9-fluorenylmethoxycarbonyl (FMOC) amino acids and a Wang resin. Peptides were conjugated to antennapedia peptide (5) containing an unprotected NH2-terminal cysteine by admixture in a 1:2 molar ratio in water under N2 gas for 24 h. The disulfide-conjugated peptides were purified by reverse-phase HPLC using a BioCAD Vision system (PerSeptive Systems), and peptide identities were confirmed by MALDI-TOF mass spectrometry.


   RESULTS
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 GRANTS
 REFERENCES
 
Myocyte decoration with {epsilon}V1-GFP. Decoration of saponin-skinned rat ventricular myocytes with {epsilon}V1-GFP resulted in a pronounced striated staining pattern that colocalized with the Z-line of the sarcomere (Fig. 1A and Ref. 36). This striation pattern was further characterized here with an emphasis on demonstrating specificity of {epsilon}V1 binding. {delta}V1-GFP, an analogous GFP construct derived from PKC-{delta} (rather than {epsilon}), gave no striation staining, suggesting that the interaction between {epsilon}V1 and the Z-line of myocytes was isoform specific (Fig. 1B). A fluorescent {epsilon}V1 construct was prepared without the use of GFP by covalent modification of lysine side chains with Alexa-568. This fluorescent {epsilon}V1 construct gave rise to a similar striation pattern to {epsilon}V1-GFP (Fig. 1C), indicating that COOH-terminal GFP did not contribute significantly to binding. The Alexa-568 probe complements {epsilon}V1-GFP by being labeled in a completely distinct manner and location, but also has advantages over GFP, including brighter fluorescence and an emission that is more spectrally separated from tissue autofluorescence. Specificity in {epsilon}V1 binding was supported by the observation that {epsilon}V1-GFP and {epsilon}V1-Alexa-568 bound reversibly, as demonstrated by both washout and fluorescence recovery after photobleaching (FRAP) experiments (see below).



View larger version (121K):
[in this window]
[in a new window]
  		Fig. 1. Representative confocal images of isolated permeabilized myocytes decorated with fluorescent PKC domains. A: 1 µM first variable domain of PKC-{epsilon} ({epsilon}V1)-green fluorescent protein (GFP). B: 1 µM {delta}V1-GFP comprising residues 1–130 from the NH2-terminal variable domain of PKC-{delta}. C: 90 nM Alexa-568-labeled {epsilon}V1. Myocytes were permeabilized with 100 µg/ml saponin. Scale bars = 10 µm.

 
The {epsilon}V1 domain has previously been shown to bind to phospholipid bilayers (3, 32), so we treated myocytes with Triton X-100 to remove cell membranes. Figure 2 shows a comparison of {epsilon}V1-GFP staining of saponin-skinned myocytes vs. Triton X-100-skinned myocytes. Saponin is thought to permeabilize membranes (17) by reacting with cholesterol to form 12- to 60-nm pores (39) with minimal disruption of plasma membrane ultrastructure or internal membrane integrity, whereas Triton X-100 causes large-scale membrane solubilization (effects on myocyte ultrastructure are shown in Fig. 3 and are described below). Despite this pronounced difference in the two skinning procedures, {epsilon}V1-GFP stains both types of skinned myocytes with a marked striation pattern (Fig. 2, A and B). In saponin-permeabilized myocytes, two other types of staining were observed, particularly at high concentrations of fluorescent {epsilon}V1. One we refer to as longitudinal streaking, which may correspond to labeling of intracellular membranous organelles (e.g., Figs. 1A and 2A). These longitudinal streaks somewhat confounded the regular sarcomeric striations at the highest probe concentrations and placed a practical upper limit on concentrations used for quantitative analyses. The other type of staining was at the intercalated disc, which was particularly prominent with the Alexa-568-{epsilon}V1 probe (Fig. 1C). Triton X-100-skinned myocytes revealed decoration of sarcomeres only at the Z-line, without detectable longitudinal streaks or intercalated disc staining (Fig. 2B).



View larger version (147K):
[in this window]
[in a new window]
  		Fig. 2. Comparison of myocyte decoration patterns with two skinning procedures. A: myocyte skinned with 100 µg/ml saponin and stained with 400 nM {epsilon}V1-GFP. B: myocyte skinned with 0.5% Triton X-100 and stained with 400 nM {epsilon}V1-GFP. Because of low contrast in this image of Triton-skinned myocytes, the myocyte ends are indicated by arrows. Scale bars = 10 µm.

 


View larger version (125K):
[in this window]
[in a new window]
  		Fig. 3. Electron micrographs of isolated myocytes subjected to various skinning methods. A: intact (not skinned) myocyte. B: myocyte skinned with 100 µg/ml saponin. C: myocyte skinned with 150 µg/ml saponin. D: myocyte skinned with 0.5% Triton X-100. Black and white arrowheads indicate T tubules in cross section; Z, Z-line. Scale bars = 1 µm.

 
Electron micrographs of skinned myocytes showed near-intact mitochondria and T tubules after 100 µg/ml saponin skinning (Fig. 3B), with a trend toward more disruption at 150 µg/ml saponin such as larger-diameter T tubules (Fig. 3C). In contrast, skinning with 0.5% Triton X-100 caused large-scale solubilization of membranes and virtually eliminated all membranes and organelles (Fig. 3D). The presence of a striated binding pattern with both skinning procedures indicated that Triton-insoluble cytoskeletal structures play a major role in PKC-{epsilon} anchoring near the cardiac Z-lines.

An important test of specific {epsilon}V1 binding is to demonstrate competition by unlabeled probes. Preincubation with an unlabeled form of {epsilon}V1 (containing no GFP or Alexa-568) reduced striated staining of both saponin-skinned and Triton X-100-skinned myocytes (Fig. 4), consistent with competition between {epsilon}V1 and {epsilon}V1-GFP for the same sites. This reduction in staining was variable from myocyte to myocyte and was typically incomplete (Fig. 4, summarized in Table 1). It is likely that a minor component of {epsilon}V1-GFP binding to the Z-line/T tubule region may involve association in a nonsaturable, nonblockable manner with phospholipid membranes.



View larger version (103K):
[in this window]
[in a new window]
  		Fig. 4. Competition between {epsilon}V1-GFP and unlabeled {epsilon}V1. A and B: saponin-permeabilized myocytes stained with 800 nM {epsilon}V1-GFP for 15 min. C and D: Triton-skinned myocytes stained with 800 nM {epsilon}V1-GFP for 10 min. In B and D, myoyctes were preincubated with 4 µM unlabeled {epsilon}V1 for 15 min before {epsilon}V1-GFP. Scale bars = 10 µm.

 

View this table:
[in this window]
[in a new window]
  		Table 1. Effects of translocation inhibitor and activator peptides on {varepsilon}V1-GFP binding

 
Myocyte decoration with {epsilon}V1 fragments. A systematic deletion strategy was used to define regions within the 144-amino acid domain of {epsilon}V1 responsible for striation staining of myocytes. A schematic diagram of deletion mutants and a structural representation showing areas of {epsilon}V1 affected by truncations are shown in Fig. 5. The goal was to create constructs lacking one or more subdomains previously proposed to mediate protein or membrane anchoring of PKC-{epsilon}, including EAVSLKPT (residues 14–21), HDAPIGYD (residues 85–92; see Ref. 7), and membrane docking loops 1 and 3 (32). The probes were expressed in Escherichia coli and purified to >75% purity as assessed by SDS-PAGE with Coomassie staining (Fig. 5C). In partially denaturing gels, GFP fluorescence was restricted to a single fluorescent band in each lane (data not shown), indicating that minor contaminating bands were nonfluorescent species. InVision His-tag protein stain also identified a single major band at the expected molecular weight for each construct (data not shown). Only construct 94–144 was difficult to express and purify in large quantities, possibly because of poor folding. Constructs were administered over a range of concentrations to saponin- and Triton X-100-skinned myocytes to identify regions of {epsilon}V1 necessary for binding.



View larger version (39K):
[in this window]
[in a new window]
  		Fig. 5. Schematic of deletion strategy. A: linear representation of subdomains of {epsilon}V1 (labeled 1–144) and truncated fragments tested for Z-line binding. B: {epsilon}V1 rendered in public domain software MolMol from the crystal structure by Ochoa et al. (32). Regions of {epsilon}V1 highlighted include the receptor for activated C kinase (RACK2) binding sequence [14-EAVSLKPT-21 (red)], loop 1 (residues 17–30; red and some gray), and the pseudoRACK sequence [85-HDAPIGYD-92 (green); equivalent to loop 3]. C: SDS-PAGE of purified GFP constructs demonstrating purities of >75% [except for GFP-(94–144) in the penultimate lane]. N, NH2 terminal; C, COOH terminal. Relative molecular weight based on molecular weight markers are on left.

 
Confocal images showing the effects of deletions on the nature and intensity of staining are shown in Fig. 6. Unexpectedly, deleting the RACK-binding sequence 14-EAVSLKPT-21 did not abolish the striated decoration of saponin-permeabilized rat cardiac myocytes. Construct GFP-(24–144) still labeled myocytes in a striated pattern, although the intensity of the striations was reduced (Fig. 6B). Also surprising was the observation that construct GFP-(1–84) showed only a low level of binding (Fig. 6C) even though it contained the putative 14-EAVSLKPT-21 RACK binding sequence and was lacking the autoinhibitory pseudoRACK sequence 85-HDAPIGYD-92. Similar results were obtained with Triton X-100-skinned myocytes (data not shown).



View larger version (111K):
[in this window]
[in a new window]
  		Fig. 6. Confocal micrographs of {epsilon}V1 fusion proteins applied to saponin-permeabilized myocytes. A: full-length {epsilon}V1-GFP [GFP-(1–144)]. B: construct GFP-(24–144) lacking the RACK2-binding 14-EAVSLKPT-21 sequence. C: construct GFP-(1–84). D: construct GFP-(85–144). E: construct GFP-(94–144). D: construct GFP-(14–93), which retained the RACK-binding sequence, the 85-HDAPIGYD-92 pseudoRACK sequence (equivalent to loop 3), and loop 1. All constructs were applied at a final concentration of 200–400 nM.

 
One rationalization of these results is that high-affinity PKC-{epsilon} anchoring might require the presence of both EAVSLKPT and HDAPIGYD sequences. This was tested by studying fragments 85–144 and 94–144, which lacked one or both of these sequences. {epsilon}V1 fragments missing 14-EAVSLKPT-21 (Fig. 6D) or both 14-EAVSLKPT-21 and 85-HDAPIGYD-92 (Fig. 6E) bound in striated distributions with similar intensity. This observation demonstrates that even a peptide segment (i.e., 94–144) lacking both EAVSLKPT and HDAPIGYD sequences displayed significant decoration at the Z-line and directed our attention to this COOH-terminal aspect of {epsilon}V1. However, poor expression limited the usefulness of the 94–144 construct (Fig. 5C). A complementary construct containing both 14-EAVSLKPT-21 and 85-HDAPIGYD-92 segments but lacking the COOH-terminus, namely fragment 14–93, expressed very well but bound most poorly of all GFP constructs tested. The GFP-(14–93) fragment appeared to preferentially decorate myocytes in longitudinal streaks (Fig. 6F), a pattern indicative of general membrane localization. Other GFP constructs, but not GFP itself (36), also showed decoration of longitudinal membrane structures when applied at high concentration. At lower concentrations, however, the clear striated pattern was the dominant visible feature for all constructs except fragment GFP-(14–93).

FFT analysis of striation patterns. To facilitate quantification of staining intensities for GFP constructs, we performed two-dimensional (2D)-FFT analysis of images of the type shown in Fig. 6. A quantitative comparison of striation intensities of all probes analyzed by 2D-FFT image analysis is given in Fig. 7A. The regular sarcomeric autofluorescence pattern of cardiac myocytes results in a background signal in the power spectrum at the same sarcomere length as the additional fluorescence signal of bound probe. This autofluorescence background places a lower limit on the detection of low-intensity signals, and is listed in Fig. 7A as "no probe." A statistical analysis of differences was carried out by a Student's t-test with significance taken as P < 0.05. In general, the data showed that the full-length probe bound more effectively than any of the truncated constructs, whereas GFP-(14–93) bound most poorly. Other probes gave intermediate quantitative binding values consistent with visual impressions of the intensities of the striated patterns in the image data. Binding of fragments of {epsilon}V1-GFP did not appear to be hampered by nonspecific binding at concentrations up to 500 nM, as striation patterns were readily and completely reversed by washout or by FRAP [with the exception of GFP-(14–93), see below]. Integrated first-order peak volumes associated with the different fragments were not systematically correlated to fragment length (R2 = 0.61; Fig. 7B), further arguing against nonspecific binding.



View larger version (32K):
[in this window]
[in a new window]
  		Fig. 7. Quantification of striation patterns by 2-dimensional (2D)-fast-Fourier transform (FFT) analysis. A: bar graph of striation intensity for each construct compared with the full-length {epsilon}V1-GFP domain. Data represent means ± SE (no. of measurements indicated in bars) of normalized peak volume obtained from the first-order peak of the transformed power spectrum (* in inset). Probe GFP-(1–144) was statistically different from all others, and all probes except 14–93 were different from "no probe" (P < 0.05). B: intensities of striated patterns plotted vs. length of the fragment (no. of residues not including GFP). Correlation coefficient for the fitted line is R2 = 0.61. C: concentration dependencies for binding of 3 truncated {epsilon}V1-GFP constructs to saponin-permeabilized myocytes [GFP-(24–144) ({triangleup}), GFP-(85–144) ({square}), and GFP-(1–84) ({lozenge})]. Binding is normalized and scaled to the full-length {epsilon}V1 domain ({circ}) included in each experiment for that purpose.

 
Estimates of affinity and binding energy. The observed differences in binding of the various fragments held across a range of concentrations. Figure 7C shows the concentration dependencies for the NH2-terminal fragment [GFP-(1–84)] and two COOH-terminal fragments [GFP-(21–144) and GFP-(85–144)]. The values shown are means ± SE minus autofluorescence background. Because the experiments were performed at different microscope settings to maintain quantitative linearity, each data point was normalized to a reference measurement of {epsilon}V1-GFP on the same experimental day. Data from all experiments were then scaled to a previously determined {epsilon}V1-GFP binding isotherm (36) and plotted as in Fig. 7C. Hyperbolic fits of the form y = ax/(b + x), where "a" was constrained to maximal GFP-(1–144) binding, gave estimates of dissociation constant (Kd) as follows: 230 nM for GFP-(1–144); 750 nM for GFP-(24–144); 880 nM for GFP-(85–144); and 4.8 µM for GFP-(1–84). Thus the data suggest that the COOH-terminal fragments GFP-(24–144) and GFP-(85–144) possess ~3-fold lower affinity than the full-length {epsilon}V1 domain, whereas the NH2-terminal GFP-(1–84) is ~20-fold lower in affinity than full length. These changes in affinity were translated into changes in free energy using the following relationship: {Delta}{Delta}G = –RT ln(Kd1/Kd2), where R is the universal gas constant, T is temperature, Kd1 is the dissociation constant for GFP-(1–144), and Kd2 is the dissociation constant for the test fragment. The losses of binding energy, or {Delta}{Delta}Gs, resulting from various deletions were estimated to be 1.2 RT for deletion of NH2-terminal sequences up to residue 23 vs. 3.0 RT for deletion of COOH-terminal sequences beyond residue 85. Thus, in this myocyte decoration assay, elimination of COOH-terminal segments of the {epsilon}V1 domain had a greater impact on affinity (by 7-fold) and on binding energy (by 2.5-fold) than did elimination of NH2-terminal segments.

Reversibility. Z-line decoration by GFP constructs was generally reversible, with complete washout occurring within several minutes (data not shown). Moreover, dilution of equilibrated probes caused relaxation to a new steady state on a time scale compatible with previously determined kinetics (half-time = 2–5 min; Ref 36). The one {epsilon}V1 fragment that failed to demonstrate reversibility was GFP-(14–93). At high concentrations, this probe showed conspicuous general membrane binding, little Z-line decoration, and incomplete washout, even after 30 min. Otherwise, all GFP-tagged {epsilon}V1 fragments washed out of myocytes fully over the course of 5–10 min.

Peptide modulators of PKC translocation. Competition studies were performed with synthetic peptides widely used to block or stimulate PKC-{epsilon} translocation (7). The inhibitory peptide EAVSLKPT only modestly reduced binding of full-length {epsilon}V1-GFP, even at the relatively high concentration of 100 µM peptide. Cell-to-cell variability prevented this effect from achieving statistical significance (summarized in Table 1). This may be an indication that membrane binding contributes significantly to {epsilon}V1 binding in this system or that EAVSLKPT represents only a portion of the binding interface, as suggested by the deletion analysis described above. Interestingly, the loop 3 peptide HDAPIGYD also did not reduce binding but tended to enhance the intensity of {epsilon}V1-GFP striation staining, consistent with the notion that this peptide may stimulate PKC-{epsilon} translocation and anchoring under some conditions (7). Cell-to-cell variability was also large for these experiments (Table 1). These peptide competition experiments generally supported the conclusions of the deletion analysis that neither 14-EAVSLKPT-21 nor 85-HDAPIGYD-92 behaved as exclusive anchoring interfaces for PKC-{epsilon} binding to cardiac Z-lines.

Functional studies with these peptide modulators provided insight into a possible cause of cell-to-cell variability. Exposure of isolated rat ventricular myocytes to a cell-permeant form of EAVSLKPT (at 500 nM) was without effect on myocyte twitches (data not shown), whereas a cell-permeant form of HDAPIGYD (at 500 nM) gave rise to a strong enhancement of the myocyte twitch amplitude and a prolongation of the twitch time course (Fig. 8). This observation is consistent with activation and translocation of PKC-{epsilon} by HDAPIGYD in this cell system (19). It is important to note that only 35% of myocytes tested showed this response to HDAPIGYD, with the remainder showing normal twitches and no inotropic or lusitropic effects. This observation suggests that the myocyte populations under investigation are more heterogeneous than originally recognized (possibly because of myocytes arising from distinct epicardial, endocardial, or other regions).



View larger version (14K):
[in this window]
[in a new window]
  		Fig. 8. Functional effects of HDAPIGYD (loop 3) peptide on cardiac myocyte twitches. Freshly isolated adult rat cardiac myocytes were electrically paced at 0.5 Hz at room temperature in 1 mM Ca2+ Ringer solution. The stimulation chamber was perfused with 500 nM Cys-HDAPIGYD conjugated to Cys-antennapedia peptide (5) through a disulfide bridge (Cys-Cys, see MATERIALS AND METHODS). Graph shows the average of 10 twitches in the same myocyte before (a) and 10 min after (b) perfusion with the peptide conjugate. A positive inotropic response was observed in 14 out of 40 myocytes (35%), with mean ± SE increases of 42 ± 13% in twitch amplitude and 19 ± 6% in twitch duration. The 26 nonresponding cells showed no change in twitch amplitude or time course.

 

   DISCUSSION
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
GRANTS
 REFERENCES
 
Two models have been advanced to account for anchoring of PKC-{epsilon} to membranes and other cellular structures via its unique 144-amino acid {epsilon}V1 domain. The work of Dorn and Mochly-Rosen (7) identified an eight-amino acid sequence within the {epsilon}V1 domain that interacted with a cloned RACK2. This sequence [EAVSLKPT at residues 14–21 of {epsilon}V1 (see Fig. 5 for location of residues in a ribbon structure)], was hypothesized to be the binding interface responsible for PKC-{epsilon} translocation and anchoring. The work of Mochley-Rosen and coworkers (7, 8, 26, 38) also provided evidence for the existence of an intramolecular interaction with another sequence (HDAPIGYD at residues 85–92) that sequestered the anchoring interface when PKC-{epsilon} was inactive. From this one might predict that deletion of 14-EAVSLKPT-21 would eliminate {epsilon}V1 binding, whereas deletion of 85-HDAPIGYD-92 would activate binding by relieving autoinhibition.

An alternative model has been proposed in conjunction with a recent report of the crystal structure of the {epsilon}V1 domain (3, 32). Several charged and hydrophobic amino acids were identified in the structure that could participate in binding of {epsilon}V1 to phospholipid membranes (3, 24, 32). In this model, loop 1 (residues 17–30) and loop 3 (residues 85–92), which are well separated in the {epsilon}V1 primary sequence, cooperate in a manner analogous to membrane binding of conventional Ca2+-sensitive C-2 domains. Although the {epsilon} isoform does not bind Ca2+, a coordinated Mg2+ may serve a similar purpose of mediating interactions with charged phospholipid head groups. If this mechanism was responsible for the observed {epsilon}V1 binding in myocytes, deletions affecting loops 1 and 3 should reduce myocyte decoration.

Taken together, the anchoring models of Mochly-Rosen et al. (26) and Ochoa et al. (32) focus primarily on two regions of the NH2-terminus of the variable domain of PKC-{epsilon} (Fig. 5 for a ribbon structure). First, loop 1, comprised of amino acids 17–30, contains most of the putative RACK2 interaction interface, 14-EAVSLKPT-21, and putative membrane docking residues within 22–30. Second, amino acids 85–92 represent both the pseudoRACK sequence (85-HDYPIGYD-92) and additional membrane docking residues (loop 3). Here, we tested these models by making NH2- and COOH-terminal deletions in the {epsilon}V1 variable domain and then quantifying their ability to bind Z-lines of cardiac myocytes.

Several important conclusions emerge from the data. First, the discrete RACK-binding sequence 14-EAVSLKPT-21 identified by Johnson et al. (16) is not likely to be the sole mediator of binding in this experimental context. Deletion of this sequence in the probe GFP-(24–144) significantly reduced but did not abolish binding. Second, dissection of {epsilon}V1 into two complementary subdomains, an NH2-terminal [GFP-(1–84)] and a COOH-terminal (85–144-GFP) fragment, revealed that both NH2- and COOH-terminal fragments decorated cardiac cells in a striated pattern, albeit less intensely than when they were integrated as full-length GFP-(1–144). It seems, therefore, that rather than being mediated by a small, discrete binding sequence, the anchoring interaction may involve multiple binding regions distributed along the {epsilon}V1 primary structure.

Such a "distributed interface" for {epsilon}V1 binding is compatible with the Ochoa et al. model involving surface loops 1 and 3, which align at one end of the molecule in the crystal structure. These loops appear to play a critical role in membrane docking via phosphatidic acid and phosphatidylserine head groups both in vitro (3) and in vivo (24). However, such a membrane docking model is an incomplete explanation of the present results, since it does not account for apparent saturation of binding of the {epsilon}V1 probe, nor does it explain the involvement of regions COOH-terminal to loop 3. Consistent with this general view of {epsilon}V1 anchoring in cardiac myocytes, we also found that the EAVSLKPT peptide blocker widely used to disrupt RACK2/PKC-{epsilon} interactions (7, 12, 13, 16, 26) sponsored modest inhibitory effects, although these did not achieve statistical significance. The loop 3 peptide HDAPIGYD also did not inhibit but tended to enhance {epsilon}V1-GFP binding in this system, consistent with its proposed role as a translocation activator (7, 8). We also observed a positive inotropic effect of the loop 3 peptide in a subset of myocytes, consistent with the expected physiological effects of stimulating PKC-{epsilon} translocation and anchoring in adult myocytes (19).

Importantly, transgenic mice expressing the same translocation inhibitor peptide (EAVSLKPT) or translocation activator peptide (HDAPIGYD) revealed profound physiological consequences in vivo despite only a 20% change in the distribution of PKC between cytosolic and particulate compartments (8, 26). Such small equilibrium shifts may not be easily resolvable in the in vitro binding assay described here. Thus a weakness of single myocyte imaging studies may be that relatively modest changes in binding are masked by relatively large variability in binding from cell to cell. In previous experiments using myocyte populations and Western blotting, we did observe inhibition of full-length PKC-{epsilon} binding to the cardiac Z-line by EAVSLKPT (albeit only 60% blockade at 50 µM peptide; see Ref. 13). This partial inhibition by the EAVSLKPT peptide is also generally compatible with the results of the present study suggesting the existence of additional determinants of Z-line anchoring.

In the myocyte decoration assay described here, fluorescent ligands may encounter many potential binding surfaces, so specificity is an important consideration. Extra caution may be necessary given the highly regular sarcomeric structure and unique composition of the Z-line/T tubule region of myocytes, which may conspire to exaggerate regular patterns in nonspecific binding or in tissue autofluorescence. A number of observations argue for the selectivity of the observed Z-line decoration. Striation pattern intensity was not correlated to fragment length (R2 = 0.61), and Z-line decoration of GFP constructs was readily reversible [with the exception of GFP-(14–93)]. Binding of {epsilon}V1-GFP was saturable, with a concentration dependence well described by the hyperbolic function y = ax/(b + x) (see Refs. 13 and 36). Binding of {epsilon}V1-GFP was also significantly reduced by preincubation with unlabeled {epsilon}V1 protein. In both Triton X-100- and saponin-skinned myocytes, the striated binding pattern was not observed for GFP alone at concentrations beyond 1 µM (36). {epsilon}V1 labeled with Alexa-568 on lysine side chains (instead of a COOH-terminal GFP) decorated the Z-line much like {epsilon}V1-GFP. A fluorescent protein fusion of residues 1–130 from the NH2-terminal variable domain of PKC-{delta}, {delta}V1-GFP, did not decorate cardiac Z-lines at concentrations up to threefold higher than used for {epsilon}V1-GFP. This isoform specificity is consistent with the finding that activated PKC-{epsilon} bound much more effectively to Z-lines of cardiac myofilaments than did activated PKC-{delta} (13, 14). Thus the weight of evidence is consistent with a specific binding site on the cardiac Z-line for the V1 domain of PKC-{epsilon}.

With regard to the likely involvement of both lipids and proteins in the mechanism of PKC-{epsilon} translocation, Souroujon et al. (40) recently described a monoclonal antibody that recognizes an epitope on the {epsilon}V1 domain that is exposed only transiently, after lipid activation but before RACK anchoring. This result suggests the existence of an intermediate state in which activated and membrane-associated PKC searches for a receptor (i.e., RACK) along the plane of the membrane (25, 27). Of these two "translocated states" of PKC-{epsilon}, the RACK-bound state but not the lipid membrane-associated state would be expected to be blocked by competitive inhibitors. Overall, the data presented here are consistent with the existence of both a specific "blockable" (perhaps anchoring protein-mediated) and a minor "nonblockable" (perhaps lipid membrane-mediated) component of {epsilon}V1-GFP binding at the cardiac Z-line/T tubule.

Strong binding of GFP-(85–144) provides evidence that a critical determinant of anchoring lies in a COOH-terminal stretch of sequence. A similar observation was reported by Lehel et al. (22), who deleted specific domains of full-length PKC-{epsilon} and examined the effects on translocation in NIH 3T3 cells. A 33-amino acid subdomain just upstream of the C-1 domain was found that strongly biased PKC-{epsilon} translocation to (unknown) cytoskeletal elements. The region identified by Lehel et al. overlaps the final 10 amino acids of the COOH-terminus of {epsilon}V1 investigated here. Clearly, models of {epsilon}V1 and PKC-{epsilon} anchoring need to be refined to account for the observation that probes encompassing either the COOH-terminus or the NH2-terminus of {epsilon}V1 decorate the Z-line of permeabilized myocytes. One possibility is that the COOH-terminal and NH2-terminal segments of {epsilon}V1 anchor to sites that are physically separate on a "saturable receptor," whereas full-length {epsilon}V1 is capable of binding these sites simultaneously. In this way, the {epsilon}V1 domain could communicate with separate and distinct aspects of the anchoring protein, or even bind to multiple members of a signaling complex (1, 43).

The crystal structure of {epsilon}V1 was used to incorporate most of the available data in a more complete model of anchoring (Fig. 9). Most importantly, Fig. 9 projects on the crystal structure how the COOH-terminal aspect of {epsilon}V1 (orange) contributes to a surface that is contiguous with the EAVSLKPT sequence (red) identified by Mochly-Rosen and colleagues (12, 16, 26). We propose that the COOH-terminal aspect of {epsilon}V1 (residues 94–144) participates in PKC-{epsilon} anchoring, either by contributing to a RACK2 interaction interface, by binding to other unidentified anchoring proteins, and/or by stabilizing the degree of the {epsilon}V1 domain and preventing inappropriate intramolecular interactions between 14-EAVSLKPT-21 and 85-HDAPIGYD-92.



View larger version (58K):
[in this window]
[in a new window]
  		Fig. 9. Model of PKC-{epsilon} anchoring at the cardiac Z-line/T tubule. Ribbon structure: A, interaction of 14-EAVSLKPT-21 (red) with receptor for activated C kinase (RACK2; see Refs. 4 and 6); B, interaction of loop 1 with membranes (3, 32) via insertion of hydrophobic 23Trp residue in the bilayer; C, interaction of loop 3 (green) with membranes (3, 32) via insertion of hydrophobic residues 89Ile,91Tyr in the bilayer while 86Asp and 92Asp coordinate an Mg2+ that in turn interacts with phospholipid head groups; D, point of convergence between 85-HDAPIGYD-92 (green; equivalent to loop 3), 14-EAVSLKPT-21 (red), and COOH-terminal residues 94–144 (orange). The site of interaction between PKC-{epsilon} and the cardiac Z-line is hypothesized to be composed of all or part of the red/orange surface. Space-filling structure: E, prominent side chain interaction between 122Glu (orange) and 19Lys (red) and F, prominent side chain interaction between 93Asp (orange) and 26Arg (gray).

 
It is of interest to know precisely how the HDAPIGYD peptide promotes translocation of PKC-{epsilon}. The suggestion by Mochley-Rosen and coworkers (8, 38, 40) was that it disrupted an autoinhibitory interaction between loop 3 and EAVSLKPT. The present results can be interpreted in the context of such an autoinhibitory interaction. In particular, the finding that the 14–93 construct does not bind cardiac Z-lines is generally compatible with enhanced interactions between 14-EAVSLKPT-21 and 85-HDAPIGYD-92 in the absence of residues 1–13 and 94–144. However, 14-EAVSLKPT-21 (red) and 85-HDIPIGYD-92 (green) do not appear to interact directly in the crystal structure of 1–144 (32), although, at one residue away, 93Asp forms a prominent ionic interaction with 26Arg farther along in loop 1 (Fig. 9F). The fact that both binding and twitch amplitudes were enhanced by exogenous application of the negatively charged HDIPGYD peptide suggests that conformational changes induced by its binding to the {epsilon}V1 domain promote a structural reorganization that facilitates anchoring. One possibility is that endogenous loops 1 and 3 move away from the core structure, permitting these loops to interact with membranes. Such conformational changes may also stabilize or make more accessible the putative anchoring interface formed by juxtaposition of 14-EAVSLKPT-21 (red) and side chains from the COOH-terminal domain (orange). In the surface rendering of {epsilon}V1, a number of interactions between these regions are quite prominent, including a salt bridge between residues 19Lys (red) and 122Glu (orange in Fig. 9E), H-bonds, and hydrophobic interactions (data not shown).

It is also of interest to know precisely what the {epsilon}V1 regulatory domain binds to at the cardiac Z-line. We consider it unlikely that {epsilon}V1-GFP binds exclusively to the T tubule lipid bilayer as the models of Corbalan-Garcia (3) and Ochoa et al. (32) might suggest. The minimum sequence that decorated Z-lines is 94–144, a segment located on the opposite side of the {epsilon}V1 domain from the Ochoa et al. membrane interface. T-tubular membrane staining typically appears punctate in the confocal microscope because of the "discontinuous" nature of the T tubule network when sampled by optical sectioning. On this basis, {epsilon}V1-GFP and its subfragments appear to bind to a more continuous cytoskeletal structure such as the Z-line itself. Moreover, {epsilon}V1-GFP decorated Z-lines of Triton X-100-skinned myocytes despite the loss of membrane ultrastructure visible by electron microscopy. Protein-protein interactions that mediate Z-line anchoring of PKC-{epsilon} remain to be fully elucidated, but early indications tend to rule out F-actin and Cypher-1 as candidate anchoring proteins at this location (13).

In conclusion, evaluation of Z-line binding of truncated fragments of the {epsilon}V1 domain (labeled at their COOH-termini with GFP) leads us to suggest refinements to current models of PKC-{epsilon} anchoring. The NH2-terminal region containing the 14-EAVSLKPT-21 RACK binding sequence probably contributes to but is not solely responsible for PKC-{epsilon} binding to Z-lines. Our data support a model in which multiple regions of the {epsilon}V1 domain dictate anchoring in myocytes, including a novel segment within residues 94–144 at the COOH-terminus of the first variable regulatory region of PKC-{epsilon}. In the crystal structure of {epsilon}V1, the 14-EAVSLKPT-21 sequence forms a contiguous surface with the 94–144 subdomain, extending tens of angstroms away from the proposed plane of the inner bilayer leaflet (32). This surface provides a cytoplasmic interaction interface through which PKC-{epsilon} can bind to adjacent proteins or protein complexes. Importantly, the previously proposed autoinhibitory 85-HDPIGYD-92 sequence forms no obvious contacts with 14-EAVSLKPT-21 in the crystal structure, but both of these peptide segments form side chain interactions with COOH-terminal side chains. This proposed COOH-terminal anchoring interface is well-positioned to complement Mg2+-dependent interactions with phospholipids and is compatible with additional membrane interactions via diacylglycerol binding to the adjacent C-1 domain of full-length PKC-{epsilon}.


   GRANTS
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
REFERENCES
 
This work was supported by National Heart, Lung, and Blood Institute Grant HL-04759 to J. W. Walker and a predoctoral fellowship from the American Heart Association to S. L. Robia (9910123Z).


   ACKNOWLEDGMENTS
 
Confocal microscopy was performed at the Keck Center for Biological Imaging at the University of Wisconsin with the assistance of Lance A. Rodenkirch. We thank Judith Maloney and Kara R. Kemnitz for technical contributions and helpful suggestions, Dr. Valentin Robu for engineering of the {delta}V1-GFP construct, and Luke L. Lestikow for assistance with {epsilon}V1 construct expression and purification. Electron microscopy was performed by Randall Massey University of Wisconsin Medical School Electron Microscopy Facility.

Present address for S. L. Robia: Dept. of Biochemistry, Molecular Biology, and Biophysics, University of Minnesota, Minneapolis, MN, 55455.


   FOOTNOTES
 

Address for reprint requests and other correspondence: J. W. Walker, Dept. of Physiology, 1300 Univ. Ave., Madison, WI 53706 (e-mail: jwalker{at}physiology.wisc.edu)

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


   REFERENCES
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 

  1. Baines CP, Song CX, Zheng YT, Wang GW, Zhang J, Wang OL, Guo Y, Bolli R, Cardwell EM, and Ping P. Protein kinase Cepsilon interacts with and inhibits the permeability transition pore in cardiac mitochondria. Circ Res 92: 873–880, 2003.[Abstract/Free Full Text]
  2. Banci L, Cavallaro G, Kheifets V, and Mochly-Rosen D. Molecular dynamics characterization of the C2 domain of protein kinase Cbeta. J Biol Chem 277: 12988–12997, 2002.[Abstract/Free Full Text]
  3. Corbalan-Garcia S, Sanchez-Carrillo S, Garcia-Garcia J, and Gomez-Fernandez JC. Characterization of the membrane binding mode of the C2 domain of PKC epsilon. Biochemistry 42: 11661–11668, 2003.[CrossRef][Medline]
  4. Csukai M, Chen CH, De Matteis MA, and Mochly-Rosen D. The coatomer protein beta'-COP, a selective binding protein (RACK) for protein kinase Cepsilon. J Biol Chem 272: 29200–29206, 1997.[Abstract/Free Full Text]
  5. De Ross D, Joliott AH, Chassaing G, and Prochiantz A. The third helix of antennapedia translocates through biological membranes. J Biol Chem 269: 10444–10450, 1994.[Abstract/Free Full Text]
  6. Dorn GW II and Brown JH. Gq signaling in cardiac adaptation and maladaptation. Trends Cardiovasc Med 9: 26–34, 1999.[CrossRef][ISI][Medline]
  7. Dorn GW II and Mochly-Rosen D. Intracellular transport mechanisms of signal transducers. Annu Rev Physiol 64: 407–429, 2002.[CrossRef][ISI][Medline]
  8. Dorn GW II, Souroujon MC, Liron T, Chen CH, Gray MO, Zhou HZ, Csukai M, Wu G, Lorenz JN, and Mochly-Rosen D. Sustained in vivo cardiac protection by a rationally designed peptide that causes epsilon protein kinase C translocation. Proc Natl Acad Sci USA 96: 12798–12803, 1999.[Abstract/Free Full Text]
  9. Endo M and Iino M. Specific perforation of muscle cell membranes with preserved SR functions by saponin treatment. J Muscle Res Cell Motil 1: 89–100, 1980.[CrossRef][Medline]
  10. Fukami K, Furuhashi K, Inagaki M, Endo T, Hantano S, and Takenawa T. Requirement of phosphatidylinositol 4,5-bisphosphate for alpha-actinin function. Nature 359: 150–152, 1992.[CrossRef][Medline]
  11. Goldspink PH, Montgomery DE, Walker LA, Uroniene D, McKinney RD, Geenen DL, Solaro RJ, and Buttrick PM. Protein kinase C epsilon overexpression alters myofilament properties and composition during the progression of heart failure. Circ Res 95: 424–432, 2004.[Abstract/Free Full Text]
  12. Gray MO, Karliner JS, and Mochly-Rosen D. A selective epsilon-protein kinase C antagonist inhibits protection of cardiac myocytes from hypoxia-induced cell death. J Biol Chem 272: 30945–30951, 1997.[Abstract/Free Full Text]
  13. Huang X and Walker JW. Myofilament anchoring of protein kinase C-epsilon in cardiac myocytes. J Cell Sci 15: 1971–1978, 2004.
  14. Huang X, Pi Y, Lokuta AJ, Greaser ML, and Walker JW. Arachidonic acid stimulates protein kinase C-epsilon redistribution in heart cells. J Cell Sci 110: 1625–1634, 1997.[Abstract]
  15. Huang X, Sreekumar R, Patel JR, and Walker JW. Response of cardiac myocytes to a ramp increase of diacylglycerol generated by photolysis of a novel caged diacylglycerol. Biophys J 70: 2448–2457, 1996.[Abstract/Free Full Text]
  16. Johnson JA, Gray MO, Chen CH, and Mochly-Rosen D. A protein kinase C translocation inhibitor as an isozyme-selective antagonist of cardiac function. J Biol Chem 271: 24962–24966, 1996.[Abstract/Free Full Text]
  17. Johnson JA, Gray MO, Karliner JS, Chen CH, and Mochly-Rosen D. An improved permeabilization protocol for the introduction of peptides into cardiac myocytes: application to protein kinase C research. Circ Res 79: 1086–1099, 1996.[Abstract/Free Full Text]
  18. Kamp TJ and Hell JW. Regulation of cardiac L-type calcium channels by protein kinase A and protein kinase C. Circ Res 87: 1095–1102, 2000.[Abstract/Free Full Text]
  19. Kang M and Walker JW. Protein kinase C delta and epsilon mediate positive inotropy in adult ventricular myocytes. J Mol Cell Cardiol 38: 753–764, 2005.[CrossRef][ISI][Medline]
  20. Kentish JC, Barsotti RJ, Lea TJ, Mulligan IP, Patel JR, and Ferenczi MA. Calcium release from cardiac sarcoplasmic reticulum induced by photorelease of calcium or InsP3. Am J Physiol Heart Circ Physiol 258: H610–H615, 1990.[Abstract/Free Full Text]
  21. Kraft AS and Anderson WB. Phorbol esters increase the amount of Ca2+, phospholipid-dependent protein kinase associated with plasma membrane. Nature 301: 621–623, 1983.[CrossRef][Medline]
  22. Lehel C, Olah Z, Jakab G, Szallasi Z, Petrovics G, Harta G, Blumberg PM, and Anderson WB. Protein kinase C epsilon subcellular domains and proteolytic degradation sites. J Biol Chem 270: 19651–19658, 1995.[Abstract/Free Full Text]
  23. Liu GS, Cohen MV, Mochly-Rosen D, and Downey JM. Protein kinase C-epsilon is responsible for the protection of preconditioning in rabbit cardiomyocytes. J Mol Cell Cardiol 31: 1937–1948, 1999.[CrossRef][ISI][Medline]
  24. Lopez-Andreo MJ, Gomez-Fernandez JC, and Corbalan-Garcia S. Simultaneous production of phosphatidic acid and diacylglycerol is essential for translocation of protein kinase Cepsilon to plasma membrane. Mol Cell Biol 14: 4885–4895, 2003.
  25. Medkova M and Cho W. Differential membrane-binding and activation mechanisms of protein kinase C-alpha and -epsilon. Biochemistry 37: 4892–4900, 1998.[CrossRef][Medline]
  26. Mochly-Rosen D, Wu G, Hahn H, Osinska H, Liron T, Lorenz JN, Yatani A, Robbins J, and Dorn GW II. Cardiotrophic effects of protein kinase C epsilon: analysis by in vivo modulation of translocation. Circ Res 86: 1173–1179, 2000.[Abstract/Free Full Text]
  27. Nalefski EA and Newton AC. Membrane binding kinetics of protein kinase C betaII mediated by the C2 domain. Biochemistry 40: 13216–13229, 2001.[CrossRef][Medline]
  28. Newton AC. Protein kinase C: structural and spatial regulation by phosphorylation, cofactors, and macromolecular interactions. Chem Rev 101: 2353–2364, 2001.[CrossRef][ISI][Medline]
  29. Nishikawa K, Toker A, Johannes FJ, Songyang Z, and Cantley LC. Determination of the specific substrate sequence motifs of protein kinase C isozymes. J Biol Chem 272: 952–960, 1997.[Abstract/Free Full Text]
  30. Nishizuka Y. Protein kinase C and lipid signaling for sustained cellular responses. FASEB J 9: 484–496, 1995.[Abstract]
  31. Oancea E and Meyer T. Protein kinase C as a molecular machine for decoding calcium and diacylglycerol signals. Cell 95: 307–318, 1998.[CrossRef][ISI][Medline]
  32. Ochoa WF, Garcia-Garcia J, Fita I, Corbalan-Garcia S, Verdaguer N, and Gomez-Fernandez JC. Structure of the C2 domain from novel protein kinase Cepsilon: a membrane binding model for Ca-independent C2 domains. J Mol Biol 311: 837–849, 2001.[CrossRef][ISI][Medline]
  33. Pi Y, Kemnitz KR, Zhang D, Kranias EG, and Walker JW. Phosphorylation of troponin I controls cardiac twitch dynamics: evidence from phosphorylation site mutants expressed on a troponin I-null background in mice. Circ Res 90: 649–656, 2002.[Abstract/Free Full Text]
  34. Ping P, Zhang J, Qiu Y, Tang XL, Manchikalapudi S, Cao X, and Bolli R. Ischemic preconditioning induces selective translocation of protein kinase C isoforms epsilon and eta in the heart of conscious rabbits without subcellular redistribution of total protein kinase C activity. Circ Res 81: 404–414, 1997.[Abstract/Free Full Text]
  35. Qiu Y, Ping P, Tang XL, Manchikalapudi S, Rizvi A, Zhang J, Takano H, Wu WJ, Teschner S, and Bolli R. Direct evidence that protein kinase C plays an essential role in the development of late preconditioning against myocardial stunning in conscious rabbits and that epsilon is the isoform involved. J Clin Invest 101: 2182–2198, 1998.[ISI][Medline]
  36. Robia SL, Ghanta J, Robu VG, and Walker JW. Localization and kinetics of protein kinase C-epsilon anchoring in cardiac myocytes. Biophys J 80: 2140–2151, 2001.[Abstract/Free Full Text]
  37. Robu VG, Pfeiffer ES, Robia SL, Balijepalli RC, Pi Y, Kamp TJ, and Walker JW. Localization of functional endothelin A receptor signaling complexes in cardiac transverse tubules. J Biol Chem 278: 48154–48161, 2004.
  38. Schechtman D, Craske ML, Kheifets V, Meyer T, Schechtman J, and Mochly-Rosen D. A critical intramolecular interaction for protein kinase C epsilon translocation. J Biol Chem 279: 15831–15840, 2004.[Abstract/Free Full Text]
  39. Seeman P. Transient holes in the erythrocyte membrane during hypotonic hemolysis and stable holes after lysis by saponin and lysolecithin. J Cell Biol 32: 55–70, 1967.[Abstract/Free Full Text]
  40. Souroujon MC, Yao L, Chen H, Endemann G, Khaner H, Geeraert V, Schechtman D, Gordon AS, Diamond I, and Mochly-Rosen D. State-specific monoclonal antibodies identify an intermediate state in epsilon protein kinase C activation. J Biol Chem 279: 17617–17624, 2004.[Abstract/Free Full Text]
  41. Takeishi Y, Ping P, Bolli R, Kirkpatrick DL, Hoit BD, and Walsh RA. Transgenic overexpression of constituitively active PKC-{epsilon} causes concentric hypertrophy. Circ Res 86: 1218–1223, 2000.[Abstract/Free Full Text]
  42. Teruel MN and Meyer T. Translocation and reversible localization of signaling proteins: a dynamic future for signal transduction. Cell 103: 181–184, 2000.[CrossRef][ISI][Medline]
  43. Vondriska TM, Zhang J, Song C, Tang XL, Cao X, Baines CP, Pass JM, Wang S, Bolli R, and Ping P. Protein kinase C epsilon-src modules direct signal transduction in nitric oxide-induced cardioprotection: complex formation as a means for cardioprotective signaling. Circ Res 88: 1306–1313, 2001.[Abstract/Free Full Text]



This article has been cited by other articles: