Lys49 Myotoxin from the Brazilian Lancehead Pit Viper Elicits Pain Through Regulated ATP Release
Total Page:16
File Type:pdf, Size:1020Kb
Lys49 myotoxin from the Brazilian lancehead pit viper elicits pain through regulated ATP release Chuchu Zhanga, Katalin F. Medzihradszkyb, Elda E. Sánchezc,d, Allan I. Basbaume, and David Juliusa,1 aDepartment of Physiology, University of California, San Francisco, CA 94158; bDepartment of Pharmaceutical Chemistry, University of California, San Francisco, CA 94158; cNational Natural Toxins Research Center, Texas A&M University-Kingsville, Kingsville, TX 78363; dDepartment of Chemistry, Texas A&M University-Kingsville, Kingsville, TX 78363; and eDepartment of Anatomy, University of California, San Francisco, CA 94158 Contributed by David Julius, January 31, 2017 (sent for review September 16, 2016; reviewed by Baldomero M. Olivera and Mark J. Zylka) Pain-producing animal venoms contain evolutionarily honed tox- species within the Crotalinae subfamily (9). These toxins are ins that can be exploited to study and manipulate somatosensory devoid of phospholipase activity due to key enzymatic site mu- and nociceptive signaling pathways. From a functional screen, we tation of Asp49 to Lys49, but promote release of ATP from have identified a secreted phospholipase A2 (sPLA2)-like protein, myotubes through an as-yet uncharacterized mechanism (10–12). BomoTx, from the Brazilian lancehead pit viper (Bothrops moo- Similarly, we show that BomoTx lacks phospholipase activity and jeni). BomoTx is closely related to a group of Lys49 myotoxins that excites a cohort of sensory neurons through a mechanism in- have been shown to promote ATP release from myotubes through volving ATP release and activation of P2X2 and P2X3 purinergic an unknown mechanism. Here we show that BomoTx excites a receptors. Furthermore, we provide pharmacological and elec- cohort of sensory neurons via ATP release and consequent activa- trophysiological evidence to support a role for pannexin hemi- tion of P2X2 and/or P2X3 purinergic receptors. We provide phar- channels in mediating toxin-evoked nucleotide release. At the macological and electrophysiological evidence to support pannexin behavioral level, injection of BomoTx into the mouse hind paw hemichannels as downstream mediators of toxin-evoked ATP re- elicits nonneurogenic inflammatory pain, thermal hyperalgesia, lease. At the behavioral level, BomoTx elicits nonneurogenic inflam- and mechanical allodynia. The latter is completely dependent on matory pain, thermal hyperalgesia, and mechanical allodynia, of purinergic signaling. Together, these findings elucidate a de- which the latter is completely dependent on purinergic signaling. tailed mechanism by which a Lys49 myotoxin produces pain as- Thus, we reveal a role of regulated endogenous nucleotide release sociated with envenomations common to Latin America (13). in nociception and provide a detailed mechanism of a pain-inducing Lys49 myotoxin from Bothrops species, which are responsible for Results the majority of snake-related deaths and injuries in Latin America. A Snake Toxin That Targets Somatosensory Neurons. To identify novel toxins that target nociceptors, we used live-cell calcium Lys49 myotoxin | ATP release | pannexin | pain | purinergic receptor imaging to screen a library of 22 snake venoms for their ability to activate primary sensory neurons cultured from rat trigeminal enoms from spiders, snakes, cone snails, and scorpions have ganglia (TG). Because many snake venoms contain cytolytic com- Vevolved to produce a vast pharmacopoeia of toxins that ponents, such as proteases and lipases (1, 14), initial screening of modulate receptor or channel function as a means of producing these samples resulted in cell injury and widespread calcium shock, hemolysis, paralysis, and pain (1, 2). As such, venoms from influx that masked the action of any specific components. To these animals represent a rich potential source of potent and se- circumvent this problem, we first processed all crude venoms lective toxins that can be exploited to study and manipulate so- matosensory and nociceptive signaling pathways. For example, toxins Significance can directly target ion channels expressed by nociceptive sensory neurons to elicit or inhibit pain. Well-validated examples include Bites from venomous snakes can inflict substantial pain and toxins that target transient receptor potential (TRP) ion channels, inflammatory tissue damage. Snake phospholipase A2 (PLA2), acid-sensing ion channels (ASICs), and Nav channels (3–8). How- “ ” and PLA2-like toxins that retain the PLA2 fold but lack enzy- ever, the toxinome is incredibly diverse, and there are many pain- matic activity, are commonly found in snake venoms. In this producing venoms for which the responsible toxins have not yet study, we identify a PLA2-like toxin (BomoTx), from the Brazilian been identified. lancehead pit viper (Bothrops moojeni), that activates a sub- In addition to targeting receptors and channels directly, population of somatosensory neurons that contribute to pain algogenic toxins could also promote the synthesis and/or release sensation. We show that BomoTx excites these neurons by of metabolites, transmitters, or second messengers that activate stimulating the release of cellular ATP through a mechanism primary afferent nociceptors. Although this could occur through involving pannexin hemichannels. Consequent activation of indiscriminate actions of venom lipases and proteases that re- purinergic receptors elicits acute pain, tissue inflammation, and lease cellular contents through tissue damage and cell death, pain hypersensitivity. Thus, we have elucidated the mechanism more specific mechanisms may also exist. Their identification of action for a toxin from Bothrops snakes, which inflict a could reveal important new pathways by which exogenous or majority of bites in Latin America. endogenous pain-producing agents mediate their effects. To identify novel pain-producing toxins, we have exploited Author contributions: C.Z., A.I.B., and D.J. designed research; C.Z. and K.F.M. performed cultured sensory neurons as a facile in vitro platform for un- research; C.Z. and E.E.S. contributed new reagents/analytic tools; C.Z., K.F.M., A.I.B., and D.J. analyzed data; and C.Z. and D.J. wrote the paper. biased screening of crude venoms, optimizing conditions to re- Reviewers: B.M.O., University of Utah, Salt Lake City; and M.J.Z., University of North duce nonspecific cell damage so as to reveal toxins having Carolina. specific cellular effects. With this approach, we have identified a The authors declare no conflict of interest. toxin (BomoTx) from the Brazilian lancehead pit viper (Bothrops Data deposition: The sequence reported in this paper has been deposited in the GenBank moojeni) found mainly in Brazil. BomoTx belongs to a group of database (accession no. KX856005). secreted phospholipase A2 (sPLA2)-like proteins that have the 1To whom correspondence should be addressed. Email: [email protected]. conserved PLA2 fold but lack enzymatic activity, and is most This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. closely related to so-called Lys49 myotoxins found in snake 1073/pnas.1615484114/-/DCSupplemental. E2524–E2532 | PNAS | Published online March 6, 2017 www.pnas.org/cgi/doi/10.1073/pnas.1615484114 Downloaded by guest on September 24, 2021 through a 50-kDa molecular mass cutoff filter to enrich for peptides inwardly rectifying current (Fig. 2A). Consistent with our pharma- PNAS PLUS and small proteins. Indeed, this eliminated background activities cological profiling by calcium imaging, extracellular ATP could in more than half of the crude venoms examined. Among these, invariably elicit currents in these neurons (Fig. 2B). Moreover, filtrates from B. moojeni (Brazilian lancehead) (Fig. 1A) robustly Ro51, a specific blocker of P2X3 and P2X2/3 ionotropic subtypes activated a subpopulation of TG neurons (Fig. 1B). Fractionation of (15), reduced toxin-evoked responses by 72.2 ± 5.3% (Fig. 2B). this crude venom by reversed-phase (C18) chromatography yielded a Does BomoTx elicit its response by activating P2X receptors single active peak (Fig. S1A) that recapitulated activity observed directly, or by promoting release of ATP? To address this with the crude venom. question, we first asked whether the toxin elicits membrane To determine the toxin’s identity, we analyzed purified toxin by currents in transfected HEK293 cells expressing P2X2 and/or MALDI-TOF mass spectrometry, revealing a single polypeptide P2X3 receptors. No responses were observed (Fig. 2C), nor did with a molecular mass of 13,835 Da (Fig. S1C). The sample was the toxin shift dose–response curves for ATP (Fig. S2A), in- then trypsin-digested and analyzed by liquid chromatography-tan- dicating that BomoTx is not a direct purinergic receptor agonist dem mass spectrometry (LC/MS/MS). Homology considerations [to or modulator. Hence, the toxin likely promotes release of ATP National Center for Biotechnology Information (NCBI) accession to activate transient responses in specific sensory neurons that nos. 17865560 and 17368325 and UniProt accession no. Q9PVE3] express purinergic receptors. and manual de novo sequencing provided a tentative full-sequence If ATP-evoked activation of P2X receptors accounts for determination, which was then used to clone the full-length (122- BomoTx sensitivity in transiently responding neurons, ectopic amino acid) cDNA from the Brazilian lancehead venom gland. expression of these channels in sensory neurons should increase Indeed, the amino acid sequence translated from the