Advertisement
JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M707728200 on September 27, 2007

J. Biol. Chem., Vol. 282, Issue 48, 35005-35017, November 30, 2007
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
282/48/35005    most recent
M707728200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
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 PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zhou, Y.
Right arrow Articles by Yang, J. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zhou, Y.
Right arrow Articles by Yang, J. J.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Identification of the Calmodulin Binding Domain of Connexin 43*Formula

Yubin Zhou{ddagger}1, Wei Yang{ddagger}, Monica M. Lurtz§, Yiming Ye, Yun Huang{ddagger}, Hsiau-Wei Lee{ddagger}, Yanyi Chen{ddagger}, Charles F. Louis§, and Jenny J. Yang{ddagger}2

From the {ddagger}Department of Chemistry, Georgia State University, Atlanta, Georgia 30303, the Proteomics Laboratory, Biotechnology Core Facility Branch, Centers for Disease Control and Prevention, Atlanta, Georgia 30333, and the §Department of Cell Biology and Neuroscience, University of California, Riverside, California 92521

Received for publication, September 14, 2007


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Calmodulin (CaM) has been implicated in mediating the Ca2+-dependent regulation of gap junctions. This report identifies a CaM-binding motif comprising residues 136–158 in the intracellular loop of Cx43. A 23-mer peptide encompassing this CaM-binding motif was shown to bind Ca2+-CaM with 1:1 stoichiometry by using various biophysical approaches, including surface plasmon resonance, circular dichroism, fluorescence spectroscopy, and NMR. Far UV circular dichroism studies indicated that the Cx43-derived peptide increased its {alpha}-helical contents on CaM binding. Fluorescence and NMR studies revealed conformational changes of both the peptide and CaM following formation of the CaM-peptide complex. The apparent dissociation constant of the peptide binding to CaM in physiologic K+ is in the range of 0.7–1 µM. Upon binding of the peptide to CaM, the apparent Kd of Ca2+ for CaM decreased from 2.9 ± 0.1 to 1.6 ± 0.1 µM, and the Hill coefficient nH increased from 2.1 ± 0.1 to 3.3 ± 0.5. Transient expression in HeLa cells of two different mutant Cx43-EYFP constructs without the putative Cx43 CaM-binding site eliminated the Ca2+-dependent inhibition of Cx43 gap junction permeability, confirming that residues 136–158 in the intracellular loop of Cx43 contain the CaM-binding site that mediates the Ca2+-dependent regulation of Cx43 gap junctions. Our results provide the first direct evidence that CaM binds to a specific region of the ubiquitous gap junction protein Cx43 in a Ca2+-dependent manner, providing a molecular basis for the well characterized Ca2+-dependent inhibition of Cx43-containing gap junctions.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Gap junctions comprise the intercellular channels that mediate the cell-to-cell transfer of small molecules, including metabolites, second messengers, and ions, in mammalian cells (1). Gap junctions are composed of two hemichannels (termed connexons) with each composed of six connexin (Cx)3 subunits. The connexon from one cell joins in mirror symmetry with another connexon of the apposing cell. To date, at least 20 connexin genes have been identified in the human genome, with connexin 43 (Cx43) the most ubiquitous connexin (2).

Gap junctions have been shown to be regulated by [Ca2+]i (3) that Peracchia and others (4, 5) have shown to be probably mediated by CaM interacting directly with the connexin proteins. We have shown previously that cell-to-cell communication in lens epithelial cell cultures is inhibited by elevated [Ca2+]i. Specifically, cell-to-cell transfer of the fluorescent dye AlexaFluor594 was half-maximally inhibited at ~300 nM [Ca2+]i in lens cell cultures, and this inhibition was prevented by preincubation of these cultures with CaM antagonists (6, 7), consistent with earlier reports that elevated [Ca2+]i increased internal electrical resistance in the lens that was prevented by preincubation with CaM antagonists (6). Indeed, this action of Ca2+ in lens cell cultures is due in part to the inhibition of Cx43, the major connexin in these cell cultures. It has demonstrated that Cx43-transfected HeLa cells exhibiting a similar Ca2+-dependent inhibition appears to be CaM-mediated (8). The rapid onset of this Ca2+-dependent inhibition of cell-to-cell communication (within seconds) suggests that this is mediated by a direct interaction of CaM with the connexin protein rather than being mediated via the action of a CaM-dependent protein kinase. Indeed, the Ca2+-dependent binding of CaM to rat Cx32 (5), fish Cx35, mouse Cx36 (9), and Cx50 (10) have been reported. Two cytoplasmic CaM binding domains, with one site (Kd = 27 nM) in the N terminus and the other site (Kd = 1.2 µM) in the C-terminal region, have been identified in Cx32 (5), whereas a single CaM-binding site (Kd = 11–72 nM) was identified in the C terminus of Cx35 and Cx36 (9).

Each Cx43 monomer consists of four highly conserved transmembrane segments, a short N-terminal cytoplasmic region, one intracellular and two extracellular loops, and a C-terminal tail (Fig. 1A). Variability in sequence homology across different connexin types is greatest in the intracellular loop and C terminus. Efforts to map the potential CaM-binding sites in Cx43 have led to conflicting results. Torok et al. (5) reported the binding of the fluorescent CaM derivative (TA-CaM) to the N-terminal (aa 1–16) region of Cx43 with a dissociation constant of 1.2 µM. However, Duffy et al. (11) were unable to detect any interaction of CaM with a peptide spanning the first 21 amino acids of the N terminus. Both groups, nevertheless, failed to detect the binding of CaM to peptides derived from the C-terminal tail (aa 314–325, 336–350, and 346–360) or the intracellular loop (aa 95–114, 123–136, and 119–144). These results suggest that the interaction of CaM with Cx43 might occur via other regions of the intracellular loop. Interestingly, by using the CaM binding data base server (12), we have predicted a potential CaM-binding site with high predictive score in the second half of the intracellular loop of Cx43 (aa 136–158) that has not been tested before (Fig. 1). In the present study, we have applied a variety of biophysical approaches to examine the binding of CaM to this sequence. Our findings strongly suggest that Ca2+ effects the inhibition of Cx43 via a Ca2+-dependent interaction between the CaM and residues 136–158 of this ubiquitous gap junction protein.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Bioinformatic Analyses and Prediction of CaM-binding Site in Connexins—The topology and orientation of transmembrane regions of rodent connexin 43 were predicted on the basis of the consensus results using four different programs, including SOSUI (13), TMHMM (14), MEMSAT (15), and HMMTOP (16). The potential CaM-binding sites were predicted using the CaM target data base based on criteria, including the distribution of hydrophobic and basic residues, the propensity to form {alpha}-helix, and the hydrophobicity of the sequence, that are common to more than 100 CaM target sequences (12).

Peptide Synthesis, Protein Expression, and Purification—The peptide Cx43-(136–158) (Ac-136KYGIEEHGKVKMRGGLLRTYIIS158-NH2) was synthesized by Sigma and purified by preparative reversed-phase high pressure liquid chromatography with purity of >95%. A randomized control peptide (Ac-LGGEYLVTMESKIHIKGKRIGYR-NH2) with the same composition of amino acids but arranged in a different order with no predicted CaM binding capacity (12) was similarly synthesized. Two other peptides with lower predictive scores, one in the N-terminal part of the intracellular loop region (Ac-86SVPTLLYLAHVFYVMRKEEKLN107-NH2) and the other in the C terminus region (Ac-224NIIELFYVFFKGVKDRVKGRSDPY247-NH2), were also synthesized (EZ Biolabs). The molecular weight of the synthetic peptides were determined by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. To mimic the native protein environment and eliminate extra charges, the designed peptides were blocked at the N termini with an acetyl group and at the C termini with an amide group.

Recombinant rat CaM was expressed in Escherichia coli strain BL21(DE3)pLysS transformed with the plasmid pET7-CaM that harbors the synthetic CaM gene (17). pET7-CaM transformed cells were grown in LB medium to obtain unlabeled CaM. 15N-Labeled CaM was expressed in SV minimal medium using 0.5 g/liter 15NH4Cl (Cambridge Isotope Laboratories) as the sole nitrogen source. Bacterially expressed CaM was purified by phenyl-Sepharose (Sigma) chromatography as previously described (18). The purity of CaM was examined by mass spectrometry or SDS-PAGE (supplemental Fig. 1). The concentration of CaM was determined by UV spectroscopy using the {epsilon}276 of 3030 M-1 cm-1 (19).

Dansyl-CaM was prepared according to the method of Johnson et al. (20) with slight modifications. Briefly, rat CaM was dansylated in the dark by mixing 1 ml of protein (1 mM) with a 5-fold molar excess of dansyl chloride (dissolved in 1:1 acetone/ethanol) in 10 mM Mops, 100 mM KCl, 1 mM CaCl2, pH 7.0, for 16 h at 4 °C. The reaction mixture was then extensively dialyzed against 10 mM Tris, 100 mM KCl at pH 7.4 to remove the residual free dansyl chloride. The modification of CaM by dansyl chloride was confirmed by electrospray ionization-mass spectrometry with an increase in the molecular mass of +233. The bound dye concentration was determined by using a {epsilon}335 value of 3980 M-1 cm-1 (20). An average of ~0.8 mol of the dansyl chromophore was incorporated per mol of CaM.

Plasmid Construction and Transient Protein Expression—The wild type (WT) connexin 43-EYFP construct (Cx43-EYFP) was a generous gift of Dr. Xiaohua Gong (University of California, Berkeley, CA). The mutant Cx43-EYFP constructs Cx43K146E,R148E-EYFP double mutant, and Cx43M147Q,L151E,I156E-EYFP triple mutant were synthesized by PCR mutational insertion using the following primers: forward (GGC AAG GTG GAA ATG GAG GGC GG) and reverse (GTG CTC TTC AAT CCC GTA CTT G) for Cx43K146E,R148E-EYFP; forward (ACC TAC GAA ATC AGC ATC CTC TTC AAG) and reverse (TCT CAG TTC GCC GCC CCT CTG TTT CAC) for Cx43M147Q,L151E,I156E-EYFP. Resultant constructs were verified by sequencing. Plasmid DNA was prepared for transfection by transformation into TOP10 competent cells (Invitrogen) and growing overnight on an LB agar supplemented with kanamycin (0.1 g/liter). A single transformed colony was grown overnight in LB medium supplemented with kanamycin (0.1 g/liter) and purified using the Qiagen (Valencia, CA) Endo-Free Plasmid Maxi Kit per the manufacturer's instructions. Plasmid concentrations were quantified using UV-visible spectroscopy (260 nm/280 nm ratio; DU 640 spectrophotometer; Beckman-Coulter, Fullerton, CA). Connexin-deficient HeLa cells were grown on glass coverslips in Dulbecco's modified Eagle's medium (with 44 mM NaHCO3, pH 7.2) supplemented with 10% (v/v) fetal bovine serum, 100 units/ml penicillin, and 0.1 mg/ml streptomycin in a humidified 37 °C incubator with 5% CO2 to about 80% confluence. Transient transfection of HeLa cells was done using Lipofectamine 2000 in OptiMEM I reduced serum medium (Invitrogen) per the manufacturer's instructions. 1 µg of plasmid DNA was used per 35-mm culture dish, and 4 µl of Lipofectamine 2000 was used per 1 µg of DNA.

Surface Plasmon Resonance Measurements—Real time binding assays were performed using surface plasmon resonance at the Centers for Disease Control and Prevention (Atlanta, GA) on a Biacore 3000 system (Biacore AB, Uppsala, Sweden). CaM (500 nM in 10 mM sodium formate, pH 3.5) was directly immobilized onto the sensor chip CM5 using an amine coupling kit as described by the manufacturer. Varying concentrations of synthetic peptide were subsequently injected over the sensor surface at a flow rate of 50 µl/min in binding buffer (5 mM CaCl2, 100 mM KCl, 50 mM Tris-HCl, pH 7.4). Two minutes later, peptide-free binding buffer was injected to monitor the dissociation process. All measurements were carried out in parallel using two cells, one with immobilized CaM and the other as blank control with the carboxylated dextran matrix deactivated. The binding of peptide to CaM-immobilized flow cells was corrected accordingly from binding to control flow cells. The sensor chip was regenerated using 10 mM glycine, pH 2.2. The binding data were analyzed using the BIAevalution software with the 1:1 Langmuir fitting model.

Circular Dichroism Spectroscopy—Circular dichroism spectra were acquired in the far UV (190–260 nm) or near UV region (250–340 nm) on a Jasco-810 spectropolarimeter at room temperature using a 1-cm path length quartz cuvette. All spectra presented were averaged from 10–20 scans. The background signals from the corresponding buffers were subtracted from the sample signals. The far UV CD spectra of the peptide in different percentages of trifluoroethanol (TFE) were obtained using a 10 µM concentration of the peptide in 10 mM Tris-HCl, 10 or 100 mM KCl, pH 7.4. In the peptide titration of the CaM experiment, 2–5-µl aliquots of the peptide stock solution (150 µM in 10 mM Tris-HCl, 10 or 100 mM KCl, pH 7.4) was gradually added into a 2-ml solution containing 1–2 µM CaM in the same buffer with 5 mM CaCl2 or 5 mM EGTA. The signals from the peptide itself were subtracted. All of the measurements were carried out in at least triplicates. The binding constants of the synthetic peptide to CaM were obtained with a 1:1 binding model by fitting Equation 1,

Formula 1(Eq. 1)
where f represents the fractional change of CD signals at 222 nm, Kd is the dissociation constant for the peptide, and [P]T and [CaM]T are the total concentrations of the synthetic peptide and CaM, respectively. The secondary structure contents of the peptides or proteins were calculated with the online secondary structure prediction server DICHROWEB that integrates analysis algorithms, such as SELCON, CONTINLL, CDSSTR, and K2D (21). Near UV CD spectra were recorded with a protein/peptide concentration of 80–100 µM in 10 mM Tris-HCl, 100 mM KCl, pH 7.4, with 5 mM Ca2+ or 5 mM EGTA.

Steady-state Fluorescence Measurements—Steady-state fluorescence spectra were recorded using a QM1 fluorescence spectrophotometer (PTI) with a xenon short arc lamp at ambient temperature. Tyrosine fluorescence was monitored using excitation at 277 nm and emission at 307 nm with 2–4-nm bandpasses. The Ca2+ binding constants were determined by titrating the CaM (8 µM) or 1:1 CaM/peptide mixture (8 µM) in 1 mM EGTA, 100 mM KCl, 50 mM Tris-HCl, pH 7.4, with 1–5-µl aliquots of 10 mM Ca2+ stock solution in the same buffer containing the same concentrations of CaM and peptide. The pH change (0.03–0.04) was minimal during the titration process. To obtain accurate Ca2+ concentrations during the titration, the Ca2+ concentration at each point was determined with the Ca2+ dye Oregon Green 488 BAPTA-5N (0.2 µM; Kd = 20 µM; Invitrogen) with excitation at 492 nm and the emission at 520 nm. The ionized Ca2+ concentration was subsequently calculated according to Equation 2,

Formula 2(Eq. 2)
where F represents the fluorescence intensity of the dye at each titration point and Fmin and Fmax are the intensities of the Ca2+-free and the Ca2+-saturated dye, respectively. The Ca2+ titration of CaM data were fit to the nonlinear Hill equation (Equation 3),

Formula 3(Eq. 3)
where f represents the relative signal change observed during the experiment, [M] is the concentration of free Ca2+, K refers to dissociation constants of Ca2+, and n is the Hill coefficient.

For dansyl-CaM fluorescence measurement, a 1-ml solution containing 1–2 µM dansyl-CaM in 10 mM Tris-HCl, 100 mM KCl, pH 7.4, with 5 mM Ca2+ or 5 mM EGTA was titrated with 5–10-µl aliquots of the peptide stock solution (10 µM) in the same buffer. Fluorescence spectra was recorded using an excitation wavelength of 335 nm and collecting the fluorescence emission between 400 and 600 nm. Slit widths were set at 4–8 nm.

NMR Spectroscopy—All NMR experiments were performed using either Varian Inova 500- or 600-MHz spectrometers. NMR spectra were acquired with a spectral width of about 13 ppm in the 1H dimension and 36 ppm in the 15N dimension at 35 °C. For the 1H,15N HSQC experiment, 0.5 mM uniformly 15N-labeled CaM was titrated with 10–20-µl aliquots of the peptide stock solution (2.1 mM) in a buffer consisting of 10% D2O, 100 mM KCl, 50 mM Tris-HCl with 10 mM EGTA or 10 mM Ca2+. The pH values of both sample solutions were carefully adjusted to 7.4 with a trace amount of 2 M KOH. NMR data were processed using the FELIX98 program (Accelrys).

Dye Transfer Assay—Confluent monolayers of HeLa cells grown on glass coverslips were loaded with the Ca2+ indicator Fura-2/AM (5 µM) in 2 ml of HBSS++ buffer (containing 1.8 mM Ca2+, with added 10 mM HEPES, 5 mM NaHCO3, pH 7.2) and then transferred to a microincubation chamber (model MSC-TD; Harvard Apparatus, Holliston, MA) as described previously (8). Imaging of intracellular Ca2+ was performed with a Nikon TE300 (Nikon Inc., Melville, NY) inverted microscope equipped with Nikon filter blocks for Fura-2 emission and AF594 optics (Chroma Technology Corp., Rockingham, VT), a Metaltek filter wheel (Metaltek Instruments, Raleigh, NC) housing excitation filters for Fura-2, a 75-watt xenon short arc lamp, and a Hamamatsu CCD digital camera (Hamamatsu Corp., Bridgewater, NJ) and supported on a vibration isolation table (Technical Manufacturing, Peabody, MA). [Ca2+]i was measured ratiometrically ({lambda}340/{lambda}380) with Fura-2 throughout each experiment in the injected cell and the cells adjacent to the injected cell, and Ca2+ concentrations were determined as described previously (7). MetaFluor software (Universal Imaging Corp., Downington, PA) was used for data collection. Micropipettes (borosilicate glass capillaries: 1-mm outer diameter, 0.75-mm inner diameter, 100-µm internal microfilament; Dagan Corp., Minneapolis, MN) were pulled on a Flaming/Brown-type pipette puller (P-87; Sutter Instruments, Novato, CA). Micropipettes had tip diameters of <1 µm and resistances of ~100–300 megaohms when filled with AlexaFluor 594 (1 mM) dissolved in deionized water. The micropipette was positioned with a low drift hydraulic micromanipulator (MW-3; Narishige, Greenvale, NY), and AlexaFluor 594 was microinjected iontophoretically using a train of 5-ms current pulses applied every 100 ms for 60 s (3-s total injection time) at ambient temperature. If the micropipette became plugged, it was replaced with a new micropipette, and the data from such a partial injection were excluded from the analysis. Current was generated with a Duo 773 (World Precision Instruments, Sarasota, FL). Current duration, magnitude, and polarity were controlled with an A310 Accupulser pulse generator (World Precision Instruments). Digitized images of AlexaFluor594 cell-to-cell transfer were recorded 2 min following the iontophoretic injection of fluorescent dye. A sustained elevation in [Ca2+]i was effected by adding 1 µM ionomycin to the medium and then 2 min later increasing the extracellular [Ca2+] from 1.8 to 21.8 mM.


Figure 1
View larger version (42K):
[in this window]
[in a new window]

  		FIGURE 1.
Membrane topology and predicted CaM binding sequences in Cx43. A, transmembrane topology and the primary sequence of Cx43. The integral membrane protein Cx43 is composed of four transmembrane regions (TM1 to -4), two extracellular loops, one cytoplasmic loop, a short N terminus, and a longer C-terminal tail. The transmembrane segments (black cylinders) were assigned using four prediction programs (13–16). The predicted high affinity CaM-binding site (aa 136–158, box II) of highest predicted score (12) is located in the second half of the intracellular loop between TM2 and TM3. Two other regions (boxes I and III) with lower predictive scores (12) are also identified. The numeric score ranges from 1 to 9, representing the probability of an accurate prediction of a high affinity CaM-binding site. B, helical wheel representation of the CaM-binding site of highest score in Cx43. The identified CaM-binding sequence Cx43-(136–158) fits the hydrophobic residue 1-5-10 subclass, where each number represents the presence of a hydrophobic residue (#). Basic charged residues are shown in black, whereas the hydrophobic residues are shown in gray.

 

   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
Identification of a CaM-binding Region in the Intracellular Loop Sequence of the Gap Junction Protein Cx43—Using a search program for CaM-binding sites (12), a putative CaM-binding region with high predictive score was identified in the intracellular loop (136KYGIEEHGKVKMRGGLLRTYIIS158) of Cx43. In addition, two other stretches of sequences with lower predictive score, one at the N-terminal part of the intracellular loop and the other at the C terminus, were detected (Fig. 1A). A common feature of the CaM-binding regions is a number of hydrophobic residues interspersed with positively charged residues (≥+3), typically forming 1-5-10, 1-8-14, or 1-16 patterns of hydrophobic residues (12, 22). These regions are normally composed of 12–30 continuous amino acids with a strong propensity to form an amphipathic helical structure. The putative CaM-binding region of Cx43, aa residues 136–158, exhibits a 1-5-10 hydrophobic residue motif (Fig. 1B, bottom) as reported in other well characterized CaM-targeting proteins, such as CaM kinase I (23), CaM kinase II (24), myristoylated alanine-rich C kinase substrate (25), and synapsin (26). Similar to previously reported CaM-binding sequences (12, 22), an {alpha}-helical wheel representation of the predicted sequence in Cx43 shows that the positively charged residues segregate on one side of the wheel, whereas the hydrophobic residues segregate on the other side when the sequence forms an {alpha}-helical structure (Fig. 1B). The randomized control peptide for residues 136–158 exhibited no CaM binding potential using the search program for CaM-binding sites (data not shown).


Figure 2
View larger version (29K):
[in this window]
[in a new window]

  		FIGURE 2.
Far UV CD spectra of the synthetic peptide Cx43-(136–158) (10 µM) with 0 (•), 10% ({blacktriangleup}), 20% ({diamond}), 40% (+), 60% ({circ}), and 80% ({square}) TFE. The inset shows the changes of molar ellipticity at 222 nm as a function of TFE concentration.

 


Figure 3
View larger version (41K):
[in this window]
[in a new window]

  		FIGURE 3.
Circular dichroism studies of the interaction between Cx43-(136–158) and CaM. A, far UV circular dichroism spectrum of CaM in the presence of 1 mM EGTA ({circ}) or 1 mM CaCl2 ({square}) and a 1:1 CaM/synthetic peptide mixture with 1 mM EGTA (•) or 1 mM CaCl2 ({blacksquare}) after subtracting the contribution from buffer and added peptides. Spectra were recorded at room temperature in 10 mM Tris, 100 mM KCl, pH 7.4. B, the far UV circular dichroism spectra of Cx43-(136–158) (dashed line) and the calculated difference spectrum (solid line) by subtracting the spectrum of Ca2+-CaM from that of the Ca2+-CaM-Cx43-(136–158) mixture with 1 mM Ca2+ in a buffer consisting of 100 mM KCl, 10 mM Tris-HCl, pH 7.4. The inset reports the relative change of the circular dichroism molar ellipticity at 222 nm as a function of synthetic peptide concentration in buffer comprising 1 mM Ca2+, 100 mM KCl, 10 mM Tris-HCl, pH 7.4. All experiments were conducted in at least duplicate. C, near UV circular dichroism spectra of CaM in the presence of 1 mM EGTA ({circ}) or 1 m M CaCl2 ({square}) and a 1:1 CaM/Cx43-(136–158) mixture with 1 mM EGTA (•) or 1 mM CaCl2 ({blacksquare}) after subtracting the contribution from buffer and added peptides. D, the near UV circular dichroism spectra of Cx43-(136–158) (dashed line) and the calculated difference spectrum by subtracting the Ca2+-CaM spectrum from the Ca2+-CaM-Cx43-(136–158) spectrum (solid line) in a buffer consisting of 1 mM Ca2+, 100 mM KCl, 10 mM Tris-HCl, pH 7.4.

 
The ability of peptide Cx43-(136–158), corresponding to Cx43 residues 136–158, to form an {alpha}-helical structure (Fig. 1B) was evaluated by monitoring its secondary structure using far UV CD in the presence of varying concentrations of TFE. In aqueous solution, the peptide Cx43-(136–158) was largely unstructured with a negative maximum at 197 nm. However, when the TFE concentration was increased to 20%, it started to adopt an {alpha}-helical structure (12%), with two major troughs at 208 and 222 nm observed in the CD spectra. The helical content increased to 40% in 40% TFE and to 55% in 80% TFE (Fig. 2). TFE, with a dielectric constant one-third that of water, is capable of strengthening intramolecular hydrogen bonding (27) and is known to induce and stabilize the intrinsic secondary structures in peptides, possibly by mimicking the hydrophobic environment of these peptide regions in the intact protein (28–30). Several CaM-binding peptides have been reported to adopt {alpha}-helical structures in TFE solvent similar to those that occur following formation of the CaM-peptide complex (31, 32). These results support our prediction that this Cx43-(136–158) peptide possesses a strong propensity to form {alpha}-helix.

Secondary and Tertiary Structure Changes Induced by Peptide Binding—As described above, the binding of CaM to CaM-binding peptides typically induces the formation of {alpha}-helical structure in these peptides (33). As seen in Fig. 3A, the addition of a 1:1 molar eq of peptide to Ca2+-CaM results in an ~10% more negative signal in the spectrum. Since the {alpha}-helical content of CaM typically does not increase upon peptide binding (33), the observed net increase in the CD signal could be reasonably attributed to CaM-bound Cx43-(136-158). The difference spectrum, obtained by subtracting the Ca2+-CaM spectrum from the Ca2+-CaM-peptide spectrum, showed two major troughs at 208 and 224 nm, which is strikingly different from the random coiled structure of the peptide alone (Fig. 3B). Deconvolution of this difference spectra revealed that the CaM-bound Cx43-(136–158) had ~35% helical structure, which is comparable with the helical content of this peptide in 40% TFE (Fig. 2). The Cx43-(136–158) binding-induced CD signal change enabled us to measure the peptide binding affinity of CaM. As shown in the inset of Fig. 3B, the signal changes are maximal when the Cx43-(136–158)/CaM ratio was ~1:1, and the fitting curves provided an apparent Kd of 100 ± 20 nM (n = 3) in 10 mM KCl and 750 ± 120 nM (n = 3) in 100 mM KCl (Table 1) in the presence of 1 mM Ca2+. The decrease in Kd at higher salt concentration suggests that electrostatic interactions might play an important role in the interaction of Cx43-(136–158) with CaM. In the presence of 1 mM EGTA, no significant difference in the far UV CD spectra was detected after the addition of a 1:1 molar eq of Cx43-(136–158) to CaM (Fig. 3A), suggesting that the peptide is unable to interact with CaM or that the binding is much weaker in the absence of Ca2+.


View this table:
[in this window]
[in a new window]

  		TABLE 1
The binding affinities of Cx43-(136–158) to CaM or dansyl-CaM in the presence of 1 mM Ca2+

 
Near UV CD spectroscopy was conducted to examine the tertiary packing around aromatic residues of CaM, Cx43-(136–158), and the CaM-Cx43-(136–158) complex (Fig. 3). Cx43-(136–158) exhibited negligible signal, but the addition of Cx43-(136–158) to Ca2+-CaM resulted in the appearance of more negative ellipticity (Fig. 3C). Both CaM and the synthetic peptide do not contain any tryptophan, but each contains two tyrosine residues. The prominent band above 275 nm is probably due to the tyrosine residues in the C terminus region of CaM (Tyr99 and Tyr138), whereas the two peaks at 262 and 268 nm are probably due to the nine phenylalanine residues in CaM (34). The difference spectrum obtained by subtracting the Ca2+-CaM spectrum from that of Ca2+-CaM-Cx43-(136–158) (Fig. 3D) showed a prominent band between 275 and 279 nm, which is most likely due to the immobilization of at least one of the two tyrosine residues in the Cx43-(136–158) peptide (which contains no tryptophan) following its binding to CaM. The contribution of signals from the CaM tyrosine residue is less likely to be dominant, although both CaM and Cx43-(136–158) contain two tyrosine residues, making it difficult to unambiguously differentiate the contribution from each macromolecule in the spectra. The Tyr99 and Tyr138 in CaM are located in the C terminus EF-loops, where the local structures of the Ca2+-bound EF-loops have not been reported to be altered upon peptide binding in the currently known CaM-peptide complex despite the rearrangement of tertiary structures. Thus, the local environment of the CaM tyrosine is less likely to be significantly altered upon Cx43-(136–158) peptide binding. No significant differences were observed between the spectra of apo-CaM and the 1:1 mixture of apo-CaM/Cx43-(136–158) (Fig. 3C), further supporting the Ca2+-dependent binding of Cx43-(136–158) to Ca2+-CaM and the resulting conformational changes in Cx43-(136–158) revealed by the far UV CD studies.


Figure 4
View larger version (19K):
[in this window]
[in a new window]

  		FIGURE 4.
Monitoring the interaction of CaM with Cx43-(136–158) by surface plasmon resonance spectroscopy. The sensorgrams represent the binding of 0.1, 1, 5, and 10 µM Cx43-(136–158) (A) or a 10 µM concentration of the randomized Cx43 peptide (B) to the CaM-immobilized chip in the presence of 5 mM Ca2+ (solid line) or EGTA (dashed line). CaM was immobilized to the sensor chip CM with a response unit of ~3500. The flow rate was set at 5 µl/min. During recording, running buffer contained 5 mM CaCl2 or 5 mM EGTA in 100 mM KCl, 50 mM Tris-HCl at pH 7.4 or the same buffer plus the peptides (solid bar). Subsequently, peptide-free buffer was injected to monitor the dissociation process.

 
Monitoring the CaM-Cx43-(136–158) Interaction by Surface Plasmon Resonance Spectroscopy—The binding of Cx43-(136–158) to CaM was further confirmed using surface plasmon resonance spectroscopy with CaM immobilized on the CM5 sensor chip. Fig. 4A shows the sensorgrams of successive peptide binding experiments. The peptide Cx43-(136–158) rapidly associated with CaM in the presence of Ca2+. At the end of peptide injection, dissociation of the complex was observed during flushing of the chip with peptide-free running buffer. From analysis of the kinetic data, we inferred that the apparent Kd for the binding of Cx43-(136–158) to CaM is ~1 µM with a koff rate of 5 x 10-4 s-1 in the presence of a saturating amount of Ca2+ and 100 mM KCl. In contrast, no significant binding of Cx43-(136–158) to CaM was detected in the presence of EGTA (Fig. 4A, dashed line). Furthermore, the control randomized Cx43-(136–158) peptide that contained the same composition of amino acids as Cx43-(136–158) did not exhibit specific binding in the presence of either Ca2+ or EGTA (Fig. 4B); the response units of these injections were not significantly higher than that of injection of the peptide through the underivatized CM5 chip surface. These results further confirm the Ca2+-dependent binding and specificity of Cx43-(136–158) binding to CaM.


Figure 5
View larger version (35K):
[in this window]
[in a new window]

  		FIGURE 5.
Interaction of Cx43-(136–158) with CaM or dansyl-CaM monitored by fluorescence spectroscopy. A, tyrosine fluorescence spectra of Cx43-(136–158) in the presence of 1 mM EGTA (dashed line) or 1 mM CaCl2 (solid line), CaM in the presence of 1 mM EGTA ({circ}) or 1 mM CaCl2 ({square}), and a 1:1 mixture of CaM/Cx43-(136–158) in 1 mM EGTA (•) or 1 mM CaCl2 ({blacksquare}). B, dansyl fluorescence spectra of dansyl-CaM (1.25 µM) in the presence of 1 mM EGTA ({circ}) or 1 mM CaCl2 ({square}) and a 1:1 mixture of CaM/Cx43-(136–158) in 1 mM EGTA (•) or 1 mM CaCl2 ({blacksquare}). Inset, the fluorescence intensity change at 490 nm plotted as a function of the peptide Cx43-(136–158)/CaM ratio. C, dansyl fluorescence spectra of Ca2+-dansyl-CaM (1.25 µM) in the absence ({square}) and presence ({blacksquare}) of the Cx43-(88–107) peptide (2.5 µM) in 1 mM CaCl2. D, dansyl fluorescence spectra of Ca2+-dansyl-CaM (1.25 µM) in the absence ({square}) and presence ({blacksquare}) of the Cx43-(222–247) peptide (2.5 µM) in 1 mM CaCl2.

 
Peptide-induced Conformational Changes Detected by Fluorescence Spectroscopy—The tyrosine fluorescence intensity of Cx43-(136–158) remained almost unchanged upon the addition of Ca2+. In contrast, the addition of Ca2+ enhanced the fluorescence intensity of both CaM and a 1:1 CaM/Cx43-(136–158) mixture (2.3- and 1.7-fold, respectively). The sum of the fluorescence intensities of Ca2+-CaM and Cx43-(136–158) nearly equaled that of the Ca2+-CaM-Cx43-(136–158) complex, suggesting that the fluorescence enhancement resulted primarily from the conformational change of CaM upon Ca2+ binding. However, the blue shift of the maximal emission peak from 307 nm to 304 nm was observed when Cx43-(136–158) bound to Ca2+-CaM (Fig. 5A), suggesting that the tyrosine environment (probably of those residues in the peptide as discussed above) became more hydrophobic after the formation of the CaM-Cx43-(136–158) complex.

The dansyl fluorescence emission maxima, at ~510 nm, is well separated from any intrinsic fluorescence arising from aromatic amino acid residues. Because of this, as well as the fluorescent dye's great sensitivity to changing environment, fluorescent dye-labeled proteins, including dansylated CaM, have been widely used in studying the effect of protein-target interaction of ions, peptides, or drugs (20, 35). As shown in Fig. 5B, Ca2+ binding to dansyl-CaM resulted in an increase of the fluorescence intensity and an emission blue shift from 510 to 497 nm, indicating that the dansyl group moved to a more hydrophobic environment. The addition of Cx43-(136–158) to Ca2+-dansyl-CaM resulted in a further 80% increase in dansyl fluorescence intensity and a further emission blue shift to 477 nm, suggesting that the dansyl groups are located in an even more hydrophobic environment in the complex. In the absence of Ca2+, the dansyl fluorescence remained nearly unaltered when Cx43-(136–158) was added to dansyl-CaM, indicating a Ca2+-dependent interaction of Cx43-(136–158) with dansyl-CaM, thus supporting a model in which the interaction of Cx43-(136–158) with CaM is Ca2+-dependent. In addition, the titration of dansyl-CaM with two other Cx43 peptides (Cx43-(86–107), corresponding to residues 86–107 in Cx43 in the N-terminal region of the intracellular loop, and Cx43-(224–247), corresponding to residues 224–247 in Cx43 in the C terminus region) that have significantly lower predictive scores for CaM binding (Fig. 1A) resulted in no significant fluorescence signal changes (Fig. 5, C and D), indicating that these peptides did not bind CaM. Titration data of CaM with Cx43-(136–158) in 1 mM CaCl2 resulted in apparent dissociation constants of 240 ± 10 nM in 10 mM KCl and 860 ± 20 nM in 100 mM KCl (Table 1), assuming a 1:1 CaM/Cx43-(136–158) binding mode (Fig. 5B, inset). These binding affinities are comparable with those obtained with far UV CD spectroscopy (Fig. 3); the differences in these values are probably attributable to the dansyl modification of CaM and/or the different methodologies used.


Figure 6
View larger version (36K):
[in this window]
[in a new window]

  		FIGURE 6.
Interaction between CaM and Cx43-(136–158) revealed by NMR. A, overlaid HSQC spectra of Ca2+-CaM (green) with Ca2+-CaM-Cx43-(136–158) (red) or Ca2+-CaM/control randomized Cx43-(136–158) peptide (blue). A subset of assigned peaks displaying significant movement upon peptide binding are framed in boxes. B, overlay of HSQC spectrum of apo-CaM (green) with apo-CaM-Cx43-(136–158) (red). C, changes in chemical shift of selected well dispersed resonances as a function of the Cx43-(136–158)/Ca2+-CaM ratio in the presence of saturating Ca2+ at physiological K+ condition. D, three-dimensional representation of perturbed residues in Ca2+-CaM (Protein Data Bank code 3CLN) upon binding of the synthetic peptide Cx43-(136–158). Residues with amide proton resonance changes over 0.05 ppm are shown in red. Calcium ions are shown in green.

 
NMR Studies on the CaM-Peptide Interaction—It was important to demonstrate that there is in fact a direct interaction between the Cx43-(136–158) and CaM. This was confirmed using NMR. As shown in Fig. 6A, in the presence of Ca2+, a number of well resolved resonances of CaM underwent chemical shift changes of greater than 0.05 ppm following the addition of Cx43-(136–158). In contrast, only insignificant changes in several residues were observed upon the addition of Cx43-(136–158) to CaM in the presence of 10 mM EGTA (Fig. 6B), again supporting the absence of any of interaction at physiological concentrations of K+ in the absence of Ca2+. In addition, the addition of the control randomized Cx43-(136–158) peptide (see "Experimental Procedures") resulted in only very minor changes in the CaM 15N-1H HSQC spectra in the presence of either Ca2+ or EGTA (Fig. 6A). This latter result would appear to exclude the possibility of nonspecific binding between the randomized peptide and CaM arising from the charge or hydrophobic interactions of certain residues, indicating the necessity of the specific arrangement of the Cx43-(136–158) peptide sequence.

The NMR resonances of Ca2+-CaM have been assigned by several groups (33, 36, 37). By following their movements during titration of CaM with the Cx43-(136–158) peptide, the well resolved resonances of the peptide-CaM complex have been unambiguously assigned. CaM has been known to activate downstream proteins by displacing autoinhibitory domains, remodeling the active sites, or inducing oligomerization in these proteins (38). The binding of CaM to target proteins and peptides has been reported to involve the N- and/or C-terminal domains of CaM (33, 38–40). Analysis of the effects of peptide binding on the backbone amide chemical shifts of CaM could shed light on the underlying mode of interaction of the Cx43-(136–158) peptide with CaM. The resonances that underwent chemical shift changes of greater than 0.05 ppm in the peptide titration were present in both the N and C termini of CaM, including residues Gly25, Thr29, Gly33, Gly40, Ala57, Gly61, Asp64, Lys77, Lys94, Gly98, Gly113, Thr117, Gly134, Asn137, Ala147, Lys148, and possibly others, since the assignment of the overlapped regions is still in progress (Fig. 6, A and D). The chemical shift movements of the residues that we were able to assign during the titration followed a similar trend (Fig. 6C), suggesting a single binding process. Furthermore, the chemical shifts saturated when the Cx43-(136–158)/CaM ratio exceeded 1.0 (Fig. 6C). These results demonstrated that the 1:1 Ca2+-dependent binding of Cx43-(136–158) to CaM induces conformational changes in both the N- and C-terminal domains of CaM.


Figure 7
View larger version (15K):
[in this window]
[in a new window]

  		FIGURE 7.
Effect of Cx43-(136–158) on the Ca2+ affinity of CaM. Shown is Ca2+ titration of CaM ({circ}), the 1:1 CaM/Cx43-(136–158) mixture (•), and the 1:1 CaM/randomized control peptide ({square}) in 100 mM KCl, 50 mM Tris-HCl, pH 7.3. The intrinsic tryrosine fluorescence emission intensity of CaM or CaM/peptide mixture (8 µM) was monitored at 307 nm with fluorescence excitation at 277 nm. The Ca2+ indicator dye Oregon Green 488 BAPTA-5N was used to calibrate the ionized Ca2+ concentration as described under "Experimental Procedures." Each titration point is indicated as an open diamond at the bottom. Shown are data from a single experiment representative of at least triplicate experiments.

 
Effect of Cx43-(136–158) on the Ca2+ Binding of CaM—To examine the effect of Cx43-(136–158) on the Ca2+ binding properties of CaM, Ca2+ titrations were performed in the presence of the Ca2+ indicator Oregon Green 488 BAPTA-5N. The Ca2+-dependent change in the intrinsic tyrosine fluorescence of CaM in the presence of Cx43-(136–158) was slightly left-shifted when compared with CaM alone (pCa50 = 5.8 ± 0.1 for the mixture versus 5.6 ± 0.1 for CaM) (Fig. 7). Furthermore, the Ca2+-dependent change in the tyrosine fluorescence of the CaM/Cx43-(136–158) mixture exhibited stronger positive cooperativity than that of CaM alone (Hill coefficient 3.3 versus 2.1) for Ca2+ binding (Table 2). It is well recognized that the Ca2+ affinity of CaM is significantly enhanced upon binding to its receptor protein as a result of a slower dissociation rate of Ca2+ from CaM-receptor complex than from CaM alone (41–44). The dissociation of Ca2+ from CaM typically leads to the inactivation or inhibition of its bound protein (40), so slowing this dissociation would ensure that the CaM-Cx43 complex remains associated even during periods of frequent oscillation of [Ca2+]i. Furthermore, a greater positive cooperativity for Ca2+ binding enables the complex to respond to a narrower window of [Ca2+]i with higher sensitivity.


View this table:
[in this window]
[in a new window]

  		TABLE 2
Effect of Cx43-(136–158) binding on the Ca2+ binding affinity of CaM Data were obtained from the Ca2+ titration curve (Fig. 7) (n = 3).

 
Physiological Effects of Knocking Out the Putative CaM-binding Site in Cx43—Physiological experiments were conducted to confirm the results obtained using various biophysical approaches. Cell-to-cell dye transfer under resting and elevated intracellular Ca2+ concentration ([Ca2+]i) in communication-deficient HeLa cells transiently transfected individually with either WT Cx43-EYFP or one of two Cx43-EYFP intracellular loop mutants (CaM-binding site knockouts) was used to demonstrate that CaM mediates the Ca2+-dependent inhibition of Cx43 by directly associating with the aa 136–158 region in Cx43.

HeLa cells transiently transfected with WT Cx43-EYFP expressed large gap junction plaques at the cell-cell interface (Fig. 8, A and D, left) as has been previously described (45); there was no significant intracellular expression of WT Cx43-EYFP, making it relatively straight forward to identify cells making gap junctions with adjacent cells (Fig. 8A). In contrast, both the Cx43K146E,R148E-EYFP and Cx43M147Q,L151E,I156E-EYFP CaM binding-deficient mutants expressed an abundance of protein when transfected in HeLa cells, most of which was in non-plasma membrane locations (Fig. 8, panels B and F and panels C and G (left), respectively), making it harder to identify cells making gap junctions with adjacent HeLa cells. Under resting [Ca2+]i conditions (1.8 mM extracellular Ca2+ concentration and no ionomycin), cell-to-cell communication (i.e. AlexaFluor 594 cell-to-cell dye transfer) was measured between each pair of cells expressing the WT Cx43-EYFP; every cell that contained EYFP fluorescence exhibited cell-to-cell transfer of injected dye, indicating that Cx43-EYFP formed functional gap junctions (Fig. 8I). In contrast, only 31% of the double mutant Cx43K146E,R148E-EYFP-transfected HeLa cells and 32% of the triple mutant Cx43M147Q,L151E,I156E-EYFP-transfected HeLa cells exhibited cell-to-cell dye transfer (Fig. 8I). These results indicated that a significant fraction of the mutant Cx43-EYFP-transfected cells that exhibited EYFP fluorescence did not express functional gap junctions. This fits with the EYFP fluorescence data (Fig. 8, B and C), suggesting that a significant fraction of the mutant Cx43-EYFP protein accumulates in non-plasma membrane locations in the transfected HeLa cells. Furthermore, mock-transfected HeLa cells did not exhibit cell-to-cell transfer of injected AlexaFluor 594 dye, confirming that the cell-to-cell transfer exhibited by the WT Cx43-EYFP- and mutant Cx43-EYFP-transfected HeLa cells is gap junction-mediated (Fig. 8D).

The addition of the membrane-permeant polyether antibiotic ionomycin (1 µM) and subsequent elevation of extracellular Ca2+ from 1.8 to 21.8 mM resulted in a significant (p < 0.001) increase of the [Ca2+]i to concentrations known to inhibit Cx43-mediated cell-to-cell dye transfer in a CaM-dependent manner (7, 8). As shown in Fig. 8I, elevation of [Ca2+]i in WT Cx43-EYFP-transiently transfected HeLa cells results in the inhibition of cell-to-cell dye transfer in all cells tested, indicating that the addition of EYFP on the C terminus of Cx43 did not affect the ability of elevated [Ca2+]i to inhibit Cx43 gap junctions. In contrast, elevation of [Ca2+]i in HeLa cells expressing the Cx43K146E,R148E-EYFP double mutant or the Cx43M147Q,L151E,I156E-EYFP triple mutant did not result in the inhibition of cell-to-cell dye transfer (57 and 67%, respectively, of attempts resulted in dye transfer). The apparent increase in the percentage of injected cells inhibiting cell-to-cell dye transfer in elevated [Ca2+]i versus resting [Ca2+]i was not statistically significant (p > 0.1). There was no cell-to-cell transfer of dye when AlexaFluor 594 dye was injected into mock HeLa-transfected cells under either resting or elevated [Ca2+]i conditions (Fig. 8H).


Figure 8
View larger version (38K):
[in this window]
[in a new window]

  		FIGURE 8.
Functional effect of a sustained elevation in intracellular [Ca2+] on CaM binding-deficient Cx43-EYFP mutants. Cell-to-cell transfer of injected AlexaFluor 594 dye between HeLa cells transiently transfected with WT Cx43-EYFP or the CaM-binding Cx43 mutants (Cx43K146E,R148E-EYFP or Cx43M147Q,L151E,I156E-EYFP) was measured in confluent monolayers of cells. A–H, images of transiently transfected WT or mutant Cx43-EYFP or mock-transfected HeLa cells. The left image shows the EYFP fluorescence of the injected and adjacent cells, the middle image is the fluorescence of EYFP plus brightfield, and the images on the right are of AlexaFluor 594 dye transfer. A and E, WT Cx43-EYFP-transiently transfected HeLa cells; B and F, Cx43K146E,R148E-EYFP-transiently transfected HeLa cells; C and G, Cx43M147Q,L151E,I156E-EYFP-transiently transfected HeLa cells; D and H, mock-transfected HeLa cells. I, summary data for cell-to-cell dye transfer (solid bars) and [Ca2+]i (hatched bars) in the absence or presence of elevated [Ca2+]i. For WT Cx43-EYFP and resting [Ca2+]i (0.075 ± 0.01 µM), 100% of injections resulted in dye transfer (n = 5); for elevated [Ca2+]i (0.911 ± 0.09 µM), 0% of injections resulted in dye transfer (n = 6). For Cx43K146E,R148E-EYFP and resting [Ca2+]i (0.129 ± 0.02 µM), 31% of injections resulted in dye transfer (n = 32); for elevated [Ca2+]i (1.77 ± 0.31 µM), 57% of injections resulted in dye transfer (n = 7). For Cx43M147Q,L151E,I156E-EYFP, resting [Ca2+]i (0.135 ± 0.06 µM), 32% of injections resulted in dye transfer (n = 22); for elevated [Ca2+]i (1.54 ± 0.24 µM), 67% of injections resulted in dye transfer (n = 6). For mock-transfected, resting [Ca2+]i (0.073 ± 0.01 µM), 0% of injections resulted in dye transfer (n = 3); elevated [Ca2+]i (1.85 ± 0.15 µM), 0% of injections resulted in dye transfer (n = 3). *, cell-to-cell dye transfer was significantly lower than cell-to-cell dye transfer determined in low [Ca2+]i; **, [Ca2+]i was significantly higher than the low [Ca2+]i value.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
Indication of in Vivo Interaction—In this report, we provide evidence that a 23-mer peptide derived from the intracellular loop region residues 136–158 of the major gap junction protein Cx43, which has the highest predictive score of any region of Cx43 for binding to CaM (Fig. 1), binds CaM with high affinity in a Ca2+-dependent manner. That intracellular CaM concentration is ~9 µM (46) indicates that such an affinity would imply that this CaM-binding region of Cx43 would likely be at least half-saturated with CaM when intracellular Ca2+ concentration is elevated to ~1 µM and that this Ca2+-dependent association accounts for the well described ability of Ca2+ to inhibit gap junctions (7). As indicated by our circular dichroism, fluorescence, and NMR data (Figs. 4, 5, 6), the stoichiometry of Cx43-(136–158) peptide/CaM binding is close to 1:1. To our knowledge, this report presents the first demonstration of a direct CaM-Cx43 interaction in vitro, which would explain our previous in vivo observation that preincubation of lens cell cultures or Cx43-transfected HeLa cells with CaM antagonists counteracts the intracellular Ca2+-dependent inhibition of Cx43-containing gap junctions (6, 7). These results clearly demonstrate that the Ca2+-dependent inhibition of Cx43 gap junctions is mediated via the association of Ca2+-CaM with the intracellular loop region of Cx43. In addition, our results support a mechanism of Cx43 regulation by CaM in which CaM first binds to Ca2+, and then the Ca2+-CaM complex associates with Cx43 to inhibit Cx43 gap junctions, rather than a stable CaM-Cx43 complex in which Ca2+ binds to the complex, which then effects the inhibition of Cx43 gap junctions.

A different model for the gating of Cx43 has been proposed by Delmar's laboratory (11), in which an intracellular gating element within the C-terminal domain of Cx43 interacts with a region of the Cx43 cytoplasmic loop in a pH-dependent manner. In that report, Duffy et al. (11) used surface plasmon resonance, ELISA, and NMR approaches to demonstrate that a portion of the C terminus of Cx43 comprising amino acids 255–382 associated with a peptide corresponding to the second half of the Cx43 intracellular loop comprising amino acids 119–144, which was significantly enhanced by low pH. The results reported here for the inhibition of Cx43 by Ca2+-CaM have an interesting corollary to this pH-dependent inhibition described by Duffy et al. (11) in that binding of a portion of the Cx43 C-terminal region or CaM to the cytoplasmic region of Cx43 inhibits this gap junction protein. However, whereas Sorgen et al. (47) reported that dimerization of the Cx43 C-terminal region may be one of the structural changes involved in the pH regulation of Cx43, our data reported here indicate that Ca2+-CaM and the intracellular loop peptide associate as a 1:1 complex (i.e. that a single Ca2+-CaM binds to Cx43 (i.e. 6 Ca2+-CaM/connexon hemichannel) to effect gap junction closure). Indeed, although there is partial overlap of the C-terminal binding domain on the intracellular loop reported by Duffy et al. (11) (residues 119–144) and the CaM binding domain (residues 136–158) reported here, it is clear that CaM will associate with a portion of Cx43 that is very close to the third transmembrane region of this connexins so it is well positioned to physically obstruct the gap junction pore.

As in any molecular study conducted in vitro, direct extrapolation between the data presented in this report using a peptide and the mechanisms of Ca2+-CaM gating of Cx43 in intact cells needs to be made with caution. However, it is worth noting that there are several examples in the literature demonstrating that two molecular domains that associate in living cells also associate as separate domains in vitro (48, 49). In the studies of CaM-target interaction, synthetic peptides have been widely used due to their excellent ability to mimic specific domains of native proteins (33, 50–54). Indeed, Peracchia et al. (4, 55) have shown that calmodulin is associated with gap junctions and plays a direct role in the chemical gating of Cx32-containing gap junctions. This was supported by data for CaM binding to Cx32 in which Török et al. (5) identified two distinct CaM binding amino acid sequences. The sequence of one of these sites in the N-terminal domain of Cx32 contains a CaM binding motif common to a large class of CaM-dependent proteins (22); notably, this N-terminal CaM binding sequence is absent in Cx43 (Fig. 1).

Indication of the Potential CaM-Cx43136–158 Binding Mode—Through its reversible binding of Ca2+ and the resultant conformational changes, CaM is capable of interacting with over 300 target proteins to regulate a range of cellular events (56). Although the sequence identity among the CaM targeted sequences is low, all of the consensus peptides for such sequences (18–26-mer peptides) possess common features, including their ability to form amphipathic {alpha}-helices, containing 2–6 positively charged residues, and their pattern of bulky hydrophobic residues that anchor the peptides to the hydrophobic cleft of CaM. Our results demonstrate that the Cx43-(136–158) sequence forms an {alpha}-helical structure in TFE as well as in the CaM-peptide complex, similar to other CaM-bound peptides (25, 30, 32, 57, 58). Two lysine and two arginine residues are contained in the Cx43-(136–148) sequence, which has a high predictive score for CaM binding. The basic residues in the N terminus of CaM-binding sequences have been proposed to ensure an antiparallel orientation of the peptide with respect to CaM (59), which is believed to optimize the electrostatic attraction between the basic residues of the peptide with negatively charged residues in CaM. In addition, the hydrophobic residues Met147-Leu151-Ile156 of the Cx43 peptide reflect the 1-5-10 type of CaM-binding pattern (Fig. 1), which is similar to the CaM-binding regions of CaMKI and CaMKII (12, 22, 57).

In addition to the formation of {alpha}-helical structure by the Cx43-(136–158) peptide, the hydrophobicity of the peptide environment as well as the tertiary arrangement of the CaM are also changed following the formation of the CaM-Cx43-(136–158) complex. As indicated by the near UV CD and tyrosine fluorescence studies, the chemical environment around the tyrosine residues is significantly perturbed following the binding of Cx43-(136–158) to CaM. Furthermore, the significant blue shift and concomitant enhancement of fluorescence intensity induced by Cx43-(136–158) binding to dansyl-CaM (Fig. 5B) suggests that the dansyl group in dansyl-CaM is shielded from the solvent and moves into a highly hydrophobic environment upon binding to Cx43-(136–158); similar changes have been observed for the binding of dansyl-CaM to other receptors (60–62). A number of the currently assigned CaM resonances in the NMR 1H-15N HSQC spectra exhibited chemical shift movements of ≥0.05 ppm. These residues were spread in both the N- and C-terminal domains of CaM as well as in the linker region, indicating that a global conformational change occurred upon binding of the Cx43-(136–158) peptide to CaM (Fig. 6D). Such global changes of amide chemical shifts have also been reported in other CaM complexes, such as Ca2+-CaM-skeletal muscle myosin light chain kinase and Ca2+-CaM-CaMKI (33, 63). Together, these observations suggest that the Ca2+-CaM-Cx43-(136–158) interaction might adopt the commonly seen wrapping-around mode of action (64).

In Vivo Functional Analysis of the Putative Cx43 CaM-binding Site—Transient expression of a Cx43-EYFP construct without and with mutations in communication-deficient HeLa cells provided a system by which to test our biophysical data in a physiologic manner. Because CaM interaction with a protein requires both an electrostatic interaction and a hydrophobic interaction, we generated two Cx43 mutants: Cx43K146E,R148E-EYFP, which knocks out the Cx43 electrostatic interaction with CaM, and Cx43M147Q,L151E,I156E-EYFP, which knocks out the hydrophobic interaction of Cx43 with CaM. Transient expression of these EYFP-tagged Cx43 constructs was consistent with the anticipated results. The WT Cx43-EYFP formed very large gap junction plaques, and dye transfer was restricted to those cells expressing EYFP fluorescence (see Fig. 8, A and E). With elevated [Ca2+]i, these junctions were no longer dye-coupled as reported by us previously with WT Cx43 that lacked the EYFP label.

Since CaM has been implicated in Cx32 assembly at two stages of oligomerization (65), it was anticipated that the Cx43K146E,R148E-EYFP and Cx43M147Q,L151E,I156E-EYFP mutants would probably encounter problems reaching the plasma membrane and perhaps even not form functional gap junctions. Indeed, although both of the mutant Cx43 proteins did express, the majority of the expressed protein was trapped inside the cell, very little of this protein appeared to reside in the plasma membrane, and no obvious gap junction plaques were observed (Fig. 8, B and C). Thus, it was not anticipated that either of the mutants would exhibit cell-to-cell transfer of injected dye as reproducibly as we observed with the Cx43-EYFP. This proved to be the case, because although in low (~1 µM) [Ca2+]i 100% of the WT Cx43-EYFP-transfected HeLa cells exhibited cell-to-cell transfer of injected dye, only ~30% of each of the mutant-transfected HeLa cells appeared to be functional and exhibited cell-to-cell transfer of injected dye (Fig. 8I). However, although elevated [Ca2+]i completely prevented cell-to-cell dye transfer between WT Cx43-EYFP-transfected HeLa cells, elevated [Ca2+]i was now unable to inhibit cell-to-cell dye transfer between HeLa cells transfected with these Cx43 mutants. Thus, knocking out the CaM binding capability of the residue 136–158 region of Cx43 abolished the Ca2+-dependent inhibition of Cx43 gap junctions, supporting our biophysical data that demonstrates that CaM mediates the Ca2+-dependent inhibition of Cx43 via its interaction with residues 136–158 in this connexin.

In summary, we have identified a CaM binding sequence in the ubiquitous gap junction protein Cx43. This sequence resides in a juxtamembrane region of the only intracellular loop of Cx43. Our results demonstrate a 1:1 Ca2+-dependent CaM-Cx43 peptide interaction with an affinity in the submicromolar range. The binding of this peptide to CaM enhances the Ca2+ affinity of CaM ~2-fold. This biophysical result was confirmed in physiologic experiments with site-directed mutations in the predicted CaM-binding region of this connexin. These results explain the molecular basis of our previously reported Ca2+-CaM-dependent regulation of both lens and Cx43-containing gap junctions via a direct interaction of CaM with this connexin protein (6, 7). The data reported here confirm that this regulation is effected via the Ca2+-dependent association of CaM with residues 136–158 of Cx43, which in turn effects a change in the structural organization of Cx43 such that gap junction permeability is significantly inhibited. Further elucidation of the structural changes within both CaM and Cx43 as well as the molecular basis of the Ca2+-dependent regulation of the other lens connexins is the subject of ongoing investigations in this laboratory.


   FOOTNOTES
 
* This work was supported by National Institutes of Health Grants EY-05684 (to C. F. L. and J. Y.) and GM62999 (to J. Y.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

Formula The on-line version of this article (available at http://www.jbc.org) contains supplemental Fig. S1. Back

1 Fellow of the Molecular Basis of Disease Area of Focus at Georgia State University. Back

2 To whom correspondence should be addressed: Dept. of Chemistry, Georgia State University, University Plaza, Atlanta, GA 30302. Tel.: 404-413-5520; Fax: 404-413-5551; E-mail: chejjy{at}langate.gsu.edu.

3 The abbreviations used are: Cx, connexin; aa, amino acids; CaM, calmodulin; Mops, 4-morpholinepropanesulfonic acid; TFE, trifluoroethanol; WT, wild type. Back



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Mathias, R. T., Rae, J. L., and Baldo, G. J. (1997) Physiol. Rev. 77, 21-50[Abstract/Free Full Text]
  2. Musil, L. S., Beyer, E. C., and Goodenough, D. A. (1990) J. Membr. Biol. 116, 163-175[CrossRef][Medline] [Order article via Infotrieve]
  3. Noma, A., and Tsuboi, N. (1987) J. Physiol. 382, 193-211[Abstract/Free Full Text]
  4. Peracchia, C., Sotkis, A., Wang, X. G., Peracchia, L. L., and Persechini, A. (2000) J. Biol. Chem. 275, 26220-26224[Abstract/Free Full Text]
  5. Torok, K., Stauffer, K., and Evans, W. H. (1997) Biochem. J. 326, 479-483[Medline] [Order article via Infotrieve]
  6. Gandolfi, S. A., Duncan, G., Tomlinson, J., and Maraini, G. (1990) Curr. Eye Res. 9, 533-541[Medline] [Order article via Infotrieve]
  7. Churchill, G. C., Lurtz, M. M., and Louis, C. F. (2001) Am. J. Physiol. Cell Physiol. 281, C972-C981[Abstract/Free Full Text]
  8. Lurtz, M. M., and Louis, C. F. (2003) Am. J. Physiol. Cell Physiol. 285, C1475-C1482[Abstract/Free Full Text]
  9. Burr, G. S., Mitchell, C. K., Keflemariam, Y. J., Heidelberger, R., and O'Brien, J. (2005) Biochem. Biophys. Res. Commun. 335, 1191-1198[Medline] [Order article via Infotrieve]
  10. Zhang, X., and Qi, Y. (2005) Arch. Biochem. Biophys. 440, 111-117[CrossRef][Medline] [Order article via Infotrieve]
  11. Duffy, H. S., Sorgen, P. L., Girvin, M. E., O'Donnell, P., Coombs, W., Taffet, S. M., Delmar, M., and Spray, D. C. (2002) J. Biol. Chem. 277, 36706-36714[Abstract/Free Full Text]
  12. Yap, K. L., Kim, J., Truong, K., Sherman, M., Yuan, T., and Ikura, M. (2000) J. Struct. Funct. Genomics 1, 8-14[CrossRef][Medline] [Order article via Infotrieve]
  13. Mitaku, S., Hirokawa, T., and Tsuji, T. (2002) Bioinformatics 18, 608-616[Abstract/Free Full Text]
  14. Krogh, A., Larsson, B., von Heijne, G., and Sonnhammer, E. L. (2001) J. Mol. Biol. 305, 567-580[CrossRef][Medline] [Order article via Infotrieve]
  15. Jones, D. T. (1998) FEBS Lett. 423, 281-285[CrossRef][Medline] [Order article via Infotrieve]
  16. Tusnady, G. E., and Simon, I. (2001) Bioinformatics 17, 849-850[Abstract/Free Full Text]
  17. Fruen, B. R., Black, D. J., Bloomquist, R. A., Bardy, J. M., Johnson, J. D., Louis, C. F., and Balog, E. M. (2003) Biochemistry 42, 2740-2747[CrossRef][Medline] [Order article via Infotrieve]
  18. Balog, E. M., Norton, L. E., Bloomquist, R. A., Cornea, R. L., Black, D. J., Louis, C. F., Thomas, D. D., and Fruen, B. R. (2003) J. Biol. Chem. 278, 15615-15621[Abstract/Free Full Text]
  19. Wallace, R. W., Tallant, E. A., and Cheung, W. Y. (1983) Methods Enzymol. 102, 39-47[Medline] [Order article via Infotrieve]
  20. Johnson, J. D., and Wittenauer, L. A. (1983) Biochem. J. 211, 473-479[Medline] [Order article via Infotrieve]
  21. Whitmore, L., and Wallace, B. A. (2004) Nucleic Acids Res. 32, W668-W673[Abstract/Free Full Text]
  22. Rhoads, A. R., and Friedberg, F. (1997) FASEB J. 11, 331-340[Abstract]
  23. Clapperton, J. A., Martin, S. R., Smerdon, S. J., Gamblin, S. J., and Bayley, P. M. (2002) Biochemistry 41, 14669-14679[CrossRef][Medline] [Order article via Infotrieve]
  24. Rosenberg, O. S., Deindl, S., Sung, R. J., Nairn, A. C., and Kuriyan, J. (2005) Cell 123, 849-860[CrossRef][Medline] [Order article via Infotrieve]
  25. Yamauchi, E., Nakatsu, T., Matsubara, M., Kato, H., and Taniguchi, H. (2003) Nat. Struct. Biol. 10, 226-231[CrossRef][Medline] [Order article via Infotrieve]
  26. Nicol, S., Rahman, D., and Baines, A. J. (1997) Biochemistry 36, 11487-11495[CrossRef][Medline] [Order article via Infotrieve]
  27. Yang, J. J., Buck, M., Pitkeathly, M., Kotik, M., Haynie, D. T., Dobson, C. M., and Radford, S. E. (1995) J. Mol. Biol. 252, 483-491[CrossRef][Medline] [Order article via Infotrieve]
  28. Lehrman, S. R., Tuls, J. L., and Lund, M. (1990) Biochemistry 29, 5590-5596[CrossRef][Medline] [Order article via Infotrieve]
  29. Brokx, R. D., Scheek, R. M., Weljie, A. M., and Vogel, H. J. (2004) J. Struct. Biol. 146, 272-280[CrossRef][Medline] [Order article via Infotrieve]
  30. Zhang, M., Yuan, T., and Vogel, H. J. (1993) Protein Sci. 2, 1931-1937[Medline] [Order article via Infotrieve]
  31. Zhang, M., and Vogel, H. J. (1994) Biochemistry 33, 1163-1171[CrossRef][Medline] [Order article via Infotrieve]
  32. Zhan, Q. Q., Wong, S. S., and Wang, C. L. (1991) J. Biol. Chem. 266, 21810-21814[Abstract/Free Full Text]
  33. Ikura, M., Clore, G. M., Gronenborn, A. M., Zhu, G., Klee, C. B., and Bax, A. (1992) Science 256, 632-638[Abstract/Free Full Text]
  34. Martin, S. R., and Bayley, P. M. (1986) Biochem. J. 238, 485-490[Medline] [Order article via Infotrieve]
  35. LaPorte, D. C., Keller, C. H., Olwin, B. B., and Storm, D. R. (1981) Biochemistry 20, 3965-3972[CrossRef][Medline] [Order article via Infotrieve]
  36. Ikura, M., Spera, S., Barbato, G., Kay, L. E., Krinks, M., and Bax, A. (1991) Biochemistry 30, 9216-9228[CrossRef][Medline] [Order article via Infotrieve]
  37. Chou, J. J., Li, S., Klee, C. B., and Bax, A. (2001) Nat. Struct. Biol. 8, 990-997[CrossRef][Medline] [Order article via Infotrieve]
  38. Hoeflich, K. P., and Ikura, M. (2002) Cell 108, 739-742[CrossRef][Medline] [Order article via Infotrieve]
  39. Elshorst, B., Hennig, M., Forsterling, H., Diener, A., Maurer, M., Schulte, P., Schwalbe, H., Griesinger, C., Krebs, J., Schmid, H., Vorherr, T., and Carafoli, E. (1999) Biochemistry 38, 12320-12332[CrossRef][Medline] [Order article via Infotrieve]
  40. Ikura, M., and Ames, J. B. (2006) Proc. Natl. Acad. Sci. U. S. A. 103, 1159-1164[Abstract/Free Full Text]
  41. Johnson, J. D., Snyder, C., Walsh, M., and Flynn, M. (1996) J. Biol. Chem. 271, 761-767[Abstract/Free Full Text]
  42. Kasturi, R., Vasulka, C., and Johnson, J. D. (1993) J. Biol. Chem. 268, 7958-7964[Abstract/Free Full Text]
  43. Brown, S. E., Martin, S. R., and Bayley, P. M. (1997) J. Biol. Chem. 272, 3389-3397[Abstract/Free Full Text]
  44. Olwin, B. B., Edelman, A. M., Krebs, E. G., and Storm, D. R. (1984) J. Biol. Chem. 259, 10949-10955[Abstract/Free Full Text]
  45. Lai, A., Le, D. N., Paznekas, W. A., Gifford, W. D., Jabs, E. W., and Charles, A. C. (2006) J. Cell Sci. 119, 532-541[Abstract/Free Full Text]
  46. Black, D. J., Tran, Q. K., and Persechini, A. (2004) Cell Calcium 35, 415-425[CrossRef][Medline] [Order article via Infotrieve]
  47. Sorgen, P. L., Duffy, H. S., Spray, D. C., and Delmar, M. (2004) Biophys. J. 87, 574-581[CrossRef][Medline] [Order article via Infotrieve]
  48. Sheng, M., and Sala, C. (2001) Annu. Rev. Neurosci. 24, 1-29[CrossRef][Medline] [Order article via Infotrieve]
  49. Mayer, B. J. (2001) J. Cell Sci. 114, 1253-1263[Abstract]
  50. Urbauer, J. L., Short, J. H., Dow, L. K., and Wand, A. J. (1995) Biochemistry 34, 8099-8109[CrossRef][Medline] [Order article via Infotrieve]
  51. Martin, S. R., Bayley, P. M., Brown, S. E., Porumb, T., Zhang, M., and Ikura, M. (1996) Biochemistry 35, 3508-3517[CrossRef][Medline] [Order article via Infotrieve]
  52. Xiong, L. W., Newman, R. A., Rodney, G. G., Thomas, O., Zhang, J. Z., Persechini, A., Shea, M. A., and Hamilton, S. L. (2002) J. Biol. Chem. 277, 40862-40870[Abstract/Free Full Text]
  53. Kranz, J. K., Lee, E. K., Nairn, A. C., and Wand, A. J. (2002) J. Biol. Chem. 277, 16351-16354[Abstract/Free Full Text]
  54. Ishida, H., and Vogel, H. J. (2006) Protein Pept. Lett. 13, 455-465[CrossRef][Medline] [Order article via Infotrieve]
  55. Peracchia, C. (2004) Biochim. Biophys. Acta 1662, 61-80[Medline] [Order article via Infotrieve]
  56. Berridge, M. J., Bootman, M. D., and Lipp, P. (1998) Nature 395, 645-648[CrossRef][Medline] [Order article via Infotrieve]
  57. Meador, W. E., Means, A. R., and Quiocho, F. A. (1993) Science 262, 1718-1721[Abstract/Free Full Text]
  58. Yamniuk, A. P., and Vogel, H. J. (2004) J. Biol. Chem. 279, 7698-7707[Abstract/Free Full Text]
  59. Osawa, M., Tokumitsu, H., Swindells, M. B., Kurihara, H., Orita, M., Shibanuma, T., Furuya, T., and Ikura, M. (1999) Nat. Struct. Biol. 6, 819-824[CrossRef][Medline] [Order article via Infotrieve]
  60. Gomes, A. V., Barnes, J. A., and Vogel, H. J. (2000) Arch. Biochem. Biophys. 379, 28-36[CrossRef][Medline] [Order article via Infotrieve]
  61. Trost, C., Bergs, C., Himmerkus, N., and Flockerzi, V. (2001) Biochem. J. 355, 663-670[Medline] [Order article via Infotrieve]
  62. Turner, J. H., Gelasco, A. K., and Raymond, J. R. (2004) J. Biol. Chem. 279, 17027-17037[Abstract/Free Full Text]
  63. Mal, T. K., Skrynnikov, N. R., Yap, K. L., Kay, L. E., and Ikura, M. (2002) Biochemistry 41, 12899-12906[CrossRef][Medline] [Order article via Infotrieve]
  64. Bhattacharya, S., Bunick, C. G., and Chazin, W. J. (2004) Biochim. Biophys. Acta 1742, 69-79[Medline] [Order article via Infotrieve]
  65. Ahmad, S., Martin, P. E., and Evans, W. H. (2001) Eur. J. Biochem. 268, 4544-4552[Medline] [Order article via Infotrieve]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 J. Biol. Chem.Home page
R. Dodd, C. Peracchia, D. Stolady, and K. Torok
Calmodulin Association with Connexin32-derived Peptides Suggests trans-Domain Interaction in Chemical Gating of Gap Junction Channels
J. Biol. Chem., October 3, 2008; 283(40): 26911 - 26920.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
282/48/35005    most recent
M707728200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
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 PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zhou, Y.
Right arrow Articles by Yang, J. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zhou, Y.
Right arrow Articles by Yang, J. J.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals   Molecular and Cellular Proteomics 
 Journal of Lipid Research   ASBMB Today 
Copyright © 2007 by the American Society for Biochemistry and Molecular Biology.
Advertisement
spacer
Advertisement
Advertisement