Conserved Structural and Sequence Elements Implicated in the Processing of Gene-encoded Circular Proteins
2004; Elsevier BV; Volume: 279; Issue: 45 Linguagem: Inglês
10.1074/jbc.m407421200
ISSN1083-351X
AutoresJ.L. Dutton, Rosemary F. Renda, Clement Waine, Richard J. Clark, Norelle L. Daly, Cameron V. Jennings, Marilyn A. Anderson, David J. Craik,
Tópico(s)Glycosylation and Glycoproteins Research
ResumoThe cyclotides are the largest family of naturally occurring circular proteins. The mechanism by which the termini of these gene-encoded proteins are linked seamlessly with a peptide bond to form a circular backbone is unknown. Here we report cyclotide-encoding cDNA sequences from the plant Viola odorata and compare them with those from an evolutionarily distinct species, Oldenlandia affinis. Individual members of this multigene family encode one to three mature cyclotide domains. These domains are preceded by N-terminal repeat regions (NTRs) that are conserved within a plant species but not between species. We have structurally characterized peptides corresponding to these NTRs and show that, despite them having no sequence homology, they form a structurally conserved α-helical motif. This structural conservation suggests a vital role for the NTR in the in vivo folding, processing, or detoxification of cyclotide domains from the precursor protein. The cyclotides are the largest family of naturally occurring circular proteins. The mechanism by which the termini of these gene-encoded proteins are linked seamlessly with a peptide bond to form a circular backbone is unknown. Here we report cyclotide-encoding cDNA sequences from the plant Viola odorata and compare them with those from an evolutionarily distinct species, Oldenlandia affinis. Individual members of this multigene family encode one to three mature cyclotide domains. These domains are preceded by N-terminal repeat regions (NTRs) that are conserved within a plant species but not between species. We have structurally characterized peptides corresponding to these NTRs and show that, despite them having no sequence homology, they form a structurally conserved α-helical motif. This structural conservation suggests a vital role for the NTR in the in vivo folding, processing, or detoxification of cyclotide domains from the precursor protein. The discovery of naturally occurring circular proteins has in recent years introduced a new topological paradigm into the field of protein structure. The cyclotides, backbone cyclic proteins of 28–35 amino acid residues isolated from plants, are the largest family of naturally occurring circular proteins (1Craik D.J. Daly N.L. Bond T. Waine C. J. Mol. Biol. 1999; 294: 1327-1336Crossref PubMed Scopus (653) Google Scholar, 2Trabi M. Craik D.J. Trends Biochem. Sci. 2002; 27: 132-138Abstract Full Text Full Text PDF PubMed Scopus (244) Google Scholar) (see Table I). Interest in these molecules has been driven by their topological novelty and associated resistance to thermal and enzymatic degradation. They were initially discovered as active components in a tea used in parts of Africa to accelerate childbirth. The uterotonic activity of the prototypic cyclotide kalata B1 was maintained after boiling of the plant Oldenlandia affinis to make the medicinal tea (3Gran L. Acta Pharmacol. Toxicol. 1973; 33: 400-408Crossref PubMed Scopus (201) Google Scholar). The remarkable stability of these proteins has been attributed to their novel structure, the cyclic cystine knot motif (4Saether O. Craik D.J. Campbell I.D. Sletten K. Juul J. Norman D.G. Biochemistry. 1995; 34: 4147-4158Crossref PubMed Scopus (376) Google Scholar, 5Craik D.J. Anderson M.A. Barry D.G. Clark R.J. Daly N.L. Jennings C.V. Mulvenna J. Lett. Peptide Sci. 2002; 8: 119-128Google Scholar) (Fig. 1), in which two disulfide bonds and the surrounding peptide sequence form an embedded ring through which a third disulfide bond is threaded. Although the cystine knot motif is found in a wide range of other peptides from sources as diverse as fungi, insects, plants, and animals (6Craik D.J. Toxicon. 2001; 39: 1809-1813Crossref PubMed Scopus (101) Google Scholar), it is only in the cyclotides that this motif is coupled with a cyclic backbone to form the cyclic cystine knot.Table IThe cyclotide family1 Both varv A and varv E were also isolated from other plant species and their sequences published under the names of kalata S and cycloviolacin O12. Open table in a new tab Fig. 1The cyclotide framework. The structure is of the prototypic cyclotide kalata B1 (PDB code 1NB1) and illustrates the region of β-sheet (broad arrows) and disulfide bonds (in ball-and-stick mode). The Cys residues are labeled with Roman numerals, and the backbone loops between them are numbered loops 1–6. The curved arrow indicates the point in the sequence where a cis-Pro residue defines the Moebius sub-class of cyclotides. The sequence of kalata B1 is given below the structure.View Large Image Figure ViewerDownload (PPT) 1 Both varv A and varv E were also isolated from other plant species and their sequences published under the names of kalata S and cycloviolacin O12. The number of cyclotide sequences has grown to over 50 in recent years (7Craik D.J. Daly N.L. Mulvenna J. Plan M.R. Trabi M. Curr. Protein Pept. Sci. 2004; 5: 297-315Crossref PubMed Scopus (162) Google Scholar). It is thought that their natural function is in plant defense (8Jennings C. West J. Waine C. Craik D. Anderson M. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 10614-10619Crossref PubMed Scopus (412) Google Scholar), but there is still relatively little known about the process by which these peptides are synthesized in plants. Our previous isolation of cDNA clones encoding cyclotides from O. affinis, from the Rubiaceae family, has shown that in contrast to well known microbially derived cyclic peptides, which typically contain fewer than 12 residues and are non-ribosomally synthesized by enzyme complexes, the cyclotides are directly encoded by genes (8Jennings C. West J. Waine C. Craik D. Anderson M. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 10614-10619Crossref PubMed Scopus (412) Google Scholar). Four clones, named Oak1 through Oak4 (O. affinis kalata-encoding clones 1–4), were isolated in that study, which until now was the only study to report nucleic acid sequences of cyclotides. The precursor proteins predicted from the Oak cDNA clones feature a signal peptide sequence and one to three cyclotide repeats corresponding to mature cyclotide domains. The predicted precursors also exhibit a highly conserved sequence near the N terminus of each of the cyclotide sequences that has been termed the N-terminal repeat (NTR). 1The abbreviations used are: NTR, N-terminal repeat; OaNTR, Oldenlandia affinis NTR; VoNTR, Viola odorata NTR; TFE, trifluoroethanol; NOE, nuclear Overhauser Effect; NOESY, nuclear Overhauser effect spectroscopy; r.m.s.d., root mean square deviation; DQF-COSY, double quantum filtered correlation spectroscopy; ECOSY, exclusive correlation spectroscopy. Until now all that was known about the genes that encode cyclotides came from a single Rubiaceae family species, and all of these clones encoded cyclotides from the so-called Moebius sub-class. The cyclotides have been classified into two sub-classes depending on whether they have a proline residue in the fifth loop as shown in Fig. 1. Moebius cyclotides have a cis proline that induces a local backbone twist, whereas bracelet cyclotides do not. In the current study we have isolated cyclotide-encoding cDNA clones from the other major plant family known to express cyclotides, the Violaceae. In particular, we have isolated four cDNA clones from Viola odorata and have found that the precursor proteins predicted from these clones feature the same general arrangement, having an endoplasmic reticulum signal peptide, an N-terminal prodomain, and one to three cyclotide domains each preceded by a conserved sequence of 25 residues at their N termini. As well as reporting the first cDNA sequences encoding cyclotides from the Violaceae family, and isolation of the first cDNA clones encoding bracelet cyclotides, this report examines the structure and function of conserved elements of sequence in the processing of mature cyclotides from their precursor proteins. Many peptides and proteins are produced through processing of larger precursor molecules. Sequences within the precursor that are not part of the mature protein may modulate protein folding and/or structure (9Braun P. Tommassen J. Trends Microbiol. 1998; 6: 6-8Abstract Full Text PDF PubMed Scopus (18) Google Scholar, 10Shinde U. Inouye M. Semin. Cell Dev. Biol. 2000; 11: 35-44Crossref PubMed Scopus (135) Google Scholar). Indeed, a diverse range of actions has been found for such propeptides, including modulation of protein folding or protein function, detoxification, and targeting. A recent study of a family of conotoxins, which are disulfide-rich proteins similar in size to the cyclotides, suggests that propeptides may play a role in the protein disulfide isomerase-catalyzed folding of precursor proteins (11Buczek O. Olivera B.M. Bulaj G. Biochemistry. 2004; 43: 1093-1101Crossref PubMed Scopus (51) Google Scholar). In this report, the sequences of the precursor proteins predicted from cDNA clones have been determined and are used to identify residues that may be important for cyclotide processing. The solution structures of synthetic peptides based on the NTRs from O. affinis and V. odorata are presented, and the significance of the NTR structure to cyclotide processing is discussed. RNA Isolation and Production of Partial Clones—Total RNA was isolated from V. odorata leaves using TRIzol® reagent (Invitrogen). The partial clones were obtained by reverse transcriptase-PCR using total RNA and the Gene Amp® Gold RNA PCR kit from PE Biosystems. The Vok1 cDNA was amplified with Kal2 and oligo(dT) primers as described by Jennings et al. (8Jennings C. West J. Waine C. Craik D. Anderson M. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 10614-10619Crossref PubMed Scopus (412) Google Scholar) (Kal2: GGGGATCCGTITGYGGIGARACITG (BamHI-VCGET)) for amplification of Moebius cyclotide sequences from O. affinis RNA. The Voc1, Voc 2, and Voc3 partial cDNAs were amplified using oligo(dT) with a degenerate forward primer, cyc1, which is complementary to a conserved sequence in the bracelet family of cyclotides (cyc1: TGTGTTTGGATACCTTGC (CVWIPC)). The PCR conditions comprised 30 cycles with the following regime: 94 °C, 1 min; 47 °C, 1 min; and 72 °C, 1 min. The PCR fragments were gel-purified and cloned into the pCR2.1 vector (Invitrogen) for sequencing. Preparation of the V. odorata cDNA Library and Isolation of Clones— Poly(A)+ RNA was isolated from total RNA (1 mg) using the PolyATract® mRNA isolation system (Promega). The mRNA (5 μg) was used to construct a cDNA library with the Lambda ZAP-cDNA synthesis kit and packaging extracts from Stratagene. Full-length cDNAs were obtained by screening the library with the [α-32P]dCTP-labeled PCR fragments. Secondary structure of the predicted precursor proteins was predicted in MacVector (12Olson S.A. Methods Mol. Biol. 1994; 25: 195-201PubMed Google Scholar) using Chou-Fasman and Robson-Garnier algorithms. Peptide Synthesis—Boc-l-amino acids were obtained from Novabiochem (Laufelfingen, Switzerland) or the Peptide Institute (Osaka, Japan); t-Boc-Pro-OCH2-PAM resin was obtained from PerkinElmer Life Sciences (Brisbane, Australia). 2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate was obtained from Richelieu Biotechnologies (Quebec, Canada). The following reagents were of peptide synthesis grade and obtained from Auspep (Melbourne, Australia): trifluoroacetic acid, N,N-diisopropylethylene, and N,N-dimethylformamide. Acetonitrile (high-performance liquid chromatography grade) was purchased from BDH (Poole, England). A peptide based on the NTR from the O. affinis clones and another based on the NTR of the V. odorata clones were assembled manually by stepwise solid-phase peptide synthesis using the in situ neutralization protocol of Boc chemistry (13Schnolzer M. Alewood P. Jones A. Alewood D. Kent S.B. Int. J. Pept. Protein Res. 1992; 40: 180-193Crossref PubMed Scopus (941) Google Scholar) starting from t-Boc-Pro-OCH2-PAM resin on a 0.25-mmol scale. The following side-chain protected amino acids were used: Leu-OH·H2O, Gly-OH, Lys(Clz)-OH, Gln(Xan-OH, Met-OH, Glu-(OcHex)-OH, Phe-OH, Thr(Bzl)-OH, Ser(Bzl)-OH, and Val-OH. The crude peptides were purified by preparative reverse-phase high-performance liquid chromatography (Vydac C18) on a Waters high-performance liquid chromatography system using a linear gradient of 0–80% acetonitrile in water and 0.1% trifluoroacetic acid over 80 min. Mass spectrometry data were obtained using an atmospheric pressure ionization electrospray mass spectrometer (PerkinElmer Life Sciences, PE-Sciex, Canada). NMR Spectroscopy—Samples for various 1H NMR measurements contained 0.2–3.5 mm peptide in 90% 1H2O/10% 2H2O at pH 3.6–5.3. A sample of the VoNTR peptide was also prepared in 20% deuterated trifluoroethanol (TFE)/70% 1H2O/10% 2H2O. For structure determination samples at 3.5 mm and pH 3.6 were used. Spectra were acquired on Bruker ARX 500- and 750-MHz spectrometers, mainly at 298 K, but for clarification of some assignments additional spectra were acquired at 280 K and 310 K. NOESY spectra were acquired with mixing times of 250 and 300 ms for OaNTR and 200, 150, and 100 ms for VoNTR. Solvent suppression in DQF-COSY and ECOSY experiments was achieved using selective low power irradiation of the water resonance during a relaxation delay of 1.8 s. In NOESY and TOCSY experiments, a modified WATERGATE sequence was used to achieve solvent suppression. Chemical shifts were internally referenced to sodium 2,2-dimethyl-2-silapentane-5-sulponate. Slow exchange experiments were performed following dissolution of the protonated peptides in 2H2O or in 20% deuterated TFE/80% 2H2O. TOCSY and one-dimensional experiments were conducted to monitor the exchange of the amide protons with the solvent. Amide proton signals that still appeared 60 min after dissolution in 2H2O were regarded as slow exchanging. Spectra were processed using XWINNMR software (Bruker) on an SGI Octane2 workstation. Distance restraints were derived from the intensities of cross-peaks in NOESY spectra recorded with mixing times of 100–250 ms. Inter-proton distance restraints and pseudo-atom corrections were applied where necessary, and empirical corrections were added to intensities associated with methyl protons (14Clore G.M. Brunger A.T. Karplus M. Gronenborn A.M. J. Mol. Biol. 1986; 191: 523-551Crossref PubMed Scopus (256) Google Scholar). Dihedral angle assignments were made from a combination of coupling constants and NOE (nuclear Overhauser effect) intensities: ϕ angle restraints were assigned based on 3JHN-Hα couplings measured from one-dimensional and DQF-COSY spectra, whereas χ1 angles were assigned based on 3JHα-Hβ couplings measured from ECOSY spectra and observed NOE intensities. Preliminary three-dimensional structures of the peptides were determined using X-PLOR version 3.851 (15Brünger A.T. X-PLOR Version 3.1 A System for X-ray Crystallography and NMR. Yale University, New Haven1992Google Scholar). Starting from a template structure with randomized ϕ and ψ angles and extended side chains, the program generated an ensemble of 50 structures using an ab initio simulated annealing protocol. These structures were subjected to high temperature dynamics (1000 K), with low initial weighting on the force constant and NOE restraints, prior to cooling to 100 K. Each structure was energy-minimized for 1000 cycles using the Powell algorithm and a refined CHARMM force-field. Final structures were calculated in CNS (16Brunger A.T. Adams P.D. Rice L.M. Structure. 1997; 5: 325-336Abstract Full Text Full Text PDF PubMed Scopus (198) Google Scholar) with energy minimization in a water shell (17Rosengren K.J. Clark R.J. Daly N.L. Goransson U. Jones A. Craik D.J. J. Am. Chem. Soc. 2003; 125: 12464-12474Crossref PubMed Scopus (218) Google Scholar). The structures were displayed using Insight II (Biosym Technologies, San Diego, CA) or MolMol (18Koradi R. Billeter M. Wuthrich K. J. Mol. Graph. 1996; 14 (29–32): 51-55Crossref PubMed Scopus (6498) Google Scholar) and analyzed using PROMOTIF (19Hutchinson E.G. Thornton J.M. Protein Sci. 1996; 5: 212-220Crossref PubMed Scopus (999) Google Scholar) and PROCHECK (20Laskowski R.A. Rullmannn J.A. MacArthur M.W. Kaptein R. Thornton J.M. J. Biomol. NMR. 1996; 8: 477-486Crossref PubMed Scopus (4474) Google Scholar). Previously we reported the isolation of four cDNA clones encoding cyclotide precursors from O. affinis, from the Rubiaceae, and prior to the current report these were the only known cyclotide-encoding clones. The O. affinis clones only encode cyclotides from the Moebius sub-class. In the current study it was of interest to isolate cDNAs from a plant species that is evolutionarily distant from O. affinis and also to isolate cDNAs encoding cyclotides from the other sub-class, namely bracelet cyclotides. V. odorata was screened, because it is both evolutionarily distant from O. affinis and a rich source of bracelet cyclotides (1Craik D.J. Daly N.L. Bond T. Waine C. J. Mol. Biol. 1999; 294: 1327-1336Crossref PubMed Scopus (653) Google Scholar, 21Svangard E. Goransson U. Smith D. Verma C. Backlund A. Bohlin L. Claeson P. Phytochemistry. 2003; 64: 135-142Crossref PubMed Scopus (57) Google Scholar). Isolation of V. odorata cDNA Clones and Comparison with the Oak Clones—Forward primers based on cyclotide sequences were used with oligo(dT) in reverse transcription-PCR to amplify cDNA-encoding cyclotides from V. odorata total leaf RNA. The partial cDNA clones were sequenced and used individually to screen a leaf cDNA library for full-length clones. Using the same forward primer that was used to isolate the Oak clones from O. affinis, which is based on loop 1 of the Moebius cyclotide kalata B1 (Fig. 1 and Table I; VCGET), we cloned a Moebius sub-class member from V. odorata. Conservation of the general structure of the genes, discussed in more detail below and shown in Fig. 2, indicates that they have evolved from an ancestral gene present in flowering plants before diversification of the asterid and rosid lineages in the eudicots. To isolate clones encoding members of the bracelet class of cyclotides, we designed a new primer complementary to a sequence that is conserved in loop 2 (Fig. 1), CVWIPC (Table I), of the bracelet cyclotides. This approach led to the identification of three cDNAs encoding bracelet cyclotides. In total, four full-length cDNA clones were isolated. The predicted protein sequences are shown in the upper panel of Fig. 2. For comparison, the earlier clones from O. affinis are shown in the lower panel of Fig. 2. Nomenclature for the V. odorata clones follows that used for the O. affinis clones, where kalata-encoding cDNA clones were named OaKx, with x related to the order of discovery. The Moebius clone encodes a precursor with a single copy of kalata B1 and two copies of kalata S and was therefore named Vok1 (V. odorata kalata encoding clone 1). The Vok1 precursor is remarkably similar in structure to the Oak4 precursor, shown in Fig. 2, with three cyclotide repeats separated by conserved NTRs, an N-terminal signal sequence, and a small C-terminal tail. The bracelet cyclotide clones, Voc1, Voc2, and Voc3 (V. odorata cycloviolacin encoding clones), encode cycloviolacin O8, cycloviolacin O11, and cycloviolacin O13, a cycloviolacin/circulin-like peptide that has not been reported previously (numbering is based on the reported peptide sequences in Table I). These clones all resemble Oak1, encoding precursors with just one cyclotide repeat, with an endoplasmic reticulum signal sequence, an N-terminal pro-domain, NTR, and small C-terminal tail. The amino acid identities of the cyclotide and NTR domains of the O. affinis and V. odorata clones relative to Oak1 and Voc1 are summarized in Table II. Cyclotide domains in the Moebius group share greater than 79% identity at the amino acid level irrespective of whether they are sequences from the Rubiaceae or Violaceae. When the Moebius and bracelet mature cyclotide domains are compared they share only 40% identity. Interestingly, the NTR sequences are not as tightly conserved, and the sequences cluster according to the plant of origin rather than the Moebius or bracelet groupings. Conservation of the N-terminal signal sequences and NTRs between Vok1 and the Voc clones is not as great as in the Oak clones, but then this is to be expected, because the Oak clones all encode kalata variants, whereas the V. odorata clones encode cyclotides from different sub-classes, i.e. kalata and cycloviolacin variants.Table IIPercentage amino acid identity of cyclotide and NTR domains with the cyclotide and NTR domains of the Oak1 and Voc1 precursorsCyclotideNTR (25Kliemannel M. Rattenholl A. Golbik R. Balbach J. Lilie H. Rudolph R. Schwarz E. FEBS Lett. 2004; 566: 207-212Crossref PubMed Scopus (39) Google Scholar)NTR (17Rosengren K.J. Clark R.J. Daly N.L. Goransson U. Jones A. Craik D.J. J. Am. Chem. Soc. 2003; 125: 12464-12474Crossref PubMed Scopus (218) Google Scholar)Oak1Voc1Oak1Voc1Oak1Voc1Moebius subclassOak1 (kalata B1)1004410041005Oak2 (kalata B6)863645185817Oak2 (kalata B3)823050135817Oak3 (kalata B7)793745135817Oak4 (kalata B2 repeat 1)823757146417Oak4 (kalata B2 repeat 2)823759136417Oak4 (kalata B2 repeat 3)823759136417Vok1 (kalata S repeat 1)9648454541Vok1 (kalata B1)100449312329Vok1 (kalata S repeat 2)96484272929Bracelet subclassVoc1 (cycloviolacin O8)4410041005100Voc2 (cycloviolacin O11)449641005100Voc3 (cycloviolacin O13)448014611741 Open table in a new tab Synthesis and Three-dimensional Structure Characterization of NTR Peptides—Because the NTR sequences are conserved within an individual plant species, it was hypothesized that these regions may play a role in cyclotide processing. Surprisingly, however, the NTRs are not conserved in sequence between species, so it was of interest to investigate possible structural conservation. Structure prediction suggested that the NTRs are helical. To experimentally verify this, two peptides were synthesized; one based on the NTR from the O. affinis clones and another based on the NTR of the V. odorata clones (named OaNTR and VoNTR, respectively). Both peptides were designed from the NTR sequences preceding a C-terminal cyclotide repeat domain and were synthesized using solid phase peptide chemistry. The sequences of these peptides are in Fig. 2. The OaNTR peptide was designed based on the conserved 19 amino acids just upstream of the cyclotide repeat in Oak2. The NTR sequence from Oak2 was used, because it has a higher homology with the other repeat fragments than the NTR of Oak1. The first three amino acids of the mature cyclotide domain, Gly-Leu-Pro, were also included in the peptide. The VoNTR peptide incorporated all of the sequence between the second and third cyclotide domains in Vok1, that is, the conserved 19 residues and the 6 remaining N-terminal residues. As with the OaNTR peptide, the first 3 residues of the adjacent cyclotide domain (Gly, Leu, and Pro) were included. NMR spectra were recorded, and the amino acid spin systems of the OaNTR and VoNTR peptides were identified using a combination of TOCSY and DQF-COSY spectra at 298 K. Sequential assignments were completed using TOCSY and NOESY spectra. Ambiguities due to overlap were resolved from additional spectra acquired at 280 K or 310 K. Secondary αH shifts, that is, the differences between observed αH shifts in the peptide of interest and those for the corresponding residues in random coil peptides, are indicative of secondary structure. In aqueous solution the C-terminal half of OaNTR exhibits negative secondary shifts typical of helix, as may be seen in Fig. 3. The sample conditions of 298 K, 3.5 mm, and pH 3.6 were chosen, respectively, to provide good dispersion, good sensitivity, and to minimize exchange broadening of amide signals. A series of one-dimensional and TOCSY spectra recorded with concentrations ranging from 0.2 to 3.5 mm and pH from 3.6 to 5.3 showed no substantial changes in αH chemical shifts, suggesting that the helical tendency is not critically dependent on solution concentration or pH over these ranges. The secondary shifts of the VoNTR peptide in water are also indicative of nascent helix in the conserved region highlighted in Fig. 3. To evaluate this helical tendency, TFE, a solvent that is known to stabilize helices in peptides that have an intrinsic propensity for helicity (22Buck M. Q. Rev. Biophys. 1998; 31: 297-355Crossref PubMed Scopus (725) Google Scholar), was added to the solution. Comparison of the secondary shifts of the V. odorata repeat fragment dissolved in water and 20% TFE, shown in Fig. 3, reveals that the addition of TFE clearly increases the helicity of the peptide. The secondary structures of the OaNTR molecule were further inferred from a qualitative analysis of sequential and medium-range NOE intensities, amide proton slow exchange information and coupling constants, which are summarized in Fig. 4. Many strong to medium strength dNN(i, i+1) and medium strength dαN(i, i+1) and some medium range NOEs were present between residues in the region of Ser-5 to Leu-21. Slow exchanging amide protons were detected from a series of one-dimensional spectra recorded up to 60 min after dissolution in 2H2O. These slow exchange protons correspond to residues Phe-12, Leu-13, and Glu-15 to Leu-21. All amide protons had exchanged by 60 min, and although this is faster than what is commonly seen for larger well structured proteins, the existence of an extensive set of observable amide protons in 2H2O indicates a high degree of structural ordering in this small peptide. Indeed the NOE pattern and slowly exchanging amide protons are strongly indicative of a helix in the region Phe-12 to Leu-21. Similarly, slow exchange and NOE data confirmed that the VoNTR peptide forms a helix over residues His-9 to Lys-20. Dissolution of the fully protonated VoNTR peptide in 2H2O identified four slow exchanging amide protons. Slow exchange experiments in the presence of 20% deuterated TFE identified the amide protons of residues Leu-10, Leu-11, and Glu-13 to Lys-21 as slow exchanging. This is consistent with the VoNTR peptide featuring a nascent helical structure in water that is stabilized by the addition of 20% TFE. The structures of the OaNTR peptide were generated using 172 inter-proton distances derived from 110 sequential, 59 medium range, and four long range NOEs, and 12 backbone (ϕ) and two side-chain (χ1) dihedral angle restraints. The residues Leu-9 to Leu-13, Met-16, Leu-18, Lys-19, and Leu-21 were assigned ϕ angles of –65 ± 30° and ϕ restraints of –100 ± 80° were applied for Ser-4, Glu-6, and Thr-7 based on additional coupling and NOE data. Stereospecific assignments of methylene protons and χ1 dihedral angle restraints (60 ± 30°) were derived for Ser-4 and Ser-5. Two NOEs were observed for Hαi-1-Hδi for Leu-21 and Pro-22, indicating that Pro-22 was in the trans conformation. The structures of the VoNTR peptide were calculated using 143 inter-proton distances, including 82 sequential and 61 medium range NOEs derived from NOESY spectra acquired with the peptide dissolved in 20% TFE/80% 2H2O. Although the COSY couplings measured for the VoNTR were inconclusive due to large linewidths, some ϕ angles could be inferred due to the distinct differences in cross-peak intensities observed in the spectrum. In particular the cross-peaks for residues His-9 to Lys-20 were weak relative to those for all other residues so the ϕ angles for these residues could confidently be assigned as –65 ± 15°. The β-methylene protons of residue Glu-3 were stereospecifically assigned, and the corresponding χ1 angle was restrained to –60 ± 30°. Sequential NOEs were observed for Hαi-1-Hδi for Lys-6 and Pro-7 and for Leu-27 and Pro-28, indicating that Pro-7 and Pro-28 were both in the trans conformation. The restraints used to calculate both structures are summarized in Fig. 4, from which it is apparent that there are a greater number of NOEs that are typical of helices found in OaNTR than in VoNTR. However, in the VoNTR spectra there is some overlap of Hα shifts, and as a result potential dα-N(i, i+4) and dα-N(i, i+3) NOEs were overlapped with sequential NOEs. The final set of OaNTR structures, generated from energy minimization in a water shell, was analyzed, and the 20 structures with the lowest overall energies were selected. The ensemble of structures is shown in Fig. 5, superimposed over the backbone atoms of residues Thr-7 to Gly-20. A ribbon depiction of a representative structure is also included in Fig. 5 and shows that the C-terminal part of the peptide, between residues Thr-7 to Gly-20 adopts an α-helical conformation. The N-terminal region is less ordered. The view down the helical axis in Fig. 5 shows that the OaNTR peptide forms an amphipathic helix, with hydrophobic residues clustered on the opposite side of the helix to the charged residues. The geometric and energetic statistics that define the family of low energy OaNTR structures are given in Table III. No significant deviations from idealized covalent geometry were observed. The superimposition of the ensemble of structures is excellent over residues Thr-7 to Gly-20, with a pairwise backbone r.m.s.d. of 0.49 ± 0.16 Å. There is slight fraying at the C-terminal and a degree of disorder near the N-terminal, with an overall pairwise backbone r.m.s.d. of 1.75 ± 0.64 Å. This reflects the general structure of the molecule, where the C-terminal half adopts a well defined α-helix and the N-terminal region is relatively disordered. Hydrogen bond restraints were not included in the structure calculations but were identified to be Phe-12 HN–Thr-8 O, Leu-13 HN–Leu-9 O, Glu-15 HN–Met-11 O
Referência(s)