Purification, Characterization, and Biosynthesis of

0 downloads 0 Views 2MB Size Report
Potassium channels comprise a family of proteins that have been classified ..... 5). After partial purification of the fusion protein and cleavage with factor X., protease, rMgTX was purified to .... Amp*-ampicillin resistance, Ori-origin. Lower panelĀ ...
THEJOURNAL OF B I O L N ~ CCHEMISTRY ~ 0 1993by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 268. No. 25, Issue of September 5, pp. 1886618874,1993 Printed in U.S.A.

Purification, Characterization, andBiosynthesis of Margatoxin, a Component of Centruroides margaritatusVenom That Selectively Inhibits Voltage-dependent Potassium Channels* (Fkceived for publication, March 25, 1993)

Margarita Garcia-CalvoS, Reid J. Leonard, Jon Novick, Scott P. Stevens, William Schmalhofer, Gregory J. Kaczorowski, and MariaL. Garciat From the Department of Membrane Biochemistry and Biophysics, Merck Research Laboratories, Rahway, New Jersey 07065

A novel peptidyl inhibitor of K+ channels has been purified to homogeneity from venom of the new world scorpion Centruroides margaritatus. Theprimary structure of this 39-amino-acid peptide, which we term margatoxin (MgTX), was determined by amino acid compositional analysis and peptide sequencing. Margatoxin potently inhibits bindingof radiolabeled charybdotoxin (ChTX) tovoltage-activatedchannelsin brain synaptic plasma membranes. Like ChTX, MgTX blocks the n-type current of human T-lymphocytes (Kv1.3 channel), but compared to ChTX, is 20-fold more potent (half-block at -50 PM), has a slower dissociation rate, and hasno effect on calcium-activated channels. To demonstrate that these characteristicsare due solely to the purified toxin, recombinant MgTX was expressed in Escherichia coli as part of a fusion protein. After cleavageand folding, purified recombinant MgTX displayed the same properties as native peptide. Replacement of the COOH-terminal histidine residue of MgTX with asparagine resulted ainpeptide a with a 10-fold reduction inpotency. This was due to faster apparent dissociation rate, suggesting that the COOH-terminal amino acid may play an important role in the binding of MgTX to the KJ.3 channel. MgTX displayssignificant sequence homology withpreviously identified K+ channel inhibitors (e.g. ChTX, iberiotoxin, noxiustoxin, and kaliotoxin). However, given its potency and unique selectivity,MgTX represents an especially useful tool with which to study the physiologic role of Kv1.3 channels.

Potassium channelscomprise a family of proteins that have been classified according to their biophysical and pharmacological characteristics. These channels modulate a number of cellular events such as muscle contraction, neuro-endocrine secretion, frequency and duration of action potentials, electrolyte homeostasis, and resting membrane potential (1).The biochemical characterization of K+ channels is underdeveloped, due to the paucity of selective high affinity probes. However, molecular biological approaches are providing insight into the structure of some of these proteins. Individual cDNA clones have been shown to encode distinct voltagedependent K+ channelsthat exhibit defined properties when expressed heterologously (2-6). Site-directed mutagenesis has

* 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 tassium fact. .f Recipient of a Fulbright Fellowship. 5 To whom correspondence should be addressed.

defined regions in these proteins that areinvolved in channel activation and inactivation (7-12), pore formation (13-15), and thatconfer sensitivity to quaternary ammonium (13, 1619) and peptidyl (20, 21) inhibitors. Functional voltage-gated K+ channels canexist as tetrameric structures (22) formed by the association of either identical or dissimilar subunits (10, 23-25). This phenomena is thought to account for the wide diversity of K+ channels found in different tissues (26-31). Despite these rapid advances in the molecular biology of K+ channels, subunitcompositions of native K+ channels and the physiologic role that particular channels play are, in most cases, still unclear. T o address these problems, potent, selective probes for channels of interest must be identified. Several peptidyl toxins, purified from scorpion venoms, interact with specific types of K+ channels. Charybdotoxin (ChTX)' is the best studied of all thesetoxins. ChTX is a 37amino-acid peptide isolated from venom of the oldworld scorpion Leiurus guinqwstriutus var. hebraeus (32). Originally described as an inhibitor of the high conductance, Ca2+activated K+ (Maxi-K) channel present in muscle and neuroendocrine cells (33), ChTX was later found to inhibit a number of different medium and small conductance Ca2+activated K+ channels(34-37), as well as a voltage-dependent K+ channel (KJ.3) that is found in neurons (38),blood cells (39-411, and osteoclasts (42). In each case, channel inhibition occurs with similar potency, in the low nanomolar range. Radiolabeled ChTX hasbeen used to characterize high affinity binding sites associated with Maxi-K channels in smooth muscle (43), and with voltage-activated potassium channels in brain (44). A related toxin, iberiotoxin (IbTX), shares68% sequence homology with ChTX and blocks selectively the Maxi-K channel (45). Other peptidyl toxins homologous to ChTX have been identified (e.g. limbatustoxin (46), kaliotoxin (47), and noxiustoxin (48)), but their selectivities are still being characterized. Venom of the new world scorpion Centruroides mrgaritatus was determined to contain an activity selectively directed against voltage-dependent K+ channels: it inhibited binding of [1261]ChTXto Kv1.3 channels in rat brain synaptosomal membranes but not to Maxi-K channels in smooth muscle sarcolemma. In the present study we report the purification from this venom of margatoxin (MgTX),itsprimary sequence, characterization as a Kv1.3inhibitor, and the expresThe abbreviations used in the paper are: ChTX, charybdotoxin; MgTX, margatoxin; rMgTX, recombinant margatoxin; [IIChTX, monoiodotyrosine charybdotoxin; IbTX, iberiotoxin; NxTX, noxiustoxin; KTX, kaliotoxin; Maxi-K channel, high-conductance, calcium activated potassium channel; KJ.3 channel, voltage dependent pochannel; PAGE, polyacrylamide gel electrophoresis; HPLC, high performance liquid chromatography; D'M', dithiothreitol; Tricine, N-tris(hydroxymethy1)methylglycine.

18866

Margatoxin Inhibition of K,1.3 Channels

18867

for each experimental point,and thedata were averaged. The standard deviation of the mean was typically less than 2%. Electrophysiological Experiments-Ionic currents were recorded from human peripheral T-lymphocytes using the whole cell patch clamp technique (50). Purified human T-cells were obtained as previously described (41), and recordings were carried out in a bath solution consisting of 125 mM NaCl, 2.5 mM KCl, 1.0 mMCaC12,2.0 mM MgClZ,10 mM HEPES (potassium salt), 10 mM glucose, 5 mM sodium-pyruvate, and 0.05% bovine serum albumin. The pH was adjusted to 7.2, and experiments were performed at room temperature EXPERIMENTAL PROCEDURES (19-22 "C). The intracellular (pipette) solution contained 140 mM Materials-Lyophilized venom from the scorpion C. margaritatus potassium-glutamate, 1mMMgC12, 10 mM HEPES (potassium salt), was obtainedfrom Miller International Venoms, Hollywood, FL. and 0.1 mM EGTA. The pHwas adjusted to 7.2. Patch pipettes were ChTX was purified from venom of the scorpion L. quinquestriatus fabricated from DAGAN (Minneapolis, MN) LG-16 glass and had var. hebraeus as previously described (32). ['TIChTX was purchased resistances of 2 to 8 MQ. Currents were measured using a DAGAN from Du Pont-New EnglandNuclear. Restriction enzymes were pur- model 3900 patch clamp. Delivery of command voltages, data acquichased from Life Technologies, Inc., and bovine blood "Restriction sition, and post acquisition analysis were aided by a microcomputer Protease Factor X." was obtained from Boehringer Mannheim. T4 running the Pclamp program suite (Axon Instruments, Foster City, DNA ligase was purchased from New England Biolabs. Reagents for CA). Ionic currents were also recorded from Xenopus oocytes that had polymerase chain reaction were obtained from Perkin Elmer Cetus. been injected with RNA transcribed in vitro from one or another of Precast 16% Tricinegels were obtained from Novex. Purification of MgTX-Lyophilized venom of the scorpion C. mar- the following K+ channel cDNA clones: K,1 (rat K,1.5),K,2 (rat garitatus was suspended in 20 mM sodium-borate, pH 9.0, at a Kv1.6), rat or human Kv3 (rat KJ.3; human K,1.3), K,4 (rat K,3.1), concentration of 5 mg/ml. After agitating the venom by vortex, the rat ISK, orDrosophila Shker-H4. Thenames of the clones are listed suspension was clarified by centrifugation at 27,000 x g for 15 min. as originally published, with the appropriate parenthetical designaThe supernatant was removed, filtered through a Millex-GV filter, tion according to a unified gene nomenclature. The KJ-4 cDNAs, as (0.20-pm pore size, Millipore), and loaded onto a MonoS fast protein well as ISK were kindly provided by R. Swanson (MRL, West Point). liquid chromatography column (HR5/5, Pharmacia LKB Biotechnol- The Shker-H4 cDNA was a gift from M. Tanouye (University of ogy Inc.), equilibrated with 20 mM sodium-borate, pH 9.0. After the California, Berkeley). Methods for oocyte isolation and microinjecabsorbance of the eluate had returned to base line, bound material tion were previously described (51,52). Currentsexpressed in oocytes was eluted from the resin with a linear gradient of NaCl (0.75 M/h) were measured using a DAGAN model CA-1 two-electrode voltage at a flow rate of 0.5 ml/min. Fractions were separated and assayed clamp. Delivery of command voltages, data acquisition, and post for their ability to inhibit [lZ61]ChTXbinding to either rat brain acquisition analysis were aided by a microcomputer running the synaptic plasma membranes or bovine aortic smooth muscle sarco- Pclamp program suite. The oocyte bathing solution consisted of 96 lemmal membranes. Fractions that blocked [1261]ChTXbinding only mM NaC1,2.5 mMKC1, 1mM MgCl,, 1.8 mM CaCl2, 10 mM HEPES, to brain membranes and thatdisplayed the highest absorbance at 280 and 0.05% bovine serum albumin. The pH was 7.2.All recordings nm were considered for further characterization. These fractions were were obtained at room temperature (19-22 "C). The intracellular loaded onto a 300-A pore size Cls reversed-phase HPLC column (0.46 microelectrodes were fabricated from DAGAN LE-16 glass in a twostage pull to a final resistance of 0.6-2 M ~ and I filled with 3 M KC1. X 25 cm) that had been equilibrated with 10 mM trifluoroacetic acid. The column was eluted with a linear gradient of isopropyl alcohol/ For both lymphocyte and oocyte experiments, peptides were diluted into the extracellular solution from 20 pM aqueous stock containing acetonitrile(2:l; 0-15%, 33min) at a flow rate of0.5 ml/min. Fractions were collected, lyophilized, and later reconstitutedwith 100 100 mM NaC1. The presence of 0.05% bovine serum albumin in the mM NaCl, 20mM Tris-HC1, pH 7.4. bath solution inhibited nonspecific loss of dilute peptide toxins in the Amino Acid Sequence Determination-Purified MgTX was re- perfusion tubing. duced, subjected to alkylation with iodoacetic acid, and repurified by Bacterial Strains and Plasmids-E. coli DH5a was used for plasmid reverse-phase chromatography as described previously for ChTX (32). propagation and strain BL21(DE3) was used for expression of the Approximately 0.5 nmol of reduced-alkylated toxin was loaded onto fusion protein (53). Plasmid pCSP105 (54) is a variant of pSR9 (55) a Porton peptide support filter, and Edman degradation was per- and was a gift from C. Miller (Brandeis University). formed using a Porton 2090 microsequencer. Phenylthiohydantoin Construction of MgTX Plasmid-The MgTX gene was constructed derivatives were analyzed using an on-line detection system. Typical using two synthetic oligonucleotides that were synthesized using an repetitive yields from at least six sequencing runs were 94%. AB1 model 391 DNA synthesizer (Applied Biosystems). The two 89Amino Acid Analysis and Determination of Extinction Coefficientmers are shown in Fig. 5, except for the five dCTP bases that were A sample containing 1 nmol of purified MgTX was subjected to acid added to the 5' end of each to facilitate cutting with restriction hydrolysis, derivatized, and the phenylthiocarbamyl amino acid de- enzymes. Codon usage was optimized for high level expression in E. rivatives identified by reversed-phase chromatography using a Pico- coli (56). The oligonucleotides were purified using Nensorb columns Tag (Waters) system. The absorbance spectrum of purified MgTX (Du Pont), annealed, filled in with dNTPs using Sequenase (U. S. was digitized in a Beckman DU 7400 UV/VIS spectrophotometer. Biochemicals), and digested with restriction enzymes using standard Protein extinction coefficients were calculated by determining the techniques (57). The resulting fragment was gel purified using Nuamino acid composition of an aliquot of MgTX and thencorrelating sieve (FMC Bioproducts) and ligated into pCSP105 using standard protein content with the recorded absorbance at 280, 235, and 215 techniques (57). The resulting construct, encoding a fusion protein nm, respectively. of T7 gene 9 and MgTX separated by a factor X. cleavage site was Polyacrylamide Gel Electrophoresis-Samples were dissolved in verified using dideoxy sequencing (58). sample buffer containing 50 mM dithiothreitol and heated at 100 "C Expression and Purification of rMgTX-E. coli BL21(DE3) harfor 5 min. Samples were loaded into precast 16% Tricine gels, and boring p6MgTX were cultured and induced with isopropyl-l-thio-j3electrophoresis was carried out in a Novex system at constantvoltage D-galactopyranoside. Purification of the fusion protein was done (100 V). After electrophoresis, the gels were stained for protein with essentially as described (54). Briefly, induced cells were pelleted at silver staining as described (49). 4,000 X g for 10 min, washed once in 50 ml of phosphate-buffered Binding Studies-The binding of ['261]ChTX to either rat brain saline, and stored overnight at -80 "C. Cells were thawed on ice and synaptic plasma membrane vesicles (44) or bovine aorticsmooth resuspended in 20 ml of 10 mM Tris-HC1, pH 8.0,50 mM NaC1,l mM muscle sarcolemmal membranes (43) was determined as previously Naz EDTA, 1 mM DTT, and protease inhibitors (100 p~ phenyldescribed. Briefly, membrane vesicles were incubated with[1261]ChTX methylsulfonyl fluoride, 1pg/ml pepstatin A, 1pg/ml leupeptin), and in the presence or absence of other added agents until equilibrium incubated with 0.5 mg/ml lysozyme. Cells were disrupted by sonicawas achieved. Separation of bound from free ligand was accomplished tion, and the lysate was centrifuged at 27,000 X g for 15 min. After by filtration through GF/C glass fiber filters (Whatman) that had removal of nucleic acids with streptomycin sulfate, the lysate was been presoaked in 0.5% polyethylenimine. Nonspecific binding was loaded onto a DEAE-Sepharose column equilibrated in 10 mM Trisdetermined in the presence of 10 nM unlabeled ChTX. Data from HC1, pH 8.0, 50 mM NaC1, 1 mM D m . After washing the column, saturation experiments were analyzed in a Scatchard representation the fusion protein was eluted with 350 mM NaCl, 10 mM Tris-HC1, to determine I 1 WM. It is another Observed Integer member of the Shaker family. Kv3.1 is a channel cloned from Aspartic acid or asparagine 1.80 2 rat (27) that appears to encode a ChTX-insensitive delayed Glutamic acid or glutamine 3.00 3 rectifier found inthymic cells (the I-current) (60). It is a Serine 2.05 2 member of the Shaw subfamily. ISK is a slowly activating K+ 3.15 3 Glycine Histidine 1 0.98 channel cloned from smooth muscle andheart (61). This ND" Arginine channel appears to underlie the IKe delayed rectifier in heart Threonine 1.97 2 muscle (62, 63). It isstructurallyunrelated totheother Alanine 3.01 3 voltage-activated K+ channel clones. The ISK transcript has Proline 4.05 4 been detected in lymphocytes (64), but no evidence of the IKs 1.02 1 Tyrosine current has been found in those cells. Valine 1 1.02 Methionine 0.93 1 Of the six cloned channels tested, three (Kv1.5, Kv3.1, and 5.92 6 Cysteine 1%) were completely insensitive to 200 nM MgTX. The Isoleucine 1.64 2 results from dose-response studies of the other three clones Leucine 1.00 1 are shown in Fig. 6B. The most sensitive channel is KJ.3, 1.04 1 Phenylalanine with an ICW of -30 PM. The next most sensitive channel, 5.94 6 Lysine Kv1.6, was half-blocked at 5 nM, while Shaker-H4 displayed ND. not detected. an IC, of 150 nM. MgTX has been shown to have no effect on the delayed rectifier of mouse pancreatic /3 cells.' When tested on Maxi-K channels from bovine aorta reconstituted in planar lipid bilayers, MgTX was without effect at 1P M . ~ Functional expression of rMgTX in E. coli allows large amounts of peptide to be made (e.g. a typical yield of 3-4 mg from 1 liter of culture is obtained). This technique also allows for the production of toxin variants with which to determine a structure-activity relationshipfor the interaction of MgTX with Kv1.3. As a first step, the carboxyl-terminal His residue of rMgTX was replaced with a Asn, the carboxyl terminus of noxiustoxin (NxTX),a closely related toxin (see below). MgTXHsgN is a weaker inhibitor of Kv1.3 than MgTX. The IC, as an inhibitor of ['"I]ChTX binding was 190 PM (Fig. a), and in electrophysiological experiments, it displayed a - 4 1Foxin1 (M) B @moVmg prot) potency of -200 PM (Fig. 6C). The reduced potency of MgFIG.4. Effect of MgTX on [1B61]ChTXbinding to K+chan- TXHsgN appears to result from a faster dissociation rate, as nels. A , rat brain synaptic plasma membrane vesicles (0,0, and A) evidenced by its rapidwash off from the channel ( tlI2< 3 min or bovine aortic sarcolemmal membrane vesicles (D) were incubated uersus tlIz >20 min for normal MgTX)in voltage clamp with 25 pM ['261]ChTX in the absence or presence of increasing experiments (not shown). Importantly, MgTXH39N still disconcentrations of native MgTX (0,W) r-MgTX (O),or M ~ T X H ~ ~ N (A),at room temperature until equilibrium was achieved. Inhibition played selectivity for Kv1.3 since it did not block the ChTXof binding was assessed relative to an untreatedcontrol. B, rat brain sensitive, Caz+-activated K+ channels in human T-lymphosynaptic plasma membrane vesicles were incubated in the presence cytes. As shown in Fig. 6D, elevating the potassium concenof increasing concentrations of ['261]ChTX in the absence (0)or tration of the bathing medium allows for simultaneous represence (0)of 25 pM MgTX, at room temperature until equilibrium cording from voltage-activated and calcium-activated potaswas achieved. Specific binding data are presented in the form of a sium channels of the T-cell. Between -120 and -50 mV, the Scatchard representation. response is nearly linear, reflecting the activity of voltageindependent channels that areopen because of the high Ca'+ recombinant material exclusively. Fig. 6A shows the effect of (1WM) in thepipette. Between -50 and 0 mV, there is a sharp 1 nM MgTX on the voltage-activated K+ current recorded rise in theamplitude of the inward current (downward deflecfrom human T-lymphocytes. Complete inhibition occurs at 1 tion), due to the activation of the voltage-sensitive (Kv1.3) nM, and recovery from block, following removal of MgTX channels. The trace crosses the zero-current level at +5 mV from the bath, proceeds slowly, with a tlIz of >20 min (not and reverses direction to become an outward current. The shown). The ICW for inhibition of the lymphocyte Kv1.3 positive limb of the response reflects the combined activity of current isbetween 30-100 p ~As. reported previously, MgTX the Ca'+-activated and voltage-activated channels. The curhas no effect on three different ChTX-sensitive, Ca'+-acti- rents active at potentials more positive than -50 mV (Kv1.3 vated K+channels of lymphocytes, even at 100 nM (41). channels) were blocked by both ChTX and M ~ T X H ~while ~N, As a preliminary investigation to examine the specificity of the calcium-activated currents (visible between -120 and -50 MgTX, we compared the sensitivity to rMgTX of six cloned mV) were sensitive only to ChTX. K+ channels, expressed in Xenopus oocytes. The channels Sequence Homology with Other K+ Channel Toxins-The tested were Shaker-H4, KJ.3, Kv1.5, Kv1.6, K"3.1, and ISK. primary amino acid sequence of MgTX was compared with Shaker-H4 is a splice variant from the original Drosophila that of other known K+ channeltoxins (Fig. 7). MgTX Shaker K+ channel gene. Kv1.3is theChTX-sensitive delayed rectifier current found in rat and human T-lymphocytes (data M. Leibowitz, personal communication. K. Giangiacomo, personal communication. from rat and human clones were indistinguishable and were

18871

Margatorin Inhibition of K J . 3 Channels

pMGTX 5.7 Kb

3

17 promotor

Factor Xa

S I

1 0

20

AAACAGTGCCTGCCGCCGTGCAAAGCTCAGttcggtcagtctgctggtgctaaa tttgtcacggacggcGGCACGTTTCGAGTCAAGCCAGTCAGACGACCACGATTT L y s G l n C y s L e u P r o P r o C y s L y s A l a G l n P h e G l y G l n S ~ r A l a G l y A l a L y s

3 0

39

H~ndlll

tgcatgaacggtaaatgcaaatgctacccgcactgataggaagctt ACGTACTTGCCATTTACGTTTACGATGGGCGTGACTATCCTTCGAA C y s M e t A s n G l y L y s C y s L y s C y s T y r P r o H i s E n d

FIG. 6. Design of synthetic MgTX gene for expression in E. coli. Upper p a n e l , the plasmid map of ~ M ~ shows T X the locations of the synthetic MgTX gene, Factor Xa cleavage site, and the T7 gene9 fusion protein. Amp*-ampicillin resistance, Ori-origin. Lower p a n e l , DNA sequence of synthetic MgTX gene. Oligonucleotides used for initial MgTX gene construction are shown in capitals. The Factor X. cleavage site is boxed. The corresponding amino acid sequence is also shown with the MgTX peptide sequence from residues Thr' to His3'.

displays 44% sequence homology with ChTX, a blocker of both Kv1.3and Maxi-K channels. MgTX shares 41% homology with IbTX and 54% homology with KTX. Both IbTX and KTX are selective Maxi-K inhibitors.The highest degree of primary sequence homology is observed between MgTX and NxTX (79%). Noxiustoxin is also a high affinity blocker of KJ.3, but itspotency is somewhat less than thatof MgTX. It is interesting to note that some of the positively charged residues shown to be important for the interaction of ChTX with the Maxi-K channel (65,661 are conserved in MgTX. In particular, Lys27of ChTX is a critical residue which must bind very close to theion conduction pathway of the channel, since neutralization at this position affects the knock-off of toxin by K+ entering thechannel from the inner pore. This residue is conserved in MgTX, and it may play a similar role in thistoxin's interaction with the Kv1.3channel. DISCUSSION

The results presented inthis paper describe the identificationand characterization of margatoxin, a novel peptide derived from venom of the scorpion C. mrgaritatw that blocks with high affinity and specificity the Kv1.3 channel. Amino acid sequence determination, mass spectroscopy, and amino acid analysis define MgTX as a39-amino-acid peptide containing 6 cysteine residues and 7 positively charged resi-

dues. Confirmation of MgTX as theactive component of the purified preparation was obtainedafter expression of the peptide in E. coli as part of a fusion protein. After cleavage and refolding, rMgTX was purified and shown to display properties identical to those of the native peptide. Over the last few years it hasbeen recognized that scorpion venoms are a rich source of peptidyl inhibitors directed against various types of K+ channels (67). Some of these peptides have been purified to homogeneity and their properties characterized. The most extensively studied of these toxins is ChTX. Although ChTX was initially discovered as an inhibitor of the Maxi-K channel, it was later observed that this toxin blocks other types of K+ channels with similar potencies, namely Kv1.3 and several intermediate and small conductance Ca2+-activated K+ channels. Therefore, caution must be exercised when using ChTX tostudy the physiological role of a given channel in a tissue of interest. Other peptidyl inhibitors, such as iberiotoxin (45) and kaliotoxin (47), have been shown to possess greater selectivity for the Maxi-K channel.These peptides will prove useful for the further characterization of the physiologic roles of the Maxi-K channel. Similarly, MgTX represents a unique tool with which to probe the function of Kv1.3. This channel has been identified as the major voltage-dependent K+ conductance pathway in peripheral human T-lymphocytes (41, 68), and several labo-

18872

200

D

P

150 R

2 700

v

=L

50 -100-80

3

0

-60 -40 -20

0

-50 -100

Test Potential (mV)

FIG. 6, Electrophysiological effects of MgTX. A, whole cellcurrent recording from a human peripheral T-lymphocyte. The membrane potential was held at -70 mV, and depolarizing voltage steps (-60 mV to +20 mV in 20 mV increments; 180 ms duration) were given 45 s apart to elicit voltage-gated currents. The upper traces ( A I )were obtained in control saline, while the lower (A2)were obtained after addition of MgTX (1 m)to thebath. No leakage correction. The voltage clamp protocol is shown as an insert. B, dose inhibition curves for MgTX block of three cloned potassium channels expressed in Xenopus oocytes. Data were collected by delivering repetitive voltage pulses to 0 mV (100 msduration) from a holding potential of -80 mV during the application of increasing concentrations of MgTX. The percent inhibition was calculated by comparing the currentelicited by the test pulse at a given concentration of MgTX uerssus that obtained in control d i n e . Each curve contains data from 2 to 4 cells. Continuous curves were generated by fitting the data toa logistic dose-response curve, with the slope factor fixed at unity. Each symbol denotes data from a different cloned channel: human or rat KJ.3 (m); rat KJ.6 (A);and Shker-H4 (a).C, the effect of M ~ T X Hon ~N rat KJ.3 channels expressed in a Xenopus oocyte.The plot shows the peak current measured at each test potential in thepresence of increasing concentrations of M ~ T X H ~The ~ Ncell . was exposed first to control saline (O), then to 50 pM (A),200 pM (0)and finally 1 nM (v)M ~ T X H ~D~, effect N . of M ~ T X H ~on S Npotassium currents in a human peripheral T-cell. The cell was dialyzed with a pipette solution containing, in mM: 140KCI, 5.2 KOH, 1 K2-EGTA, 1 MgCh, and 0.9 CaCl,, (free Ca2+calculated as -1 g ~ ) .The bathing solution was 145 KCI, 2 MgC12, 1 CaGI2. The pH was 7.2 inside and out. Expected K+ reversal potential = +3 mV. Currents were elicited by a voltage ramp (-120 mV to f 100 mV 8 240 mV/s). Each trace is the average of 3 ramp currents. The tmce marked I was taken in bathing solution. Trace 2, after addition of ChTX (20 REA) to thebath. Trace 3, ChTX was removed and replaced by M~TXHSN ( 2 5 nM). Trace 4, ChTX (20 nM) plus M ~ T X H ~(25 S NnM) were added together.

IT I I N V K C T S P K O C L P P C K

FIG. 7. Sequence homologies between MgTX and other scorpion toxins. Comparison of amino acid sequences of margatoxin (MgTX),noxiustoxin (NrTX),kaliotoxin (KTX),charybdotoxin (ChTX), and iberiotoxin (IbTX). The toxins were aligned using the PILEUP program (Genetics Computer Group) which uses a simplification of the progressive alignment method of Feng and Doolittle (83). Residues identical to MgTX are bod.An additional gap was inserted into the KTX alignment to maximize sequence identity with MgTX.

ratories have demonstrated that ChTX can prevent T-cell activation driven by Ca2+-dependentpathways (39,691. However, human T-cells contain, in addition to Kv1.3, several distinct small conductance Ca2+-activatedK+ channels that are also blocked byChTX (41),making this toxin inadequate for assigningthe role of any specific channel in the control of T-cell proliferation. The recent demonstration that MgTX depolarizes human T-cells (41), and prevents activation and proliferation (69),has clarified the role of KJ.3 in setting the membrane potential of resting human T-cells. MgTX is closely related (79% identity) to NxTX. Noxiustoxin is a high affinity blocker of KJ.3 (6), but also inhibits the delayed rectifier K+ channel of squid axon (70) and the Maxi-K channel from skeletal muscle (71), although the affinity of those interactions is low. Of the othertoxins sharing significant homologies with MgTX,ChTX blocks both Kv1.3

and Ca2+-activatedK+ channels, while IbTX and kaliotoxin (KTX) are selectivefor Ca2+-activatedK+ channels. The conservation of position of the cysteines among all these toxins suggests a common disullide bridging pattern, asshown for ChTX and IbTXwhose solution structures were resolved by two-dimensional NMR (72-76). It is interesting to note that 2 of the proline residues of MgTX reside in a region of the molecule that has an a-helix conformation in ChTX and IbTX. How these residues affect the structure of MgTX should be resolved by 'H NMR spectroscopy. Site-directed mutagenesis studies with ChTX have identified Lys2, as a critical residue for vol~ge-dependentblock of the Maxi-K channel. Since this lysine is highly conserved among all of the toxins, it will be interesting to determine its contribution to block of Kv1.3 by MgTX. Kv1.3 is blocked with high affinity by ChTX, and a single

Margatoxin Inhibition

of K,1.3 Channels

18873

Li, M., Jan, Y. N., and Jan,L. Y. (1992)Science 267,1225-1230 point mutation, Glf7'Phe, abolishes toxin block (77). Since 23. 24. Christie, M. J., North, R. A., Oshorne, P. B., Douglass, J., and Adelman, J. P. (1990)Neuron 4,405-411 the interaction between MgTX and fl2'I]ChTX is not strictly 25. McCormack, K., Li,J. W., Iverson, L. E., and Rudy, B. (1990) Biochem. competitive, their sites of interaction with Kv1.3 are not Biophys. Res. Commun 171,1361-1371 expected to be identical, although they may overlap. Site- 26. Wei, A., Covarrubias, M., Butler, A,, Baker, K., Pak, M., and Salkoff, L. (1990)Science 248,599403 directed m u ~ e n e s i studies s with Kd.3 should reveal which 27. Luneau, C. J., Williams, J. B., Marshall, J., Levitan, E. S., Oliva, C., Smith, J. S.,Antanavage, J., Folander, K., Stein, R. B., Swanson, R., Kaczmarek, residues of the channel are critical for MgTX binding. L. K., and Buhrow, S. A. (1991)Proc. Natl. Acad. Sei. U. S. A. 88,3932Several contradictory observations have been maderegard3936 ing the activity of ChTX as a blocker of Sfaaker and related 28. Kamb, A., Tseng-Crank, J., and Tanouye, M. A. (1988)Neuron 1,421-430 Rudy, B.(1988)Neuroscience 26,729-749 K+ channels, with the results obtained dependingon the 29. 30. Ponga, 0.(1989)Pflugers Arch. Eur. J. Physiol. 414,571-575 source of toxin used (78-80). The high affinity block of 31. Stuhmer, W.. Ruppersberg, J. P., Schroter, K. H., Sakmann, B., Stocker, M.. Gieae. K.P.. Perschke,. A...Baumann.. A.,. and Ponps. EMEO - . 0.(1989) . voltage-gated channels other than Kv1.3 is due, in fact, to J. 8,3235-3244 other components of L. guinquestriatus venom not present in 32. Gimenez-Gallego, G., Navia, M. A, Reuben, J. P., Katz, G. M., Kaczorowski, G. J., and Garcia, M. L. (1988)P m . NatL Acod. Sci U. S. A. 86. material prepared by the method of ~ ~ e n e z - ~ ~(81). lego 3B9-3333 . . ~ ~ The necessity to verify the activity of components purified 33. Miller, C., Moczydlowski, E., Latorre, R., and Phillips, M. (1985)Nature 313,316-318 from venomsand other natural sources cannot be overstated. 34. Hermann A., and Erxleben, C. (1987)J. Gee. PhysioL 90, 27-47 In this respect, the use of solid-phase peptide synthesis (72, 35. Castle, N.' A., and Strong,P. N. (1986)FEBS Lett. 209,117-121 D., Cecchi, X., Spalvins, A., and Caneasa, M. (1988)J. Membr. BioL 82) or recombinant expression techniques (54) have been very 36. Wolff, 106,243-252 useful, not only in producing large amounts of the molecule 37. Luechesi, K., Ravindran, A,, Young, H., and Moczydlowski, E. (1989)J. Membr. BWL 109,269-281 of interest, butalso in confirming the chemical identity of the 38. Schweitz, H., Bidard, J. N., Maes, P., and Lazdunski, M. (1989)Biochempresumed activecomponent of the toxin preparation. Followistry 28,9708-9714 M.. Lee. S. C.. and Deutsch.~. C. 11989)Proc. NatL Acad. Sei U. S. A. ing the procedures described in this study, it is feasible to 39. Price. 86; ioi71-io175~ produce 3-4 mg of homogeneous and biologically activetoxin 40. Sanda,S. B., Lewis, R. S., and Cahalan, M. D. (1989)J. Gen. PhysWL 93, 1nnl -1 ".n74from 1 liter of E. coli culture by a modification of methods 41. Leonard, R. J., Garcia, M.L., Slaughter, R. S., and Reuben, J. P. (1992) previously publishedfor the biosynthesis of ChTX (54). The P m . NatI. A d . SCE.U. S. A. 8 9 . 1 ~ 4 - 1 ~ Arkett, S . A., Dixon, S. J., and Sims, S. M. (1992)J. PhysioL ( L o n d . ) 468, 42. identification of ligands that interact specifically witha given 633-653 channel is critical for defining its pharmacology and physio- 43. Vazquez, J:, Feigenbaum, P., Katz, G., King, V. F., Reuben, J. P., Ro Contancm, L., Slaughter, R. S., Kaczorowski, G. J., and Garcia, M. logical role.For example, MgTXis a powerful toolfor defining (1989)J. Bio!. Chem. 264,20902-20909 the role that Kv1.3 plays in the regulation of immunorespon- 44. Vazquez, J., Fezgenbaum, P., King, V. F., Kaczorowski, G. J., and Garcia, M. L. (1990)J. Bid. Chem. 266,15564-15571 siveness in human T - l ~ p h ~Within ~ s . the family of K+ 45. Galvez, A., Gimenez-Gallego,G., Reuben, J. P., Roy-Contancin, L., Feigenchannels, only a few members have been shown to be sensitive baum, P., Kaczorowski,G., and Garcia, M. L. (1990)J. Biol. Chem. 265, 11083-11090 to low concentrations of the identified peptidyl toxins (67). 46. Novick, J., Leonard, R. J., King, V.F., Schmalhofer, W., Kaczorowski, G. Many other K+channels do not have a well defined pharmaJ., and Garcia, M. L. (1991)BW hys J. 69,78 47. Crest, M., Jac uet, G., Gola, M., % F u k , H., Benslimane,.A., Rochat, H., cology. It seemslikely that venoms will continue to yield Mansuelle, %.,and Martin-Eauclmre, M. F. (1992)J. Bml. Chem. 267, selective inhibitors and that their identification will help to 1640-1647 48. Possani, L. D., Martin, B. M., and Svendaen, I. B. (1982)Cartsberg Res. elucidate the special rolesof individual channels. Commun. 47, 285-289 " "

i:

49. Oakley, B. R., Kirsch, D. R., and Morris, N. R. (1980)A d Biochem. 106, 361-363 50. Cahalan, M., and Neher, E. (1992)Me~hodsEnzymL 207,3-14 51. Goldin, A. L. (1991)Methods Cell Biol. 36,487-509 52. Swanson, R.,and Folander K. (1992)Methods Entymol. 207,310-319 53. Studier, F. W., and MoffatL, B. A. (1986)J. MOL BioL 189,113-130 54. Park, C. S., Hausdorff, S. F., and Miller, C. (1991)Proc. Natl. A d . Sei. REFERENCES U. S. A. 68,2046-2050 1. Hille, B. (1992)Ionic Channek ofExcitable Membranes, Sinauer Associates, 55. Howell, M. L., and Blumenthal, K. M. (1989)J. Biol. Chem. 264, 1526815273 Inc., Sunderland, MA 56. Sharp, P. M., Cowe, E., Higgins, D. G., Shields, D.C., Wolfe, K.H., and 2. Timpe, L. C Schwan, T. L., Tempel, B. L., Papaeian, D. M., Jan, Y.N., Wright, F. (1988)Nucleic Acids Res. 16,8207-8211 and Jan,2. Y. (1988)Nature 331,143-145 3. Aldrich, R.W., Hoshi, T., and Zagotta, W. N. (1990)Cold Spring Harbor 57. Ausubel, F. M., Brent, R., Kingaton R. E., Moore, D. D., Seidman, J. G., Smith, J. A., and Struhl,X. (eds) i1989) Current Protccols in Molecular S y q . QurMEWL . 6419-27 Biology, John Wiley & Sons,Inc., New York 4. Hausdorff, S. F., Goldstein, S. A., Rushin, E. E., and Miller, C. (1991) 58. Sanger, F., Nicklen, S., and Coulson, A. R. (1977)Proc. Natl. Acad. Sei. Biochemistry 30,3341-3346 U. S. A. 74,5463-5467 5. Pongs, 0.(1992)Trends PhormacoL Sci 13,359-365 59. Phil1 son, L H., Hice, R. E., Schaefer, K., LaMendola, J., Bell, G. I., 6. Swanson, R., Marshall, J., Smith, J. S., Williams, J. B., k y l e , M. B., Nep8on, D: J., and Steiner, D. F. (1991)Proc. Natl. Acad Sei. U. S. A. Folander, K., Luneau, C. J., Antanavage, J., Oliva, C., Buhrow, S, A., 88,53-57 Bennett, C., Stsin, R. B., and Kaczmarek, L. K. (1990)Neuron 4, 92960. Chandy, K. G., Ghanshani, S., Dethlefs, B., McPherson, J., Wasmuth J 939 Gutman, G. A., Cahalan, M. D., and Grimmer, S. (1992)Biophys. J. 7. IveGon, L. E., and Rudy, B. (1990)J. Neurosci 10,2903-2916 A q K 7' 8. Busch, A. E., Hurst, R. S., North, R. A., Adelman, J. P., and Kavanaugh, 61. Folander, K., Smith, J. S., Antanavage, J., Bennett, C., Stein, R. B., and M. P. (1991)Biochem Bio hys Res Commun. 179,1384-1390 Swanson, R. (1990)Proc. Natl. Acad. Sei. U. S. A. 87,2975-2979 9. Oliva, C., Wiedmann, R., g n n e t t , C., Folander, K., Swanson, R., and 62. Sanguinetti, M. C.,and Jurkiewicz, N.K. (1991)Am. J. Physiol. 260, Smith, J. (1991)Soc. Neumci. Abst. 17,775 H393-H399 10.Iaacoff, E. Y., Jan, Y. N., and Jan, L. Y. (1990)Nature 346,530-534 63. Sanguinetti, M. C.,and Jurkiewicz, N. K. (1990)J. Gen. Physwl. 96, 19511. Stocker, M Stuhmer, W., Wittka, R., Wang, X., Muller R., Ferns, A., 215 and Pon& 0.(1990)Proc. NatL Aeod Sci. U. S. A. 87,8903-8907 64. Attali, B., Romey, G., Honore, E., Schmid-Alliana, A., Mattei, M.G., 12. Jan, L. Y., and Jan, Y. N. (1992)Annu Rev. Physior 54,535-555 13. Yelien, G., Jurman, M. E., Abramson, T., and MacKinnon, R. (1991) Leaage, F., Ricard, P., Barhanin, J., and Lazdunski, M. (1992)J. Biol. Chem. 267,8650-8657 Science 251,939-942 14. K m h , G. E.,Drewe, J. A., Hartmann, H. A., Taglialatela, M., de Biasi, 65. Park, C. S., and Miller, C. (1992)Neuron 9,307-313 66. Park, C. S., and Miller, C. (19921Biochemistry 31,7749-7755 M., Brown, A. M., and Joho, R. H. (1992)Neuron 8,499-505 67. Garcia, M. L., Galvez, A., Garcia-Calvo, M., King, V. E'., Vazquez, J., and 15. Kirsch, G. E., Drewe, J. A., Taglialatela, M., Joho, R. H., DeBiaai, M., Hartmann, H. A., and Brown, A.M. (1992)Bioph s J 62,136-43 Kaczorowski, G. J. (1991)J. Bioeng. Biomembr. 23,615-646 16. Kavanaugh, M. P., Hurst, R. S., Yakel, J., Vamum, D., Adelman, J. P., 68. Lee, S. C., Levy, D. I., and Deutach, C. (1992)J. Gen. PhysioL 99,771-793 and North, R. A. (1992)Neuron 8,493-497 69. Lin, C., Boltz, R C., Blake, T.,Nguyen, M., Talenta, A., Fischer, P. 17. Heginbotham L., and MacKmnon, R. (1992)Neuron 8,483-491 Springer. M., Sigal, N., Slaughter, R. S., Garcia, M., Kaczorowski, G. J.: 18. Kavanau h M. P., Varnum, M.D., Oshorne P. B Christie M. J Busch and Koo, G. C. (1993)J. Exp. Med. 177,637-646 A. E.,fdhman, J. P., and North,R. A. (li91)J.'BioL Chek. 266,7583: 70. Carhone, E., Prestipino, G., Spadavecchia, L. Franciolini, F., and Possani, 7587 L. D. (19873Pflugers. Arch. Eur. J. Physw!. 408,423-431 19. MacKinnon, R., and Yellen, G. (1990)Science 260,276-279 71. Valdivia, H. H., Smith, J. S.,Martin, B. M., Coronado, R., and Possani, L. 20. MacKlnnon, R., Heginbotham, L., and Abramson, T. (1990)Neuron 6, D. (1988)FEBS. Lett. 226,280-284 767-771 72. Lambert, P., Kuroda, H., Chino, N., Watanahe, T. X., Kimura, T., and 21. MacKmnon R., and Miller, C. 11989)Science 246,1382-1385 Sakakibara, S. (1990)B k h e m . Ewphys. Res. Commun 170,684490 22. MacKinnon: R. (1991)Nuture 360,232-235 73. Bontems, F., humeatand, C., Boyot, P., Gilquin, B., Doljansky, Y., Menez,

A c k ~ ~ ~ ~ n t s -thank W Deborah e Zink for massspectral analysis of MgTX, McHardy Smith for assistance with peptide sequencing, and Pat Hardiman for helpin preparing the manuscript.

61;

'I""

k

18874 A., and Toma,F. (1991 Bur J B k h e m . 196,19-28 74. Bontems, F Roumestan~C., Giiwin, B., Menez, A., and Toma, F. (1991) Science &4,1521-1523 75. Bontems F Giiquin, B., hUmes*d, c., Menes A., and Toma, F.(1992) BiocheLLttry31 7756-7764 76. Johnson, B. A. and Su ,E.E. 1992)Biochemistry 31,8151-8159 77. Ohva C. Wiedmann Gan R. Folander K., Bennett, C., Swanson, R. 'and Smith J. (i99i)E' ' hya. $. 69,450k 78. Mackinnon, R . , h n h a r t , P%., and White, M.M. (1988)Neuron 1,9971001

1.

79. Goh, J. W., Kelley, M. E. M., Pennefather, P. S.,Chicchi, G. G., Cascieri, M.A., Garcia, M. L,and Kaczorowski, 0. J. (1992)Brain Res. 691, 165-170 80. oliva, c., ~ ~ l K., ~and&Smith, ~ J., (1991)~iophy~. J . 59,450~ 81. Garcia-Calvo, M., Leonard, R. J., Giangiacomo, K., McManua, 0.B., Kaczorowski, G. J., and Garcia, M.L. (1992)Biophys. J. 61,377 82. Sugg,E.E.,Garcia, M. L., Reuben, J. P., Patchett, A. A., and Kacmrowski, G. J. (1990) J. Bwl. Chem. 266,18745-18748 83. Feng, D.F.,and Doclittle, Ft. F. (1987)J. MOL Evol. 26,351-360