Protein Engineering, Vol. 15, No. 3, 161-167,
March 2002
© 2002 Oxford University Press
Design of ETB receptor agonists: NMR spectroscopic and conformational studies of ET7–21[Leu7, Aib11, Cys(Acm)15]
1 Department of Biochemistry, Conway Institute of Biomolecular and Biomedical Research, University College Dublin, Belfield, Dublin 4, Ireland, 2 Department of Chemistry, University of Edinburgh, West Mains Road, Edinburgh EH9 3JJ 3 Department of Pure and Applied Chemistry, University of Strathclyde, Thomas Graham Building, 295 Cathedral Street, Glasgow G1 1XL, UK
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionConclusionReferences |
|---|
In a previous report we have shown that the endothelin-B receptor-selective linear endothelin peptide, ET-1[Cys (Acm)1,15, Ala3, Leu7, Aib11], folds into an
-helical conformation in a methanol-d3/water co-solvent [Hewage et al. (1998) FEBS Lett., 425, 234–238]. To study the requirements for the structure–activity relationships, truncated analogues of this peptide were subjected to further studies. Here we report the solution conformation of ET7–21[Leu7, Aib11, Cys(Acm)15], in a methanol-d3/water co-solvent at pH 3.6, by NMR spectroscopic and molecular modelling studies. Further truncation of this short peptide results in it displaying poor agonist activity. The modelled structure shows that the peptide folds into an -helical conformation between residues Lys9–His16, whereas the C-terminus prefers no fixed conformation. This truncated linear endothelin analogue is pivotal for designing endothelin-B receptor agonists.
Keywords: endothelin-1/ETB agonist/molecular modelling/NMR/solution conformation
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionReferences |
|---|
Endothelin-1 is generated from a 203-residue peptide known as preproendothelin by a dibasic endopeptidase. This enzyme cleaves the prepropeptide to a 38-residue peptide (human) known as big-endothelin or proendothelin. Big-endothelin is then cleaved by an endothelin-converting-enzyme to afford the biologically active 21-residue peptide, endothelin-1. The endothelins (Yanagisawa et al., 1988
; Inoue et al., 1989) vasoactive intestinal contractor (Saida et al., 1989) sarafotoxins (Takasaki et al., 1988; Bdolah et al., 1989) and bibrotoxins (Becker et al., 1993) are known to be a family of potent vasoconstrictor peptides. These peptides are characterized by 21 amino acid residues arranged uniquely in a bicyclic fashion via formation of disulfide bridges between cysteine residues 1–15 and 3–11. The tail, hydrophobic amino acid C-terminal residues, 16–21, are considered to be important for the expression of its biological activities (Saeki et al., 1992; Doherty et al., 1993). Two main endothelin receptor sub-types (ETA and ETB) have been cloned and expressed from a variety of sources (Arai et al., 1990; Sakurai et al., 1990; Adachi et al., 1991; Skamoto et al., 1991) and a third endothelin receptor (ETC) has also been identified (Karne et al., 1993). The existence, distribution and binding affinity of these receptor sub-types are known to vary. The receptor binding of ETA is more highly specific for ET-1 and ET-2 than for ET-3, whereas receptor binding of ETB shows similar affinity with ET-1, ET-2 and ET-3. The ETC receptor selectively binds ET-3 (Karne et al., 1993). Previous studies, which were carried out to assess the structural requirements for specific receptor binding, have shown that the entire 21 amino acid sequence is required for ETA receptor binding, whereas N-terminal residues 1–9 have been shown to be non-essential for ETB receptor recognition (Saeki et al., 1991).
Peptides of the endothelin family are widely distributed throughout a variety of tissues and organs such as vascular and non-vascular smooth muscle, nervous tissue, heart, kidney and mesangial cells and adrenal glands. The pharmacological evaluations and structure–activity relationship studies of endothelins have been extensively reviewed (Doherty, 1995; Huggins et al., 1993). There have been many reports consisting of a combination of high resolution NMR studies and molecular modelling techniques deployed to determine the tertiary structures of peptides belonging to the endothelin family (Reily and Dunbar, 1991; Mills et al., 1992; Atkins et al., 1995; Hewage et al., 1997).
Increasing evidence for the involvement of endothelin peptides in a variety of human diseases has prompted a major effort in structure–activity relationship studies and drug design based on peptides from the endothelin family. Our previous reports of linear, modified, synthetic endothelin analogues (Hewage et al., 1998a,b) showed that the disulfide bridges associated with maintaining the structural integrity of the natural peptide are neither important for the formation of helical secondary structure nor essential for the biological activity of ETB receptor. In this study, we present the solution structure of the ETB receptor selective agonist, the peptide ET7–21[Leu7, Aib11, Cys(Acm)15] (designated here as LJP33), by NMR-based molecular modelling.
ET-1: CSCSSLMDKECVYFCHLDIIW
peptide LJP2: CSASSLLDKEXVYFCHLDIIW
peptide LJP33: LDKEXVYFCHLDIIW
The disulfide bridges of ET-1 are between C1–C15 and C3–C11. The Cys1 and Cys15 of peptide LJP2 and Cys15 of peptide LJP33 are protected by the acetamidomethyl group (Acm) and X denotes -aminoisobutyric acid.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionReferences |
|---|
Peptide LJP33 was synthesized (Ramageet al., 1996
) in our laboratory on an adapted ABI 430A peptide synthesizer using Fmoc chemistry, purified by column chromatography and characterized by HPLC, electrospray MS and AAA according to published procedures (Hewage et al., 1998a). NMR spectroscopy
All NMR experiments were performed using a 5 mm probe head on a Varian VXR 600S NMR spectrometer operating at a 1H resonance frequency of 599.945 MHz. Peptide LJP33(5 mg), was dissolved in d3-methanol/H2O co-solvent (1:1) except for amide proton exchange studies when d4-methanol/D2O (1:1) was used as solvent. All NMR experiments were performed on the sample at a pH of 3.6 and a temperature of 298 K. One-dimensional NMR data were recorded with a pre-saturation delay of 1.5 s. A total of 256 transients were acquired over a 7 kHz spectral width into 35K data points. The data were zero filled to 65 536 data points prior to apodization using an optimized shifted squared sinebell function followed by Fourier transformation and phase correction. Two-dimensional phase-sensitive TOCSY (Bax and Davis, 1985) and NOESY (Kumar et al., 1980) data sets were acquired with a pre-saturation delay of 1.5 s. The 2D-TOCSY pulse sequence was used with an 80 ms mixing time and two 2 ms trim pulses. A series of 2D-NOESY data sets, with mixing times of 40, 80, 150, 200 and 300 ms, were collected and evaluated for the effects of spin-diffusion. The residual water resonance was saturated during preparation and mixing periods. A total of 16 and 32 transients were acquired for each of 2x256 t1 increments (hypercomplex acquisition; States et al., 1982) into 2048 complex data points for the 2D-TOCSY and 2D-NOESY data sets, respectively. Other acquisition parameters for both data sets were an acquisition time of 0.146 s and a spectral width of 7 kHz. Fourier transformation in F2 was carried out without zero filling whereas data in F1 were zero filled to 1024 data points prior to Fourier transformation. All time domain data were apodized in both dimensions using an optimized shifted squared sinebell window function. The real Fourier transformation was carried out on 1024x2048 data points. All the two-dimensional spectra were acquired on a non-spinning sample.
Variable temperature studies of peptide LJP33 were carried out on a solution of 1 mg of the peptide using a standard one-dimensional 1H NMR pulse sequence. Spectra were recorded at temperature intervals of 4° from 290 to 298 K and 5° from 298 to 318 K continuously. Amide proton–deuterium exchange experiments were carried out soon after dissolution of 1 mg of the peptide and then at regular time intervals. For standard 1D 1H NMR spectra a relaxation delay of 1.5 s was used only during the preparation period to eliminate the residual HOD solvent resonance. All data were internally referenced to the residual 1H NMR resonance of CD2HOH defined as 3.30 p.p.m. Varian's VNMR software (version 4.1) was used for all data acquisition and preliminary data processing.
Molecular modelling
Structure calculation studies were carried out using the Tripos molecular modelling software SYBYL version 6.1. (Tripos Software, 1994). 2D-NOESY data at 150 ms were converted and processed using the TRIAD module of SYBYL. Peak picking was carried out manually. Volume integrals were calculated for all cross-peaks within the TRIAD module which were then used for generating lower and upper distance constraints. The lower distance bound was in all cases set to the van der Waals radius of 1.8 Å. Upper distance bounds were defined by volume integral values of 2.8 (strong), 3.6 (medium) and 5.0 Å (weak). The threshold values of upper distance bounds were established using the known sequential distances of dNN and dN (Billeter et al., 1982). Pseudoatoms with appropriate distance corrections were employed for protons which could not be stereospecifically assigned (Wüthrich et al., 1983). Partially overlapped signals were ignored in the structure calculation. Slow amide proton exchange rates were observed for some NHs. Constraints between these protons and associated backbone carbonyl oxygen atoms were introduced after the first phase of the structure calculations, when the fold of the peptide chain emerged. Hydrogen bonding partners were identified using the standard model for polypeptide -helicies. These inter-atomic distance constraints were set in the range 1.8–2.0 and 2.7–3.0 Å for the H–O distance and N–O distances, respectively (Williamson et al., 1985). Torsion angle (
) constraints were calculated using the expression 3JHN = 6.4cos2
– 1.4cos + 1.9 (Pardi et al., 1984). 3JHN values were measured directly from the high resolution 1H NMR spectrum of peptide LJP33.
The DIANA package (distance geometry algorithm for NMR applications; Güntert et al., 1991) was used to generate random starting structures based on an initial structure built within SYBYL. The 169 distance constraints obtained from NOESY cross-peak volumes were used as input for the DIANA calculations. Six constraints for hydrogen bonds and 13 torsional angle constraints were also incorporated. All distance and torsional constraints were included in the DIANA calculations. Atomic distances were constrained using the force constant kNOE = 1 kcal mol–1 Å–2 and torsions were constrained using kDihed_c = 0.01 kcal mol–1 deg–2. DIANA was used to calculate 300 structures from random starting conformations and the calculation produced 31 acceptable structures dependent on the final target function value. Distances, which were predetermined by the covalent geometry of the molecule or by conformations which violated the constraints, were regarded as irrelevant by the DIANA program. These constraints were eliminated during the calculation to yield 112 structurally important constraints defined as modified bounds.
Structures from DIANA were subsequently constrained by these modified distance bounds. Inter-atomic distances were then constrained using a higher force constant of kNOE = 10 kcal mol–1 Å–2 and torsions were constrained using the same force constant of kDihed_c = 0.01 kcal mol–1 deg–2. Structures were then subjected to 200 steps of conjugate-gradient energy minimization. Higher energy structures were initially minimized using an atom-by-atom Simplex minimization. The Tripos 5.2 force field with the energy minimizer MAXIMIN 2 was used in the minimization procedure. Other parameters for constraining covalent geometry were kBond = 600 kcal mol–1 Å–2, kAngle = 0.02 kcal mol–1 deg–2 and kTor = 0.2 kcal mol–1 deg–2. At this stage, 10 structures which showed least violated constraints and lowest energy were chosen for further calculations.
Energy minimized structures were then refined using a dynamical simulated annealing method (DSA), this being a process by which local energy minima are overcome by successive heating and cooling cycles of the molecular model. DSA calculations began by holding the temperature constant at 1000 K for a period of 5400 fs. During the annealing time of 900 fs, the temperature was reduced `stepwise' until a minimum temperature of 100 K was reached. The Boltzmann scaling of atomic velocities was chosen from a random number. This process completed the first cycle. Ten cycles of DSA were run for each starting structure. The conformations obtained by DSA were further minimized with 200 steps of conjugate-gradient energy minimization. The minimization parameters used were the same as those described for the DIANA calculated structures.
The energy minimized conformations were subjected to a final molecular dynamics (MD) quenching calculation without changing either distance or torsional force constants. The MD calculations were performed in the gas phase. The initial atomic velocities were chosen from a random distribution at 1000 K and the dynamic trajectory (100 fs) was followed for 20 ps in 1 fs steps. These calculations were carried out under NTV ensemble conditions with a 10 fs coupling factor for the temperature. The MD calculations were repeated three times for each conformation using different initial values for random number seed and these three different conformers were then averaged to obtain the averaged conformation. After removing all the distance and torsional experimental energy barriers, these averaged conformers were finally subjected to 500 steps of conjugate-gradient energy minimization as described previously.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionReferences |
|---|
Structure determination
The sequence of 15 residue synthetic linear peptide, peptide LJP33, includes nine unique residues, Lys9, Glu10, Aib11, Val12, Tyr13, Phe14, Cys15, His16 and Trp21. The fingerprint region of the 2D-TOCSY spectrum (Figure 1) provided a basis for the identification of individual residue spin-system resonances and showed 13 resonances for NH/H excluding N-terminal Leu7 and Aib11 resonances. An additional resonance was observed for the side chain protecting group (Acm) of Cys15. The magnetization transfer from NH through the entire spin-systems were observed where possible for all amino acid residues in the 2D-TOCSY spectrum.
|
The unique amino acid residues Lys9, Glu10 and Val12 were clearly identified on the basis of their individual signals in the 2D-TOCSY spectrum. The chemical shifts of methyl resonances were used to identify the leucine and isoleucine residues. The two aspartic acid residues and four aromatic residues could not be distinguished at this stage due to the similarity of their spin system. Identification of the non-aromatic and aromatic residues, which show the same spin pattern, was made via the ßH to aromatic proton resonance connectivities present in the 2D-NOESY spectrum. Discrimination of duplicated residues were made by direct comparison of the 2D-TOCSY and 2D-NOESY spectra. The unique amino acid residues were used as starting points for the identification of sequence-specific resonance assignments. The backbone fingerprint region of the 2D-NOESY spectrum is shown in Figure 2
. The chemical shift assignments of peptide LJP33 1H resonances which were made according to the following procedure are given in Table I.
|
|
The unique residue Lys9 showed connectivities to its neighbours, residues Asp8 and Glu10. The methyl resonances of the residue Aib11 showed a prominent cross-peak to its neighbour Glu10. The
H/NH backbone connectivity between N-terminal residue Leu7 and the Asp8 residue was prominently visible in the fingerprint region of the 2D-NOESY spectrum thereby completing the N-terminal part of the sequential assignment between residues Leu7 and Aib11. Sequential connectivity from the unique residue Val12 to Tyr13 was clearly identifiable. The strong clear resonance of Tyr13ßH/Phe14NH helped to distinguish the Phe14 residue. The backbone connectivities of C-terminal residues Asp18/Ile19, Ile19/Ile20 and Ile20/Trp21 were also clearly identifiable. The backbone connectivities of two adjacent residues, Cys15 and His16, thence Leu17 were clearly distinguished in the 2D-NOESY fingerprint region discriminating the remaining leucine residue, Leu7 thus completing the sequence-specific assignment of the peptide LJP33.
The remainder of the 2D-NOESY spectrum was then searched for secondary structure correlations. The clear, long-range iH/Ni + 3H connectivities of Lys9/Val12, Val12/Cys15 and Tyr13/His16 were identified in the fingerprint region of the 2D-NOESY spectrum. Correlations between Glu10/Tyr13 could not be distinguished due to the presence of overlapping cross-peaks (Figure 2
). Strong NiH/Ni + 1H backbone connectivities and weak iH/Ni + 1H connectivities confirmed the -helical pattern between Lys9 and His16. The clear iH/ßi + 3H connectivities between residues Lys9/Val12, Glu10/Tyr13, Val12/Cys15 and Tyr13/His16 appeared in the H–ßH cross-peak region of the 2D-NOESY spectrum, which also confirmed the -helical nature of the structure. Smaller 3JNH coupling constants (Table I) were observed for residues Lys9–His16. Figure 3 summarizes the inter-residue nOe connectivity patterns of peptide LJP33.
|
On dissolving the sample in a CD3OD/D2O co-solvent, all of the NH proton resonances were initially observable except for those from N-terminal residues 1–3. The amide protons of the C-terminal residues Ile20 and Trp21 were observed even after 25 min. All of the observed amide protons, except for Aib11, Tyr13 and Phe14 had exchanged with D2O after 6 h. This observation strongly supports the presence of a hydrogen bonded segment. No amide protons were observable 10 h after dissolving the sample in the exchanging solvent.
The temperature dependence of backbone amide proton chemical shifts was measured over the temperature range 290–313 K. A structured state was apparent from the very small temperature coefficients of Aib11 
/T = –1.4 p.p.b. K–1 and Phe14 /T = –2.1 p.p.b. K–1. This suggested that amide protons in the helical region were involved in hydrogen bonding.
The final structures obtained from the structure calculation of peptide LJP33 by DG, DSA followed by MD were in good agreement with the experimentally derived constraints. The calculated structures remain within a single family of conformations. The major structural feature of peptide LJP33 was a highly-ordered right-handed -helix from Lys9–His16. The structure calculations reveal that the C-terminal tail region of residues Leu17–Trp21 appears to have no well-defined conformation in aqueous co-solvent, consistent with the fewer distance constraints obtained from this region.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionReferences |
|---|
The current study described here lies in the context of a wider project designed to develop more potent peptidic drugs. Much of this study to date has focused on the structural requirements for endothelin receptor binding. Previously we have examined full length linear analogues of the peptide ET-1, namely LJP1 (Hewage et al., 1999a
), LJP2 (Hewage et al., 1998a), LJP26 (Hewage et al., 1998b) and LJPC (Hewage et al., 1999b). The solution structures of each of these analogues have been compared with that of ET-1 itself, determined under identical conditions in our laboratory (Hewage et al., 1997). Each of the linear analogue structures have displayed essentially -helical characteristics along the length of the sequences except at N- and C-termini, which were unstructured in each case. We have previously argued that the solvent environment for such studies should be hydrophobic in nature to reflect the probable environment in which peptidic drugs and antibiotics would normally act, namely in a membrane or at a protein surface (Hewage et al., 1999b).
Previous NMR studies of endothelin peptides under different solution conditions have also revealed helical structural features. However, these have usually involved fewer residues. Our previous work on ET-1 (Hewage et al., 1997) and modified, synthetic endothelin peptide analogues (Hewage et al., 1998a,b) studied under identical solution conditions as those reported here have also shown -helical conformations in the same residue region of the polypeptide sequence. The nOe patterns observed at both the N- and C-termini regions were also in good agreement with the previously published work. The NMR and molecular modelling studies of the parent linear ET-1 analogue, peptide LJP2, revealed a structure characterized by an -helix lying between residues Ser5 and His16. Biological activity studies have revealed that peptide LJP2 displays a potency against the ETB receptor which is a factor of 10 greater than that of the parent peptide ET-1 (Jiang, 1996). For this reason truncated analogues of LJP2 were synthesized in a bid to determine which features of the sequence and structure were responsible for this enhanced potency. The truncation proceeded from the N-terminus of the peptide, since it has been shown that most of the C-terminal residues were essential for activity (Leonard et al., 1995). Peptide LJP33, described in this report, retained the potency found for peptide LJP2, but was without the same level of receptor selectivity. Therefore, it was of interest to us to know whether the solution structure of this material differed in any way from that of peptide LJP2.
Therefore, we have solved the solution structure of the truncated ETB selective agonist, ET7–21[Leu7, Aib11, Cys (Acm)15] (LJP33) in an aqueous methanol co-solvent by means of a combination of 1H NMR and molecular modelling studies. Although peptide LJP33 has a rather low molecular weight, good 2D-NOESY spectra, with strong negative nOes, were obtained which is consistent with a slowly tumbling macromolecule of defined tertiary structure. Linear correlation of nOe cross-peak intensities up to 150 ms of the NOESY mixing time was observed. This behaviour may be due to the fact that the effective volume of the hydrated molecule led to a longer correlation time. Evidence for the prevalence of a secondary helical conformation was revealed by the observation of iH/Ni + 3H and strong iH/ßi + 3H interactions found in the 2D-NOESY spectrum. The slow exchange of Tyr13, Phe14 amide protons suggested that the helix was stabilized by hydrogen bonds, occurring between residues Lys9/Tyr13 and Glu10/Phe14. Small temperature coefficients of the amide protons of Aib11 and Phe14 indicated shielding from solvent exchange.
All NMR spectroscopic studies and molecular modelling calculations of peptide LJP33 yielded a well defined -helical conformation between residues Lys9 and His16. The mode of convergence of the calculated structures is demonstrated by the RMSD values calculated for all atoms in the helical region, which was found to be 1.84 ± 0.26, compared with backbone only atoms, for which the RMSD figure was found to be 0.54 ± 0.16 (Table II). The C-terminal region of the polypeptide chain beyond residue His16 may exist as an extended random structure as indicated by random coil 3JNH coupling constants (>7 Hz) and a series of strong iH/Ni + 1H nOes.
|
We observed that the backbone conformation of the helical portion of the peptide (residue Lys9–His16) was almost identical to the same portion of both ET-1 and LJP2, whereas the C-terminus shows variations. The mobile, N-terminal two residue segment of the peptide appears to be in equilibrium, which is expected to be the case for such a molecule. Relative to peptide LJP2, the N-terminus of LJP33 was found to be slightly more ordered, which is a reflection of the point at which the truncation was made, namely in the first turn of the
-helix of the extended peptide LJP2. Figure 4 shows a representative family of 10 structures resulting from the combination of structure calculations and energy minimizations on peptide LJP33. Solution structures of ET-1 and LJP2 are shown in Figure 5 for comparison purposes.
|
|
The crystal structure of ET-1 (Janes et al., 1994
) has been described as having an N-terminal extended ß-strand with a bulge between residues 5 and 7 followed by a hydrogen-bonded loop between the carbonyl oxygen of residue 7 and the amide proton of residue 11. The residues 12–15 of endothelin form an irregular -helix whereas the C-terminal tail (residues 16–21) forms an ordered -helical structure. In addition, the crystal structure of the tail portion is neither more flexible nor more disordered than the globular head region (residues 1–15). Both NMR and crystal structures show somewhat similar hydrogen bond patterns in the middle of the peptide sequence. Thus, the X-ray and NMR structures of ET-1 do have common features but also differ significantly in detail, especially the C-terminal residues (16–21). The structural difference between the crystal and NMR structures can be attributed to the solvent effects which play a vital role in NMR studies. A detailed comparison of X-ray and NMR structural features of human ET-1 (Wallace et al., 1995) indicated that the experimental conditions and environmental factors also contribute to the major differences of the X-ray and NMR structures.
Structure–activity relationship studies of this peptide can be considered in terms of the structural features revealed by our model. Changing Met7 of ET-1 to Leu and also replacing Cys11 by the Aib residue did not alter the formation of an -helical conformation. According to general peptide chemistry, Aib is known to introduce -helical characteristics into a peptide backbone structure. Previous studies, where Met7 was replaced by alanine (Dalgarno et al., 1992) or norleucine (Aumelas et al., 1991), induced very little conformational changes relative to ET-1. Most of the biologically active amino acid residues reported in the literature for these peptides lie in the helical region of the peptide (De Castiglione et al., 1992; Galantino et al., 1992; Hunt et al., 1991). Residues 10 (Glu) and 14 (Phe), which are positioned on the same side of the calculated model, are important residues for biological activity in rat pulmonary artery (Nakajima et al., 1989). Thus, aromaticity at position 14 plays an essential role for biological activity and also for folding of the peptide.
Although the C-terminal region of the calculated structures of ET-1 showed no preferred conformation, most of the residues present in this region are known to be important for biological activity (Doherty et al., 1991; Leonard et al., 1995). The N-terminal residues of ET-1, namely Ser2, Ser4, Ser5 and Leu6, have been shown to be relatively tolerant of substitutions. Pharmacological evaluation of this region found fewer biologically important residues compared with the other regions of the peptide (Saeki et al., 1991; Watanabe et al., 1991). Although the N-terminal residues are not critical for agonist activity, they may be responsible for receptor selectivity, as displayed by the differences between peptides LJP2 and LJP33. We speculate that it is the extended -helical structure of peptide LJP2 compared with peptide LJP33 and not the specific peptide sequence, which is responsible for this observed selectivity.
|
Conclusion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionReferences |
|---|
We have determined the solution conformation of a linear endothelin analogue, ET7–21[Leu7, Aib11, Cys(Acm)15] (peptide LJP33) in 1:1 aqueous methanol co-solvent which is also an ETB receptor selective agonist. The results, obtained under these experimental conditions, indicate that the peptide adopts a defined structure. The helical nature of the peptide is more likely to reflect more closely the conformation of the peptide in a hydrophobic environment, such as that in a membrane or at a receptor site. Our study clearly shows that the deletion of the N-terminal part of the peptide plays no vital role in the ability of the peptide to form secondary structural features in solution. The study is consistent with the findings that the N-terminal residues are non-essential for agonist activity at the ETB receptor but are probably responsible for the loss of receptor selectivity relative to peptide LJP2. Although we have shown that peptide LJP33 is
-helical, it is clearly the length of the -helix (two full turns for LJP33 compared with three full turns for other linear ET-1 analogues) that is important for ETB receptor selectivity.
|
Notes |
|---|
4 To whom correspondence should be addressed. E-mail: chandralal.hewage{at}ucd.ie
|
Acknowledgments |
|---|
This work was performed while C.M.H. and J.A.P. were research scientists at the Department of Chemistry, University of Edinburgh, UK. We thank the EPSRC for 600 MHz NMR facilities and the Wellcome Trust for molecular modelling facilities at the above address. We are grateful to the Commonwealth Association and the British Council for funding to C.M.H. We also thank K.Shaw and B.Wigham for technical assistance with synthesis. We gratefully acknowledge Parke Davis Pharmaceuticals for financial support and bioassays of peptides LJP2 and LJP33.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionReferences |
|---|
Adachi,M., Yang,Y.-Y., Furuichi,Y. and Miyamoto,C. (1991) Biochem. Biophys. Res. Commun., 180, 1265–1272.[CrossRef][Web of Science][Medline]
Arai,H., Hori,S., Aramori,I., Ohkubo,H. and Nakanishi,S. (1990) Nature, 348, 730–732.[CrossRef][Medline]
Atkins,A.F., Martin,R.C. and Smith,R. (1995) Biochemistry, 34, 2026–2033.[CrossRef][Medline]
Aumelas,A., Chiche,L., Mahe,E., Le-Nguyen,D., Sizun,P., Berthault. P. and Perly,B. (1991) Int. J. Pept. Protein Res., 37, 315–324.[Web of Science][Medline]
Bax,A. and Davis,D.G. (1985) J. Magn. Reson., 65, 355–360.[Web of Science]
Bdolah,A., Wollberg,Z., Fleminger,G. and Kochva,E. (1989) FEBS Lett., 256, 1–3.[CrossRef]
Becker,A., Dowdle,E.B., Hechler,U., Kauser,K., Donner,P. and Schleuning, W.-D. (1993) FEBS Lett., 315, 100–103.[CrossRef][Web of Science][Medline]
Billeter,M., Braun,W. and Wüthrich,K. (1982) J. Mol. Biol., 155, 321–346.[CrossRef][Web of Science][Medline]
Dalgarno,D.C., Slater,L., Chackalamannil,S. and Senior,M.M. (1992) Int. J. Pept. Protein Res., 40, 515–523.[Web of Science][Medline]
De Castiglione,R., Tam,J.P., Liu,W., Zhang,J.-W., Galantino,M., Bertolero,F. and Vaghi,F. (1992) In Smith,J.A. and Rivier,J.E. (eds), Peptides: Chemistry and Biology; Proceedings of the 12th American Peptide Symposium. ESCOM, Leiden, The Netherlands, pp. 402.
Doherty,A.M., Cody,W.L., Leitz,N.L., DePue,P.L., Taylor,M.D., Rapundalo,S.T., Hingorani,G.P., Major,T.C., Panek,R.L. and Taylor,D.G. (1991) J. Cardiovasc. Pharmacol., 17, S59–S61.[CrossRef]
Doherty,A.M. et al. (1993) J. Med. Chem., 36, 2585–2594.[CrossRef][Web of Science][Medline]
Doherty,A.M. (1995) In Weiner,D.B. and Williams,W.B. (eds), Chemical and Structural Approaches to Rational Drug Design. CRC Press, Ann Arbor, pp. 85–123.
Galantino,M., De Castiglione,R., Tam,J.P., Liu,W., Zhang,J.-W., Cristiani,C. and Vaghi,F. (1992) In Smith,J.A. and Rivier,J.E. (eds), Peptides: Chemistry and Biology; Proceedings of the 12th American Peptide Symposium. ESCOM, Leiden, The Netherlands, pp. 404.
Güntert,P., Braun,W. and Wüthrich,K. (1991) J. Mol. Biol., 217, 517–530.[CrossRef][Web of Science][Medline]
Hewage,C.M., Jiang,L., Parkinson,J.A., Ramage,R. and Sadler,I.H. (1997) J. Pept. Sci., 3, 415–428.[CrossRef][Web of Science][Medline]
Hewage,C.M., Jiang,L., Parkinson,J.A., Ramage,R. and Sadler,I.H. (1998a) FEBS Lett., 425, 234–238.[CrossRef][Web of Science][Medline]
Hewage,C.M., Jiang,L., Parkinson,J.A., Ramage,R. and Sadler,I.H. (1998b) J. Biomol. Struct. Dyn., 16, 425–435.[Web of Science][Medline]
Hewage,C.M., Jiang,L., Parkinson,J.A., Ramage,R. and Sadler,I.H. (1999a) J. Pept. Res., 53, 223–233.[CrossRef][Web of Science][Medline]
Hewage,C.M., Jiang,L., Parkinson,J.A., Ramage,R. and Sadler,I.H. (1999b) Neurochem. Int., 35, 35–45.[CrossRef][Web of Science][Medline]
Huggins,J.P., Pelton,J.T. and Miller,R.C. (1993) Pharmacol. Ther., 59, 55–123.[CrossRef][Web of Science][Medline]
Hunt,J.T., Lee,V.G., Stein,P.D., Hedberg,A., Liu,E.C.K., McMullen,D. and Moreland,S. (1991) Bioorg. Med. Chem. Lett., 1, 33–38.
Inoue,A., Yanagisawa,M., Kimura,S., Kasuya,Y., Miyauchi,T., Goto,K. and Masaki,T. (1989) Proc. Natl Acad. Sci. USA, 86, 2863–2867.
Janes,R.W., Peapus,D.H. and Wallace,B.A. (1994) Nat. Struct. Biol., 1, 311–319.[CrossRef][Web of Science][Medline]
Jiang,L. (1996) PhD Thesis, University of Edinburgh, UK.
Karadi,R., Billeter,M. and Wüthrich,K. (1996) J. Mol. Graph., 14, 51–55.[CrossRef][Web of Science][Medline]
Karne,S., Jayawickreme,C.K. and Lerner,M.R. (1993) J. Biol. Chem., 268, 19126–19133.
Kumar,A., Ernst,R.R. and Wüthrich,K. (1980) Biochem. Biophys. Res. Commun., 95, 1–6.[CrossRef][Web of Science][Medline]
Leonard,D.M., Reily,M.D., Dunbar,J.B., Holub,K.E., Cody,W.L., Hill,K.E., Welch,K.M., Flynn,M.A., Reynolds,E.E. and Doherty,A.M. (1995) Bioorg. Med. Chem. Lett., 5, 967–972.[CrossRef]
Mills,R.G., O'Donoghue,S.L., Smith,R. and King,G.F. (1992) Biochemistry, 31, 5640–5645.[CrossRef][Medline]
Nakajima,K. et al. (1989) Biochem. Biophys. Res. Commun., 163, 424–429.[CrossRef][Web of Science][Medline]
Pardi,A., Billeter,M. and Wüthrich,K. (1984) J. Mol. Biol., 180, 741–751.[CrossRef][Web of Science][Medline]
Ramage,R. et al. (1996) In Epton,R. (ed.), Proceedings of the Fourth International Symposium: Innovation and Perspectives in Solid Phase Synthesis and Combinatorial Libraries. Mayflower Scientific Ltd, Birmingham, UK, pp. 1–10.
Reily,M.D. and Dunbar,J.B. (1991) Biochem. Biophys. Res. Commun., 178, 570–577.[CrossRef][Web of Science][Medline]
Saeki,T., Ihara,M., Fukuroda,T., Yamagiwa,M. and Yano,M. (1991) Biochem. Biophys. Res. Commun., 179, 286–292.[CrossRef][Web of Science][Medline]
Saeki,T., Ihara,M., Fukuroda,T., Yamagiwa,M. and Yano,M. (1992) Biochem. Int., 28, 305–312.[Web of Science][Medline]
Saida,K., Mitsui,Y. and Ishida,N. (1989) J. Biol. Chem., 264, 14613–14616.
Sakurai,T., Yanagisawa,M., Takuwa,Y., Miyazaki,H., Kimura,S., Goto K. and Masaki,T. (1990) Nature, 348, 732–735.[CrossRef][Medline]
Skamoto,A., Yanagisawa,M., Sakurai,T., Takuwa,Y., Yanagisawa,H. and Masaki,T. (1991) Biochem. Biophys. Res. Commun., 178, 656–663.[CrossRef][Web of Science][Medline]
States,D.J., Haberkorn,R.A. and Ruben,D.J. (1982) J. Magn. Reson., 48, 286–292.[Web of Science]
Takasaki,C., Tamiya,N., Bdolah,A., Wollberg,Z. and Kochva,E. (1988) Toxicon, 26, 543–548.[Medline]
Tripos Molecular Modelling Software (1994) SYBYL Version 6.1. Tripos Associates, St Louis, MO, USA.
Wallace,B.A., Janes,R.W., Bassolino,D.A. and Krystek,S.R. (1995) Protein Sci., 4, 75–83.[Web of Science][Medline]
Watanabe,T.X., Itahara,Y., Nakajima,K., Kumagaye,S., Kimura,T. and Sakakibara,S. (1991) J. Cardiovasc. Pharmacol., 17, S5–S9.
Williamson,M.P., Havel,T.F. and Wüthrich,K. (1985) J. Mol. Biol., 182, 295–315.[CrossRef][Web of Science][Medline]
Wüthrich,K., Billeter,M. and Braun,W. (1983) J. Mol. Biol., 169, 949–961.[Web of Science][Medline]
Yanagisawa,M., Kurihara,H., Kimura,S., Tomobe,Y., Kobayashi,M., Mitsui,Y., Yazaki,Y., Goto,K. and Masaki,T. (1988) Nature, 332, 411–415.[CrossRef][Medline]
Received August 31, 2001; revised November 29, 2001; accepted December 7, 2001.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||