Portal de Periódicos da CAPES

Conformational analysis and stability of collagen peptides by CD and by1H- and13C-NMR spectroscopies

2000; Wiley; Volume: 53; Issue: 1 Linguagem: Inglês

10.1002/(sici)1097-0282(200001)53

1097-0282

Roberto Consonni, Lucia Zetta, Renato Longhi, Lucio Toma, Giuseppe Zanaboni, Ruggero Tenni,

Meat and Animal Product Quality

BiopolymersVolume 53, Issue 1 p. 99-111 Free Access Conformational analysis and stability of collagen peptides by CD and by 1H- and 13C-NMR spectroscopies Roberto Consonni, Roberto Consonni Lab NMR, Istituto di Chimica delle Macromolecole, CNR, Via Ampère 56, 20131 Milano, ItalySearch for more papers by this authorLucia Zetta, Corresponding Author Lucia Zetta [email protected] Lab NMR, Istituto di Chimica delle Macromolecole, CNR, Via Ampère 56, 20131 Milano, ItalyLab NMR, Istituto di Chimica delle Macromolecole, CNR, Via Ampère 56, 20131 Milano, ItalySearch for more papers by this authorRenato Longhi, Renato Longhi Istituto di Biocatalisi e Riconoscimento Molecolare, CNR, Via M. Bianco 9, 20131 Milano, ItalySearch for more papers by this authorLucio Toma, Lucio Toma Dipartimento di Chimica Organica, via Taramelli 10, University of Pavia, 27100 Pavia, ItalySearch for more papers by this authorGiuseppe Zanaboni, Giuseppe Zanaboni Dipartimento di Biochimica “A.Castellani,” University of Pavia, Via Taramelli 3b, 27100 Pavia, ItalySearch for more papers by this authorRuggero Tenni, Ruggero Tenni Dipartimento di Biochimica “A.Castellani,” University of Pavia, Via Taramelli 3b, 27100 Pavia, ItalySearch for more papers by this author Roberto Consonni, Roberto Consonni Lab NMR, Istituto di Chimica delle Macromolecole, CNR, Via Ampère 56, 20131 Milano, ItalySearch for more papers by this authorLucia Zetta, Corresponding Author Lucia Zetta [email protected] Lab NMR, Istituto di Chimica delle Macromolecole, CNR, Via Ampère 56, 20131 Milano, ItalyLab NMR, Istituto di Chimica delle Macromolecole, CNR, Via Ampère 56, 20131 Milano, ItalySearch for more papers by this authorRenato Longhi, Renato Longhi Istituto di Biocatalisi e Riconoscimento Molecolare, CNR, Via M. Bianco 9, 20131 Milano, ItalySearch for more papers by this authorLucio Toma, Lucio Toma Dipartimento di Chimica Organica, via Taramelli 10, University of Pavia, 27100 Pavia, ItalySearch for more papers by this authorGiuseppe Zanaboni, Giuseppe Zanaboni Dipartimento di Biochimica “A.Castellani,” University of Pavia, Via Taramelli 3b, 27100 Pavia, ItalySearch for more papers by this authorRuggero Tenni, Ruggero Tenni Dipartimento di Biochimica “A.Castellani,” University of Pavia, Via Taramelli 3b, 27100 Pavia, ItalySearch for more papers by this author First published: 21 January 2000 https://doi.org/10.1002/(SICI)1097-0282(200001)53:1 3.0.CO;2-DCitations: 13AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinkedInRedditWechat Abstract Four small type I collagen CNBr peptides containing complete natural sequences were purified from bovine skin and investigated by CD and 1H- and 13C-nmr spectroscopies to obtain information concerning their conformation and thermal stability. CD showed that a triple helix was formed at 10°C in acidic aqueous solution by peptide αl(I) CB2 only, and to lesser extent, by α1(I) CB4, whereas peptides α1(I) CB5 and α2(I) CB2 remained unstructured. Analytical gel filtration confirmed that peptides α1(I) CB2 and α1(I) CB4 only were able to form trimeric species at temperature between 14 and 20°C, and indicated that the monomer = trimer equilibrium was influenced by the chaotropic nature of the salt present in the eluent, by its concentration, and by temperature variations. CD measurements at increasing temperatures showed that α1(I) CB2 was less stable than its synthetic counterpart due to incomplete prolyl hydroxylation of the preparation from the natural source. 1H- and 13C-nmr spectra acquired in the temperature range 0–47 and 0–27°C, respectively, indicated that with decreasing temperature the most abundant form of α1(I) CB2 was in slow exchange with an assembled form, characterized by broad lines, as expected for the triple-helical conformation. A large number of trimer cross peaks was observed both in the proton and carbon spectra, and these were most likely due to the nonequivalence of the environments of the three chains in the triple helix. This nonequivalence may have implications for the aggregation of collagen molecules and for collagen binding to other molecules. The thermal transition from trimer to monomer was also monitored by 1H-nmr following the change in area of the signal belonging to one of the two β protons of the C-terminal homoserine. The unfolding process was found to be fully reversible with a melting temperature of 13.4°C, in agreement with CD results. The qualitative superposition of the melting curves obtained by CD for the peptide bond characteristics and by nmr for a side chain suggests that triple-helical backbone and side chains constitute a single unit. © 2000 John Wiley & Sons, Inc. Biopoly 53: 99–111, 2000 INTRODUCTION The triple helix motif, which is characteristic of collagens, is present in only a few other proteins.1 Three polypeptide chains intertwine to form a rodlike triple-helical conformation, made possible by the repetition of tens or hundreds of triplets in the form (Gly–X–Y), which are characterized by the presence of a glycine every third residue. Short triple helices (5–8 triplets per chain) are formed only when more stabilizing amino acids are incorporated or if the three chains are covalently linked at one extremity.2-6 The physiologically relevant feature of the collagen triple helix is its propensity to self-associate, giving rise to supramolecular structures (fibrils, networks, etc.),7 and to associate with many ligands (see, e.g., Ref. 8). The correct functioning of the triple helix is affected by virtually any mutation, giving rise to inherited human connective tissue disorders, each being heterogeneous from a clinical, genetic, and biochemical point of view. Due to the lack of detailed knowledge of the binding patterns between collagens and other connective tissue macromolecules, the relationship between genotype and phenotype has still not been fully elucidated. The structure and function of the triple helical motif have been the subject of several recent investigations that have usually used very repetitive collagenlike imino acid-rich model peptides. The main results have concerned the influence of the nature and position of amino acids on their propensity toward the triple-helical conformation and stability (reviewed by Brodsky and Ramshaw9). Moreover, a detailed picture of the molecular and crystal structure and the highly ordered hydration shell was recently obtained from an x-ray study of a peptide mimicking single glycine substitutions found in many collagenopathies.10-12 In the present study we have further extended our results on type I collagen CNBr peptides as model peptides with fully natural sequences. Previous work showed that the solution of each peptide is a complex system for the molecular species peptides can form (unfolded species or monomers, trimers, and for some peptides, aggregates of trimers), in terms of conformation, thermodynamics, and kinetics of the equilibria between species.13 In particular, we studied the lower thermal stability of peptide trimers according to Privalov's analysis of collagen stability15 and concluded that it was due to a less extended/ordered hydration shell per unit length than the parent type I collagen.14 However, there is disagreement concerning the fact that Hyp also has a stabilizing effect in nonaqueous solvents2 or that 4-trans-fluoroproline has an even greater stabilizing effect than Hyp.16 These facts point to the importance that inductive effects of the hydroxyproline (Hyp) hydroxyl group may have on collagen stability. The results we present here concerning the four shortest (30–47 residues) CNBr peptides derived from type I collagen (Scheme 1) show that assembly into a triple-helical conformation is possible in an acidic aqueous solution only for those peptides with the highest imino acid content. Of all type I collagen CNBr peptides, peptide α1(I) CB2 was found to be the most stable per unit length. This peptide was also found to be suitable for detailed nmr studies in both its monomeric and trimeric conformation. Scheme 1Open in figure viewerPowerPoint Nonstandard symbols: O, hydroxyproline; J, hydroxylysine; M*, homoserine/homoserine lactose. Sequences are from Refs. 42 and 48. Previous x-ray and nmr studies on collagen and peptides were recently reviewed by Mayo.17 In particular, Torchia et al.,18 by means of a 13C-nmr study on peptide α1(I) CB2, found a high degree of backbone mobility for the unstructured monomeric peptide, due to segmental motion; the assembled trimeric form showed much slower motions, associated with local mobility of some side-chain carbons with the relevant exception of Hyp residues. MATERIALS AND METHODS Materials Column PEP-RPC HR5/5 (1 mL bed volume) was from Pharmacia. All reagents were high performance liquid chromatography (HPLC) or analytical grade. N-Fmoc-O-t-butyl-L-trans-4-hydroxyproline, N-Fmoc-amino acid derivatives, reagents, and resin for peptide synthesis were from Novabiochem. Preparation of the Small CNBr Peptides from Type I Collagen Type I collagen CNBr peptides α1(I) CB2 (36-mer, residues 4–39 of the triple-helical domain), α1(I) CB4 (47-mer, residues 40–86), α1(I) CB5 (37-mer, residues 87–123), and α2(I) CB2 (30-mer, residues 328–357) were purified from acid-soluble type I collagen using a combination of molecular-sieve chromatography followed by reverse-phase chromatography. Preparation of type I collagen from calf skin, its cleavage with cyanogen bromide, and the gel filtration step on Bio Gel A 0.5m were performed as previously described.13 On Bio Gel A 0.5m, the small peptides eluted as three, mostly overlapping peaks with a Kav of about 0.71–0.79. The order of elution was in agreement with the above reported size. The peaks were collected separately, desalted in 0.1M acetic acid by ultrafiltration (1000 MW cutoff), and freeze-dried. A fourth peak (Kav = 0.85), which was better resolved, was found to contain the composite peptide CB1 bound to N-telopeptide, with only some molecules showing lysine on amino acid analysis (data not shown). Each preparation was subjected to reverse-phase high performance liquid chromatography (RP-HPLC) on a PEP-RPC HR5/5 column, essentially following the manufacturer's guidelines. To explain briefly: two solvents were used: 0.1% (v/v) trifluoroacetic acid (TFA) in water (solution A) and 0.1% (v/v) TFA in acetonitrile (solution B). The column was run at 0.5 mL/min and equilibrated in solution A. The samples were dissolved in the same solution at a concentration of 1–2 mg/mL, denatured for 2 min at 50°C, clarified by centrifugation and immediately injected. A 5-min wash with solution A was followed by a steep 5-min gradient from 0 to 10% B. Elution was followed at 214 nm and was achieved by a 50-min gradient from 10 to 20 or 25% B, depending on the peptide. Small variations between runs were performed in order to optimize the separations. All peaks were collected separately and freeze- dried. The samples were analyzed using several methods: hydroxyproline content per mg of solid was determined according to Huszar et al.19; amino acid analysis; capillary electrophoresis with the micellar electrokinetic chromatographic procedure, essentially as described by Zanaboni et al.20 These analyses showed that the order of hydrophobicity of peptides was α2(I) CB2 < α1(I) CB5 < α1(I) CB4 < α1(I) CB2, and that heterogeneity existed for all peptides derived from α1(I) chain, since they eluted from the PEP-RPC HR5/5 column as two peaks (CB4 and CB5) or three peaks (α1(I) CB2). One reason for this behavior could be that the presence of two forms of the C-terminal homoserine are in equilibrium at the acidic pH used, i.e., the undissociated COOH form and its lactone. In particular in the case of α1(I) CB2, we also found that the Pro/Hyp molar ratio in the three peaks increased from ∼1.1 to ∼1.4 to ∼2.0 with increasing hydrophobicity. Homologous fractions were pooled. In the case of CB2, the two initial peaks only were pooled; the more hydrophobic peak was eliminated as its amount was low. The sequence of the 6–10 N-terminal residues gave a single sequence and an estimated purity of ≥95% for all peptides. Matrix-assisted laser desorption mass spectrometry (MALDI-TOF/MS) on a Voyager-RP Biospectrometry Workstation (PerSeptive Biosystems, Framingham MA) on the CB2 pool confirmed the heterogeneity by giving several signals, the two major ones with a mass of 3298.0 and 3280.0 Da (calculated 3312.5 and 3294.5 Da for the peptide with C-terminal homoserine and its lactone, respectively). Therefore, the difference between determined and calculated mass is about the mass of an oxygen atom. CB2-MG Synthesis, Purification, and Characterization The 37-mer amidated peptide was assembled onto a Rink Amide MBHA resin by solid-phase method on an automated Applied Biosystem Model 433A peptide synthesizer, using N-(9-fluorenylmethoxycarbonyl)-protected amino acids and 2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium tetrafluoborate) as activating reagent. The synthesis proceeded without difficulties during all 36 coupling cycles. After completion of the synthesis, the peptide was deblocked and cleaved from the resin with mixture K containing 2% of triisopropylsilane.21 The peptide was purified (homogeneity > 95%) employing semipreparative RP-HPLC using a gradient of acetonitrile in water containing 0.1% TFA. Purity was verified by analytical RP-HPLC in the same conditions on a Whatman PartiSphere C-18 column. The observed mass of the peptide, analyzed by MALDI-TOF/MS performed as reported above, was 3399.22 0.5 Da (calculated 3398.6 Da, average isotope composition). CD Analysis CD spectroscopy and the determination of thermal stability through CD measurements in the temperature range 0.5–58°C were performed as described earlier.13 NMR Analysis 13C-nmr spectra were acquired at 67.9 MHz on a Bruker AM-270 spectrometer, by using a 28.7 mM sample of CB2-MG in 50 mM tetradeuteroacetic acid, at pH 2.9 and at different temperatures in the range 0–27°C. The 13C-nmr spectra were obtained by proton decoupling using a standard pulse sequence and the following parameters: spectral width 20 kHz; recycle delay 32 s; from 2500 to 5000 transients were recorded by using 64K data points, with a spectral resolution of 0.6 Hz; a 1.2 Hz line-broadening factor was applied before Fourier transformation. 1H-nmr spectra were acquired on a Bruker DMX-500 spectrometer, by using CB2 (6 mM), CB2-MG (28.7 mM), and CB4 (2.6 mM) in 50 mM tetradeuteroacetic acid, pH 2.9, and at different temperatures in the range 0–47°C. Bidimensional nmr experiments, total correlated spectroscopy (TOCSY)22, 23 and nuclear Overhauser effect spectroscopy (NOESY)24, 25, were performed at 13.5 and 23°C. The two-dimensional (2D) spectra were acquired over 4K data points and 512 t1 increments in the absorption mode with time-proportional phase increments26 for quadrature detection in the t1 dimension. Water saturation was achieved either by low-power irradiation during the relaxation delay introduced between scans or using gradients.27 A total of 16 transients were collected for each t1 increment. Mixing times of 90 and 300 ms were employed for TOCSY and NOESY experiments, respectively. 2D experiments were processed on a SGI INDY workstation using the XWINMR program provided by Bruker. The data set was resolution enhanced using a π/2 shifted sine-bell function prior to Fourier transformation and zero filling in F1. For both 1H- and 13C-nmr spectra chemical shifts were referred to sodium trimethylsilyl(2,2,3,3,-2H4)sulfonate. The curve fitting of the area variations as a function of temperature was performed with a sigmoid logistic function: the inflection point represents the midpoint temperature of the monomer/trimer conformational transition. Other Procedures Analytical gel filtration in nondenaturing conditions was performed as earlier described13 without column thermostating and at room temperature (13.5–18.5°C). The eluent contained 50 mM acetic acid and one of the following salts: 50 mM NaCl, 50 mM Na2SO4, 50 mM NaH2PO4, 166 mM Na2SO4, and 800 mM (NH4)2SO4. The samples were preequilibrated in 0.1M acetic acid plus the salt for at least 1 h before the analysis. RESULTS CD Analysis Acid-soluble type I collagen from calf skin was the source of four small CNBr peptides containing between 30 and 47 residues: peptides CB2, CB4, and CB5 derived from the α1(I) chain and CB2 was from the α2(I) chain. For simplicity, the parent chain will be reported only in the case of α2(I) CB2. A 37-mer synthetic analogue of α1(I) CB2, differing only in the C-terminal sequence that contained Met–Gly instead of homoserine and encompassing residues 4–40 was also prepared and named CB2-MG. This synthetic peptide was synthesized to have a sample that contained the natural sequence but that could be easily converted to the peptide prepared from skin collagen, which was not heterogeneous unlike the natural peptide (see Materials and Methods); we also needed enough peptide to enable 13C-nmr spectroscopy. CD spectroscopy performed at 10°C in 0.1M acetic acid showed that trimers were formed by peptide CB2 and, to a lesser extent, CB4 only, whereas peptides CB5 and α2(I) CB2 remained unstructured (Figure 1A). The unique CD spectrum of collagen is characterized by a small positive band at 221 nm and a large negative band at 197 nm.28, 29 The CD spectrum of CB2 at 10°C at a concentration of 66 μg/mL in 0.1M acetic acid exhibited a positive peak centered at 222 nm with associated molar ellipticity close to 1270 deg cm2 dmol−1 and a more pronounced negative peak at 198 nm (−26550 deg cm2 dmol−1). The whole spectrum was similar to the spectrum of native type I collagen with differences only in the intensity of the signal, in line with results from previous investigations.30, 31 The same was valid for CB2-MG, which showed a slightly greater molar ellipticity. Helicity for these small peptides was, however, lower than helicity of larger type I collagen CNBr peptides.13 The CD molar ellipticity at 221 nm was used as a function of temperature to monitor the thermal denaturation of the triple helical conformation of CB2. At low temperatures only the trimer was present in significant concentrations, whereas at temperatures higher than 20°C only the monomer state was populated. The melting temperature was found to be about 15°C for CB2, and about 2°C higher for its synthetic counterpart, CB2-MG (Figure 1B). The thermodynamic parameters characterizing the denaturation profile of CB2 and CB2-MG (Table I) were derived from classical van't Hoff treatment of the denaturation curve. Figure 1Open in figure viewerPowerPoint (A) CD spectra. Peptides were dissolved in 0.1M acetic acid and equilibrated for ≥7 days at 4°C before the analysis. CD spectra were recorded at 10°C. The inset is an enlargement of the positive peak centered at 221 nm. Under the same conditions, acid-soluble type I collagen shows at 221 nm a mean residue ellipticity of 7873 (±6%) deg cm2 dmol−1 at 221 nm.13 (B) Melting profiles. The thermal stability of peptides was followed by CD measurements at 221 nm.13 The peptides were dissolved in 0.1M acetic acid and equilibrated for ≥7 days at 4°C prior to analysis. Table I. Thermodynamic Parameters for Peptides α1(I) CB2 and Its Synthetic Counterpart CB2-MGa Peptide Tm (°C) ΔH° (kcal mol−1) ΔS° (kcal K−1 mol−1) ΔG° (kcal mol−1) CD measurements α1(I) CB2b 14.6 −84 −0.25 −7.6 CB2-MG 16.6 −68 −0.22 −8.5 NMR measurements on α1(I) CB2c Signal of homoserine 13.4 −69 −0.22 −3.4 a Melting temperatures refer to trimers; other changes are formally computed for the monomer to trimer transition. b Mean of two determinations. The thermodynamic values from CD measurements were obtained as described by Engel et al.2 These authors have stated that the uncertainty of ΔH° and ΔS° is about 15% and of ΔG° is about 5%. We estimate a larger uncertainty on the small peptides we analyzed, also because of their low Tm. c In order to compute these data, a normalization step was performed by dividing the area of nmr signal shown by homoserine at about 2.4 ppm in trimeric species only, by the area of the signals for leucine δ, δ′ protons at 0.8 ppm. The area of these δ proton signals remained constant during the thermal transition (see text). In contrast, the tendency for CB4 to form a triple helix in 0.1M acetic acid was very low. CD spectra demonstrated that this peptide was barely triple helical at 10°C at concentrations of 65–800 μg/mL in 0.1M acetic acid (Figure 1A), and no clear transition was observed in the temperature range 4–20°C (Figure 1B). Analytical gel filtration analyses in acidic conditions in the presence of different salts confirmed that only CB2 and CB4 were able to form trimeric species. The monomer = trimer equilibrium is influenced by the type of salt present in the eluent, by its concentration, and by variations of temperature in the range 14–20°C. For example, the percentage of trimers was as high as 55% for CB2 and 35% for CB4, at 14.5°C in the presence of 0.8M ammonium sulfate (Figure 2). It should be noted that the results of the analyses reported in Figure 2 are purely indicative, because the temperatures of analysis could be within the denaturation range of the peptide and thus any difference in temperature could greatly influence the percentage of the trimeric species. Figure 2Open in figure viewerPowerPoint Nondenaturating analytical gel filtration. The analysis was performed on the four small peptides and on α1(I) CB2,4 for comparison, with a previously described procedure13 on a Superose 12 column. The eluent contained 50 mM acetic acid and 50 mM NaCl (first row); 50 mm Na2SO4 (second row); 800 mM (NH4)2SO4 (third row). Room temperature (within 0.5°C) during the run is indicated for each chromatogram. Legend: t, trimeric species; m, monomeric species, asterisk, sample solvent. Where feasible, the percentage of trimers is indicated (arrow). All these results demonstrated that peptide CB2 was suitable for nmr investigations in either the unassembled or the trimeric form: the percentage of trimers could be increased by operating at low temperatures and the complexity and repetitivity of trimeric species were within the range covered by nmr. NMR Spectroscopy At 30°C two sets of nmr resonances could be observed for CB2. The major set contained all trans Gly–Pro and X–Hyp peptide bonds, while the minor set (about 5%) most likely represented conformations in which the above mentioned peptide bonds were cis (Figure 3).32 No investigations have been done to further characterize this minor component, due to its low amount. Figure 3Open in figure viewerPowerPoint The 500 MHz 1H-nmr spectra of 6 mM 1(I) CB2 dissolved in 50 mM tetradeuteroacetic acid, at pH 2.9 and different temperatures. Asterisks indicate the signal characteristic of the triple helix which belongs to one of the two β protons of the C-terminal homoserine. Filled triangles indicate minor conformations (about 5%) in which Gly–Pro and X–Hyp peptide bonds are cis. Two-dimensional nmr spectra performed at 23 and 13.5°C gave the sequence-specific assignments for the residues of CB2 in its monomeric form: all linear amino acids (except for serines) and few glycines were assigned. In contrast, no prolines or hydroxyprolines were specifically assigned because of the significant signal overlap. Only fragment Q18–G19–F20–Q21–G22, devoid of Pro or Hyp, was sequentially assigned (Figure 4). The C-terminal homoserine was assigned by comparison with the spectrum of the free amino acid in acidic conditions and on the basis of scalar connectivities observed in the TOCSY spectrum. By titrating a strongly basic solution of homoserine with hydrochloric acid, it was possible to observe the simultaneous presence of the linear and lactonic forms of the amino acid (not shown). At acidic conditions the cyclic lactonic form prevailed. This finding suggested that in the acidic conditions used the C-terminal homoserine of CB2 occurs mainly in its lactonic form. Figure 4Open in figure viewerPowerPoint The 500 MHz 2D spectra of 6 mM α1(I) CB2 dissolved in 50 mM tetradeuteroacetic acid, at pH 2.9: 2D TOCSY at (A) 13.5°C and (B) 23°C; (C) 2D NOESY at 13.5°C. Arrows in A and B indicate the signals that belong to the C-terminal homoserine in the trimeric species only. Arrows in C indicate the sequence assignment of fragment Q18–G22. In the temperature range 0–47°C the most abundant form of CB2 (Figure 3) was found to be in slow exchange with an assembled form, characterized by broader lines, as expected for the triple helical conformation. By comparing the TOCSY spectra acquired at 13.5°C (Figure 4A) and 23°C (Figure 4B), several residues exhibited an increasing amount of trimer cross peaks at the lower temperature. In particular, a large number of trimer cross peaks (probably six) were observed for the two alanines: it is likely that the nonequivalent environment surrounding each alanine gives rise to small differences in the chemical shift of the alanyl amide protons in the three chains of the trimer, in agreement with the observations by Fan et al.33 on an 15N-enriched peptide. In addition to the alanyl residues, the chemical shift of other amino acids are also sensitive to the nonequivalence of the environments of the three chains, especially those of Leu-8, Arg-6, and of the glutamine and glutamic acid residues, as indicated in the fingerprint region of the TOCSY spectrum (Figure 4B). In the α region, splittings of the glycyl signals were difficult to be detected because of a general broadening. Interestingly, 13C-nmr spectra acquired at different temperatures in the range 0–27°C (Figure 5), also showed an equilibrium between two slowly exchanging forms, particularly manifest in the aliphatic region between 20 and 30 ppm, where the methyl carbons of Ala-14, Ala-32, and Leu-8 resonate, according to Torchia assignments.18 At 27°C, the four methyl groups show, as expected, four signals, one for each carbon (Figure 5A); these signals strongly decrease in intensity with decreasing temperature, but they are still present at 0°C (Figure 5C). Below 27°C, as observed in the proton spectra, a new form appears characterized by more than one peak for each methyl carbon, particularly well resolved in the higher field methyl resonance of Leu-8 (Figure 5B). Figure 5Open in figure viewerPowerPoint The 67.9 MHz 13C- nmr spectra of 28 mM CB2-MG dissolved in 50 mM tetradeuteroacetic acid at pH 2.9: methyl (left) and aromatic (right) at (A) 27°C, (B) 16°C, and (C) 0°C. At lower temperatures, the proton spectrum of CB2 showed an additional broad resonance at about 2.4 ppm (Figure 3), well isolated, not overlapping with other peaks. In the TOCSY spectrum acquired at 13.5°C, this peak showed the peculiar correlation of a proton belonging to a spin system characteristic of the homoserine residue (not reported). The homoserine peak at about 2.4 ppm deserved particular attention since it was well resolved and not overlapping with any other resonance in the whole range of temperatures from 0 to 47°C. It was thus used to monitor the melting transition of the triple-helical conformation. In recent investigations on collagen model peptides, proton nmr spectroscopy has shown that the side chains of individual residues are sensitive to triple-helical environments.32, 53-55 By comparing the nmr spectra of CB2 (Figure 6B) and CB4 (Figure 6C) in 50 mM perdeuteroacetic acid at pH 2.9 and 13.5°C, close to the melting temperature of CB2, only a very low amount of signal was observed for CB4 at about 2.4 ppm relative to one of the homoserine β-protons in the trimer conformation, suggesting a melting point transition value lower than 0°C. This feature prevented for CB4 a complete nmr study of the triple-helical form in aqueous solution. Figure 6Open in figure viewerPowerPoint The 500 MHz 1H-nmr spectra of (A) 28.7 mM synthetic CB2-MG, (B) 6 mM α1(I) CB2, and (C) 2.6 mM α1(I) CB4 dissolved in 50 mM tetradeuteroacetic acid, at pH 2.9 and 13.5°C. The synthetic peptide CB2-MG, devoid of homoserine and comprising the C-terminal segment MG, present in the natural sequence of type I collagen α1(I) chain, was also analyzed. By comparing the one-dimensional spectra of CB2 with those of the synthetic peptide at variable temperatures from 0 to about 50°C, the resonance at 2.4 ppm (Figure 6A), characteristic of a homoserine β proton in the trimeric form, was substituted by two well-separated resonances belonging to the methionine protons. As observed for CB2, as well for the synthetic peptide, the area of these peaks decreased to zero upon heating. Thermal Transition and Thermodynamic Parameters The determination of the thermal stability by CD at acidic pH was possible only for CB2 and its synthetic 37-mer analogue (Figure 1B). The thermodynamic data for the stability of the trimers of the two preparations were determined within the framework of the two state model for trimer = monomer15 and with the mathematical procedure of Engel et al.2 The re