Amphibian peptides that inhibit neuronal nitric oxide synthase The isolation of lesueurin from the skin secretion of the Australian Stony Creek Frog Litoria lesueuri Jason Doyle 1 , Lyndon E. Llewellyn 1 , Craig S. Brinkworth 2 , John H. Bowie 2 , Kate L. Wegener 2 , Tomas Rozek 2 , Paul A. Wabnitz 2 , John C. Wallace 3 and Michael J. Tyler 4 1 Australian Institute of Marine Science, Townsville MC, Queensland, Australia; 2 Departments of Chemistry, 3 Molecular Biosciences and 4 Environmental Biology, The University of Adelaide, South Australia Two neuropeptides have been isolated and identi®ed from the secretions of the skin glands of the Stony Creek Frog Litoria l esueuri. The ®rst of these, the known neuropeptide caerulein 1.1, is a common constituent of anuran skin secretions, a nd has t he sequence p EQY(SO 3 )TGWMDF- NH 2 . This n europeptide i s s mooth m uscle active, an anal- gaesic more potent t han morphine and is also thought t o be a hormone. The second neuropeptide, a new peptide, has been named lesueurin and has the primary structure GLLDILKKVGKVA-NH 2 . Lesueurin shows no signi®- cant antibiotic or anticancer activity, but inhibits the for- mation of the ubiquitious chemical messenger nitric oxide from neuronal nitric oxide synthase (nNOS) at IC 50 (16.2 l M ), and is the ®rst amphibian peptide reported to show inhibition of nNOS. As a consequence of this activity, we have tested other peptides previously isolated from Australian amphibians for nNOS inhibition. There are three groups of peptides that inhibit nNOS (IC 50 at l M concen- trations): these are (a) the citropin/aurein type peptides (of which lesueurin is a member), e.g. citropin 1.1 (GLFDVIKKVASVIGGL-NH 2 )(8.2l M ); (b) the frenatin type peptides, e.g. frenatin 3 (GLMSVLGHAVGNVLG GLFKPK-OH) (6.8 l M ); and (c) the caerin 1 peptides, e.g. caerin 1.8 (GLFGVLGSIAKHLLPHVVPVIAEKL-NH 2 ) (1.7 l M ). From Lineweaver±Burk plots, t he mechanism of inhibition is revealed as noncompetitive with respect to the nNOS substrate arginine. When the nNOS i nhibition tests with the t hree p eptides outlined above w ere c arried out in the presence of increasing concentrations of Ca 2+ calmodulin, the inhibition dropped by  50% in each case. In addition, these peptides also inhibit the activity of calcineurin, another enzyme that requires the presence of the regulatory protein Ca 2+ calmodulin. It i s p roposed that the amphibian peptides inhibit nNOS by interacting with Ca 2+ calmodulin, and as a consequence, blocks the attachment of this protein to the calmodulin domain of nNOS. Keywords: amphibians; Litoria lesueuri; neuropeptides; nNOS inhibition; Ca 2+ calmodulin interaction. Amphibians have rich chemica l arsenals that form an integral part of their defence system, and also assist with the regulation of dermal physiological action. In res ponse to a variety of stimuli, host defence compounds are secreted from specialized glands onto the dorsal s urface and into the gut of t he amphibian [1±4]. A number of different types o f bio-active peptides have been identi®ed from t he glandular skin secretions of Australian anurans o f the Litoria genus, including (a) n europeptides of the caerulein family [5±8], and (b) wide-spectrum antibiotics, e.g. the caerin peptides from green tree frogs of the genus Litoria [6±8], the citropins from the tree frog Litoria citropa [9,10], and the aureins from the bell frogs Litoria aurea and Litoria raniformis [11]. Amongst the most a ctive of t hese are n europeptide caeru- lein 1.1, and the antibiotics caerin 1 .1, citropin 1.1 and aurein 1.2: caerulein 1.1 pEQGY(SO 3 )TGWMDF-NH 2 ; caerin 1.1 G LLSVLGSVAKHVLPHVVPVIAEHL-NH 2 ; citropin 1.1 GLFDVIKKVASVIGGL-NH 2 ;aurein1.2 GLFDIIKKIAESF-NH 2 . Aurein 1.2 contains only 13 a mino-acid residues, and is the smallest peptide from an anuran reported to have signi®cant antibiotic activity. The aurein peptides have also been shown to exhibit anticancer activity in tests carried ou t by the N ational Cancer Institute ( NCI, Washington DC, USA) [12]. The s olution s tructures o f t he antibiotic (and anticancer active as appropriate) peptides shown above have been investigated by NMR spectroscopy. In tri¯uoroethanol/ water mixtures, caerin 1.1 adopts two well-de®ned helices (Leu2 to Lys11 and from Val17 to His24) separated by a hinge region of less-de®ned helicity and greater ¯exibility, with hydrophilic and hydrophobic residues occupying well- de®ned zones [ 13]. The central hinge region is n ecessary for optimal antibiotic activity [13]. Similar NMR studies of Correspondence to J. H. Bowie, Department of Chemistry, The University of Adelaide, South Australia, 5005. Fax: + 61 08 83034358, Tel.: + 61 08 83035767, E-mail: john.bowie@adelaide.edu.au Abbreviations: FAD, ¯ avin adenine dinucleotide, oxidized form; FMN, ¯avin mononucleotide; IC 50 , concentration (of peptide) which causes 50% inhibition; MS/MS, mass spectroscopy/mass spectro- scopy; NADPH, nicotinamide adenine nucleotide phosphate, reduced form; NCI, National Cancer Institute (Washington); nNOS, n euronal nitric oxide synthase. (Received 18 June 2001, revised 8 October 2001, accepted 23 October 2001) Eur. J. Biochem. 269, 100±109 (2002) Ó FEBS 2002 citropin 1.1 [9] and aurein 1.2 [11] show that these peptides adopt conventional amphipathic a helical structures, a feature commonly found in membrane-active agents [1±4,8]. Interaction occurs at the m embrane s urface with the c harged, a nd normally basic peptide adopting an a helical conformation and a ttaching itself to charged, and normally anionic sites on the lipid bilayer. This ultimately causes disruption of norm al membrane function leading to lysis of t he bacterial or c ancer cell [14±16]. Many Australian anurans that we have studied conform to the model outlined above in t hat t hey have a v ariety of host d efence peptides in skin glan ds including a neuropep- tide that acts on smooth muscle, and at least one powerful wide-spectrum antibiotic peptide such as those d escribed above [8]. However there are some species of anuran that divert markedly from this scenario. For example, the Red Tree Frog Litoria rubella [17±19], and the related species Litoria electrica [20] excrete neither antimicrobial peptides nor neuropeptides such as caerulein. Instead, they release large quantities o f small peptides, called tryptophyllins [21], onto t heir skin, wh ich are thought to be neurotransmitters or neuromodulators [22], at least partly because they s how structural similarity to the human brain endomorphins (e.g. YPWG-NH 2 ) [23]. The sequences of two tryptophyllins are as follows: tryptophyllin L1, FPWL-NH 2 ; tryptophyllin L2, pEFPWL-NH 2 . Even more unusual are the marsh frogs o f the Limno- dynastes genus. These produce only minute amounts of anionic skin peptides, none of which are post-translationally modi®ed, or s how neur opeptide or antimicrobial activity [24,25]. A particular example, dynastin 1 (from Limnodyn- astes interioris) [ 24] has t he sequence GLLSGLGL-OH. In this paper, we describe the isolation, sequence deter- mination, and activities of t wo bioactive p eptides f rom the skin glands of the stony creek frog Litoria l esueuri. We have not studied a member of the L. lesueuri complex of frogs previously, and we have now found that it does not produce antimicrobial peptides, unlike the majority of the other members of the genus Litoria. Instead, it produces a neuropeptide t hat i nhibits the formation of nitric oxide by neuronal n itric oxide synthase (nNOS). We next examined other p eptides isolated earlier from a nurans of t he genus Litoria. This has led to the discovery of three types of amphibian peptides that inhibit nNOS. MATERIALS AND METHODS Preparation of skin secretions L. lesueuri was h eld by the back legs, the skin moistened with deionized water, and the granular dorsal glands situated on the back were stimulated by means of a bipolar electrode of 21G p latinum attached to a Palmer Student Model electrical stimulator. The electrode was rubbed gently i n a circular manner on the particular gland (under study) of the animal, using 10 V and a pulse duration of 3 ms [26]. The resulting secretion was washed from the f rog with d eionized water (50 mL), t he mixture diluted with an equal volume of methanol, centrifuged, ®ltered through a Millex HV ®lter unit (0.45 lm), and lyophilized. This w ork conforms with the Code of Practice for t he Care and Use of Animals for Scienti®c Purposes (1990) and the Prevention of Cruelty to Animals Act (1985), a nd was approved by the University of Adelaide Animal Ethics Committee. HPLC separation of peptide material HPLC separation of the s kin secretion was achieved using a VYDAC C18 HPLC column (5lm, 300 A Ê ,4.6´ 250 mm) (Separations Group, Hesperia, CA, USA) equilibrated with 10% acetonitrile/aqueous 0.1% tri¯uoroacetic acid. The lyophilized mixture (0.5 mg) was dissolved in deionized water (20 lL) and i njected into the c olumn. The e lution pro®le was generated using a linear gradient produced by an ICI DP 800 Data Station controlling two LC1100 HPLC pumps, increasing from 10 to 75% acetonitrile over a period of 60 min at a ¯ow rate of 1 mLámin )1 . T he eluant was monitored by ultraviolet absorbance at 214 nm using an ICI LC-1200 variable wavelength detector (ICI Australia, Melbourne, Australia). The r esultant HPLC trace shows two major peaks (Fig. 1). The two fractions were collected, concentrated and d ried in vacuo. T he ®rst major fraction (Fig. 1B) (50 lg) contained the known neuropeptide caeru- lein 1.1, identi®ed b y HPLC a nd mass spectrometry [27]. The second major f raction (Fig. 1C) contained 10 lgofa new peptide, called l esueurin. Sequence determination of lesueurin Electrospray mass spectrometry. Electrospray mass spec- tra were determined u sing a F innigan LCQ ion t rap mass spectrometer. Puri®ed fractions from the HPLC s eparation were dissolved in m ethanol/water (1 : 1 , v/v) a nd infused into the electrospray source at 8 lLámin )1 . Electro spray conditions were as follows: n eedle potential 4.5 kV, tube lens 60 V, heated capillary 200 °C and 30 V, sheath gas ¯ow 30 p.s.i. Mass spectra were acquired with the automatic gain control on, a maximum time of 400 ms, and averaging over three microscans. Mass spectrometric sequencing was carried by the MS/MS method using B and Y + 2 fragmentations [28]. Amino-acid sequencing. Automated Edman sequencing of lesueurin was performed by a standard procedure a s described p reviously [29] using an applied Biosystem 492 Fig. 1. H PLC separation of the glandular secretion of Litoria lesueuri. B, caerulein 1.1; C, lesueurin. A, n onpep tide material. Ó FEBS 2002 Amphibian peptides that inhibit nNOS (Eur. J. Biochem. 269) 101 procise sequencer equipped with a 900-A data analysis module. The best results were ob tained using a disc o f immobilon ®lm treated with bioprene in ethanol, onto which t he peptide was absorbed fr om aqueous acetonitrile (90%). The disc was pierced several times with a razor blade to aid t he ¯ow of s olvent. Preparation of synthetic lesueurin Lesueurin w as synthesized ( by Mimotopes, Clayton, Vic- toria, Australia) using L -amino acids via the standard N-a-Fmoc method (full d etails including protecting g roups and deprotection have b een reported rece ntly [30]). Syn - thetic lesueurin w as shown to be i dentical to the n atural lesueurin by e lectrospray mass spectrometry, and HPLC. Bioactivity assays Antimicrobial testing. Synthetic lesueurin was t ested for antibiotic activity by the M icrobiology D epartment o f the Institute of Medical and Veterinary Science (Adelaide, Australia) by a standard method [31]. T he method used involved the measurement of inhibition zones ( produced by the a pplied peptide) o n a thin agarose plate containing the microorganisms under study. T he microorganisms u sed in this procedure were Bacillus cereus, Escherichia coli, Leuco- nostoc lactis, Listeria innocua, Micrococcus luteus, Pasteu - rella multocida , Staphylococcus aureus, Stap hylococcus epidermidis and St reptococcus uberis. Lesueurin s howed no activity at MIC values below 100 lgámL )1 against any of these organisms. Anticancer activity testing. Synthetic lesueurin showed no activity below 10 )4 M in the Ô60-human tumour line testing programÕ of the US NCI (Washington) [12]. Neuronal nitric oxide synthase inhibition. Inhibition of nNOS was measured by monitoring the conversion of [ 3 H]arginine to [ 3 H]citrulline. In brief, this involved incuba- tion of a homogenate o f r at cerebella (which had endogenous arginine removed b y ion exchange chromatography) i n a reaction buffer (33 m M Hepes, 0.65 m M EDTA, 0.8 m M CaCl 2 ,3.5 lgámL )1 calmodulin, 670 l M b-NA DPH, 670 l M dithiothreitol, pH 7.4) containing 20 n M [ 3 H]arginine (NEN Life Sciences, Boston, MA, USA). The n NOS inhibitor, N x -nitro- L -arginine (1 m M ) was used to measure back- ground radioactivity. Reactions were terminated after 10 min with 50 lLof0.3 M EGTA. An a liquot (50 lL) of this quenched reaction mixture was transferred to 50 lLof 500 m M Hepes (pH 5.5). AG50W-X8 (Na + form) resin (100 lL) was added to separate [ 3 H]arginine from [ 3 H]cit- rulline. After repeated vortexing, this slurry was centrifuged at 1200 g for 10 min, and 70 lL of supernatent was removed and the [ 3 H]citrulline measured b y scintillation c ounting. Peptides selected for f urther examination t o d etermine the mechanism of i nhibition were assayed in t he same reaction buffer as used for initial screening except that it contained 30 n M [ 3 H]arginine supplemented with 0.3±13.3 m M argi- nine. Peptide concentrations used are given in the legend t o Fig. 4. Data analysis for nNOS studies. Peptide inhibition curves were ®tted using the curve-®tting routin e of SIGMAPLOT (SPSS, Chicago, IL, USA) with the variation o f the Hill equation: fmol [ 3 H]citrulline production  1/(1 + [inhibi- tor]/IC 50n ), where I C 50 is the c oncentration a t w hich the peptide causes 50% inhibition and n is the slope of the curve and can be considered as a pseudo Hill coef®cient [32]. Lineweaver±Burk plots [33] were generated using SIGMA- PLOT (SPSS, Chicago, I L, USA). Calcineurin assay. This assay was performed following the manufacturer's protocol [34] with minor modi®cations. Calcineurin was diluted to 0.036 UálL )1 in enzyme dilution buffer (50 m M Tris, 0.5 mgámL )1 BSA, pH 7.4). Peptides (5 lL) were assayed i n duplicate in a reaction mixture containing 16 m M p-nitrophenylphosphate, 0.4 mgámL )1 BSA, 0.8 m M NiCl 2 ,4lgámL )1 calmodulin in Tris buffer (40 m M , pH 7.4). The reaction was started with the addition of the enzyme (5 lL; 0.1 U ). All samples were assayed at 30 °C with positive controls containing water in the place of the test sample and negative controls containing no enzyme. Absorbance (A) readings at 405 nm were taken after 30 min with readings av eraged, a djusted t o the change in A per minute and corrected for background A (negative control). Percent of control w as then calculated as (A 405 test sample / A 405 positive control ) ´ 100. RESULTS L. lesueuri, usually called either Lesueur's Frog or the Stony Creek Frog, has varied do rsal colouration ranging fromyellowtobrown,withablackheadstripefromthe snout to the tympanum [35]. The animal ranges from 37 to 63 mm in length, and is often found in the vicinity of rocky streams in coastal regions from north of Queens- land to eastern Victoria. It is reported that there are t wo distinct populations of this frog, one con®ned to north- eastern Queensland, the other in New South Wales and Victoria. Whether these are two different subspecies of L. lesueuri or two different species has not yet been determined [36]. AsinglespecimenofL. lesueuri , collected at Atherton, Queensland, was used in this study. The electrical stimula- tion method [26] was used to elicit secretion from the granular skin glands. The animal was not harmed in this study. Less than 0.5 mg of material was obtained following work up of the secretion, and this contained less than 100 lg of peptide material. This is an unusually small amount of peptide material. It is normal for a species of the genus Litoria to contain 5 mg of peptide material in the glandular secretion; some frogs, such as Litoria splendida, secrete more than 100 mg of peptide material [8]. HPLC separation (Fig. 1) revealed t wo major p eptide components. The ®rst, and m ajor component was identi®ed as caerulein 1.1 from its electrospray mass spectrum and HPLC behaviour [27]. This potent smooth m uscle agent, analgaesic and hormone is often a m ajor peptide component of species of the g enus Litoria [8]. The minor component is a 13-residue peptide that we have called lesueurin. Only 10 lgofthismaterialwas available for study. The MS/MS of the MH + ion of lesueurin is shown in Fig. 2. As this mass spectrum does not differentiate between isomeric Leu or Ile or isobaric Lys and Asn, we have con®rmed the identity of these f our residues using the automated Edman procedure [29]. The automated 102 J. H. Bowie et al. (Eur. J. Biochem. 269) Ó FEBS 2002 Edman procedure identi®es all but the last amino acid in the lesueurin sequence. This residue, C-terminal A la-NH 2 ,is clearly identi®ed by mass spectrometric fragmentation data (Fig. 2). A synthetic sample of lesueurin was prepared, and shown to be identical with natural lesueurin using HPLC and mass spectrometry. The sequence of lesueurin is GLLDIL KKV GKVA- NH 2 . Lesueurin h as no antibiotic activity (at minimum inhibitory concentration values below 100 lgámL )1 ) against the nine bacteria we use in our test regime. Also, lesueurin exhibited no cytotoxity at concentrations less than 10 )4 M against any of the 60 human tumour lines in the NCI test regime. Lesueurin was, however, shown to inhibit nNOS with an IC 50 of 16.2 l M (Table 1) . This Fig. 2. Ele ctrospray mass spectrum (MS/MS) of the MH + ion of lesueurin. Masses shown in this spectrum are nominal m asses. Data from B fragmentations give sequence information from the C-terminal end of lesueurin (see schematic arrows above the spectrum). D ata from Y + 2 f ragmentations give sequence information from the N-terminal end of the peptide ( see schematic arrows below the spectrum). For a review of fragmentations of MH + ions of peptides see [28]. This method does not distinguish between isomeric Leu and Ile ( nominal mass 113) and isobaric Lys and Gln (nominal mass 128). These residues have been dierentiated using the automated Edman procedure [29]. The correct residues are shown in this ®gu re. Table 1. n NOS Inhibition by amphibian peptides. Amino-acid sequences, so urce species and na ming convention used where a peptide has been named previously. Peptides with minimal or no eect are h ighlighted with respective concentration used listed in f ootnotes below. All peptides in Group 1 are amphipathic a helices in solution [10,11]. C aerin 1 peptides in solution have a helices from r esidues 1±1 1 a nd 18±25, with a ¯exible hinge region including residues 12±17 [13]. Name Source Sequence IC 50 (l M ) Hill slope Net charge Inactive or weakly active peptides Rubellidin 3.1 a L. rubella [17] IEFFT-NH 2 NA NA 0 Trytophyllin L3.1 b L. rubella [17] FPWP-NH 2 NA NA +1 Electrin 2.1 c L. electrica [20] NEEEKVKWEFPDVP-NH 2 NA NA )2 Caeridin 3 d L. caerulea [8] GLFDAIGNLLGGLGL-NH 2 NA NA 0 Maculatin 1.3 e L. eucnemis f GLLGLLGSVVSHVLPAITQHL-NH 2 NA NA +1 Inhibitor Group 1 Lesueurin L. lesueuri g GLLDILKKVGKVA-NH 2 16.2 1.5 +3 Aurein 1.1 L. aurea [11] GLFDII KK I AES I-NH 2 33.9 2.0 +1 Citropin 1.1 L. citropa [10] GLFDVIKKVASVIGGL-NH 2 8.2 1.6 +2 Aurein 2.2 L. aurea [11] GLFDIVKKVVGALGSL-NH 2 4.3 2.5 +2 Aurein 2.3 L. aurea [11] GLFDIVKKVVGIAGSL-NH 2 1.8 1.7 +2 Aurein 2.4 L. aurea [11] GLFDIVKKVVGTLAGL-NH 2 2.1 3.1 +2 Inhibitor Group 2 Not named L. dahlii f GLLGSIGNAIGAFIANKLKP-OH 3.2 2.2 +3 Frenatin 3 L. infrafrenata [8] GLMSVLGHAVGNVLGGLFKPKS-OH 6.8 1.4 +3 Splendipherin L. splendida [67] GLVSSIGKALGGLLADVVKSKGQPA-OH 8.5 1.3 +3 Inhibitor Group 3 Caerin 1.1 L. splendida [6] GLLGVLGSIAKHVLPHVVPVIAEHL-NH 2 36.6 1.4 +1 Caerin 1.10 L. chloris [8] GLLSVLGSVAKHVLPHVVPVIAEKL-NH 2 41.0 0.6 +2 Caerin 1.6 L. chloris [8] GLFSVLGAVAKHVLPHVVPVIAEKL-NH 2 8.5 1.7 +2 Caerin 1.8 L. chloris [8] GLFKVLGSVAKHLLPHVVPVIAEKL-NH 2 1.7 3.7 +3 Caerin 1.9 L. chloris [8] GLFGVLGSI AKHVLPHVVPVIAEKL-NH 2 6.2 2.2 +2 a Inactive at 13.3 lgámL )1 . b Inactive at 33.3 lgámL )1 . c Inactive at 66.7 lgámL )1 . d Inactive at 133.3 lgámL )1 . e Caused 46.4% inhibition at 31.4 l M . Full IC 50 determination was not possible because of solubility. f Brinkworth, C., Bowie, J.H., Wallace, J.C. & Tyler, M.J. unpublished work. g Present paper. Ó FEBS 2002 Amphibian peptides that inhibit nNOS (Eur. J. Biochem. 269) 103 discovery of previously unreported pharmacological activity of an amphibian peptide p rompted us to t est for similar bioactivity of amphibian peptides that we have investigated previously. Table 1 lists the sequences of the natural peptides t ested, both active and inactive towards nNOS. The solubility of t hese peptide s and the maximum tolerable amount of the peptide solvent in the nNOS assay affected the maximum concentration tested for several of the peptides. Those peptides found to inhibit nNOS from the primary screen were then titrated to determine t heir IC 50 parameters. T hese results are given in Table 1 together with the slope of the inhibition curves and various physical properties. Three examples of the inhibition curves, including that generated by lesueurin, are shown in Fig. 3. Four model peptides were selected from the inhibitor groups (Table 1) to further examine t he mode of action by which they were inhibiting nNOS. Lesueurin and citro- pin 1.1 (both from inhibitor group 1), f renatin 3 (inhibitor group 2) and c aerin 1 .9 (inhibitor group 3) all produced Lineweaver±Burk plots consistent with a noncompetitive mode of inhibition (Fig. 4) with respect to the nNOS substrate arginine. Thus inhibition is not mediated by direct action upon the arginine-catalysing site. Rat cerebellar nNOS may retain endogenous calmodulin rendering it unsuitable f or a Michaelis±Menten study of enzyme kinet- ics, as was carried out with ar ginine. An experimental procedure was used to gauge the potential for selected peptides to inhibit nNOS by displacing Ca 2+ calmodulin from the calmodulin b inding domain of nNOS. In this p rocedure, the nNOS inhibition experiments were carried out with citropin 1.1, frenatin 3, and caerin 1.9 with added Ca 2+ calmodulin to determine the in¯uence o f the calmodulin on th e inhibition of nNOS. These experi- ments measured nNOS activity in the presence of 142.9 lgámL )1 of the a ctive p eptide (about 10-fold greater than the IC 50 value of each peptide), with the Ca 2+ calmodulin concentration in the assay buffer increased 100- Fig. 3. Inhibition of nNOS exempli®ed by (A) lesueurin (circles) and frenatin 3 (squares), and (B) citropin 1.1 (inverse triangles) and caerin 1.9 (triangles). These p ept ides represent each of the groups of peptide inhibitors listed in Table 1. Curves are drawn to the Hill equa- tion, and the values f or the I C 50 and slope of the curve are given in Table 1. Fig. 4. Li neweaver±Burk plots derived from changes in enzyme velocities in the presence of increasing amounts of the peptides citropin 1.1 (A), lesueurin (B), caerin 1.9 (C)andfrenatin3 (D). Enzyme k inetics data o btained in the absence of peptide are depicted by ®lled cir- cles, and re present the uninhibited re action. Open circles show t he e ect o f t he ci tropin 1.1, lesueurin, caerin 1.9 and frenatin 3 at 4.1, 24.7, 5.1 a nd 6.1 l M , respectively. Filled inverse triangles show the eect when the concentra- tion of citropin 1.1, lesueurin, caerin 1.9 and frenatin 3 are increased to 8.2, 49.4, 10.3 and 12.2 l M , respectively. 104 J. H. Bowie et al. (Eur. J. Biochem. 269) Ó FEBS 2002 fold to 350 lgámL )1 . For citropin 1.1, the inhibition decreased from 94 to 58% in the presence of Ca 2+ calmod- ulin. T he results for frenatin 3 and caerin 1.9 are 8 3±52%, and 84±53%, r espectively. Selected peptides were tested for t heir ability t o inhibit calcineurin, another calmodulin dependent enzyme. Lesueurin at 7 4 l M (4.6-fold g reat er tha n t he I C 50 against nNOS) inhibited c alcineurin by 33%. C itropin 1 .1 reduced calcineurin a ctivity b y 34% at 31 l M (3.8-fold h igher t han the IC 50 against nNOS) but at double this concentration (i.e. 7.6-fold greater than its IC 50 against nNOS), calcineurin was inhibited by 96%. Frenatin 3 inhibitied calcineurin by 38% at 46 l M , a concentration that is 6.8-fold g reater than the IC 50 against nNOS. Finally, caerin 1.9 inhibited c alcineurin by 48.1% at 19.3 l M (4.8-fold more concentrated than its IC 50 against nNOS). DISCUSSION Lesueurin is the name we have given to a new 13-residue basic peptide isolated from L. lesueurii. The sequence shows reasonable homology to that o f aurein 1.1, an antimicrobial and anticancer peptide isolated from L. aurea [11] (Table 2). There are some critical differences in the hydrophilic residues o f the two peptides w ith lesueurin having Lys11 whereas aurein 1.1 has Glu11. Even so, we predicted that lesueurin should show similar antimicrobial and anticancer activity to that of aurein 1.1. Surprisingly, lesueurin neither exhibits antibiosis below 100 lgámL )1 , or cytotoxity below 10 )4 M in the 60 human tumour lines in the NIC test program. Thus, L. lesueuri, as with only a minority of species of the g enus Litoria (see introduction), has no host defence a ntibiotic against microbial pathogens. To discover whether lesueurin has other biological properties, it was subjected to the bioactive molecule discovery p rogram of the Australian Institute of Marine Science. It effectively inhibitednNOSwithanIC 50 value of 16.2 l M with a slope of 1.5. This is the ®rst instance of an amphibian peptide that inhibits the f ormation of the chemical messenger nitric oxide. Testing of other am phibian peptides indicated that there are three well-de®ned groups of basic peptides that i nhibit nNOS, and the results are summarized in Table 1. These are: (a) the aurein/citropin group of peptides, of which lesueurin is a member. Most of these (lesueurin is a notable exception) are membrane active peptides, which show potent antibiotic acitivity, and in the case of the aureins, signi®cant anticancer activity. These peptides are amphi- pathic ahelices, as evidenced by solution NMR studies on aurein 1.2 [11] and citropin 1 .1 [9]; (b) the frenatin 3 type peptides, molecules that are characterized by a C-terminal CO 2 H group together with two lysine residues near the C-terminus. T hese peptides show little or no antibiotic and/ or anticancer activity; (c) those caerins 1, particularly those containing Phe3. These molecule s are also potent mem- brane-active antibiotics, and NMR studies show they have two a helical reg ions separated by a central ¯exible h inge region [13]. How do these three seemingly unrelated groups of peptides inhibit t he formation of n itric oxide by nNOS? The three nitric oxide synthases, namely neuronal, endothelial and inducible, are highly regulated enzymes responsible for the synthesis of the signal molecule nitric oxide. They are amongst th e most c omplex enzymes known (for nNOS see [ 37,38]). By a complex sequence involving binding sites for a number of cofactors including heme, tetrahydrobiopterin, F MN, FAD and NADPDH, nNOS converts arginine to citrulline, releasing the short-lived but reactive radical nitric oxide [39,40]. Nitric oxide synthases are composed of two domains: (a) the catalytic oxygenase domain t hat binds heme, tetrahydrobiopterin and the substrate arginine, and (b) the electron-supplying reductase domain that binds NADPH, FAD a nd FMN. Communi- cation between the oxygenase and reductase domains is determined by the regulatory enzyme calmodulin that interacts a t a speci®c site between the t wo domains. In the cases of the nNOS and e NOS isoforms, the calmodulin is regulated b y intracellular Ca 2+ , b ut not for i NOS [41±44]. Dimerization of the oxygenase do main is necessary for catalytic activity [39,40]. The signi®cant departure of the Hill slope of all of the inhibition curves from unity was the ®rst indication that these peptides are not acting directly upon the arginine substate site of nNOS [33]. A Hill slope  1 would indicate an interaction between a single active enzyme element with a single substrate. As Hill slopes > 1 are obtained, some other interaction must be causing the inhibition of nNOS. To con®rm this suspicion and to elucidate the general mech- anism of inhibition, an analysis was conducted of the kinetics of the nNOS inhibition reaction at different inhibitor concentrations. Lineweaver±Burk plots are shown in Fig. 4 for representatives of each of the inhibitor groups listed in Table 1, namely, lesueurin as the original nNOS inhibitor found, citropin 1.1, anoth er member of inhibitor group 1, frenatin (inhibitor group 2) and caerin 1.9 (inhibitor group 3). The fact that the regression lines plotted for each inhibitory peptide shown in Fig. 4 all intercept at a common point on the X-axis of t hese plots is typical of noncompetitive inhibition, and so these peptides are unlikely to directly involve the arginine substrate site [33]. In its simplest de®nition, noncompetitive inhibition is when an inhibitor binds at a site other than the active site, changing the e nzyme±substrate af®nity. These four peptides, lesueurin, citropin 1.1, frenatin 3 and c aerin 1 .9, a nd pr obably t he o ther active amphibian peptides of inhibitor groups 1±3, m ust therefore inhibit t he formation of nitric o xide by either blocking one or more of the cofactor s ites on nNOS or by some chemical modi®ca- tion reaction with nNOS that alters the activity of the enzyme. A n obvious example o f b locking a cofactor s ite would be if the amphibian peptide reacts with the regulatory enzyme Ca 2+ calmodulin, t hus changing the t hree-dimen- sional structure and preventing its attachment to the calmodulin binding site on nNOS. There are examples of small basic peptides, often a helices, being captured and enclosed within Ca 2+ calmodulin, and as a consequence changing the three-dimensional shape of the calmodulin [45±48], but nNOS deactivation by these peptides has, to Table 2. Se quen ce identity of lesueurin and aurein 1.1. Peptide Sequence Lesueurin GLLDILKKVGKVA-NH 2 Aurein 1.1 GLFDIIKKIAESI-NH 2 Ó FEBS 2002 Amphibian peptides that inhibit nNOS (Eur. J. Biochem. 269) 105 our knowledge, not been tested. There are also e xamples where certain small p eptides (one of which mirrors part o f the sequence of the calmodulin binding site of rat nNOS) preferentially interact with Ca 2+ calmodulin, impeding the interaction of Ca 2+ calmodulin with its binding site on nNOS, and, as a c onsequence, preventing the formation of nitric oxide ( IC 50 values at l M concentrations) [49,50]. The possibility that the mechanism of nNOS inhibition by the amphibian peptides does i nvolve complex formation of peptides with Ca 2+ calmodulin is supported by the following experiments. The inhibition of nNOS by selected peptides (citropin 1.1, frenatin 3 and caerin 1 .9) is reduced by the addition of Ca 2+ calmodulin to the assay buffer. The maximum r eduction of inhibition under t he experimental conditions used is 50%. The enzyme Ca 2+ calmodulin regulates not only nNOS but also a number of other enzymes including calcineurin. If the active amphibian pep tides are i ndeed interacting with Ca 2+ calmodulin, they should also inhibit the activity of calcineurin. The four model peptides lesueurin, citropin 1.1, frenatin 3 and caerin 1. 9, all inhibit the activity of calcineu- rin, but at concen trations lower t han t hose obtained f or nNOS (Table 1). Even so, the f act that all four peptides inhibit both n NOS and calcineurin enzymes, provides credence to the proposal that the amphibian peptides are affecting the Ca 2+ calmodulin interaction with nNOS. This is an interesting observation because sequences of the Ca 2+ calmodulin binding sites of nNOS and calcineurin are quite different ( see below [51]), even though they have been classi®ed a s belonging t o the s ame class o f enzymes [52]. The sequences are nNOS (rat, human) and calcineurin A (rat, human), respectively are a s follows: KRRAIGFKKLAEAVKFSAKLM RKEIIRNKIRAIGKMARVFSVLR. Currently, although three-dimensional structures o f c er- tain domains of nitric oxide synthase isoforms are known [53±57], we are not aware of the three-dimensional s tructure of the calmodulin binding sites of any isoform being reported. Thus, we cannot make meaningful comparisons of common structural motifs that might underlie calmodulin binding to the e nzymes. Most nNOS inhibitors are small organic molecules t hat are either analogues of the arginine substrate such as N-nitro- L -arginine [58] or s pecies that pr event calmodulin conjugation such as tamoxifen [59] and m elatonin [60]. Protein and peptidic inhibitors of nNOS are few. Apart from those w e have m entioned above [45,46], caveolin-1, a structural protein component of plasmalemmel caveolae, inhibits endothelial NOS but not neuronal NOS or indu- cible NOS [61]. Residues 82±101 of caveolin-1 in the form of glutathione S-transferase f usion p roteins h ave been shown to attenuate the activities of all three isoforms of NOS [62]. A novel 10-kDa protein inhibitor of nNOS, known as PIN, destabilizes the native d imer form of nNOS preventing it from functioning [63±65]. Proteins such as presynaptic density proteins-93 and -95 and synophin bind to n NOS via a specialized domain called PDZ domain, which s erves as cellular localization signals, but these proteins do not modulate nNOS activity [66]. The basic amphibian peptides that inhibit nNOS show little homology of sequence ( a) between the three active groups of peptides shown in Table 1, (b) with the sequence of amino acid residues of the Ca 2+ /calmodulin binding domain of nNOS, (c) w ith t he sequence of Ca 2+ /calmodulin itself, and (d) with o ther peptides or proteins which inhibit nNOS. Thus we are unable, at this time, to predict from homology, any features of the active amphibian peptides that allow them to inhibit nNOS. There is, however, a signi®cant homology of sequences within each of the active groups shown in Table 1. Consider group 1 (Table 1) as an example. All peptides of group 1 have a post-translational mo di®cation in that the N-terminal group is -CONH 2 . Lesueurin has high homol- ogy w ith some of the citropins and aureins. It i s s everal residues shorter that aureins 2.2, 2.3, 2.4 and citropin 1.1, all of which have IC 50 values more potent than that of lesueurin. Aurein 1.1 has the same number of residues as lesueurin but a l ower overall positive charge, and is not as effective as lesueurin in inhibiting nNOS. The length of these peptides seems t o b e a factor in determining t heir activity. Another important feature conserved in this group of peptides is the cen tral Lys-Lys pair and the b asic nature of the peptides. It appears that the length of the peptide and this pair of basic r esidues are important in determ ining the magnitude of the nNOS inhibition. All three groups of nNOS active amphibian peptides (Table 1) are unique new molecular p robes f or the functioning of nNOS. We intend to carry out further s tructural s tudies on each of the three groups of active peptides listed in Table 1, i n order to probe the r elationship b etween sequence, solution str ucture (determined by NMR) a nd nNOS inhibition. The ®nal question to be addressed is what role do t hese active peptides have in the amphibian integument? The glandular secretion of the animal is exuded onto the skin (and into the gut) when the animal is stressed, sick, or under attack. Logically, there are two possibilities, either the nNOS inhibitor is playing some regulatory role in the animal and/or it is acting as part of the primary host defence arsenal against predators. Consider the ®rst possibility. The nNOS used in this investigation is from rat. Although the sequence of nNOS in anurans has not been determined, i t is likely to be very similar to that of rat nNOS as the sequences of nNOS so far determined (from man, rat, rabbit and snail) are very similar [51]. W e have s hown that L. lesueuri secretes two neuro- peptides onto its skin. T he ®rst of these, caerulein 1.1 is a known host defen ce peptide o f many a nurans [1±4,8]. It has a multifaceted role including smooth muscle activity (at ngákg )1 body weight), which makes it a powerful toxin against predators, it has potent analgaesic properties and it is also thought to be a hormone involved in the hibernation cycle of a nurans [67]. The o ther n europeptide, lesueurin, inhibits the formation of nitric oxid e by nNOS. Nitric oxide has many roles in animals and plays important roles in t he nervous, muscular, cardiovascular and immune systems [68]. In anurans, nitric oxide is already known to be involved in sight [ 69], reproduction [70] and modulation o f gastric acids [71]. It is possible that together with caerulein 1.1, lesueurin has a role in e ither stress control and/or temper- ature control. The other scenario is that nNOS inhibitors are front line defence c ompounds. W e have now identi®ed nNOS i nhib- itors as major skin peptides in 10 out of 12 studied species of frogs of the genus Litoria. These compounds are all active at micromolar-concentrations. A predator ingesting even a small amount of the anuran skin secretion could be seriously 106 J. H. Bowie et al. (Eur. J. Biochem. 269) Ó FEBS 2002 affected if even part of its nitric oxide messenger capability is reduced. The predator could either be large, or small (bacteria have recently been shown to contain NOS [72±75]). In conclusion, most frogs of the genus Litoria secrete a cocktail of bioactive peptides onto their skin, some of which are c ytotoxic and antibiotic, and others w hich regulate the neuronal isoform of the enzyme N OS. We propose t hat this effect on nNOS is mediated by allosteric modulation of arginine catalysis, through an effect on Ca 2+ calmodulin binding to nNOS. We do not believe this is unexpected bioactivity. The peptides that inhibit the formation of n itric oxide from nNOS either p lay a role in the fundamental physiology o f the animal, and/or are part o f t he defence arsenal to combat attack by predators, both small and large. ACKNOWLEDGEMENT Amphibian experiments and peptide s yntheses were funded by t he Australian Research Council. REFERENCES 1. Bevins, C.L. & Zaslo, M. (1990) Peptides from frog skin. Ann. Rev. Biochem. 59, 395±414. 2. Lazarus, L.H. & Attila, M. (1993) T he toad, ugly and venomous, wears yet a precious j ewel in his skin. Prog. Neurobiol. 41, 473± 507. 3. Erspamer, V. (1994) Bioactive secretions of the amphibian integu- ment. In Amphibian Biology. The Integument.(Heatwole,H.,ed.), pp. 178±350. 4. Barra, D. & Simmaco, M. (1995) Amphibian skin: a promising resource for antimicrobial pe ptides. Trends. Biochem. 13, 205±209. 5. Anastasi, A., Erspamer, V . & Endean, R. (1968) The structure of caerulein. Arch. Biochem. Biophys. 125, 57±63. 6. Stone,D.J.M.,Waugh,R.J.,Bowie,J.H.,Wallace,J.C.&Tyler, M.J. (1992) The structures of the caerins and caeridin 1 from Litoria splendida. J. Chem. S oc. Perkin Trans. 1, 3173±3179. 7. Wabnitz, P.A., Bowie, J.H. & Tyler, M.J. (1999) Caerulei n like peptides f rom the skin glands o f the Australian Blue M ountains tree frog Litoria citropa. Part 1. Sequence determination using electrospray mass spectrometry. Rapid Commun. Mass Spectrom. 13, 2498±2502. 8. Bowie, J.H., Wegener, K.L., Chia, B.C.S., Wabnitz, P.A., Carver, J.A., Tyler, M.J. & Wallace, J.C. (1999) Host defence antibacterial peptides from the skin secretions of Australian amphibians. Protein Peptide Lett. 6, 259±269. 9. Wegener, K.L., Wabnitz, P .A., Bowie, J.H., Carver, J.A., W allace, J.C. & Tyler, M.J. (1999) Host defence peptides from the skin glands of the Australian Blue Mountains tree frog Litoria citropa. The antibacterial citropin 1 peptides. The solution structure of citropin 1.1. Eur. J. Biochem. 265, 627±635. 10. Wabnitz, P .A., Bowie, J .H., W allace, J.C. & Tyler, M.J. (1999) The a ntibiotic citropin peptides from the Australian tree frog Litoria citropa. Structure determination using electrospray mass spectrometry. Rapid Commun. Mass Spectrom. 13, 1724± 1732. 11. Rozek, T ., Wegener, K.L., Bowie, J.H., Olver, I.N., Carver, J.A., Wallace, J.C. & Tyler, M.J. (2000) The antibiotic and anticancer active aurein peptides from the Australian bell frogs Litoria aurea and Litoria raniformis. The solution structure of aurein 1.2. Eur. J. Biochem. 267, 5330±5341. 12. Monks, A., Scudiero, D., Skehan, P., Shoemaker, R., Paull, K., Vistica, D., Hose, C., Langley, J., Cronise, P. & Vaigro-Wol, A. (1991) Feasibility of a high-¯ux anticancer drug screen using a diverse panel of cultured human tumor cell lines. J. Natl Cancer Inst 83, 757±766. See also http://dtp.nci.nih.gov 13. Wong,H.,Bowie,J.H.&Carver,J.A.(1997)Thesolutionstruc- ture and activity o f caerin 1 .1, an a nt ibiotic peptide from the Australian tree frog Litoria splendida. Eur. J. Biochem. 247, 545± 557. 14. S hai, Y.C. (1995) Molecular recognition between m embrane- scanning peptides. Trends Biochem. Sci. 20, 460±464. 15. B echinger, B . (1997) Structure and function of channel form ing peptides; magainins, cecropins, melittin and alamethicin. J. Membr. Biol. 156, 197±211. 16. Matsuzaki, K. (1998) Magainins as paradigm for the mode of pore f orming polypeptides. Biochim. Biophys. Acta 1376, 391±400. 17. Steinborner, S.T., Gao, C., Raftery, M.J., Waugh, R.J., Blumenthal, T ., Bowie, J.H., W allace, J.C. & Ty ler, M.J. (1994) The structures of four tryptophyllin and three rubellidin peptides from the Australian red tree frog Litoria r ubella. Aust. J. C hem. 72 , 2099±2108. 18. Steinborner, S.T., Wabnitz, P.A., Bowie, J.H. & Tyler, M.J. (1996) The application of mass spectrometry to the study of evo- lutionary tre nds in amphibians. Rapid C ommun. Mass Spectrom. 10, 92±97. 19. Steinborner, S.T., Wabnitz, P.A., Waugh, R.J., Bowie, J.H., Gao, C., Tyler, M.J. & Wallace, J.C. (1996) The structures of new peptides from Litoria rubella. Aust. J. Chem. 49, 955±963. 20. Wabnitz, P.A., Bowie , J.H., Tyler, M .J. & Wallace, J.C. (1999) Peptides from the skin glands of the A ustralian buzzing tree frog Litoria electrica. Comparison with the skin peptides of Litoria rubella. Aust. J. Chem. 52, 639±649. 21. Erspamer, V., Melchiorri, P., Erspamer, G.F., Montecucchi, P.C. & de Castiglione, R. (1985) Phyllomedusa skin: a large factory and storehouse of a variety of active peptides. Peptides 6, 7±12. 22. Renda, T.D., -Este, L., Bua, R., Usellini, L., Capella, C., Vaccaro, R. & Melchiorri, P. ( 1985) Tryptophyllin-like immu- noreactivity in rat adenohypophysis. Peptides 6, 1 97±202. 23. Zadina, J.E., Hackler, L., Ge, L.J. & Kastin, A.J. (1997) A potent and selective endogenous agonist for the mu-opiate receptor. Nature 38 6 , 499±500. 24. Raftery, M.J., Bradford, A.M., Bowie, J.H., Wallace, J .C. & Tyler, M.J. (1993) The structures of the dynastins from the Banjo frogs Limno dynastes interioris. L . dumerilii an d L. terraereginae. Aust.J.Chem.46, 833±842. 25.Bradford,A.M.,Raftery,M.J.,Bowie,J.H.,Wallace,J.C.& Tyler, M.J. (1993) The Structures of the dynastins and ¯etch erin from Limnodynastes salmini and Limnodynastes ¯etcheri. Aust. J. Chem. 46, 1235±1244. 26. Tyler, M.J., Stone, D.J.M. & Bowie, J.H. (1992) A novel method for the release and collection of dermal glandular secretions from the skin of frogs. J . Pharmacol. Toxicol. Meth 28, 199±200. 27. Wabnitz, P.A., Bowie, J.H. & Tyler, M .J. (1999) Caerulein type peptides f rom the skin glands of the Australian Blue Mountains tree frog Litoria citropa. Part 1. Sequence determination using electrospray mass spectrometry. Rapid. Commun. Mass Spectrom. 13, 2498±2502. 28. B iemann, K. & Martin. S. ( 1987) Mass spectrometric determin- ation of the sequences of peptides and proteins. Mass Spectrom. Rev. 6, 1±76. 29. Hunkapiller, M.W., Hewick, R.M., Drewer, W.J. & Hood, L.E. (1983) High sensitivity sequencing with a gas-phase sequencer. Methods Enzymol. 91, 399±406. 30. Maeji, N .J., Bray, A .M., V alerio, R .M. & Wang, W. (1995) Lar ger scale multi-pinpeptide synthesis. Pe pt. Res. 8, 33±38. 31. Jorgensen, J.H., Cleeland, W.A., Craig, G., Doern, M., Ferraro, J., Finegold, C.M., Hansen, S.L., Jenkins, S.G., Novick, W.J., Pfaller, M.A., Preston, D.A., Reller, L.B. & Swenson, J.M. (1993) Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically, 3rd approved standard. Natl Committee Clin. Lab. Stand . 33, 1±12. Document M7±A3. Ó FEBS 2002 Amphibian peptides that inhibit nNOS (Eur. J. Biochem. 269) 107 32. Doyle, D.D., Gua, Y., Lustig, S.L., Satin, J., Rogart, R.B. & Fozzard, H.A. (1993) Divalent cation competition with [ 3 H]saxi- toxin b indin g to tetrodotoxin-resistant and ± sensitive sodium channels. A two-site structural model of i o n/toxin interaction. J. General Physiol. 101 , 153±182. 33. Fersht. A. Enzyme Stru cture and Mec h anism,W.H.Freeman, New York, NY, USA. 34. Anonymous. (2000) Protein phosphatase-2B. Technical B ulletin 534.Promega,Madison.WI,USA. 35. Dume  ril. A.M.J. & Bibron, G. (1841) Erpe  tologie Ge  ne  rale Ou Histoire Nature lle Comple Á te des Reptiles, Paris. 8, 792. 36. Barker, J., Grigg, G.C. & Tyler, M.J. (1995) In A F ie ld Guide t o Australian Frogs. p. 68. Surrey Beatty and Sons. Chipping Norton, New South Wales, Australia. 37. Bredt, D.S., Hwang, P.M., Glatt, C.L., Lowenstein, C., Reed, R.R. & Snyder, S.H. (1991) Cloned and expressed nitric oxide synthase structurally resembles cytochrome P-450 reductase. Nature 351, 714±718. 38. Marletta, M.A. ( 1993) Nitric oxide s ynthase and m echanism. J. Biol. Chem. 268, 12231±12234. 39. Marletta, M.A. (1994) Nitric oxide synthase: aspects c oncerning structure and catalysis. Cell 78, 927±930. 40.Pfeier,S.,Mayer,B.&Hemmens,B.(1999)Nitricoxide:a protective or pathogenic molecu le. Angew. Ch em. I ntl E d. Engl. 38 , 1715±1731. 41. Forsen, S. (1995) Calcium induced structural changes and domain autonomy in calmodulin. Nat. Struct. Biol. 2, 777±783 and refer- ences cited th erein. 42. Venema, R.C., Sayegh, H.S., Kent, J.D. & Harrison, D.G. (1996) Identi®cation, characterisat ion and comparison of the calmodulin- binding domains of nitric oxide s ynthases. J.Biol. C hem. 271, 6435±6440. 43. Lee, S.J. & Stull, J.T. (1998) Calmodulin-dependent regulation of inducible a nd ne uronal nitric oxide synthase. J. Biol. Chem. 273, 27430±27437. 44. Matsuda, H. & Iyanagi, T. (1999) Calmodulin activated intra- molecular transfer between the two ¯avins of neuronal nitric oxide synthase. Biochim. Biophys. Acta 1473, 345±355. 45. Malencik, D.A. & Anderson, S .R. (1982) Binding of simple pep- tides, hormones and neurotransmitters by calmodulin. Biochem- istry 21, 3480±3486. 46. Ikura, M., Kay, L.E., Krinks, M. & Bax, A. ( 1991) Triple reso- nance multidimensional NMR study of calmod ulin complexed with the binding domain of skeletal muscle myosin light-chain kinase: indication of a comformational change i n the central helix. Biochemistry 30, 5498±5504. 47. Ikura, M., Clore, G.M., Gronenborn, A.M., Zhu, G., Klere, C.B. & Bax, A. (1992) Solution structure o f a c almodulin-target peptide complex by multidimensiona l NMR. Science 256, 632±638. 48. Scaloni, A., Miraglia, N., Orru, S., Amodeo, P., Motta, A., Marino, G. & Pucci, P . ( 1998) T opolo gy of t he calmo du lin- melittin complex. J. Mol. Biol. 277, 945±958. 49. Matsubara, M., Hayashi, N., Titani, K. & Taniguchi, H. (1997) Circular dic hroism and 1 H NMR studies on the structures of peptides derived f rom t he calmodulin-binding domains of indu- cible and endothelial NOS in solution and in complex with calmodulin. Nascent alpha-helical structures are stabilized by calmodulin both in the p resence and absence o f C a 2+ . J. Biol . Chem. 272, 23050±23056. 50. Watanabe,Y.,Hu,Y.&Hidaka,H.(1997)Identi®cationofa speci®c a mino acid c luster in t he calmodulin binding d omain of nNOS. FE BS Lett. 403,75. 51. For the sequences of various NOS enzymes together with the positions of the binding domains of the cofactors C a 2+ calmodu- lin. S WISS-PRO http://au.expasy.org/cgi-bin/sprot-search-de 52. Rhoads, A .R. & F riebe rg, F. ( 1997) Sequence motifs f or cal- modulin recognition. FASEB J. 11, 331±342. 53. Crane, B.R., Arvai, A.S., Gachhui, R., Wu, C.O., Ghosh, D .K., Getzo, E.D., Stuehr, D.J. & Tainer, J.A. The structure of nitric oxide synthase oxygenase domain and inhibitor complexes. Science 278, 425±431. 54. Raman, C.S., Li, H.Y., Martasek, P., Krai, V., Masters, B.S.S. & Poulos, T.L. ( 1998) Crys tal s tructure of constitutive endothelial oxide synthase; A paradigm for pterin function involving a novel metal center. Cell 95, 939±950. 55. Crane, B.R., Arvai, A .S., Ghosh, D.K., Wu, C.O., Getzo, E.D., Stuehr, D.J. & Tainer, J.A. ( 1998) Structure of nitric oxide syn- thase oxygenase dim er with p terin and substrate. Science 279, 2121±2126. 56. Fischmann, T.O., Hruza, A., Fosseta, J.D., Lunn. C.A., Dolphin. E., Prongay, A.J., Reichert, P., Lundell, D.J., Naruda, S.K. & Weber, P.C. (1999) Structural characterisation of nitric oxide synthase isoforms reveals striking active-site conservation. Nat. Struct. Biol. 6, 233±242. 57. Tochio, H., Zhang, O., M andal, O., Li, M. & Zhang, M.J. (1999) Solution structure of the e xtended neuronal nitric oxide synthase PDZ domain complexed with an associated peptide. Nat. Struct. Biol. 6, 417±421. 58. Moore, P. & Handy, R. (1997) Selective i nhibitors o f n euronal nitric oxide synthase ± is no NOS really good NOS f or the n ervous system? Trends Pharmacol. Sci. 272, 15583±15586. 59. Renodon, A., Bou cher, J .L., Sari, M.A., Delaforge, M., O uazzani, J. & Mansuy, D. (1997) Stro ng inhibition of nNOS by the cal- modulin antagonist a nd anti-estrogen drug tamoxifen. Bioc hem. Pharmacol. 54 , 1109±1114. 60. Leon, J., Macias, M., Escames, G., Camacho, E., Khaldy, H., Martin, M ., Espinosa, A., Gallo, M.A. & Acuna-Castroviejo, D. (2000) Structure related inhibition of calmodulin-dependent nNOS activity by melatonin and synthetic kynurenines. Mol. Pharmacol. 58 , 967±975. 61. Michel, J.B., Feron, O.D.S. & Michel, T. (1997) Reciprocal reg- ulation of endothelial NOS by C a 2+ calmodulin and caveolin. J. Biol. Chem. 272, 15583±15586. 62. Garcõ  a-Carden Ä e, G., Martasek, P., Sue Siler Master, B., Skidd, P., Couet, J., Li, S., Lisanti, M. & Sessa, W. (1997) Dissecting the interaction between NOS and caveolin. J. Biol. Chem. 272, 25437± 25440. 63. Jarey, S. & Snyder, S. (1996) PIN: an associated protein i nhibitor of nNOS. Science 274, 774±777. 64. Tochio, H., Ohki, S., Zhang, O., Li, M. & Zhang, M.J. ( 1998) Solution structure of a protein inhibitor or neuronal n itric oxide synthase. Nat. Structure Biol. 5, 965±969. 65. Fan, J.S., Zhang, O.A., Li, M., Tochio, H., Yamazaki, T., Shimizu, M. & Zhang, M.J. (1998) Protein inhibitor of neuronal nitric oxide synthase, PIN, binds to a 17-amino acid residue fragment of the enzyme. J. Biol. Chem. 273, 33472± 33481. 66. Brenman, E., Ch ao, S., Gee, S.H., McGee, A.W., Craven, S.E., Santillano, D.R., Wu, Z., Huang, F., Xia, H., P eters, M.F., Froehner, S.C. & B redt, D.S. (1996) Interaction of NOS with the postsynaptic density protein PSD-95 and alpha 1-syntroph in mediated by PDZ domains. Cell 84, 757±767. 67. Wabnitz, P.A., Bowie, J.H., Tyler, M.J., Wallace, J.C. & Smith, B.P. (2000) Dierences in the skin peptides of the male and female Australian m agni®cent tree frog Litoria splendida ± a three year monitoring survey: the discovery o f t he male sex pheromone splendipherin together with phe 8 caerulein a nd the antibiotic caerin 1.10. Eur. J. Biochem. 267, 269±275. 68. Ghosh,S.,Wolan,D.,Adak,S.,Crane,B.R.,Kwon,N.S.,Tainer, J.A., Getzo, E.D. & Stuehr, D.J. (1999) Mutational analysis of the tetrahydrobiopterin-binding site in inducible nitric oxide syn- thase. J. Bio l. Chem. 274, 24100±24112. 69. Renteria, R .C. & Constantine-Paton, M. (1999) N itric o xide i n the retinotectal s ystem: a signal but no t a retrograde messenger 108 J. H. Bowie et al. (Eur. J. Biochem. 269) Ó FEBS 2002 during map re®nement and segregation. J. Neuroscience 99, 7066±7076. 70. Gobbetti, A. & Zerani, M. (1999) Hormonal a nd cellular brain mechanism regulating amplexus of the m ale and f emale water frog Rana esculenta. J. Neuroendocrinol. 11 , 589±596. 71. Molero, M., Hernandez, I .M., Lopo, P., Cardenas, P., Romero, R. & Chacin , J. (1998) Modu lation by nitric oxide of gastric acid secretionintoads.Acta Physiol. Scand. 164, 229±236. 72. Choi, W.S., Chang, M .S., Han, J.W., Hong, S.Y. & Lee, H.W. (1997) Identi®cation o f NOS in Staphylococcus aureus. Biochem. Biophys. Res. Commun. 237, 554±558. 73. Morita,H.,Yoshikawa,H.,Sakata,R.,Nagata,Y.&Tanaka,H. (1997) Synthesis of nitric oxide from the two equivalent guanidino nitrogens of L -arginine b y Lactobacillus f ermentum. J. Bacteriol. 179, 7812±7815. 74. Choi, W.S., Seo, D .W., Chang, M.S., Han, J.W., Paik, W.K. & Lee, H.W. (19 98) Methylesters of L -arginine and N-nitro- L -argi- nine induce n itric o xide s ynthase i nStaphyloccoccus aureus. Bio - chem. Biophys. Res. Commun. 246, 431±435. 75. Er,H.,Turkoz,Y.,Ozerol,I.H.&Uzmez,E.(1998)EectofNOS inhibition in experimental Pseudo monas keratitis.rabbits.Eur. J. Opthalmol 8, 137±141. Ó FEBS 2002 Amphibian peptides that inhibit nNOS (Eur. J. Biochem. 269) 109 . Amphibian peptides that inhibit neuronal nitric oxide synthase The isolation of lesueurin from the skin secretion of the Australian Stony Creek Frog Litoria lesueuri Jason Doyle 1 , Lyndon. by nNOS? The three nitric oxide synthases, namely neuronal, endothelial and inducible, are highly regulated enzymes responsible for the synthesis of the signal molecule nitric oxide. They are amongst. produce antimicrobial peptides, unlike the majority of the other members of the genus Litoria. Instead, it produces a neuropeptide t hat i nhibits the formation of nitric oxide by neuronal n itric oxide synthase