Eur J Biochem 269, 6009–6019 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.03278.x Monitoring the structural consequences of Phe12 fi D-Phe and Leu15 fi Aib substitution in human/rat corticotropin releasing hormone Implications for design of CRH antagonists Georgios A Spyroulias, Spyridon Papazacharias, George Pairas and Paul Cordopatis Department of Pharmacy, University of Patras, Greece A new human/rat CRH analogue has been synthesized using the Fmoc/tBu solid-phase synthetic protocol The sequence of the new peptide differs from the original in two positions, 12 and 15, at which the native amino acids L-phenylalanine 12 and L-leucine 15 have been replaced by the nonprotein amino acids D-phenylalanine and a-aminoisobutyric acid (Aib), respectively The high resolution three-dimensional solution structure of [D-Phe12, Aib15]CRH has been determined by 688 distance constraints (656 meaningful NOE and 32 H-bonds distance limits) and 21 angle constraints A family of 40 energy-minimized conformers was obtained with average ˚ ˚ rmsd of 0.39 ± 0.16 A and 0.99 ± 0.13 A for backbone and heavy atoms, respectively, and distance penalty ˚ functions of 0.42 ± 0.03 A2 The NMR data acquired in a solvent system of water/trifluoroethanol (34%/66%, v/v) revealed that this 41-polypeptide adopts an almost linear helical structure in solution with helical content which reaches an 84% of the residues Structural analysis confirmed the existence of two helical peptide fragments The first was comprised of residues Ile6–Arg16 and the second of residues Glu20–Ile40, forming an angle of 34.2° The structural differences with respect to the native peptide have been identified in the region D-Phe12–Glu20 where double substitution at positions 12 and 15 seems to perturb the elements of the native 35-residue helix These structural rearrangements promote non-native intramolecular interactions in the region of the molecule between either the hydrophobic side-chains of D-Phe12, Aib15 and Leu18, or the charged groups of the residue pairs Arg16– Glu20 and His13–Glu17 being responsible for changes in hormonal functionality This CRH analogue currently exhibits lack of any activity Corticotropin releasing hormone (CRH) is a 41-amino acid neuropeptide characterized by the presence of an isoleucine-amide in its C-terminal end (Scheme 1) It is produced mainly in the hypothalamus and acts as the principal neuroregulator of the secretion of adrenocorticotropic hormone (ACTH), b-endorphin and other proopiomelanocortin products of the anterior pituitary gland CRH is considered to be the key hormone in the control of the hypothalamic–pituitary–adrenal axis function and is responsible for the endocrine, autonomic, immunological and behavioural responses of mammalians to stress [1–3] The hormone interacts at a nanomolar level with a transmembrane receptor belonging to the Gs-protein coupled family CRH binds to membrane homogenates from the anterior pituitary, revealing a high-affinity component of 0.1 nM and a low-affinity binding site at 20 nM [4] There is a wide distribution of CRH receptors throughout the human body, indicating that the hormone is probably implicated in a variety of actions, including inflammatory phenomena, the effects of various stressors, pregnancy and labour, food intake, thermoregulation, etc [5–11] The role of CRH has not yet been elucidated in every possible situation and new evidence is constantly appearing, establishing an even more complicated involvement in many biological functions Moreover, inappropriate or disturbed CRH neuronal activity manifests itself in various neuropsychiatric disorders such as affective disorders, anorexia nervosa and Alzheimer’s disease [12] The great variety of functions mediated by CRH, or in which the hormone plays an active role, led to an effort to determine the structure–activity relationships which could clarify its action Thus, an a-helical conformation of high amphiphilicity was found to be the major characteristic of the molecule’s tertiary structure Residues 4–8 were found to be essential for receptor recognition, while residues 9–41 are responsible for binding and exertion of action [13] The study of this molecule has led to the development of agonists Correspondence to P Cordopatis, Department of Pharmacy, University of Patras, GR-265 04, Greece Fax: + 30 2610997 714, Tel.: + 30 2610997 713, E-mail: pacord@upatras.gr Abbreviations: CRH, corticotropin releasing hormone; Aib, a-aminoisobutyric acid; TFE, 2,2,2-trifluoroethanol; DQF-COSY, Double-quantum-filtered phase sensitive correlated spectroscopy; WATERGATE, water suppression by gradient-tailored excitation; REM, restrained energy minimization Note: atomic coordinates of the 40 best and the mean energy minimized structures have been deposited in the Protein Data Bank (accession codes 1goe and 1go9, respectively) Note: a web site is available at http://www.pharmacy.upatras.gr (Received 22 July 2002, accepted 13 September 2002) Keywords: corticotropin releasing hormone; NMR; solid phase synthesis; solution structure; synthetic analogue Ó FEBS 2002 6010 G A Spyroulias et al (Eur J Biochem 269) Scheme Primary structure of various natural CRH peptides and antagonists of CRH, which can be either peptidic or nonpeptidic These analogues have the potential to be pharmacological tools for intervention in pathological conditions such as Cushing’s syndrome, inflammatory diseases, pre-eclampsia or preterm labour, anxiety and depression and a potentially great number of other disorders We have previously described the synthesis and the biological evaluation of a new series of CRH analogues, which showed that most analogues tested were deprived of any significant agonist effects [14] These findings prompted us to study the conformational features of the analogue [D-Phe12, Aib15]CRH through one- and two-dimensional J-correlated 1H NMR spectroscopy Structure calculations were performed using simulated annealing protocols and restrained energy minimization (REM) Structural characteristics imposed by the introduction of the novel residues in the sequence of the molecule and the differences between the CRH analogue and the native structure [15] are also discussed The major goal of our study is the solution structure determination of a CRH analogue devoid of any agonist activity, a fact particularly relevant to the further design and development of CRH antagonists with wide-ranging clinical applications in psychiatric disorders MATERIALS AND METHODS Sample preparation The CRH analogue was synthesized using Fmoc/tBu chemistry by solid-phase peptide synthesis and was purified and characterized via reversed phase HPLC and ESI-MS [14] The CRH analogue was dissolved to a final concentration of 2–2.5 mM in H2O/2,2,2-trifluoroethanol (TFE)-d2 (CORTEC, Paris, France; 34%/66%, v/v) to record oneand two-dimensional NMR spectra, and the pH was adjusted to 3.8 For amide exchange experiments, the peptide was lyophilized and dissolved in D2O/TFE-d3 (34%/66%, v/v) NMR spectroscopy NMR spectra were recorded on Bruker’s 500 and 600-MHz spectrometers in 298K and 310K One- and two-dimensional double-quantum-filtered-COSY [16], TOCSY [17,18] and time-proportional phase incrementation-NOESY [19,20] spectra were acquired in the 10–14 p.p.m spectrum window using PRESATURATION [21] and WATERGATE [22] pulse sequences for H2O suppression Spectra were calibrated according to the characteristic methylene resonance of TFE-d2 at 3.926 p.p.m TOCSY experiments were carried out with a mixing time of 80–100 ms, whereas NOESY experiments were recorded using 100 and 200 ms mixing times All two-dimensional spectra were acquired using a spectral width of 6410.26 Hz (10.7 p.p.m) and consisted of 2048 data points in the F2 dimension, 16–32 transients and 720–1024 complex increments in the F1 dimension Raw data were multiplied in both dimensions by a pure cosine-squared bell window function and Fouriertransformed to obtain 2048 · 1024 or 2048 · 2048 real data points A polynomial base-line correction was applied in both directions NMR data processing was performed using the standard Bruker software package on a Silicon Graphics O2 workstation The two-dimensional maps were analysed on Silicon Graphics O2 or PentiumIII PC-Linux computers with the aid of the XEASY program (ETH, Zurich) [23] ă NOE constraints, 3JHNHa and CaH chemical shift analysis A total of 1868 NOESY cross-peaks were assigned in both dimensions and their intensities were converted into upper limits of interatomic distances [24], yielding 928 unique NOE constraints (23.1 per assigned residue) JHNHa were determined using the Protocol of D S Wishart et al [25] Twenty-one out of 28 measured J were calculated in the range of 4.8–5.6 Hz or > 8.0 Hz [26] and introduced to the calculations Seven 3JHNHa were found in the range 6–7 Hz and were not used in structural calculations [27] Slowly exchanging amides were identified using the lyophilized original sample in D2O/TFE-d3 solution, through two-dimensional NOESY spectra over a time period of 4–5 h Chemical shift analysis was performed according to the approach established by Wishart et al [28,29] and shift differences between the Ha-protons of the human/rat CRH analogue and their values for coil conformation were extracted No correction for the solvent was applied to the initial CaH shifts values drawn from the TOCSY/NOESY maps [30] Structure calculations and refinement Structure was calculated using the NMR data acquired at 310 K, the temperature at which native human/rat CRH solution structure was also determined Of 928 constraints, 656 were found meaningful (16.4 per assigned residue) in DYANA structural calculations [31] Appropriate pseudoatom corrections were applied to methylene and methyl hydrogens that were not stereospecifically assigned [32] a-Aminoisobutyrilo (Aib) residue is derived from a valine where the Ca of the Aib corresponds to valine Ca and Cb1/Cb2 of the Aib corresponds to valine Cc1/Cc2 Sixteen hydrogen bonds involving nonexchangeable HN amide protons, with an occurrence of 87.5% (in 35 out of 40 structures), were used as additional structural constraints in the final stage of DYANA calculation ˚ Upper (2.40 A) distance limits between HN and O atoms involved in hydrogen bonds, together with upper ˚ ˚ (3.30 A) and lower (2.60 A) distance limits between the corresponding N and O atoms, were set for each one of the 16 hydrogen bonds Fourteen pairs of diastereotopic protons (among them the two methyls of Aib15) were stereospecifically assigned through the GLOMSA program [24] Each one of the 40 best structures (out of 400 calculated) in terms of DYANA target function (larger NOE constraint Ó FEBS 2002 NMR structure of a CRH synthetic analogue in TFE/H2O (Eur J Biochem 269) 6011 ˚ violation 0.09 A) has been refined through REM (AMBER 5.0 [33], SANDER [34]) The whole procedure followed protocols previously applied to the solution structure determination of larger biomolecules, such as metallopro˚ teins [35] A force constant of 133.76 kJỈmol)1 A2 was applied for the distance constraints Structural calculations were performed on IBM RISC 6000 or PentiumIII PC-Linux computers Quality assessment of the structure PROCHECK [36] and PROCHECK-NMR [37] programs were used to check the stereochemical quality of the peptide structure and provide the elements of the secondary structure, as well as other statistical parameters RESULTS NMR spectra–proton assignment The spin patterns of 39 out of 41 residues have been identified through analysis of the TOCSY spectrum (Fig 1), whereas Ser1 and Leu27 proton resonances have not been assigned at this stage The latter have been identified in NOESY maps recorded at two different temperatures due to their sequential dipolar connectivities while the former have remained undetectable in all two-dimensional maps at both temperatures Proton chemical shifts are given in Table S1 Secondary structure Fig 2A shows the short and medium range NOEs observed for the backbone and CaH protons Sequential HN–HN connectivities for stretches up to two amino acids were observed for residues 6–13, 14–19, 20–28, 29–31, and 31–40 HN–HN connectivities involving the HN resonances of Ser7/ Thr11, Phe12/Leu14, Val18/Met21, Leu19/Glu20 and Gln26/Ala28/Gln30, could not be detected due to their degenerated shifts The same is true of some Ha–HN medium-range connectivities, where the Ha resonances of Ala28/Gln29 and Arg35/Lys36 have degenerated chemical shifts Eighteen 3JHNHa (Fig 2B, upper panel) were found to be £ 5.6 Hz while three (Glu2, Glu3 and Val18) were found to be in the range 8.0–8.5 Hz 3JHNHa for D-Phe12 Fig 1H-1H two-dimensional TOCSY 600MHz NMR of the fingerprint region of Ha–HN protons recorded on a 10.7-p.p.m spectrum width (H2O/TFE-d2 34%/66% v/v, at pH = 3.8, T = 310 K) The number of the amino acid in the CRH sequence, to which the Ha–HN connectivity belongs, is noted was found 6.9 Hz and six (Leu10, Thr11, Leu19, Gln26, Ile39 and Ile40) exhibited values in the range 6.0–6.6 Hz Chemical shift difference analysis between the observed CaH shift values and the corresponding random coil values is presented in Fig 2B (lower panel) and provides strong evidence for the conformational preference of the majority of amino acids towards the helical configuration The only tetrapeptide in which Ha values present large positive differences from random coil values is the N-terminal Glu2-Pro5/Ile6 region (Ile6 Dd Ha 0.06 p.p.m) A three-residue fragment, Ala31–Asn34, exhibited slightly positive shift differences (in the range Dd Ha 0.009–0.045 p.p.m) while the remaining 32 residues in the region Ser7–Ile41 exhibited negative Ha shift differences in the range 0.011–0.524 p.p.m According to data reported previously the conformation of that fragment seems to be different in ovine and human CRH, and for this reason its structure has been discussed in the literature [38] (Scheme 1) However, despite the positive Ha shift difference, Ala31–Asn34 Ha chemical shift values compared with the averaged Ha shifts for helical and b-sheet conformation [28] and they exhibited a smaller deviation from a-helical rather than b-strand values (see Fig 2C) Tertiary structure An overall evaluation of NOE, 3JHNHa and Ha chemical shifts suggests a helical conformation for the peptide fragment Ile6–Ile40 and an extended structure for the N-terminal hexapeptide Structure was calculated using NMR data acquired at 310 K (as in native CRH [15]) The resulting DYANA family of 40 structures has rmsd values ˚ (calculated for residues 6–41) of 0.56 ± 0.24 A and ˚ 1.40 ± 0.20 A for backbone and heavy atoms, respectively ˚ The target function lies in the range 0.10–0.12 A2 The final REM family exhibits pair-wise rmsd values of ˚ ˚ 0.56 ± 0.23 A (backbone), 1.42 ± 0.19 A (heavy atoms) ˚ for the 40 structures and 0.39 ± 0.16 A (backbone), ˚ (heavy atoms) for the mean structure The 0.99 ± 0.13 A NOE distance penalty function for the family of 40 ˚ ˚ structures is 0.42 ± 0.03 A2 and 0.44 A2 for the mean structure The corresponding values for H-bond and angle ˚ ˚ constraints are 0.016 ± 0.01 A2 (40 structures), 0.012 A2 ˚ ˚ (mean) and 0.014 ± 0.006 A2 (40 structures), 0.005 A2 (mean), respectively (see also Table 1) and in both cases the 6012 G A Spyroulias et al (Eur J Biochem 269) penalty function is around 30 times smaller than that of NOEs The NOE-derived distance constraints are reported in Fig 2D while rmsd per residue for the final family is given in Fig S1 Fig 3A shows the family of 40 structures and Table contains a summary of statistical data indicating the quality of the 40 models and the average energyminimized structure The polypeptide chain of the human/rat CRH analogue seems to fold to an almost linear tertiary structure with the main features being the high degree of helical character and a small bend along the principal molecular axis As far as the ill-defined N terminus is concerned, all data (chemical shift index, 3JHNHa, and u vs w plots, Fig S2) imply a conformational preference for an extended structure The unambiguous conformational preference for helical structure for the C-terminal 10-peptide is manifested by numerous helix-diagnostic NOEs (see also Fig 2A), and the JHNHa values Strong evidence for the Glu39–Ile41 tripeptide’s compact structure was obtained from the intense NOESY peaks of the two –CONH2 amide protons which, in turn, give rise to a network of HN–HN cross-peaks throughout in this region The numerous OiHN type iỵ H-bonds identied after the nal DYANA calculation (Ala28 O¢–HN His32, Gln29 O¢–HN Ser33, His32 O¢–HN Lys36, Ser33 O¢–HN Leu37, and Arg35 O¢–HN Glu39 in 40 out Ĩ FEBS 2002 of 40 DYANA models) further support the continuous C-terminal helix structure Monitoring the Phe12 fi D-Phe and Leu15 fi Aib substitution The NOE pattern of the CRH analogue involving HN, Ha and Hb protons of D-Phe12, was constituted only by dNN(i,i + 2) and daN(i,i + 2) connectivities, while connectivities diagnostic for a-helix such as daN(i,i + 3), dab(i,i + 3) and daN(i,i + 2) were not observed (Fig 2A) The Aib15 residue possesses two diastereotopic methyl groups, with the pro-R methyl replacing the Ha proton at Ca carbon of an L-valine and the pro-S group corresponding to the side chain (Cb of an L-valine) The protons of the two CbH3 groups were found to be chemically and magnetically nonequivalent and to yield two different resonances in the region 1.70–1.90 p.p.m Moreover, the stereospecific assignment of these methyl groups was achieved through analysis of the observed spin couplings and the intensities of dbN(i,i + 1), dbN(i,i + 2) and dbN(i,i + 3) NOEs Thus, the Cb1 methyl group which resonates at 1.861 p.p.m is assigned to the pro-S because it exhibits NOEs of low intensity in respect with Cb2 (1.716 p.p.m) methyl, of the type dbN(i,i + 2) and dbN(i,i + 3) [39] Fig Analysis of NMR-derived data (A) Schematic representation of the sequential and medium-range NOE connectivities involving NH, a- and b-protons The thickness of the bar indicates the intensity of NOEs The average, over the 40 DYANA structures, secondary structure elements are also reported (B) Measured values of 3JHNHa coupling constants in Hz (upper panel) and Ha chemical shift differences from random coil values (lower panel) (C) Comparison of Ha chemical shift values of [D-Phe12, Aib15]CRH with the corresponding Ha chemical shifts for helical ( ) and b-strand (m) configurations (D) Number of meaningful NOE constraints per residue for [D-Phe12, Aib15]CRH used in the structural calculations White, grey, and dark grey bars represent, respectively, intraresidue, sequential, and medium-range connectivities Long-range connectivities were not observed Ó FEBS 2002 NMR structure of a CRH synthetic analogue in TFE/H2O (Eur J Biochem 269) 6013 Fig (Continued) Distribution of electrostatic charges on the surface and along the principal axis of the molecule In the synthetic CRH analogue, as in native peptide, there are seven negatively charged (six Glu and one Asp) and four positively charged residues Analysis of the distribution of the electrostatic potential on the surface of the molecule reveals the ampiphilic character of the helix (Fig 4A) Schematic representation of the charge distribution along and perpendicularly to the [D-Phe12, Aib15]CRH axis is presented at Fig 4B and C Non-native interaction between charged groups in this synthetic analogue is discussed below DISCUSSION Analysis of the calculated structures performed by the PROCHECK program provides quality data for the calculated CRH models (Table 1) Neither the 40 models nor the mean structure possess a residue in disallowed regions of Ramachandran plot (Table and Fig S2) According to the PROCHECK-NMR [37] program, the average of the 40 models and the mean minimized structure of the synthetic analogue [D-Phe12, Aib15]CRH, contain up to 84% helical structure This value is either somewhat larger than that of the 76–78% regarding the helicity of the native peptide in solution containing 66–100% of TFE [15,40], or identical to the 84% for studies carried out in 40% TFE in an H2O mixture [38] According to studies of CRH binding to single bilayer egg phosphatidylcholine vesicles, the development of the intramolecular interaction with the cell membrane is accompanied by the induction of a a-helical structure and this conformation characterizes the biologically active form of the peptide [41] TFE assists amphipathic peptides with high helical propensity, such as CRH, to reach the maximum helical content at around 40% TFE [42], since it diminishes the exposure of CO and NH groups in the solvent [43], favouring hydrogen bond formation [44] This apparently occurred in the case of the currently studied CRH molecule and the same has been reported earlier for ovine and human/rat-CRH [38,41] In calculated models, the CRH a-helix consists of 34 residues in regions 6–18 and 20–40 and this is fully consistent with the diagram of sequential dipolar connectivities (Fig 2A), the 3Js and OiHN type H-bonds The structure iỵ of the Gln30Ile40 fragment has been determined with high resolution and exhibits remarkably low rmsd values ˚ (0.28–0.54 A) These data: (a) suggest a conformational difference with the structureless C terminus of ovine CRH [38] and (b) contradict the helix-stop signal identified for the native human/rat CRH in the residue fragment 31–34 and the ill-defined terminal pentapeptide 37–41 [15,38] Leu19 interrupts the helical structure in all structures composing the family of 40 models and also in the mean Ó FEBS 2002 6014 G A Spyroulias et al (Eur J Biochem 269) Table Statistical analysis for the REM a and ỈREMỉa structures of [D-Phe12, Aib15]CRH ÆREMæ REM ˚ RMS violations per experimental distance constraints (A) Intraresidue (229) Sequential (193) Medium-range (234) Long-range (0) Total (656) Mean number of violations per structure Intraresidue Sequential Medium-range Long-range Total ˚ Mean no of NOE violations > 0.3 A ˚ Largest residual NOE distance violation (A) ˚ Mean distance penalty function (A2) b Statistics of other structural constraints H-bond constraintsc (32) ˚ Mean distance penalty function (A2) / Constraints from 3JHNHa (21) Average torsion penalty function (kJỈmol)1) AMBER energy (kJỈmol)1) Structural analysis Residues in disallowed regions (%) Residues in generously allowed regions (%) Residues in allowed regions (%) Residues in most favorable regions (%) Overall G-factor 0.0184 0.0195 0.0274 0.0000 0.0224 ± ± ± ± ± 0.0015 0.0017 0.0017 0.0000 0.0011 9.78 10.18 14.65 00.00 34.60 0.000 0.245 0.400 ± ± ± ± ± ± 1.56 1.53 2.01 0.00 2.80 0.00 ± 0.03 0.0173 0.0192 0.0313 0.0000 0.0235 9.0 10.0 18.0 00.0 37.0 00.0 0.222 0.440 0.016 ± 0.010 0.012 0.014 ± 0.006 )1706.70 ± 111.6 0.005 )1768.52 0.0 0.2 9.4 90.4 )0.16 ± 0.04 0.0 0.0 13.5 86.5 )0.11 a REM indicates the energy-minimized family of 40 structures and ỈREMỉ the mean energy-minimized structure b Numbers in parenthesis indicate the number of meaningful upper distance limits per class c Two distance limits per each H-bond constraint (see Materials and methods) structure However, in various DYANA models, helical structure is also interrupted for two up to five residues in the 6/9–13/15 region Destabilization of the helix in fragment 9–13 could be discussed in terms of the introduction of a D-amino acid at position 12 [45] Destabilization caused by D-amino acid depends on the nature of its side chain (charged or not, bulky or not) and D-phenylalanine has been found to present one of the highest a-helix disturbing propensities [46] Loss of NOEs between D-Phe and sequential residues in the CRH analogue was observed as in a similar kind of L-/D-amino acid substitution [45,46], and suggests slight helix destabilization in the region of Leu10, Thr11 and D-Phe12 For this fragment, experimental coupling constant values are measured between and Hz (Fig 2B), suggesting conformational averaging over the contributing conformations [27] On the other hand, the helix-promoter/stabilizer Aib, which replaced Leu15, increases the helix propensity of a peptide fragment towards either a- or 310-helix by restriction of the backbone conformational freedom [47–49] This limitation is imposed by the steric hindrance that arises from its gem-dimethyl group The torsion angles for the mainchain conformation of Aib15 in the family of 40 energy minimized structures are calculated as u ¼ )59.6° ± 1.5 and w ¼ )31.9 ± 1.5, while for the mean energy minimized structure they are )56.5° and )31.0°, respectively, close to those of Aib conformational energy minima (u ¼ ) 58° and w ¼ )34° [50]) According to these values, the Aib u torsion angle deviates less from a- than from 310-helix typical values ()55.2° and )49.9°, respectively [48]) while the w torsion angle is practically the same as the ideal w-value ()34°) for 310 helical configurations ()52.2° for a- helix [48]) As far as the surrounding amino acids in the His13–Glu17 fragment are concerned, His13 and Glu17 w angles are found slightly out of the range of values for a-helix structure, while w angles for Leu14 and Arg16 are considerably close to the range of values for 310 helix Additionally, Leu14 possess a u torsion angle almost typical for the 310 helix In this region of the polypeptide chain, the secondary structure is characterized also by the OiHN hydrogen iỵ bonds, found in all 40 DYANA structures in His13–Arg16 and Leu14–Glu17, which are indicative of a 310-, rather than a pure a-helical structure Nevertheless, H-bonds Oi HN , iỵ typical in a-helices, are also observed in Leu14–Val18 and Aib15–Leu19 All of the above data imply that the His13–Leu19 backbone structure is probably a conformational average bearing features of a pure a-helix and 310-helix structure This is in agreement with the ÔambihelicityÕ behaviour pattern stated by Basu et al [51,52] according to which the a- and 310-helical forms are almost equivalent energetically when only one or two Aib residues replace other amino acids in a peptide sequence Furthermore, a noticeable feature of the [D-Phe12, Aib15]CRH molecule is a characteristic bend occurring Ó FEBS 2002 NMR structure of a CRH synthetic analogue in TFE/H2O (Eur J Biochem 269) 6015 Fig (A) The family of 40 structures calculated for the [D-Phe12, Aib15]CRH analogue and (B) the mean energy minimized [D-Phe12, Aib15]CRH structure and conformation of selected residues in the site where double substitution has been performed Figures were generated with the program MOLMOL [55] in the 12–16 region which is a striking difference from the native CRH molecule This feature is apparently related to the contemporary incorporation of a D- and an a,adialkylo amino acid in close proximity within the CRH sequence Besides its tendency to form constrained helices, Aib is also known for its ability to bend helical peptides [53] Analogous helical structure perturbation has been observed where either a sequential double replacement of D-amino acid [46] or a double incorporation of Aib residue in a tripeptide fragment [53] has been carried out These features in the CRH analogue are supported by numerous, well-defined NOE cross-peaks between the side-chains of the residues in the 12–22 region (see Fig 3B) Ha, Hb and Cc2 methyl protons of Thr11 give rise to NOE interactions with Aib15 methyls while a NOE signal was also observed between Thr11 Cc2 methyl and Aib15 HN proton Aib15 Cb1 methyl (1.861 p.p.m) participates in a NOE network involving the backbone or side-chain protons of the residues D-Phe12 (Ha, phenyl ring Hd and He), His13 (HN), Leu14 (HN) and Val18 (HN, Cc1 methyl) On the 6016 G A Spyroulias et al (Eur J Biochem 269) Ó FEBS 2002 Ó FEBS 2002 NMR structure of a CRH synthetic analogue in TFE/H2O (Eur J Biochem 269) 6017 other hand the Cb2 methyl (1.716 p.p.m) gives rise to intense NOE cross-peaks with practically all Val18 protons (HN, Ha, Hb, Cc1 and Cc2 methyls) Additionally, dipolar interactions between HN, Ha, Cc1 and Cc2 Val18 and HN and Cb Ala22 protons have been observed This distortion of the linear structure of human/rat CRH [15] in the region of Aib residue probably acts as a helix-breaking signal in the region of Val18–Leu19 (3JHNHa 8.0 and 6.0 Hz, respectively) justifying the nonhelical backbone conformation of Leu19 observed in all calculated models and average structures The question raised at this point is whether or not these NOEs are consistent with an extended helix such as that of the native CRH As no coordinates for the native CRH NMR structure were available, a direct comparison could not be performed However, we executed an additional structural calculation excluding the Aib15 NOE-derived constraints In the absence of Aib15 constraints, the fragment His13–Leu19 becomes linear and in some models the fragment His13–Glu17 possesses a pure a-helical conformation However, the extended CRH native, Ile6–Ile40 helix, is divided in this synthetic analogue, into two independent helical fragments, Ile6–Val18 and Glu20–Ile40, respectively These two 13-residue N- and 21-residue C-terminal fragments form an angle of 34.2° (Fig 3B) This bend, in concert with the His13–Aib15 3- and Aib15–Leu19 4-residue turns, help the D-Phe12 aromatic ring to approach the vicinal Aib methyl while Val18 main and side-chain approach the other Aib methyl group (Fig 3B) The average distance between the vicinal Val18 and Aib15 ˚ methyls is measured as 3.206 A in the mean structure, while ˚ the other methyl of Aib15 is found 3.817 and 3.903 A from b d the aliphatic C and the vicinal H aromatic proton of D-Phe12, respectively Assisted by this bend, D-Phe12, Aib15 and Val18 form a highly hydrophobic core in the middle of the CRH sequence, a region that is particularly important for receptor binding and activation [41,54] On the opposite side, the Ôexternal partÕ of this helical fragment, the extended side chains of Arg16 and Glu20, as well as those of His13 and Glu17, are oriented towards the same direction and are relatively close The Oe1 and Oe2 atoms of Glu20 are 3.726 ˚ and 1.983 A, respectively, from the imine He proton and ˚ 3.644 and 5.121 A from the proximal guanidyl Hg11 and Hg12 protons of Arg16 (Fig 3B) On the other hand, the ˚ Oe2 atom of Glu17 is found at 5.485 A, far from the Hd2 proton of the imidazole ring of His13 The above-mentioned side-chain interactions between His13–Glu17 and Arg16–Glu20 are remarkably different in the calculated models without Aib15 NOEs When the fragment His13–Leu19 is linear, the interatomic distances between Arg16 He, Hg11/Hg12 and Glu20 Oe1/Oe2 become ˚ on average larger than 5.0–6.0 A, whereas when the 310helix conformation becomes a pure a-helix, the His13– Fig (A) Surface electrostatic potentials of CRH analogue, displayed with MOLMOL [55] and schematic representation of the charge residues’ distribution along (B) and (C) perpendicularly to the axis of the molecule is also presented The molecular surface is colour coded: potential less than )10 kT are illustrated in red while potential larger than 10 kT in blue and neutral (0 kT) potential are white Glu17 atoms (Hd2 and Oe2, respectively) are found at ˚ distances larger than 6.5–7.0 A The above data could probably be indicative for a native-like linear CRH structure and suggest a conformational difference between the parent and the synthetic molecule CONCLUDING REMARKS – BIOLOGICAL IMPLICATIONS The NMR data acquired and analysed in 66% TFE suggest that the synthetic analogue [D-Phe12, Aib15]CRH adopts a rather helical structure (84% of residues), which is considered to be the active conformation of the hormone [38,40] The changes engineered in the sequence of the peptide, either through the nature of the side chain of the new amino acids or through their non-native conformational features, increase the peptide–CRH complex stability with the following consequences: they (a) differentiate the helical content and/or characteristics (a- or 310-helix); (b) alter the hydrophobicity of the molecule in the 16–20 region which is considered the ÔinterfaceÕ between the 6–19 region, where the side-chains of the hydrophobic part of the helix play a key role in receptor binding and activation [54], and the 20–41 region, where side-chains have a prominent role in structural conservation rather than in function [54]; and (c) disturb the linearity of the molecule in the region Ile6–Ile40 Helix bending is a structural feature that has not been observed in the solution structure of the native peptide human/rat CRH [15] and has not been implied by studies of other CRH sequences [38] Although the peptide bonds in the Aib15–Glu20 region could be accommodated without severely perturbing the helical character of both fragments, the helix breaking/bending effect is strongly coupled with reorientation of the residue side-chains in the amphipathic region 12–20 Indeed, helix bending diminishes the exposure of the hydrophobic core of Phe12, Aib15 and Val18 to the solvent, while establishing the structural basis for the side-chain interaction of the charged residues Arg16 and Glu20 The biological impact of the Aib15 and D-Phe12 modifications on the synthetic CRH analogue is illustrated by the peptide binding affinity enhancement towards the CRH receptor and its modified biological response More specifically, this CRH synthetic analogue seems to effectively bind to the receptor pocket while its biological activity is extinguished when its effect on the release of catecholamines from the PC12 rat pheochromocytoma cell line is compared with that of native CRH Consequently, these properties imposed on CRH comprise the first and possibly the most important achievement in engineering a pharmacologically interesting compound with properties antagonistic to CRH receptors Analysis of the non-native structural features provides valuable information towards the design and synthesis of new molecules with higher binding affinity and enhanced stability against biological degradation and possibly biological activity The impact of Aib introduction in the position of residues with similar or different hydrophobic character, in the amphipathic region Leu14–Glu20, was studied in silico before the synthesis of new analogues Conformational studies and biological activity evaluation of new CRH analogues are in progress in our laboratory Ó FEBS 2002 6018 G A Spyroulias et al (Eur J Biochem 269) ACKNOWLEDGEMENTS General Secretariat of Research and Technology of Greece (P.C) and Marie Curie Research Training Grant – Contract No HPMF-CT1999-00344 (G.A.S) are acknowledged for financial support EC’s Access to Research Infrastructures Action of the Improving Human Potential Program (PARABIO, Contract No HPRI-CT-1999-00009) and CERM (University of Florence) are also acknowledged for access to NMR instrumentation and to powerful computational systems 15 16 17 REFERENCES Vale, W., Spiess, J., Rivier, C & Rivier, J (1981) Characterization of a 41-residue ovine hypothalamic peptide that stimulates secretion of corticotropin and beta-endorphin Science 213, 1394– 1397 Vale, W., Rivier, C., Brown, M.R., Spiess, J., Koob, G & Swanson, A (1983) Chemical and biological characterization of corticotropin releasing factor Recent Prog Horm Res 39, 245– 270 Orth, D (1992) Corticotropin-releasing hormone in humans Endocr Rev 13, 164–191 De Souza, E.B., Insel, T.R., Perrin, M.H., Rivier, J., Vale, W.W & Kuhar, M.J (1985) Differential regulation of corticotropinreleasing factor receptors in anterior and intermediate lobes of pituitary and in brain following adrenalectomy in rats Neurosci Lett 56, 121–128 Perrin, M., Donaldson, C., Chen, R., Blount, A., Berggren, T., Bilezikjian, L., Sawchenko, P & Vale, W (1995) Identification of a second corticotropin-releasing factor receptor gene and characterization of a cDNA expressed in heart Proc Natl Acad Sci USA 92, 2969–2973 Kishimoto, T., Pearse, R.V II, Lin, C.R & Rosenfeld, M.G (1995) A sauvagine/corticotropin-releasing factor receptor expressed in heart and skeletal muscle Proc Natl Acad Sci USA 92, 1108–1112 Lovenberg, T.W., Liaw, C.W., Grigoriadis, D.E., Clevenger, W., Chalmers, D.T., De Souza, E.B & Oltersdorf, T (1995) Cloning and characterization of a functionally distinct corticotropinreleasing factor receptor subtype from rat brain Proc Natl Acad Sci USA 92, 836–840 Chen, R., Lewis, K.A., Perrin, M.H & Vale, W.W (1993) Expression cloning of a human corticotropin-releasing-factor receptor Proc Natl Acad Sci USA 90, 8967–8671 Karalis, K., Sano, H., Redwine, J., Listwark, S., Wilder, R.L & Chrousos, G.P (1991) Autocrine or paracrine inflammatory actions of corticotropin-releasing hormone in vivo Science 254, 421–423 10 Makrigiannakis, A., Margioris, A.N., Le Goascogne, C., Zoumakis, E., Nikas, G., Stournaras, E., Psychoyos, A & Gravanis, A (1995) Corticotropin-releasing hormone (CRH) is expressed at the implantation sites of early pregnant rat uterus Life Sci 57, 1869–1875 11 Zoumakis, E., Makrigiannakis, A., Margioris, A.N., Stournaras, E & Gravanis, A (1996) Corticotropin releasing hormone (CRH) in normal and pregnant uterus: physiological implications Front Biosci 1, e1–8 12 Owens, M.J & Nemeroff, C.B (1991) Physiology and pharmacology of corticotropin-releasing factor Pharmacol Rev 43, 425– 473 13 Gilligan, P.J., Robertson, D.W & Zaczek, R (2000) Corticotropin releasing factor (CRF) receptor modulators: progress and opportunities for new therapeutic agents J Med Chem 43, 1641– 1660 14 Papazacharias, S., Pairas, G., Kouimtzoglou, E., Dermitzaki, E., Manessi-Zoupa, E., Gravanis, A & Cordopatis, P (2001) Synthesis and biological evaluation of new h/rCRH peptide 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 analogs In Peptides 2000 (Martinez, J & Fehrentz, J.-A., eds), pp 933–934 EDK, Paris, France Romier, C., Bernassau, J.-M., Cambillau, C & Darbon, H (1993) Solution structure of human corticotropin releasing factor by 1H NMR and distance geometry with restrained molecular dynamics Protein Eng 6, 149–156 Piantini, U., Sorensen, O.W & Ernst, R.R (1982) Multiple quantum filters for elucidating NMR coupling networks J Am Chem Soc 104, 6800–6801 Braunschweiler, L & Ernst, R.R (1983) Coherence transfer by isotropic mixing: application to proton correlation spectroscopy J Magn Reson 53, 521–528 Bax, A & Davis, D.G (1985) MLEV-17-based two-dimensional homonuclear magnetization transfer spectroscopy J Magn Reson 65, 355–360 Marion, D & Wuthrich, K (1983) Application of phase sensitive ă two-dimensional correlated spectroscopy (COSY) for measurements of 1H-1H spin-spin coupling constants in proteins Biochem Biophys Res Commun 113, 967–974 Jeener, J., Meier, B.H., Bachmann, P & Ernst, R.R (1979) Investigation of exchange processes by two-dimensional NMR spectroscopy J Chem Phys 71, 4546–4553 Macura, S., Wuthrich, K & Ernst, R.R (1982) The relevance of ă J cross-peaks in two-dimensional NOE experiments of macromolecules J Magn Reson 47, 351–357 Piotto, M., Saudek, V & Sklenar, V (1992) Gradient-tailored excitation for single-quantum NMR spectroscopy of aqueous solution J Biomol NMR 2, 661–666 Eccles, C., Guntert, P., Billeter, M & Wuthrich, K (1991) Eă ă cient analysis of protein 2D NMR spectra using the software package EASY J Biomol NMR 1, 111–130 Guntert, P., Braun, W & Wuthrich, K (1991) Ecient compuă ă tation of three-dimensional protein structures in solution from nuclear magnetic resonance data using the program DIANA and the supporting programs CALIBA, HABAS and GLOMSA J Mol Biol 217, 517–530 Wang, Y., Nip, A.M & Wishart, D.S (1997) A simple method to quantitatively measure polypeptide JHNHa coupling constants from TOCSY or NOESY spectra J Biomol NMR 10, 373–382 Smith, L.J., Bolin, K.A., Schwalbe, H., MacArthur, M.W., Thorton, J.M & Dobson, C.M (1996) Analysis of main chain torsion angles in proteins: prediction of NMR coupling constants for native and random coil conformations J Mol Biol 255, 494–506 Wuthrich, K (1986) NMR of Proteins and Nucleic Acids Wiley, ă New York Wishart, D., Sykes, B & Richards, F (1991) Relationship between nuclear magnetic resonance chemical shift and protein secondary structure J Mol Biol 222, 311–333 Wishart, D., Sykes, B & Richards, F (1992) The chemical shift index: a fast and simple method for the assignment of protein secondary structure through NMR spectroscopy Biochemistry 31, 1647–1651 Merutka, G., Dyson, H.J & Wright, P.E (1995) ÔRandom coilÕ 1H chemical shifts obtained as a function of temperature and trifluoroethanol concentration for the peptide Series GGXGG J Biomol NMR 5, 14–24 Guntert, P., Mumenthaler, C & Wuthrich, K (1997) Torsion ă ă angle dynamics for NMR structure calculation with the new program DYANA J Mol Biol 273, 283–298 Wuthrich, K., Billeter, M & Brown, W (1995) Pseudo-structures ă for the 20 common amino acids for use in studies of protein conformations by measurements of intramolecular proton-proton distance constraints with nuclear magnetic resonance J Mol Biol 169, 949–961 Pearlman, D.A., Case, D.A., Caldwell, J.W., Ross, W.S., Cheatham, T.E., Ferguson, D.M., Seibel, G.L., Singh, U.C., Weiner, Ó FEBS 2002 34 35 36 37 38 39 40 41 42 43 44 45 46 NMR structure of a CRH synthetic analogue in TFE/H2O (Eur J Biochem 269) 6019 P.K & Kollman, P.A (1997) AMBER 5.0 University of California, San Francisco Pearlman, D.A., Case, D.A., Caldwell, G.C., Siebel, G.L., Singh, U.C., Weiner, P & Kollman, P.A (1991) AMBER 4.0 University of California, San Francisco Banci, L., Bertini, I., Luchinat, C & Turano, P (2000) Solution Structures of Hemoproteins In The Porphyrin Handbook (Kadish, K.M., Smith, K.M & R.Guilard, R., eds), pp 323–350 Academic Press, San Diego Laskowski, R.A., MacArthur, M.W., Moss, D.S & Thornton, J.M (1993) PROCHECK – a program to check the stereochemical quality of protein structures J Appl Crystallogr 26, 283–291 Laskowski, R.A., Rullmann, J.A.C., MacArthur, M.W., Kaptein, R & Thornton, J.M (1996) AQUA and PROCHECK-NMR: programs for checking the quality of protein structures solved by NMR J Biomol NMR 8, 477–486 Dathe, M., Fabian, H., Gast, K., Zirwer, D., Winter, R., Beyermann, M., Schumann, M & Bienert, M (1996) Conformational ă dierences of ovine and human Corticotropin releasing hormone Int J Peptide Protein Res 47, 383–393 Pallai, P.V., Mabilia, M., Goodman, M., Vale, W & Rivier, J (1983) Structural homology of corticotropin-releasing factor, sauvagine, and urotensin I: circular dichroism and prediction studies Proc Natl Acad Sci USA 80, 6770–6774 Goodman, M & Listowsky, I (1962) Conformational aspects of synthetic polypeptides VI Hypochromic spectral studies of oligoc-methyl-L-glutamate peptides J Am Chem Soc 84, 3770–3771 Lau, S.H., Rivier, J., Vale, W., Kaiser, E.T & Kezdy, F.J (1983) Surface properties of an amphiphilic peptide hormone and of its analog: corticotropin-releasing factor and sauvagine Proc Natl Acad Sci USA 80, 7070–7074 Luo, P & Baldwin, R (1997) Mechanism of helix induction by trifluoroethanol: a framework for extrapolating the helix-forming properties of peptides from trifluoroethanol/water mixtures back to water Biochemistry 36, 8413–8421 Cammers-Goodwin, A., Allent, T.J., Oslick, S.L., McClure, K.F., Lee, J.H & Kemp, D.S (1996) Mechanism of stabilization of helical conformations of polypeptides by water containing trifluoroethanol J Am Chem Soc 118, 3082–3090 Jaravine, V.A., Alexandrescu, A & Grzesiek, S (2001) Observation of the closing of individual hydrogen bonds during TFEinduced helix formation in a peptide Protein Sci 10, 943–950 Krauser, E., Bienert, M., Schmieder, P & Wenschuh, H (2000) The helix-destabilizing propensity scale of D-amino acids: The influence of side chain steric effects J Am Chem Soc 122, 4865– 4870 Rothemund, S., Beyermann, M., Krause, E., Krause, G., Bienert, M., Hodges, R.S., Sykes, B.D & Sonnichsen, F.D (1995) Strucă 47 48 49 50 51 52 53 54 55 ture effects of double D-amino acid replacements: a nuclear magnetic and circular dichroism study using amphipathic model helices Biochemistry 34, 12954–12962 Karle, I.L & Balaram, P (1990) Structural characteristics of alpha-helical peptide molecules containing Aib residues Biochemistry 29, 6747–6756 Toniolo, C & Benedetti, E (1991) Structures of polypeptides from a-amino acids disubstituted at the a-carbon Macromolecules 24, 4004–4009 Toniolo, C & Benedetti, E (1991) The polypeptide 310-helix Trends Biochem Sci 16, 350–353 Burgess, A.W (1994) Designing amino acids to determine the local conformations of peptides Proc Natl Acad Sci USA 91, 2649–2653 Basu, G., Bagghi, K & Kuki, A (1991) Conformational preference of oligopeptides rich in a-aminoisobutyric acid I Observation of a 310/a-helical transition upon sequence permutation Biopolymers 31, 1763–1774 Basu, G & Kuki, A (1992) Conformational preference of oligopeptides rich in a-aminoisobutyric acid II A model for the 310/ahelix transition with composition and sequence sensitivity Biopolymers 32, 61–71 Gibbs, N., Sessions, R.B., Williams, P.B & Dempsey, C.E (1997) Helix bending in alamethicin: molecular dynamics simulations and amide hydrogen exchange in methanol Biophys J 72, 2490– 2495 Kornreich, W.D., Galyean, R., Hernandez, J.F., Craig, A.G., Donaldson, C.J., Yamamoto, G., Rivier, C., Vale, W & Rivier, J (1992) Alanine series of ovine corticotropin releasing factor (oCRF): a structure-activity relationship study J Med Chem 35, 1870–1876 Koradi, R., Billeter, M & Wuthrich, K (1996) MOLMOL: a ă program for display and analysis of macromolecular structures J Mol Graph 14, 51–55 SUPPLEMENTARY MATERIAL The following material is available from http://www blackwell-science.com/products/journals/suppmat/EJB/ EJB3278/ EJB3278sm.htm Table S1 Chemical shifts (p.p.m) of the protons of the residues in the [D-Phe12, Aib15]CRH Table S2 Experimental NOESY cross-peak intensities and corresponding upper distance limits Fig S1 (A) Rmsd per residue for the 40 structures for backbone (black bars) and heavy atoms (grey bars), and (B) of the mean structure Fig S2 Data extracted from PROCHECK analysis  ... 2002 of 40 DYANA models) further support the continuous C-terminal helix structure Monitoring the Phe12 fi D-Phe and Leu15 fi Aib substitution The NOE pattern of the CRH analogue involving HN, Ha and. .. Arg16 and Glu20 The biological impact of the Aib1 5 and D -Phe12 modifications on the synthetic CRH analogue is illustrated by the peptide binding affinity enhancement towards the CRH receptor and. .. for the final family is given in Fig S1 Fig 3A shows the family of 40 structures and Table contains a summary of statistical data indicating the quality of the 40 models and the average energyminimized