HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 28, 21121-21129, July 14, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/28/21121    most recent M002986200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Zühlke, R. D. Articles by Reuter, H. Search for Related Content PubMed PubMed Citation Articles by Zühlke, R. D. Articles by Reuter, H. Social Bookmarking                      What's this?

Ca²⁺-sensitive Inactivation and Facilitation of L-type Ca²⁺ Channels Both Depend on Specific Amino Acid Residues in a Consensus Calmodulin-binding Motif in the_1C subunit^*

Roger D. Zühlke^§, Geoffrey S. Pitt^¶, Richard W. Tsien^¶, and Harald Reuter

From the  Department of Pharmacology, University of Bern, 3010 Bern, Switzerland and the ^¶ Department of Molecular and Cellular Physiology, Stanford University Medical School, Stanford, California 94305-5426

Received for publication, April 9, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

L-type Ca²⁺ channels are unusual in displaying two opposing forms of autoregulatory feedback, Ca²⁺-dependent inactivation and facilitation. Previous studies suggest that both involve direct interactions between calmodulin (CaM) and a consensus CaM-binding sequence (IQ motif) in the C terminus of the channel's _1C subunit. Here we report the functional effects of an extensive series of modifications of the IQ motif aimed at dissecting the structural determinants of the different forms of modulation. Although the combined substitution by alanine at five key positions (Ile¹⁶²⁴, Gln¹⁶²⁵, Phe¹⁶²⁸, Arg¹⁶²⁹, and Lys¹⁶³⁰) abolished all Ca²⁺ dependence, corresponding single alanine replacements behaved similarly to the wild-type channel (77wt) in four of five cases. The mutant I1624A stood out in displaying little or no Ca²⁺-dependent inactivation, but clear Ca²⁺- and frequency-dependent facilitation. An even more pronounced tilt in favor of facilitation was seen with the double mutant I1624A/Q1625A: overt facilitation was observed even during a single depolarizing pulse, as confirmed by two-pulse experiments. Replacement of Ile¹⁶²⁴ by 13 other amino acids produced graded and distinct patterns of change in the two forms of modulation. The extent of Ca²⁺-dependent facilitation was monotonically correlated with the affinity of CaM for the mutant IQ motif, determined in peptide binding experiments in vitro. Ca²⁺-dependent inactivation also depended on strong CaM binding to the IQ motif, but showed an additional requirement for a bulky, hydrophobic side chain at position 1624. Abolition of Ca²⁺-dependent modulation by IQ motif modifications mimicked and occluded the effects of overexpressing a dominant-negative CaM mutant.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

As part of its ubiquitous role in controlling Ca²⁺-dependent cellular processes, calmodulin (CaM)¹ is an important regulator of many types of ion channels, both voltage- and ligand-gated (1-5). Prominent among these are L-type Ca²⁺ channels, which open in response to membrane depolarization and activate contractions of heart cells, trigger hormone secretion from endocrine cells, and control gene transcription in neurons. It is generally agreed that the passage of Ca²⁺ ions through L-type channels exerts profound feedback effects on the subsequent opening and closing of their pore-forming _1C subunit (6, 7). Both inhibitory and facilitatory forms of autoregulation have been described, and both appear to involve CaM. The inhibitory form, Ca²⁺-dependent inactivation, has long been recognized to promote the shutting off of channel conductance during a maintained depolarization (8, 9). Hypotheses about the involvement of CaM in Ca²⁺-dependent inactivation have ranged widely, the earliest considering direct phosphorylation of the channel by CaM kinase II or its dephosphorylation by the CaM-dependent phosphatase calcineurin (10-13). The enhancing form of autoregulation, Ca²⁺-dependent facilitation, is seen most prominently in L-type channels when basal Ca²⁺ is elevated (14, 15) or when the channels are repeatedly activated by trains of depolarizations (16-18). Facilitation has also been attributed to an action of CaM kinase II (18-22), the strongest evidence coming from studies with constitutively active CaM kinase II applied to excised patches (23).

Recent experiments have implicated direct binding of CaM to L-type channels as a key step in both types of autoregulation. As first shown by Zühlke and Reuter (24), partial deletion of a consensus CaM-binding IQ motif in the cytoplasmic C-terminal tail of _1C eliminated Ca²⁺-dependent inactivation. Ca²⁺-dependent binding of CaM to peptides containing the IQ motif (25-27) has also been demonstrated. Overexpression of CaM mutants with defective Ca²⁺-binding sites exerts a dominant-negative effect on both Ca²⁺-dependent inactivation (25-27) and facilitation (25). Mutations in the IQ motif that eliminate the CaM binding also prevent both forms of modulation (25).

Although all of these observations support the involvement of Ca²⁺-dependent CaM binding to the L-type channel in both inactivation and facilitation, they raise the question of how both forms of autoregulation might arise from a single Ca²⁺ sensor (25). This complexity exceeds that found in any other CaM-regulated ion channel so far. As a first step toward understanding the dual action of CaM, we have examined the contributions of individual amino acids of the IQ motif to Ca²⁺-sensitive inactivation and facilitation of L-type channels expressed in Xenopus oocytes. A comprehensive series of alanine substitutions were carried out at five positions (Ile¹⁶²⁴, Gln¹⁶²⁵, Phe¹⁶²⁸, Arg¹⁶²⁹, and Lys¹⁶³⁰), followed up by systematic replacement of the critical residue Ile¹⁶²⁴ by 13 other amino acids. Analysis of the electrophysiological properties of the various mutant channels supported the following general conclusions. First, Ca²⁺-dependent inactivation and facilitation coexist as distinct processes, separable by different kinetics of onset and decay and emphasized to differing extents by various voltage protocols. Second, when both Ca²⁺-dependent gating mechanisms are operative, Ca²⁺-dependent inactivation can mask an underlying facilitation (here we use "inactivation" or "facilitation" to refer to underlying mechanisms rather than the net behavior that results from these processes working in opposition). Third, although both processes require CaM binding to the IQ motif, inactivation and facilitation depend on significantly different structural determinants within this motif.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

cDNA Construction and Site-directed Mutagenesis-- The L-type Ca²⁺ channel _1C,77 subunit, which we call 77wt, was originally cloned from human fibroblasts (28) and subsequently assembled in an expression-competent vector to form the 77wt-containing plasmid pHLCC77 (29). This cDNA construct was the basis for the creation of all mutant Ca²⁺ channels described in this paper. As template for site-directed mutagenesis, we used the plasmid p77NB (24), which contains an NsiI-BamHI fragment from pHLCC77 that encodes the entire C-terminal end of 77wt. We used different polymerase chain reaction strategies for introducing single or multiple amino acid mutations into the IQ motif of 77wt (see Fig. 1A). The mutant channels 77IQ/2A and 77FRK/3A were constructed using the QuikChange site-directed mutagenesis kit (Stratagene GmbH, Zürich, Switzerland) according to the manufacturer's instructions. A 220-bp BglII fragment containing the mutations was then transferred into p77NB before subcloning a Van91I-BamHI fragment harboring the mutated BglII segment into full-length pHLCC77. The 77IQFRK/5A mutant was constructed by triple-fragment ligation of a 400-bp Van91I-RsaI fragment from 77IQ and a 1850-bp RsaI-BamHI fragment from 77FRK/3A into pHLCC77. The construction of 77I/A, 77I/E, and 77I/V has already been described (25). The introduction of additional amino acid substitutions at Ile¹⁶²⁴ generally followed the same procedure. To replace Ile¹⁶²⁴ by Met, Ser, Arg, Phe, Leu, Cys, and Trp, an antisense oligonucleotide was synthesized that contained the degenerate codon WKS (W is A, T; K is G, T; S is G, C) in place of the wild-type ATC codon. In addition, an adjacent silent mutation (TTC to TTT) disrupted an EcoNI restriction site, which was used to screen for mutations. With this mutant oligonucleotide, we performed the polymerase chain reaction and used the amplified mutant 400-bp Van91I-RsaI fragment for triple-fragment ligation as described above, although with a wild-type RsaI-BamHI fragment instead of a 77FRK fragment. The remaining amino acid replacements at position 1624 (Ile to Gly, Ile to Thr, and Ile to Tyr) as well as the construction of 77IQ/CA were performed accordingly using individual non-degenerate oligonucleotides. Finally, the alanine substitutions at Phe¹⁶²⁸, Arg¹⁶²⁹, and Lys¹⁶³⁰ were introduced by a similar strategy. The amplified 615-bp RsaI-AatII fragment containing the individual mutations was used together with a wild-type Van91I-RsaI fragment for triple-fragment ligations directly into pHLCC77. All mutations were confirmed by sequencing the entire amplified restriction fragment in the final cDNA construct. The synthesis of oligonucleotides and the cDNA sequencing were performed by a commercial laboratory (Microsynth AG, Balgach, Switzerland). Restriction enzymes were purchased from New England Biolabs Inc. (Bioconcept, Allschwil, Switzerland), MBI Fermentas (Mächler AG, Basel, Switzerland), and Roche Molecular Biochemicals (Rotkreuz, Switzerland). DNA isolation and purification systems were from Promega (Catalys AG, Wallisellen, Switzerland) and QIAGEN (Basel).

In Vitro RNA Transcription and Microinjection into Xenopus Oocytes-- Mutant and wild-type Ca²⁺ channel cDNAs were linearized with BamHI prior to in vitro RNA transcription with an Ambion T7 mMessage mMachine kit (AMS Biotechnology, Lugano, Switzerland). The auxiliary Ca²⁺ channel subunits ₁ (30) and ₂ (31) as well as the calmodulin mutant CaM(3) (4) were transcribed using 25 units of SP6 RNA polymerase (Stratagene GmbH) with the T7 mMessage mMachine kit after appropriate linearization of the template cDNAs. For microinjection into Xenopus oocytes, cRNAs encoding wild-type or mutant Ca²⁺ channel _1C subunits and auxiliary channel subunits were mixed at equimolar concentrations and diluted with diethyl pyrocarbonate-treated water. Injection pipettes were pulled with a DMZ Universal Puller (Zeitz Instruments GmbH, Augsburg, Germany) and had tip openings of 20-25 µm. Using a Nanoject Microinjector (Drummond Scientific Co., Bromall, PA), we routinely injected 23 nl of cRNA mixture (0.05-0.33 pmol/µl) into freshly isolated and defolliculated Xenopus oocytes (32). Optimal Ca²⁺ channel expression depended on the total amount of injected cRNA and was obtained 4-7 days after injection. For the coexpression of Ca²⁺ channels with CaM(3) it was necessary to inject large amounts of CaM(3) cRNA (up to 200 fmol) at least 3 days prior to Ca²⁺ channel cRNA injections to obtain a strong time- and dose-dependent dominant-negative effect by CaM(3) on L-type Ca²⁺ currents (I_Ca). Microinjected Xenopus oocytes were kept at 17-19 °C in a sterile Barth's medium containing 88 mM NaCl, 1 mM KCl, 0.82 mM MgSO₄, 0.33 mM Ca(NO₃)₂, 0.41 mM CaCl₂, 2.4 mM NaHCO₃, and 10 mM HEPES (adjusted to pH 7.5 with NaOH) and supplemented with 100 units/ml penicillin and 100 µg/ml streptomycin (Roche Molecular Biochemicals).

Electrophysiological Recordings-- Three to seven days after injection of Ca²⁺ channels, oocytes were used to analyze electrophysiological properties of the expressed Ca²⁺ channels with a standard two-electrode voltage clamp configuration using an oocyte clamp OC-725C amplifier (Warner Instrument Corp., Hamden, CT) connected through a CED 1401 A/D interface (CED, Cambridge, United Kingdom) to a personal computer. Before recording whole cell I_Ba or I_Ca, oocytes were injected with 23-46 nl of 100 mM BAPTA solution (pH 7.4) to minimize contaminating Ca²⁺-activated Cl currents. The injection of higher amounts of BAPTA resulted in a decrease in I_Ca and in a reduction of Ca²⁺-dependent inactivation.² The tip diameter of the BAPTA injection pipettes was only 10-15 µm to minimize leakage after injections. During I_Ba recordings, oocytes were constantly superfused with a solution containing 40 mM Ba(OH)₂, 50 mM NaOH, 1 mM KOH, and 10 mM HEPES (adjusted to pH 7.4 with methanesulfonic acid). To evoke I_Ca, the bath solution was switched to a solution containing Ca(NO₃)₂ instead of Ba(OH)₂. In general, I_Ba and I_Ca were obtained in the same oocyte, and oocytes initially expressing I_Ba < 500 nA were not used.

Current recordings have been acquired and, where possible, analyzed with CED Patch and Voltage Clamp software. Ionic currents were filtered at 0.5 kHz and sampled at 2 kHz. Additional analysis of current properties as well as statistical analysis of current parameters were done using MS Excel 97 (Microsoft Corp., Redmond, WA) and GraphPAD Prism 3.0 (GraphPAD Software for Science, San Diego CA). All values are given as means ± S.E. In some cases, error bars were smaller than graphical symbols and thus not visible (e.g. Fig. 3A). In general, statistical comparisons between mean values of parameters were done by paired Student's t test. Where appropriate, unpaired t tests or analysis of variance has been used.

Quantitative Measurements of CaM-Peptide Interactions-- Purified CaM (Calbiochem) was labeled with dansyl chloride (Molecular Probes, Inc., Eugene, OR) as recommended in the manufacturer's protocol. Emission fluorescence spectra of dansyl-CaM alone and after titration of peptides were obtained on a Perkin-Elmer LS 50 B luminescence spectrometer. Wild-type and mutant peptides were synthesized at the Protein and Nucleic Acid Center at Stanford University (Stanford, CA) and were resuspended in water. All experiments were performed with 33 nM dansyl-CaM in buffer containing 150 mM NaCl, 50 mM Tris-HCl (pH 7.4), and 1 µM free Ca²⁺ (buffered with HEDTA) to which concentrated solutions of peptide were added. The excitation wavelength was 340 nm, and both the excitation and emission band passes were set at 15 nm. The fraction of bound dansyl-CaM was calculated from the fractional increase in fluorescence intensity at 480 nM using the relationship f_b = (I_m  I_f)/(I_b  I_f), where I_m is the measured fluorescence, I_f is the fluorescence when no peptide is added, and I_b is the fluorescence when all dansyl-CaM is bound. K_d values were determined using the Hill equation f_b = 1/(1 + 10^{(log k_dX)k}), where X is the peptide concentration in moles/liter and k is the Hill coefficient.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Alanine Substitutions within the IQ Motif-- Our previous analysis (25) was based on mutations at a single position, isoleucine 1624, the first of 12 residues in a consensus IQ motif in the C-terminal tail of the _1C subunit (residues 1624-1635) (Fig. 1A). Those experiments led to the conclusion that both Ca²⁺-dependent inactivation and facilitation depend upon CaM binding to the IQ motif. We have tested this idea further with an extensive series of mutations and detailed kinetic analysis. Previous work has shown that Ca²⁺-dependent inactivation can be eliminated by deletion of the first eight amino acids of the IQ motif, but not the last four residues (24). Therefore, we focused our attention on the first five conserved positions in the consensus IQ motif (Fig. 1A, underlined boldface letters). Currents carried by Ba²⁺ and Ca²⁺ ions were recorded in the same oocytes to assess the effects of Ca²⁺ entry on the decline of current during a standard depolarizing pulse to +20 mV. Fig. 1B compares the Ca²⁺ dependence of current decay in a series of alanine mutations with that found in 77wt channels. Ca²⁺-dependent inactivation was prominent with the 77wt channel and with the mutants 77Q/A, 77F/A, 77R/A, and 77K/A. On the other hand, the Ca²⁺ dependence of decay was barely detectable with 77FRK/3A and completely absent with 77I/A, 77IQ/2A, and 77IQFRK/5A. The widely ranging Ca²⁺ dependence was the most striking feature of this series of mutants, although variations in the rate of Ba²⁺ current decay were also discernible and deserve further study in their own right. The examples displayed in Fig. 1B were representative of pooled data from many oocytes, shown in Fig. 1C. To assess the effect of Ca²⁺-dependent inactivation in a more quantitative fashion, we measured residual Ca²⁺ and Ba²⁺ currents at 100 ms, expressed them as a fraction of peak I_Ca (r_100Ca) and I_Ba (r_100Ba), and calculated the ratio as the Ca²⁺-dependent fraction, f₁₀₀ = (r_100Ba/r_100Ca)  1. 

View larger version (30K): [in this window] [in a new window]   Fig. 1.   Single and multiple alanine substitutions of conserved amino acids in the IQ motif of 77wt have distinct effects on Ca2+-dependent inactivation. A, shown is the alignment of a C-terminal region with the CaM-binding IQ motif (boldface letters; amino acids 1624-1635) in 77wt and its alanine substitution mutants. B, shown are macroscopic ion currents through wild-type (77wt) and mutant Ca2+ channels expressed in Xenopus oocytes evoked by a 400-ms depolarizing test pulse from Vh 90 to + 20 mV. IBa and ICa were recorded in the same oocytes and were scaled to peak IBa. C, as a quantitative measure for Ca2+-dependent inactivation, we calculated the normalized Ca2+-dependent fraction of inactivation, f100 = (r100Ba/r100Ca)  1, where r100Ba and r100Ca are the residual current fractions of IBa and ICa at 100 ms, respectively. The number of experiments was 4-11. *, p < 0.01 (paired Student's t test).

A particularly interesting result was obtained with the 77IQ/2A mutant. In this case, I_Ca declined considerably more slowly than I_Ba, a reversal of the usual difference in decay. In this respect, the 77IQ/2A construct was unique among the mutations tested. Two possibilities for the interpretation of the Ca²⁺ sensitivity of 77IQ/2A may be considered. The mutation alters the expression of Ca²⁺ sensitivity, to retard inactivation rather than to accelerate it, or the mutation changes the balance between two distinct processes, inactivation (which decreases current in a Ca²⁺-dependent way) and facilitation (which increases current in a Ca²⁺-dependent way). To decide between these possibilities, one may consider the results of multiple-pulse experiments. If only inactivation operates, be it slow or fast, currents evoked by a train of closely spaced pulses must become progressively smaller than the current evoked following a long quiescence. If, on the other hand, inactivation were counteracted by a genuine facilitatory process (strong enough to counteract inactivation), currents during repeated depolarizing pulses could gradually increase.

Accordingly, the various alanine replacement mutants were tested with a pulse train protocol (40 identical pulses from 90 to +20 mV for 50 ms, applied at 0.5 and 3.3 Hz). As plotted in Fig. 2A, the peak inward currents during successive depolarizations were normalized to currents evoked during the first depolarization so that facilitation would appear as an increase above unity. With Ba²⁺ as charge carrier, no facilitation was observed in any of the channel constructs. Likewise, with Ca²⁺ as charge carrier, facilitation could not be detected for either 77wt or 77IQFRK/5A. In contrast, the mutant channel 77IQ/2A displayed a very prominent facilitation of I_Ca such that currents grew pulse by pulse until they reached an amplitude at the end of the train nearly twice as large as after a long rest. This facilitation was seen only at 3.3 Hz, not at 0.5 Hz. A much smaller degree of facilitation was evident with 77FRK/3A, once again only at the higher pulse frequency, not at the lower one. Fig. 2B summarizes pooled data from experiments with pulse trains for the same series of alanine substitutions shown in Fig. 1. 77IQ/2A stands out as the construct with the greatest degree of facilitation, followed by 77I/A and then 77FRK/3A. Focusing mainly on 77IQ/2A, a working hypothesis is that the slower decay of I_Ca relative to I_Ba (Fig. 1B) arises from the same underlying facilitatory mechanism as expressed in experiments with pulse trains (Fig. 2A). In agreement with this hypothesis, a smaller but significant retardation of I_Ca inactivation relative to I_Ba was also found for 77I/A, which also exhibits marked facilitation (25).

View larger version (29K): [in this window] [in a new window]   Fig. 2.   Strong frequency-dependent facilitation of ICa can be observed only in 77I/A and 77IQ/2A. A, to test mutant Ca2+ channels for frequency-dependent ICa facilitation, a train of 40 50-ms test pulses from Vh 90 to +20 mV were applied at low (0.5 Hz; circles) and high (3.3 Hz; squares) frequencies. Peak IBa (open symbols) and ICa (closed symbols), normalized to peak currents of the first test pulse of each train, are plotted for 77wt (n = 5-7) and the multiple alanine substitution mutants 77IQ/2A (n = 8-12), 77FRK/3A (n = 5-8), and 77IQFRK/5A (n = 4-6). B, the normalized ICa changes at the end of each train of 40 test pulses are summarized for all alanine substitution mutants. The number of experiments was 3-12. *, p < 0.01 (paired Student's t test).

The next question was whether facilitation is specific to particular variants of 77wt or whether it is a widespread property of L-type channels, but is masked to varying degrees by inactivation. To try to dissect the contributions of facilitation and inactivation, we used a conventional two-pulse protocol. Systematic variations in the interval between the two identical depolarizing pulses were carried out to emphasize possible differences in the rate of relaxation of inactivation and facilitation (Fig. 3A). Effects specifically dependent on Ca²⁺ entry were highlighted by comparison of the size of I_Ca relative to I_Ba at each recovery interval. The simplest result was obtained with the construct 77IQFRK/5A. The time courses of fractional recovery of I_Ca and I_Ba were virtually identical at all interpulse intervals, consistent with the exposure of an inactivation process that was independent of Ca²⁺ entry. In contrast, in the strongly facilitating mutant 77IQ/2A, fractional recovery of I_Ca proceeded with a strikingly biphasic time course, rising to a peak of 1.4 at an interval of 0.1 s before returning much more slowly back toward 1. The increase above unity is yet another expression of the facilitation process; here, the added kinetic information is that facilitation is much slower to subside than inactivation. In the case of 77FRK/3A, there is only a small overshoot above 1.0 and a much faster decay (Fig. 3A), whereas mutant 77I/A showed facilitation and decay rates between those of 77IQ/2A and 77FRK/3A (data not shown). Evidently, the extent of facilitation and the kinetics of its decay are both affected by the various alanine replacements.

View larger version (28K): [in this window] [in a new window]   Fig. 3.   The presence of ICa facilitation in 77wt and most alanine substitution mutants can be unveiled by repriming experiments. A, fractional recovery from inactivation was obtained by a two-pulse protocol. The length of the conditioning prepulse from Vh 90 to +20 mV was varied for each experiment to reach a residual current of ~10% during inactivation. After increasing pulse intervals, peak currents evoked by test pulses to +20 mV were calculated as fractions of peak current during the conditioning prepulse and are plotted against the pulse intervals. B, maximal differences (max) between fractional recovery of ICa and IBa for each alanine substitution mutant are plotted. All but one mutant Ca2+ channel (77IQFRK/5A) exhibited significant facilitation of ICa. The number of experiments was three to seven. *, p < 0.01 (paired Student's t test).

The observations with mutant channels provide a useful perspective for interpreting the results with the wild-type channel itself. Fractional recovery of 77wt currents never exceeded unity with either Ca²⁺ or Ba²⁺ as charge carrier. At interval times shorter than ~0.05 s, the fractional recovery of I_Ca fell short of that of I_Ba, as expected for Ca²⁺-dependent inactivation. However, recovery of I_Ca clearly exceeded that of I_Ba at times longer than ~0.05 s. The difference was sustained over the 0.5-s interval examined in these experiments, although by definition, it must have decayed to zero at even longer intervals that allowed restoration of the resting state. For 77wt, the maximal difference in fractional recovery of I_Ca versus I_Ba was >0.2 (Fig. 3B), typical of the single alanine mutants listed in Fig. 1A. These differences are expressions of an underlying facilitatory process present in almost all alanine constructs, 77IQFRK/5A being the notable exception. Although the facilitation is not generally overt, it was readily uncovered by examination of differences in fractional recovery.

Could the differential effects of recovery from inactivation between I_Ca and I_Ba in 77wt simply be explained by differences in the abilities of the two divalent ions to screen surface charges (33, 34)? This is unlikely because with conventional two-pulse protocols (prepulse durations of 2 s), the isochronic inactivation curve of I_Ba was shifted to more positive potentials than that of I_Ca, with half-maximal inactivation voltages for I_Ba and I_Ca of 6.1 and 15.2 mV, respectively (24). The difference in midpoint voltages of inactivation was opposite to what would be required to explain the faster recovery of I_Ca relative to I_Ba.

Importance of Hydrophobic Amino Acids in the First Position of the IQ Motif-- The preceding survey of the effects of alanine replacement at several positions along the IQ motif pointed to Ile¹⁶²⁴, the first position of the IQ motif, as the most important residue among those studied. Relative to other individual alanine replacements, I1624A had a particularly strong effect on Ca²⁺-dependent inactivation (Fig. 1C) and, probably not coincidentally, revealed the greatest amount of facilitation with pulse trains (Fig. 2B). Accordingly, we proceeded to examine the consequences of replacing Ile¹⁶²⁴ with a series of different amino acids (Fig. 4). The electrophysiological results obtained with these constructs, 13 in all, are illustrated by current traces in panel A and by quantitative indices in panels C-E. Currents carried by Ca²⁺ or Ba²⁺ were evoked at a test potential of +20 mV in the same oocytes and normalized by scaling peak I_Ca to peak I_Ba. Together with 77wt, the mutant channels are ordered according to the hydrophobicity of amino acid 1624. Fig. 4B gives the index of hydrophobicity according to the Kyte-Doolittle scale (35).

View larger version (38K): [in this window] [in a new window]   Fig. 4.   Ca2+-dependent inactivation and facilitation of isoleucine mutants are dependent upon the hydrophobic character of the substituting amino acid. A, 13 amino acid residues have been substituted for Ile1624, the first amino acid in the IQ motif, and ionic currents through expressed mutant channels were recorded in Xenopus oocytes as described in the legend to Fig. 1B. B, representative pairs of IBa and ICa were aligned according to the hydrophobicity of the amino acid present at position 1624. C-E, the Ca2+-dependent fraction of inactivation (f100), the maximal difference of fractional recovery from inactivation (max), and the normalized ICa changes during trains of test pulses at 3.3 Hz, respectively, have been calculated and are plotted as described in the legends to Figs. 1C, 3B, and 2B, respectively. The numbers of experiments were 3-11 (C), 3-7 (D), and 2-12 (E). *, p < 0.01 (paired Student's t test).

As shown in Fig. 4A, Ca²⁺-dependent inactivation was prominent in all the channels with relatively hydrophobic residues at position 1624 over the range of hydrophobicity values between isoleucine and methionine. However, this type of inactivation disappeared virtually completely for alanine, glycine, and the increasingly hydrophilic amino acid replacements. This finding was further supported by pooled data for the Ca²⁺-dependent fraction of inactivation, f₁₀₀ (Fig. 4C), using the same analysis illustrated in Fig. 1C. f₁₀₀ was not significantly different from zero for the more hydrophilic residues. These results indicate that Ca²⁺-dependent inactivation is highly sensitive to the hydrophobicity of the first amino acid in the IQ motif.

Facilitation of I_Ca also varied systematically with alterations of the amino acid at position 1624. One expression of this can be found in the Ca²⁺ dependence of fractional recovery in two-pulse experiments (Fig. 4D). The maximal positive difference between fractional recovery of I_Ca and I_Ba was greater for the more hydrophobic residues, with values consistently >0.25 for amino acids with a positive hydrophobic index, dropping sharply to <0.1 for residues with a negative index. It is notable that the first group ends with alanine, whereas the second begins with glycine: this puts the cutoff between strong and weak Ca²⁺ dependence of fractional recovery at a different position along the amino acid series than the sudden disappearance of significant Ca²⁺-dependent inactivation.

Both aspects of Ca²⁺ dependence, inactivation and facilitation, would be expected to participate strongly in governing the overall dependence of peak I_Ca on repetitive stimulation. The collected data in Fig. 4E show that this was the case for the series of Ile¹⁶²⁴ replacements. The mutant that stands out is the 77I/A replacement. In this mutant, trains of depolarizing pulses (3.3 Hz) caused a >50% growth in peak current with Ca²⁺ (but not with Ba²⁺) as charge carrier (25). In contrast, the changes in peak Ca²⁺ current were much smaller in magnitude for the other mutants (<20%). The unusual response of 77I/A makes sense in light of the properties previously described. The alanine mutant lacks the pronounced Ca²⁺-dependent inactivation of its more hydrophobic counterparts (Fig. 4C), but it retains the strong Ca²⁺ dependence of facilitation, expressed by two-pulse measurements of fractional recovery (Fig. 4D). The combination of changes can account for the overt facilitation seen with trains of depolarizations.

Effects of Modifications at Position 1625-- Cysteine replacement of isoleucine produced a channel (77I/C) that was unusual in displaying prominent Ca²⁺-dependent inactivation as well as overt facilitation of I_Ca over the course of repetitive depolarizations (Fig. 5). This provided a welcome opportunity to test the importance of the residue adjoining Ile¹⁶²⁴, glutamine 1625. As already pointed out, when introduced in the context of I1624A, the modification Q1625A strongly favored facilitation over inactivation (Fig. 2B). To determine if this held true more generally, we prepared the mutant channel 77IQ/CA for sake of comparison with 77I/C. Ca²⁺-dependent inactivation was significantly reduced and facilitation at 3.3 Hz almost doubled with 77IQ/CA relative to 77I/C (Fig. 5). Thus, the relative change between 77I/C and 77IQ/CA (Fig. 5D) was like that between 77I/A and 77IQ/2A. Although this incremental effect of the 77Q/A modification was similar, it is worth noting that it alone reduced only the ratio of inactivation (Fig. 1C) and did not enhance facilitation (Figs. 2B and 3B). Evidently, the 77Q/A replacement has an additive effect on facilitation, but only if Ile¹⁶²⁴ is already replaced with a less hydrophobic residue.

View larger version (28K): [in this window] [in a new window]   Fig. 5.   Mutant channel 77I/C exhibits both Ca2+-dependent inactivation and frequency-dependent facilitation. An additional alanine substitution for the adjacent Gln1625 induces a strong reduction of Ca2+-dependent inactivation and, like 77IQ/2A, enhances frequency-dependent facilitation. A, representative IBa and ICa for 77I/C and 77IQ/CA. B, Ca2+-dependent inactivation (f100) of 77IQ/CA (n = 9), when compared with 77I/C (n = 6), is strongly inhibited. C and D, frequency-dependent ICa facilitation of 77I/C and 77IQ/CA. Both mutants lack facilitation when stimulated at 0.5 Hz, but show statistically significant ICa facilitation at 3.3 Hz (77I/C, n = 3-10; 77IQ/CA, n = 4-9; p < 0.05 by analysis of variance and Bonferroni's test). In 77IQ/CA, frequency-dependent facilitation was almost doubled compared with 77I/C (p < 0.05 by analysis of variance and Bonferroni's test).

Expression of an Enzymatically Inactive CaM Mutant Inhibits Both Inactivation and Facilitation-- In a previous publication (25), a calmodulin mutant that contained alanines instead of aspartates in three of the four coordination sites for Ca²⁺ binding, CaM(3) (4), was coexpressed in oocytes together with L-type Ca²⁺ channel subunits. This produced a strong dominant-negative effect on both Ca²⁺-dependent inactivation (in 77wt) and Ca²⁺-dependent facilitation (in 77I/A). Similar effects on Ca²⁺-dependent inactivation in _1C were observed with other CaM mutants expressed in a mammalian cell line (26). In the present study, we extended these results by examining the impact of CaM(3) on a wider set of channel constructs at different voltages. The stringent predictions to be tested were as follows: first, that the mutant calmodulin would attenuate the Ca²⁺/CaM-sensitive modulation, be it inactivation or facilitation; and second, that the mutant CaM would cause no change in the voltage dependence of the modulation, which presumably arises from gating of Ca²⁺ influx itself. As a strategy, this dominant-negative approach provided a way of discriminating between effects thought to be dependent on a constitutively resident CaM (25-27) and actions of CaM on enzymes such as CaM kinase II, which does not bind CaM in its resting state and requires Ca²⁺-bound CaM for activation (36).

The functional effects of CaM(3) coexpression on 77wt, 77I/C, 77I/E, and 77IQ/2A channels were tested with a classical two-pulse voltage protocol (Fig. 6A). During a conditioning prepulse lasting 100 ms, the membrane potential was stepped from 90 to 40 mV or to progressively more positive voltages in 10-mV increments. After a single prepulse and a 50-ms interval at 90 mV, a second pulse (test pulse) was imposed to a fixed level of +20 mV. I_Ba and I_Ca evoked by the test pulse were both recorded from the same oocytes. The peak currents at test pulse depended on the extent to which Ca²⁺ influx was activated by the prepulse. This dependence was bell-shaped in the case of I_Ca (Fig. 6A, right), but not so with I_Ba (see Ref. 24). I_Ca at test pulse was smallest in magnitude (least negative) when the current during the prepulse was largest (Fig. 6A, right). To correct for any voltage-dependent inactivation, we focused on the difference I_Ca  I_Ba, measured at the test pulse following identical prepulses. This parameter was calculated after scaling the peak I_Ba values by a factor that equalized I_Ba and I_Ca at test pulse after prepulses to 40 mV and then plotted against the prepulse potential (Fig. 6C). The difference was bell-shaped for the Ca²⁺-dependent component of inactivation (channels 77wt and 77I/C) and U-shaped for facilitation (channel 77IQ/2A). This procedure was done in oocytes that were injected with cRNAs for the wild-type or mutant channels alone (controls) or, in another set of experiments, together with cRNA for CaM(3) (Fig. 6B). CaM(3) produced a large and significant reduction in the size of the net Ca²⁺-dependent modulation, regardless of whether inactivation or facilitation was dominant (Fig. 6, C and D). The blockade by a functionally inactive CaM reinforced the idea that inactivation and facilitation are both strongly dependent on an action of indwelling native CaM. In contrast to the other mutants, construct 77I/E, the mutant that exhibited neither Ca²⁺-dependent inactivation nor facilitation, supported currents whose behavior was not further affected by coexpression of CaM(3) (Figs. 7, B-D). Thus, alteration of the CaM-binding domain occluded the dominant negative effect of the CaM mutant.

View larger version (35K): [in this window] [in a new window]   Fig. 6.   Ca2+-dependent inactivation and facilitation can be reduced by coexpressing Ca2+ channels with a mutant calmodulin, CaM(3), which lacks three out of four Ca2+-binding sites. A, the two-pulse protocol (left) was used to investigate the dependence of ICa inactivation and facilitation on Ca2+ influx and the corresponding normalized peak ICa (right) during prepulses (PP) and the following test pulse (TP). B, shown are the normalized current traces of 77wt and of three mutant channels (77I/C, 77I/E, and 77IQ/2A) with distinct Ca2+-dependent properties and both prepulses and test pulses to +20 mV for control ICa (), ICa in channels coexpressed with CaM(3) (), and, for comparison, control IBa (dashed line). C, differences of control peak currents at test pulse (ICa  IBa) () show bell-shaped Ca2+-dependent inactivation components for 77wt and 77I/C, U-shaped Ca2+-dependent facilitation for 77IQ/2A, and a lack of any Ca2+-dependent effect for 77I/E. The Ca2+ dependence was greatly reduced by coexpression of CaM(3) (). D, as a measure for Ca2+ sensitivities, the maxima of ICa  IBa were plotted. Significant Ca2+ sensitivity was observed for 77wt (n = 6-12), 77I/C (n = 4), and 77IQ/2A (n = 2-5), but not for 77I/E (n = 4). *, p < 0.05 (unpaired t test). Co, control.

View larger version (30K): [in this window] [in a new window]   Fig. 7.   Ca2+-sensitive facilitation unmasked after recovery from inactivation depends on Ca2+/CaM. A-C, the fractional recovery from inactivation was determined as described in the legend to Fig. 3. The time courses of repriming were plotted for 77wt (A), 77I/C (B), and 77IQ/2A (C). The clear facilitation of ICa () compared with IBa () exhibited by all three channels was strongly reduced or eliminated when measured after coexpressing CaM(3) (). D, the maximal differences (max) between fractional recovery of ICa and IBa with (open bars) and without (closed bars) CaM(3) are plotted. The number of experiments was two to seven. Co, control.

If frequency-dependent facilitation and the increased rate of recovery from inactivation represent two different expressions of facilitation with the same underlying mechanism, then the latter should also be suppressed by coexpression of CaM(3) with channel proteins. Indeed, this was true for channels 77wt, 77I/C, and 77IQ2A (Fig. 7). Irrespective of whether recovery of I_Ca from inactivation exceeded unity (77IQ/2A, 77I/C) or not (77wt), the differences in the rates of fractional recovery between I_Ca and I_Ba, the index of facilitation, disappeared with coexpression of dominant-negative CaM(3). This finding strongly supports our notion that the accelerated rate of fractional recovery of I_Ca is a genuine expression of facilitation, even if this positive modulation is often concealed by rapid Ca²⁺-dependent inactivation.

Relationship between CaM Binding and Channel Function-- Several groups (25-27) have directly demonstrated CaM binding to sequences of the C terminus of _1C that contain the IQ motif. CaM binds to an IQ motif peptide in a 1:1 stoichiometry, and the interaction depends on the presence of Ca²⁺ (25). In the present experiments, we studied the interaction of dansylated CaM with peptides encompassing the entire IQ motif of 77wt (_1C sequence from Tyr¹⁶¹⁹ to Gly¹⁶³⁸) or with peptides corresponding to _1C mutants. In the presence of 1 µM free Ca²⁺, the increase in the intensity of emission at 480 nm, an index of CaM-peptide interactions, was enhanced with increasing peptide concentration. The binding curves of the 77wt peptide and its mutants are compared in Fig. 8. Relatively small changes in the binding affinities were observed with 77IQ/2A, 77I/V, 77I/C, 77I/A, and 77I/F, whereas CaM-peptide affinities were increasingly reduced with 77I/T, 77FRK/3A, 77IQFRK/5A, and 77I/E. The measured dissociation constants (K_d) ranged from 10^7.45 for 77IQ/2A to 10^5.25 for 77IQFRK/5A, a variation of at least 2 orders of magnitude.

View larger version (23K): [in this window] [in a new window]   Fig. 8.   Mutations within the IQ motif alter the affinity of CaM-peptide interactions. Interactions between CaM and wild-type or mutant IQ peptides were assayed by examining the fractional increase in dansyl-CaM fluorescence at 480 nM and plotting the calculated fraction of bound dansyl-CaM against the peptide concentration. Solid lines are fits of the data to the Hill equation. For more details, see "Experimental Procedures." The number of experiments was two to four.

Using the CaM-peptide interactions as a gauge of how strongly the Ca²⁺·CaM complex might interact with the corresponding region of the channel, we went on to examine the possible relationship to different forms of Ca²⁺-dependent modulation. Fig. 9 shows the available indices of Ca²⁺-dependent facilitation (panel A) and inactivation (panel B), each plotted against log(K_d), a measure of the strength of the binding interaction. For facilitation (Fig. 9A), we used the maximal difference in fractional recovery, max (I_Ca  I_Ba) (Figs. 3B and 4D). This index showed a monotonic relationship with the strength of binding, ranging from 77I/E and 77IQFRK/5A, which bound weakly and showed no facilitation, to 77IQ/2A, which bound most tightly and displayed the strongest facilitation. For inactivation (Fig. 9B), we used the parameter f₁₀₀, which is zero when currents carried by Ca²⁺ and Ba²⁺ share the same course and increases in relationship to how much more rapidly I_Ca inactivates relative to I_Ba (Figs. 1B and 4C). Data for 77wt and mutants with relatively hydrophobic residues at position 1624 (Fig. 9B, closed symbols) showed a monotonic relationship between the extent of inactivation and the binding affinity, similar to that found for facilitation. However, this was not the case for the constructs containing more hydrophilic residues in place of Ile¹⁶²⁴, namely 77I/A, 77IQ/2A, 77I/T, 77I/E, and 77IQFRK/5A (Fig. 9B, open symbols). In each case, no significant inactivation was detected (Fig. 4C) regardless of the strength of binding.

View larger version (31K): [in this window] [in a new window]   Fig. 9.   Correlations of CaM-peptide interaction affinities (Kd values) with Ca2+-dependent facilitation and inactivation. The maximal differences (max) between fractional repriming of ICa and IBa after inactivation (A) and the Ca2+-dependent fraction of inactivation (f100) (B) are plotted against log(Kd), a measure of the strength of CaM-peptide interactions. Solid lines are exponential fits through data points presented by the closed symbols.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Two Distinct Forms of Ca²⁺/CaM-dependent Regulation of L-type Channels-- The present experiments firmly establish the importance of Ca²⁺/CaM binding to the IQ motif as a common point of regulation of both inactivation and facilitation, in extension of our previous work (25). Inactivation and facilitation have long been known to modify L-type Ca²⁺ channel function (6), but up until recently, these processes had been regarded as completely unrelated, in mechanism as well as outcome. On the contrary, by testing effects of a systematic series of amino acid substitutions along the length of the IQ motif with electrophysiological recordings and binding experiments, we have demonstrated that CaM interaction with this motif is necessary for both forms of Ca²⁺-dependent modulation, facilitatory as well as inhibitory. The detailed structural analysis led us to the following general conclusions. Ca²⁺-dependent inactivation and facilitation coexist as distinct processes, separable by their different kinetics of onset and decay. In wild-type channels, inactivation is more rapid to develop but also faster to subside than facilitation. Thus, voltage protocols that focused on channel behavior at longer times after the initial depolarization emphasized facilitation. On the other hand, when both Ca²⁺-dependent gating mechanisms are operative, Ca²⁺-dependent inactivation can easily mask an underlying facilitation. Although both processes require CaM interaction with the IQ motif and are abolished when binding is sufficiently weakened, inactivation and facilitation depend on significantly different structural determinants within this motif. Systematic amino acid substitutions at the first position in the IQ motif revealed that the hydrophobicity of the side chain is a key factor in promoting Ca²⁺-sensitive inactivation.

Various Forms of Modulation Exhibited by Specific Mutants-- The role of CaM binding to the IQ motif was highlighted by the wide range of channel properties seen with specific channel variants. The mutants 77IQFRK/5A and 77I/E illustrated how the channel behaved when CaM binding to the IQ motif was greatly weakened (K_d > 100-fold greater than for binding to the corresponding 77wt peptide). No Ca²⁺-dependent modulation was detected with any of the voltage clamp protocols. As tested with 77I/E, the suppression of both inactivation and facilitation by modification of the IQ domain eliminated any further effect of CaM(3) (Fig. 6), indicating that both interventions interfere with a common pathway(s) for signal transduction. There was no evidence for effects of CaM independent of the CaM-IQ domain interaction.

Of all the mutants we have examined, 77IQ/2A represents the most extreme case of one type of modulation dominant over the other. In two-pulse experiments, I_Ca evoked by the first depolarizing pulse was followed by over-recovery of I_Ca to an ~40% greater amplitude during the second pulse. The overshoot cannot be accounted for by inactivation in its previously described forms (8, 37), which would predict a monotonic return of I_Ca to its rested-state amplitude. Thus, there can be no doubt about the existence of facilitation, not merely inactivation.

Ca²⁺ current facilitation would be easy to miss in wild-type channels because the recovery in two-pulse experiments is monotonic and because I_Ca fails to grow when the oocytes are stimulated with trains of pulses. However, careful consideration of differences in fractional recovery with Ca²⁺ or Ba²⁺ as charge carrier suggests that Ca²⁺-dependent facilitation is present. Ca²⁺-dependent inactivation outweighs facilitation over short recovery intervals (<0.05 s), only to be overtaken by facilitation at longer intervals. Indeed, the maximal difference in fractional recovery of I_Ca versus I_Ba was >20%, pointing to an underlying facilitatory process of significant magnitude, even if it is not overtly seen because of concurrent inactivation. This conclusion was strongly reinforced by effects of CaM(3), which almost completely nullified the extra recovery of Ca²⁺ currents relative to Ba²⁺ currents (Fig. 7).

Distinctive Structural Determinants for Inactivation and Facilitation-- Because Ca²⁺-sensitive inactivation and facilitation were jointly eliminated by preventing CaM binding, either through overexpression of CaM(3) or by mutation of the IQ domain, we have concluded that the same CaM molecule may act as a sensor for both processes (25). We have obtained some useful clues as to how different forms of modulation might arise from a common signal transduction step through closer examination of their structural requirements. The extensive series of isoleucine 1624 replacements was particularly informative. Ca²⁺-dependent inactivation showed a remarkable correlation with the hydrophobicity (and possibly bulkiness) of the amino acid side chain at position 1624 (Fig. 4). This requirement was not simply secondary to changes in the affinity for Ca²⁺/CaM; the 77I/A substitution at position 1624 or the 77IQ/2A replacement at positions 1624 and 1625 spared high affinity CaM binding to the model peptide, yet largely eliminated Ca²⁺-dependent inactivation. Facilitation also varied with the identity of the amino acid at position 1624, although this was not so clearly separable from changes in CaM affinity. Focusing on inactivation, it appeared that this process depended jointly on binding of Ca²⁺/CaM plus some further step. One possible interpretation is that side chains of amino acids at these positions were partially exposed, even in the presence of CaM, and may somehow have been recognized by other regions of the channel. Ample evidence exists for involvement of other sequences between transmembrane segment IVS6 and the IQ motif, including the so-called EF-hand motif. Its deletion eliminates Ca²⁺-dependent inactivation (24, 38), and mutations in the F-helix of this motif effectively reduce Ca²⁺-sensitive current decay (39). On the other hand, facilitation was found to depend upon phosphorylation of the L-type channel or an associated protein by the protein kinase CaM kinase II (19-23). What remains unclear is whether and how the direct interaction between CaM and the channel might be linked to such phosphorylation.

Functional Implications-- We have studied two opposing modulatory effects on L-type Ca²⁺ channel activity that are set in motion by the same basic step in signal transduction, the Ca²⁺-dependent binding of CaM to the IQ domain. Kinetic distinctions between inactivation and facilitation are useful not only for distinguishing between these processes, but also for developing a rationale for their coexistence. In the heart, the kinetics of Ca²⁺-dependent inactivation seem well suited for moment-to-moment regulation within a cardiac action potential or shortly thereafter, whereas facilitation seems to operate as a multibeat phenomenon, responsive to stimulus frequency. Localization of CaM at the C terminus of _1C puts it in a good position to sense the build-up of Ca²⁺ in the general vicinity of the cytoplasmic mouth of the channel. A single locus of initiation for inactivation and facilitation offers possible functional advantages as a simple and efficient control mechanism. From this starting point, bifurcating signaling to different effector branches leaves open ample possibilities for further modulatory control. It is interesting to compare our recordings from oocytes, where facilitation may be masked by inactivation, with data from heart cells, where overt facilitation has often been seen with repeated trains of depolarizations (16-18, 40, 41) or with a two-pulse protocol such as shown in Fig. 3 (42). This leads us to suspect that powerful mechanisms may exist for changing the relative weights of inactivation and facilitation, acting beyond the CaM-IQ motif interaction, the joint initiator of these processes.

	ACKNOWLEDGEMENTS

We thank J. Adelman and J. Maylie for the CaM(3) cDNA and H. van Hees for technical assistance.

	FOOTNOTES

^* This work was supported by Swiss National Science Foundation Grants 31-56904.99 (to R. D. Z.) and 31-29862.90 (to H. R.), by a Pfizer postdoctoral fellowship (to G. S. P.), and by National Institutes of Health Grants HL03743 (to G. S. P.) and NS24067 and GM58234 (to R. W. T.).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.

^§ To whom correspondence should be addressed: Pharmakologisches Inst., Universität Bern, Friedbühlstrasse 49, CH-3010 Bern, Switzerland. Tel.: 41-31-632-32-90; Fax: 41-31-632-49-92; E-mail: roger.zuehlke@pki.unibe.ch.

Published, JBC Papers in Press, April 21, 2000, DOI 10.1074/jbc.M002986200

² R. D. Zühlke, unpublished observations.

	ABBREVIATIONS

The abbreviations used are: CaM, calmodulin; bp, base pair; BAPTA, 1,2-bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid Na₄ salt; dansyl, 5-dimethylaminonaphthalene-1-sulfonyl; HEDTA, N-hydroxyethylethylenediaminetriacetic acid.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit