jm501765v (2)

Download jm501765v (2)

Post on 16-Feb-2017

160 views

Category:

Documents

1 download

Embed Size (px)

TRANSCRIPT

  • Engineering Potent and Selective Analogues of GpTx-1, a TarantulaVenom Peptide Antagonist of the NaV1.7 Sodium ChannelJustin K. Murray, Joseph Ligutti, Dong Liu, Anruo Zou, Leszek Poppe, Hongyan Li,

    Kristin L. Andrews, Bryan D. Moyer, Stefan I. McDonough, Philippe Favreau,# Reto Stocklin,#

    and Les P. Miranda*,

    Departments of Therapeutic Discovery, Neuroscience, and Pharmacokinetics & Drug Metabolism, Amgen Inc., One AmgenCenter Drive, Thousand Oaks, California 91320, United StatesTherapeutic Discovery and Neuroscience, Amgen Inc., 360 Binney Street, Cambridge, Massachusetts 02142, United States#Atheris Laboratories, Case Postale 314, CH-1233 Bernex, Geneva, Switzerland

    *S Supporting Information

    ABSTRACT: NaV1.7 is a voltage-gated sodium ion channel implicated by human genetic evidence as a therapeutic target for thetreatment of pain. Screening fractionated venom from the tarantula Grammostola porteri led to the identification of a 34-residuepeptide, termed GpTx-1, with potent activity on NaV1.7 (IC50 = 10 nM) and promising selectivity against key NaV subtypes (20and 1000 over NaV1.4 and NaV1.5, respectively). NMR structural analysis of the chemically synthesized three disulfide peptidewas consistent with an inhibitory cystine knot motif. Alanine scanning of GpTx-1 revealed that residues Trp29, Lys31, and Phe34

    near the C-terminus are critical for potent NaV1.7 antagonist activity. Substitution of Ala for Phe at position 5 conferred 300-foldselectivity against NaV1.4. A structure-guided campaign afforded additive improvements in potency and NaV subtype selectivity,culminating in the design of [Ala5,Phe6,Leu26,Arg28]GpTx-1 with a NaV1.7 IC50 value of 1.6 nM and >1000 selectivity againstNaV1.4 and NaV1.5.

    INTRODUCTIONVoltage-gated sodium channels (VGSCs or NaVs) initiate andpropagate action potentials in excitable cells such as central andperipheral neurons, cardiac and skeletal muscle myocytes, andneuroendocrine cells.1 Structurally, they consist of anapproximately 260 kDa -subunit and associated smaller -subunits.2 The -subunit has four domains (IIV), eachdomain containing six transmembrane helices (S1S6). TheS5S6 domains govern the main aspects of ion permeation,and domains including most prominently fixed charge withinthe S4 transmembrane -helix transduce depolarizing voltagesinto physical opening of the channel. The family of VGSCsconsists of nine known subtypes (NaV1.1NaV1.9). Thesesubtypes show tissue specific localization and functionaldifferences with NaV1.1, NaV1.2, and NaV1.3 found principallyin the central nervous system, NaV1.6 located both centrallyand peripherally, and NaV1.7, NaV1.8, and NaV1.9 expressedprimarily in the peripheral nervous system.3 NaV1.4 is presentin skeletal muscle, and NaV1.5 is found predominantly incardiac muscle.4 Three VGSCs (NaV1.5, NaV1.8, and NaV1.9)are resistant to blockade by the sodium channel blocker

    tetrodotoxin (TTX),5 demonstrating subtype specificity withinthis gene family.A role for the NaV1.7 channel in pain perception was

    established by clinical gene-linkage analyses that revealed gain-of-function mutations in the SCN9A gene that encodes the -subunit of NaV1.7 channels as the etiological basis of inheritedpain syndromes such as inherited erythromelalgia andparoxysmal extreme pain disorder.6 Loss-of-function mutationsresult in the complete inability to sense any form of pain.7

    Global deletion of SCN9A in mice abolishes perception ofthermal, mechanical, inflammatory, and chemical pain,8 andcell-specific deletion reduces responsiveness to several forms ofpain.9 On the basis of such evidence, decreasing NaV1.7 channelactivity in peripheral sensory neurons has been proposed as aneffective pain treatment.10 A role for NaV1.7 in itch also issuggested by clinical genetics.11 Broad NaV antagonists, such asTTX, lidocaine, bupivacaine, phenytoin, lamotrigine, andcarbamazepine, have been shown to be useful for attenuating

    Received: November 13, 2014Published: February 6, 2015

    Article

    pubs.acs.org/jmc

    2015 American Chemical Society 2299 DOI: 10.1021/jm501765vJ. Med. Chem. 2015, 58, 22992314

    pubs.acs.org/jmchttp://dx.doi.org/10.1021/jm501765v

  • pain in humans and animal models but have a variety of sideeffects due to a lack of isoform specificity.12 A primarychallenge in the development of a NaV1.7 antagonist as atherapeutic is attaining sufficient selectivity against NaV1.5,expressed in cardiac tissue, and NaV1.4, in skeletal muscle, so asnot to impair normal cardiac and skeletal muscle function.13

    Spider venoms contain many peptide toxins that targetvoltage-gated ion channels, including KV, CaV, and NaVchannels, and have been useful tools to study channel structureand function.14 Two well-characterized examples of NaV1.7inhibitory peptides that display different NaV selectivity profilesand promiscuities toward other voltage-gated ion channelfamilies are Huwentoxin-IV (HWTX-IV) from the venom ofthe Chinese bird spider Selenocosmia huwena15 and Protoxin-II(ProTxII), isolated from the tarantula Thrixopelma pruriens.16

    Like many other spider toxins, these two peptides conform tothe inhibitory cystine knot (ICK) peptide structural motif17 andinhibit channel activation by binding to the voltage sensor andlocking the channel in a closed conformation. HWTX-IV,ProTxII, and two other reported NaV1.7 inhibitory peptides, -conotoxin KIIIA18 from cone snail venom and centipede toxinpeptide -SLPTX-Ssm6a,19 have been prepared and charac-terized in our lab for comparison of their biologic activities.Herein we report our identification and characterization of

    GpTx-1, a known antagonist of TTX-sensitive sodiumchannels,20 from the venom of the tarantula spider

    Grammostola porteri.21 GpTx-1 was first reported as a CaVchannel blocker after isolation from the venom of the closelyrelated Chilean tarantula Grammostola rosea and named GTx1-15 (UniproKB: accession no. P0DJA9).22 It was later identifiedin the venom of Paraphysa scrofa (Phrixotrichus auratus).23 Onthe basis of its potency and desirable NaV subtype selectivityprofile, we selected GpTx-1 as a lead in our effort to developtherapeutically useful NaV1.7 peptides. We describe a significantpeptide medicinal chemistry effort to investigate the GpTx-1structureactivity relationships and engineer analogues withimproved levels of NaV1.7 potency and selectivity against theimportant off-target NaV isoforms NaV1.4 and NaV1.5.

    RESULTS AND DISCUSSIONHigh-Throughput Screening of Venom Fractions. To

    identify a novel peptide inhibitor with NaV potency, 84 venomfractions from the tarantula Grammostola porteri (AtherisLaboratories, Switzerland, Melusine ref. MLU-020007) werescreened for activity against NaV1.7 (Figure 1). A 384-wellIonWorks Quattro (IWQ) platform, which evaluates receptorinhibition with a population patch clamp, was utilized for itshigh-throughput screening capability. Several venom fractionswith significant (>80% inhibition of peak current) NaV1.7inhibitory activity were identified, the first of which was fraction31. A second aliquot of this fraction was tested in the NaV1.7and NaV1.5 IWQ assays to confirm the activity of the hit and

    Figure 1. (A) Reversed phase (RP) HPLC fractionation of crude venom extracted from Grammostola porteri. The tick marks along the x-axisrepresent time slices of fractionation. (B) Activity of the isolated venom fraction in the NaV1.7 IonWorks Quattro (IWQ) assay. Fraction 31(indicated with rectangular box) contained a major peak in the RP-HPLC chromatogram that exhibited >80% inhibition of peak current in the ionchannel assay and was later identified as GpTx-1.

    Journal of Medicinal Chemistry Article

    DOI: 10.1021/jm501765vJ. Med. Chem. 2015, 58, 22992314

    2300

    http://dx.doi.org/10.1021/jm501765v

  • evaluate selectivity. All samples were tested for potency onsodium channels with electrophysiology to give a directmeasure of receptor inhibition. The validated hit fraction wasthen analyzed by high-resolution electrospray ionization (ESI)and matrix-assisted laser desorption ionization time-of-flight(MALDI-TOF) mass spectrometry (MS), which indicated thatthe fraction was a mixture of at least four distinct peptidespecies (Figures 2 and 3, respectively). The active fraction wasthen separated by reversed phase (RP) HPLC, and thecorresponding subfractions were screened for activity in theNaV1.7 and NaV1.5 IWQ assays. Subfraction 11 was the majorpeak in the RP-HPLC chromatogram and showed >90%inhibition of NaV1.7 activity (Figure 4). Deconvolution ofsubfraction 11 by Edman degradation and MS/MS sequencingrevealed the primary peptide sequence of GpTx-1 (1, Figure 5).GpTx-1 is a 34 residue, C-terminally amidated polypeptidecontaining six cysteine residues engaged in three disulfidebonds and is a putative member of NaSpTx family 1.24 Toconfirm its identity and activity, synthetic GpTx-1 waschemically synthesized using Fmoc solid-phase peptide syn-thesis (SPPS) to generate the linear peptide sequence, whichwas then oxidatively folded, purified by RP-HPLC to producethe final product (Figure 6), and tested.25 A coelution of thesynthetic and native products (1:1) was observed, confirmingthe authenticity of the synthetic versus native product (seeSupporting Information).Results of Electrophysiology Studies. Chemically

    synthesized GpTx-1 (1) was characterized in a manualelectrophysiology whole-cell patch clamp assay using human

    clones of several NaV subtypes (Figure 7). To test for inhibitionor stabilization of as many channel gating states as possible,doseresponse curves were measured with voltage clamped toholding potentials that imposed steady 20% fractionalinactivation. The IC50 values of NaV1.8, NaV1.7, NaV1.5,NaV1.4, and NaV1.3 inhibition for GpTx-1 were 12.2 2.2,0.0044 0.0020, 4.20 0.09, 0.301 0.041