Structure of the Model Peptides of Bombyx m ori Silk-Elastin Like

Structure of the Model Peptides of Bombyx mori Silk-Elastin Like Protein ... Citation data is made available by participants in Crossref's Cited-by Li...
0 downloads 0 Views 182KB Size
Download PDF
Biomacromolecules 2004, 5, 744-750

744

Structure of the Model Peptides of Bombyx mori Silk-Elastin Like Protein Studied with Solid State NMR Kosuke Ohgo,† Tracie L. Kurano,‡ Kristin K. Kumashiro,‡ and Tetsuo Asakura*,† Department of Biotechnology, Tokyo University of Agriculture and Technology, Koganei, Tokyo 184-8588, Japan; and Department of Chemistry, University of Hawaii at Manoa, 2545 McCarthy Mall, Honolulu, Hawaii 96822 Received September 13, 2003; Revised Manuscript Received December 8, 2003

The peptides (AG)6(VPGVG)(AG)7 and (AG)5(VPGVG)2(AG)5 are models for a new type of protein with both composition and properties such as Bombyx mori silk and elastin. In this paper, we report the solidstate NMR results for these samples and related peptides; the structures after dialysis of the 9 M LiBr aqueous solution and after treatment with formic acid were determined and compared. The detailed structural analyses were performed using deconvolution subroutines assuming Gaussian line shapes for the Ala Cβ peaks of the (AG)n sequences in these peptides. The peptide (AG)6(VPGVG)(AG)7 took the silk II structure after the dialysis, which is in contrast to the silk I form of (AG)15 after the same treatment. However, a drastic structural change of the (AG)n sequences was observed for (AG)5(VPGVG)2(AG)5; the fraction of distorted β-turn was 81% after the dialysis, but the distorted β-sheet became dominant (84%) after treatment with formic acid. The local structures of the Gly residue of the VG units in the elastin-like subunits, (VPGVG) and (VPGVG)2, were the distorted structures with a distribution of the torsion angles, which was derived from the 2D spin diffusion NMR spectral pattern of (AG)5VPG[1-13C]V[1-13C]GVPGVG(AG)5. Observation of this distribution of the Gly residue was independent of the treatment, dialysis or formic acid. Introduction The synthesis of new peptides with various combinations of specific sequence motifs from fibrous proteins are expected to yield variations in the fiber’s properties and may provide an opportunity to design unique biomaterials. For example, new biopolymers have been designed on the basis of the structural and functional information of silk fibroins and other fibrous proteins.1 Cappello et al.2 and later our group3 produced Bombyx mori silk-elastin like polymers (SELPs) with tandem repeats of B. mori silk-like (GAGAGS) and elastin-like (VPGVG) peptide blocks using genetic engineering methods. These SELP copolymers self-assemble to form gels through an irreversible exothermic event. Megeed et al.4 showed the potential of these polymers for the localized, controlled delivery of low-molecular-weight and macromolecular drugs. Systematic studies have shown that the gelation, biodegradation, swelling, hydration, temperature-sensitivity, and transport of solutes can be controlled through rational design of the polymer sequence at the genetic level, as well as with the control of the composition, concentration, and gelation time of the polymers. Thus, these copolymers are interesting biomaterials, and it will be important to study the detailed structure in the molecular design of these copolymers. The two well-known isoforms of the sequence, (GAGAGS)n, of B. mori silk fibroin are known as silk I (structure in the solid state before spinning) and silk II (structure in the * To whom correspondence should be addressed. Phone & Fax: (+81)423-83-7733. E-mail, [email protected]. † Tokyo University of Agriculture and Technology. ‡ University of Hawaii at Manoa.

solid state after spinning). The insoluble fraction after chymotripsin cleavage of the silk fibroin, Cp fraction5,6 and the alternating copolymer (Ala-Gly)15 adopt the silk I structure when the samples have been dissolved in 9 M LiBr and then dialyzed against water. The conformational characteristics of the silk I form have recently been identified as a repeated type II β-turn structure, using solid-state NMR methods; the torsion angles were Ala(φ, ψ) ) (-60°, 130°) and Gly(φ, ψ) ) (70°, 30°), and the turn was stabilized by intramolecular hydrogen bond formation.7,8 On the other hand, the silk II structure has been characterized to be heterogeneous in structure, a mixture of two antiparallel β-sheet chains with different relative orientations with distorted β-turn regions.9,10 Thus, our earlier work has established the feasibility and power of solid-state NMRbased approaches for the characterization of these two very different structures in the silk I and silk II peptides. In contrast to the silk peptides, less is known about the structure of elastin-based peptides. The principal protein component of the elastic fiber is elastin, and it is found in abundance in ligaments, arteries, skin, and lung tissue. Elastin is assembled from its soluble precursor, tropoelastin.11,12 One of the most prominent amino acid sequences of tropoelastin is (VPGVG)n, where n is 10 in porcine13 and bovine14 tropoelastins. Upon raising the temperature, solutions of synthetic high-molecular-weight poly(VPGVG) self-assemble into filamentous and fibrous structures.15,16 Subsequently, cross-linked poly(VPGVG) was shown to be an entropic elastomer.17 Urry’s extensive studies on the polypentapeptide poly(VPGVG) have been used to propose the β-spiral, a structure comprised of a regular arrangement of type II

10.1021/bm034355x CCC: $27.50 © 2004 American Chemical Society Published on Web 01/20/2004

Structure of B. mori Silk-Elastin Like Protein

β-turns.18-21 However, recent studies on this polymer in solid-state indicate other structural models may be more likely. Rodriguez-Cabello et al.22 suggested an amorphous structure with no definite conformation by vibrational Raman analysis. We reported that there is no indication of β-spiral structure for the Val14, Gly15, and Val16 residues in the model peptide (VPGVG)6 with solid-state NMR.23 As an additional consideration, Tamburro has hypothesized that the elastin structure includes a “conformational ensemble” or a population distribution of conformers instead of a regular or highly ordered structure. Indeed, recent solid-state NMR data on a related elastin polypentapeptide, poly(LGGVG)n, provide strong support for this scenario.24 In this paper, we synthesized two silk-elastin like peptides, (AG)6(VPGVG)(AG)7 and (AG)5(VPGVG)2(AG)5, together with the silk-like peptides, (AG)6, (AG)9, (AG)12, and (AG)15, and characterized their structural features with solid-state editing methods and 2D spin-diffusion NMR under offmagic-angle-spinning (OMAS). The line widths under appropriate conditions are a marker for homogeneity of the chemical environment(s) of those residues, and the 2D spindiffusion experiments provide geometric constraints, rather than just stating the conditions of the experiments. In the case of (AG)15, a silk I structure was formed after dialysis of the 9 M LiBr aqueous solution and a silk II structure was formed after treatment with formic acid. The influence of these treatments on the local structure of the model peptides of B.mori silk-elastin like protein was also studied. Finally, the 13C doubly label peptide (AG)5VPG[1-13C]V[1-13C]GVPGVG(AG)5 was also synthesized for targeted structural studies of the Gly15 residue with 2D spin diffusion NMR. Experimental Section Peptide Synthesis. The peptides, (AG)6(VPGVG)(AG)7, (AG)5(VPGVG)2(AG)5, (AG)6, (AG)9, (AG)12, and (AG)15, were synthesized by solid-phase Fmoc chemical synthesis using a Pioneer peptide synthesizer (Applied Biosystems). The doubly-13C-labeled peptide, (AG)5VPG[1-13C]V[1-13C]GVPGVG(AG)5 was prepared for 2D spin-diffusion NMR experiments.7 The crude peptides from the TFA cleavage were dissolved in 9 M LiBr and then dialyzed against water for 3 days. The precipitated peptide, (AG)6(VPGVG)(AG)7, was centrifuged and dried. However, the peptide (AG)5(VPGVG)2(AG)5 was soluble in water after dialysis, and then the peptide was lyophilized. The two peptides (AG)6(VPGVG)(AG)7 and (AG)5(VPGVG)2(AG)5 were dissolved in formic acid and then dried by evaporation for NMR experiments.9,10 13C CP/MAS NMR. The 13C CP/MAS NMR experiments were performed on a Chemagnetics Infinity 400 MHz spectrometer with a 13C operating frequency of 100.04 MHz. Samples were spun at a rate of 10 kHz ((5 Hz). A 50 kHz radio frequency field strength was used for decoupling with a decoupling period of 12.8 ms. A π/2 pulse width of 5 µs with 1 ms CP contact time was employed. Phase cycling was used to minimize artifacts. 13C chemical shifts were calibrated through the adamantane methylene peaks observed at 28.8 ppm relative to tetramethylsilane at 0 ppm.

Biomacromolecules, Vol. 5, No. 3, 2004 745

Spectral Editing Methods. Solid-state spectral editing experiments25 utilizing cross-polarization with depolarization (CPD) and CP with polarization inversion (CPPI) were performed on a Varian UnityINOVA 400 NMR spectrometer equipped with Chemagnetics double-resonance 4 mm MAS probe. Cross-polarization with depolarization (CPD) is used to identify nonprotonated and methyl carbons, whereas crosspolarization with polarization-inversion (CPPI) selects the methylene carbons. Optimal spectral editing parameters were determined as 140 µs (CP and CPD) or 1600 µs (CPPI) spinlocking time on a 1H, a maximum of 450 µs for spin-locking on the 13C, cross-polarization times of 40 µs (“short contact time”) or 1.5 ms (“long contact time”), a depolarization time of 200µs, and a polarization inversion time of 32 µs.26 2D Spin Diffusion NMR under Off Magic Angle Conditions. The 2D spin-diffusion NMR spectra were obtained with Varian UnityINOVA 400 NMR spectrometer and 7 mm Jakobsen-type double-tuned MAS probe at off magic angle condition (θm - 9°) at room temperature. The sample spinning rate was 6 kHz((3 Hz). The scaling factor of the 2D spin-diffusion spectra were 1/2(3 cos2(θm - 7.8°) - 1) ) +0.20. The mixing time of 2 s was optimized for spin diffusion between intramolecular specific carbon atoms of selectively isotope-labeled Val and Gly residues, however, for no spin-diffusion between intermolecular carbon atoms.7,27,28 The contact time was set to 2ms, and the variableamplitude CP technique was used.29 Results and Discussion 1. Structure of the (AG)n Sequences in the Peptides, (AG)6(VPGVG)(AG)7 and (AG)5(VPGVG)2(AG)5. Figure 1 shows 13C CP/MAS NMR spectra of (AG)15 (a,b), (AG)6(VPGVG)(AG)7 (c,d), and (AG)5(VPGVG)2(AG)5 (e,f). NMR data for these samples were measured after dialysis (a, c, e) and after treatment with formic acid (b, d, f). One of the more dramatic spectral differences is noted in the comparison of (AG)15 as it undergoes the treatment of formic acid, transforming it from silk I (a) to silk II (b).9,10 Differences are noted in all regions of the 13C CP/MAS spectrum; that is, representative 13C chemical shifts are observed for Gly CR (43.2 ppm), Gly CdO (169.9 ppm), Ala CR (50.7 ppm), Ala Cβ (16.5 ppm), and Ala CdO (176.8 ppm) in the silk I form, whereas in the silk II form, Gly CR (42.4 ppm), Gly CdO (169.1 ppm), Ala CR (48.7 ppm), Ala Cβ (16.7, 19.6, 22.2 ppm), and Ala CdO (171.8 ppm) are measured. There are no significant changes in isotropic chemical shifts and relative intensities after treatment with formic acid for the peptide (AG)6(VPGVG)(AG)7. Namely, the structure was already silk II after dialysis of the 9 M LiBr solution against water, which is in contrast to silk I of (AG)15 after the same treatment. Also, as expected, there are no significant differences in the 1D 13C CP/MAS NMR data. In contrast, the structure of (AG)5(VPGVG)2(AG)5 after dialysis of the 9 M LiBr solution against water is different from the silk II structure of (AG)6(VPGVG)(AG)7. The peaks from Ala and Gly residues in spectrum (e) become broader; for example, the line width of the Ala CR peak is 230 Hz in

746

Biomacromolecules, Vol. 5, No. 3, 2004

Ohgo et al.

Figure 2. 13C CP/MAS NMR spectra of (AG)6, (AG)9, (AG)12, and (AG)15 after dialysis of the 9 M LiBr solution against water. Only the expanded methyl peaks were shown.

Figure 1. 13C CP/MAS NMR spectra of (AG)15 after dialysis of the 9 M LiBr solution against water (a) and after treatment with formic acid (b), (AG)6(VPGVG)(AG)7 after dialysis of the 9 M LiBr solution against water (c) and after treatment with formic acid (d), and (AG)5(VPGVG)2(AG)5 after dialysis of the 9 M LiBr solution against water (e) and after treatment with formic acid (f). The symbol ssb means spinning sideband.

spectrum (c) and 390 Hz in spectrum (e). Moreover, the chemical shift of the Ala Cβ peak changes from 19.7 ppm in spectrum (c) to 18.1 ppm in spectrum (e). After treatment with formic acid, the peaks become sharper; for example, the line width of the Ala Ca peak is 240 Hz in spectrum (f). The change in the line width was also remarkable for the carbonyl peaks where a resolved peak was observed in spectrum (f). Detailed discussion on the local structure is included below. To examine why the (AG)n sequences in these peptides might assume either silk II or a distorted β-sheet structure after dialysis of the 9 M LiBr solution against water, (AG)6, (AG)9, and (AG)12 were synthesized. After dialysis of the 9

M LiBr solution against water, the 13C CP/MAS NMR spectra were observed and shown in Figure 2. The Ala Cβ region was expanded. The Ala Cβ chemical shifts were 16.5, 16.6, and 16.7 ppm for (AG)6, (AG)9, and (AG)12, respectively, which is in agreement with the chemical shift, 16.5 ppm, for (AG)15. The line widths were 99, 124, and 112 Hz for (AG)9, (AG)12, and (AG)15 but 192 Hz for (AG)6. Thus, the structure of the former peptides was silk I (repeated β-turn structure).7 However, the breadth of the (AG)6 peak reflects the disorder in this peptide, whereby its structure might best be described as a “distorted β-turn”, in which the torsion angles are distributed over a range of values but with an average that corresponds to the torsion angles of the silk I form. Therefore, the Ala-Gly alternative copolypeptides (AG)n form (distorted) β-turn structures when n g 6. However, (AG)n sequences, (AG)6 or (AG)7, in these peptides form a β-sheet structure, which is different from the predicted β-turn structure. The origin of this discrepancy might be considered as follows. It has been said that the Pro-Gly sequence may form a type II β-turn structure with high probability in proteins.30 In the GiAi+1Gi+2Ai+3 sequence with silk I structure, intramolecular hydrogen bonds can be formed between Glyi CdO and Alai+3 N-H. The presence of the PG sequence prevents the (AG)n sequences in (AG)6(VPGVG)(AG)7 from forming the repeated type II β-turn structure. Our previous work with (AG)15 included spectroscopic data on (AG)9AA(AG)5, in which only the 20th residue was changed from Gly to Ala; after dialysis, this peptide adopted a typical silk ΙΙ structure.31 This result demonstrated that the structural change from silk I to silk II occurs by the

Structure of B. mori Silk-Elastin Like Protein

Biomacromolecules, Vol. 5, No. 3, 2004 747

Figure 4. Deconvolutions by Gaussian functions of the Ala Cβ peak in the 13C CP/MAS NMR spectrum of (AG)15 with silk II form published previously.10

Figure 3. Spectral editing of (AG)5(VPGVG)2(AG)5 after treatment with formic acid. (a) Full CP/MAS spectrum. (b) Methylene subspectrum acquired with CPPI. (c) Nonprotonated and methyl subspectrum acquired with CPD.

displacement of only one Gly residue to the Ala residue. Therefore, the effect of the presence of one VPGVG sequence on the structure of (AG)6(VPGVG)(AG)7 seems to have the similar effect as the AA sequence in (AG)9AA(AG)5. 2. Detailed Analysis of the Alanine Cβ Peak in the CPD Spectra. Toward the detailed structural analysis of these peptides of B. mori silk-elastin like protein, particular attention was drawn to the methyl region. To identify these peaks more clearly, solid-state spectral editing experiments were applied. Figure 3 shows representative editing data of the peptide (AG)5(VPGVG)2(AG)5 after treatment with formic acid. The full 13C CP/MAS NMR spectrum illustrates the degree of overlap, particularly in the aliphatic region. The methodology of Wu and co-workers is based on the differences in CP dynamics and spin temperatures for the various types of carbons25 and has proven robust enough for these applications to polypeptides.26 In the present study, these editing methods were applied for spectral simplification, specifically, to obtain an unobscured line shape for the methyl carbons. The editing results were also useful to provide additional support for unambiguous peak assignments. Crosspolarization with polarization inversion (CPPI) was used to select for the methylene carbons, namely, Gly CR, Pro Cβ, Pro Cγ, and Pro Cδ carbons in our target systems. The crosspolarization with depolarization (CPD) experiment identifies the backbone carbonyl and methyl carbons; the latter include AlaCβ and Val Cγ carbons. Methylene and methine peaks are suppressed in the CPD experiment, and the simplified

spectra reflect the efficacy of the method. However, the methyl peaks contain contributions from Val Cγ as well as the Ala Cβ peak, and additional work is required to extract information on the environment of these side chains. In our previous study,23 the peak with 320 Hz width at 18.9 ppm was assigned to the Val Cγ nuclei in the 13C CP/ MAS NMR spectrum of (VPGVG)6. Therefore, these chemical shift and line width values were assumed for the ValCγ peak in the deconvolution of the CPD spectra, as described below. As shown in Figure 4, the Ala Cβ peak of (AG)15 with silk II structure is best-fit with three Gaussian line shapes with isotropic chemical shifts of 22.2 ppm (27%), 19.6 ppm (46%), and 16.7 ppm (27%). The peak at 16.7 ppm was assigned to a “distorted β-turn” structure as mentioned above. The other two peaks at 19.6 and 22.2 ppm were assigned to antiparallel β-sheet structures.10 Namely, methyl groups of Ala residues pointing in the same direction were assigned to the peak 22.2 ppm, whereas opposing methyl groups of adjacent sheets were assigned to the peak 19.6 ppm. Although some of the methyl features only require the three peaks, the additional one at 21.1 ppm is required to fit the data for other samples, for example, (AG)3YG(AG)2VGYG(AG)3Y(AG)3 after treatment with formic acid.10 On the basis of these assignments, the Ala Cβ peaks of the CPD spectra of the peptides were deconvoluted assuming Gaussian line shapes as shown in Figure 5. In the deconvolution process, the chemical shift of the Val Cγ peak was fixed as 18.9 ppm, but its line width changes from 310 to 338 Hz depending on the samples. The chemical shifts and line widths of three peaks from Ala Cβ carbons of (AG)6(VPGVG)(AG)7 also change slightly between two treatments. In the deconvolution of (AG)5(VPGVG)2(AG)5, the distorted β-sheet (20.7 or 21.2 ppm) and distorted β-turn (16.5 or 16.9 ppm) were assumed. In addition, the deconvolution results of the methyl peaks in Table 1 satisfy the requirement within experimental error. Based on stoichiometry, the relative populations of Ala Cβ and Val Cγ carbons are 13:4 for (AG)6(VPGVG)(AG)7 and 10:8 for (AG)5(VPGVG)2(AG)5. The ratios obtained from these experiments very closely match the predicted values. The results of the deconvolution are summarized in Table 1. After removal of the ca. 18.9 ppm contribution of Val Cγ

748

Biomacromolecules, Vol. 5, No. 3, 2004

Ohgo et al. Table 1. Chemical Shifts (in ppm from TMS), Line Width, and the Relative Intensity of the Deconvoluted Peaks of Ala Cβ Carbons in Four Samplesa

sample (VPGVG)6 (AG)6(VPGVG)AG)7 after dialysis

(AG)6(VPGVG)AG)7 after formic acid treatment

(AG)5(VPGVG)2AG)5 after dialysis (AG)5(VPGVG)2AG)5 after formic acid treatment a

Figure 5. Cross polarization with depolarization spectra (only methyl peaks are shown) of (AG)6(VPGVG)(AG)7 after dialysis of the 9 M LiBr solution against water (a) and after treatment with formic acid (b), and (AG)5(VPGVG)2(AG)5 after dialysis of the 9 M LiBr solution against water (c) and after treatment with formic acid (d). The Val Cγ peaks predicted from the spectrum of (VPGVG)6 are shown as dotted lines. The deconvolutions by Gaussian functions were performed.

from the methyl peak, the detailed analysis of the structure of the (AG)n sequences is possible. Concerning the peptide (AG)6(VPGVG)(AG)7, the treatment with formic acid tends to increase the fraction of the distorted β-turn structure slightly and approaches 2:1 as the relative intensity ratio of the two peaks at 19.6 and 22.2 ppm.9,10 The spectral change in the Ala Cβ peak is drastic for (AG)5(VPGVG)2(AG)5 as the peptide undergoes treatment with formic acid. The fraction of the distorted β-turn and

relative after removal δ width intensity of the ValCγ (ppm) (Hz) (%) peak

18.9 16.6 18.9 19.7 22.2 16.8 18.9 20.0 22.2 16.5 19.0 20.7 16.9 18.9 21.2

320 322 310 210 286 398 310 222 218 358 338 326 194 314 346

19 25 31 25 26 24 31 19 45 44 11 9 45 46

25 42 33 34 41 25 81 19 16 84

The data for the Val Cγ peak are in italic.

Figure 6. 13C CP/MAS NMR spectra of (AG)6(VPGVG)(AG)7 after dialysis of the 9 M LiBr solution against water (a) and after treatment with formic acid (b), and (AG)5(VPGVG)2(AG)5 after dialysis of the 9 M LiBr solution against water (c) and after treatment with formic acid (d). The spectrum of (VPGVG)623 was also shown as dotted lines. The symbol ssb means spinning sideband.

the distorted β-sheet reverses; the fraction of the distorted β-turn was 81% after the dialysis, but the distorted β-sheet became dominant (84%) after treatment with formic acid. Namely, the treatment with formic acid induces the structural transition from distorted β-turn to distorted β-sheet for the (AG)5 sequences in the peptide. This effect is similar to the case of (AG)15 where the structural change occurs from silk I to silk II by the treatment with formic acid.10 3. Structure of the (VPGVG)n Sequence in the Peptides, (AG)6(VPGVG)(AG)7 and (AG)5(VPGVG)2(AG)5 Studied With 13C Chemical Shifts. Figure 6 shows the expanded region from 0 to 80 ppm of the 13C CP/MAS NMR spectra of these peptides and (VPGVG)6. To discuss the local

Structure of B. mori Silk-Elastin Like Protein

Biomacromolecules, Vol. 5, No. 3, 2004 749

CR carbons, based on our previous studies of (VPGVG)6.23 In general, the chemical shifts of CR and Cβ carbons in proteins show conformational dependence. The relationships between the chemical shifts and torsion angles are constructed from the empirical database and are plotted on the Ramachandran maps as the chemical shift maps.32,33 The chemical shifts derived from (VPGVG)n sequences in these peptides are essentially the same as those of (VPGVG)6, indicating that the structure of (VPGVG)n sequences is the same as that of (VPGVG)6. 4. Structure of the Sequence (VPGVG)2 in the Peptide, (AG)5(VPGVG)2(AG)5 Studied with 2D Spin-Diffusion NMR under OMAS. In previously reported work, we have successfully applied 2D spin-diffusion NMR under offmagic-angle-spinning (OMAS) conditions to characterize the structures of the silk fibroins and the related model peptides through the determination of the torsion angles of the backbone chain.7,9,27,28 In this manner, the 2D spin-diffusion NMR method was applied to the analysis of (AG)5(VPGVG)2(AG)5. Figure 7 shows 2D spin-diffusion spectra of (AG)5VPG[1-13C]V[113 C]GVPGVG(AG)5 after dialysis of the 9 M LiBr solution against water (Figure 7a) and after treatment with formic acid (Figure 7b). The spectral patterns show strong dependence of the torsion angles of the Gly15 in the peptide. Notably, if the torsion angles are approximately (-150°, 150°), that is, in the region of β-sheet structure, only the diagonal ridge should be observed.28 However, off-diagonal patterns are observed, and therefore, the local structure of the Gly15 in the sequence, (VPGVG)2, is not β-sheet. To further structural analysis, the 2D spin-diffusion NMR spectrum of elastin-like model peptide, (VPGVG)2VPG[113 C]V[1-13C]G(VPGVG)3, reported previously23 was shown in Figure 7c. The spectral pattern was essentially the same. The pattern of (VPGVG)2VPG[1-13C]V[1-3C]G(VPGVG)3 is most consistent with a distribution of torsion angles about the Gly15 residue. Specifically, no single or well-ordered conformation can be used to fit this pattern. Conclusions

Figure 7. 2D spin-diffusion NMR spectra (only the carbonyl region was expanded) of (AG)5VPG[1-13C]V[1-13C]GVPGVG(AG)5 after dialysis of the 9 M LiBr solution against water (a) and after treatment with formic acid (b). The specrum of (VPGVG)2VPG[1-13C]V[1-13C]G(VPGVG)3 after lyophilization was also shown.23

structure of the elastin-like sequences, VPGVG or (VPGVG)2, the emphasis of this discussion is directed to the small peaks. The small peaks observed at 25, 30, 57, and 59 ppm are assigned to the Pro Cγ, Val Cβ, Pro Cβ, Val CR, and Pro

The solid-state NMR spectra of the peptides (AG)6(VPGVG)(AG)7 and (AG)5(VPGVG)2(AG)5 after dialysis of the 9 M LiBr aqueous solution and after treatment with formic acid were observed. The former peptide took the silk II structure after the dialysis, which is in contrast to the silk I form of (AG)15 after the same treatment. Only minor structural changes were observed by further treatment with formic acid. However, a drastic structural change of the (AG)n sequences was observed for (AG)5(VPGVG)2(AG)5 between two treatments. The local structures of the Gly residue of the VG units in the elastin-like subunits (VPGVG) and (VPGVG)2 were the distorted structures with a distribution of the torsion angles, which was derived from the 2D spin diffusion NMR spectral pattern of (AG)5VPG[1-13C]V[1-13C]GVPGVG(AG)5. Observation of this distribution of the Gly residue was independent of the treatment, dialysis or formic acid. Hereafter, the local structure of other residues

750

Biomacromolecules, Vol. 5, No. 3, 2004

of the elastin-like subunits should be examined using similar solid-state NMR coupled with 13C isotope labeling of the peptides. Acknowledgment. T.A. acknowledges the support from the Asahi Glass Foundation and the Insect Technology Project, Japan. Grant support for K.K.K. from the NIH (RR16453) is also acknowledged. References and Notes (1) Arcidiacono S.; Mello C.; Kaplan D. L.; Cheley S.; Bayley H. Purification and characterization of recombinant spider silk expressed in Escherichia coli. Appl. Microbiol. Biotechnol. 1998, 49, 31-38. (2) Cappello, J.; Crissman, J. W.; Crissman, M.; Ferrari, F.; Textor, G.; Wallis, O.; Whitledge, J. R.; Zhou, X.; Burman, D.; Aukerman, L.; Stedronsky, E. R. In-situ self-assembling protein polymer gel systems for administration, delivery, and release of drugs. J. Controlled Release 1998, 53, 105-117. (3) Yao, J.; Asakura, T. Synthesis and structural characterization of silklike materials incorporated with an elastic motif. J. Biochem. 2003, 133, 147-154. (4) Megeed, Z.; Cappello, J.; Ghandehari, H. Genetically engineered silkelastinlike protein polymers for controlled drug delivery. AdV. Drug DeliVery ReV. 2002, 54, 1075-1091. (5) Saito, H.; Tabeta, R.; Asakura, T.; Iwanaga, Y.; Shoji, A.; Ozaki, T.; Ando, I. High-resolution C-13 NMR-study of silk fibroin in the solid-state by the cross-polarization magic angle spinning methods conformational characterization of silk-I and silk-II type forms of Bombyx-mori fibroin by the conformation-dependent C-13 chemicalshifts. Macromolecules 1984, 17, 1405-1412. (6) Asakura, T.; Kuzuhara, A.; Tabeta, R.; Saito, H. Conformation Characterization of Bombyx mori Silk Fibroin in the Solid State by High-Frequency 13C Cross Polarization-Magic Angle Spinning NMR, X-ray Diffraction, and Infrared Spectroscopy. Macromolecules 1985, 18, 1841-1845. (7) Asakura, T.; Ashida, J.; Yamane, T.; Kameda, T.; Nakazawa, Y.; Ohgo, K.; Komatsu, K. A repeated β-turn structure in poly(Ala-Gly) as a model for silk I of Bombyx mori silk fibroin studied with twodimensional spin-diffusion NMR under off magic angle spinning and rotational echo double resonance. J. Mol. Biol. 2001, 306, 291305. (8) Asakura, T.; Yamane, T.; Nakazawa, Y.; Kameda, T.; Ando, K. Structure of Bombyx mori silk fibroin before spinning in solid-state studied with wide-angle X-ray scattering and 13C cross-polarization/ magic angle spinning NMR. Biopolymers 2001, 58, 521-525. (9) Asakura, T.; Yao, J.; Yamane, T.; Umemura, K.; Ulrich, A. S. Heterogeneous Structure of Silk Fibers from Bombyx mori Resolved by 13C Solid-State NMR Spectroscopy. J. Am. Chem. Soc. 2002, 124, 8794-8795. (10) Asakura, T.; Yao, J. 13C CP/MAS NMR study on structural heterogeneity in Bombyx mori silk fiber and their generation by stretching Protein Sci. 2002, 11, 2706-2713. (11) Debelle, L.; Alix, A. J. P. The structures of elastins and their function. Biochimie 1999, 81, 981-994. (12) Debelle, L.; Tamburro, A. M. Elastin: molecular description and function. Int. J. Biochem. Cell Biol. 1999, 31, 261-272. (13) Sandberg, L. B.; Leslie, J. G.; Leach, C. T.; Alvarez, V. L.; Torres, A. R.; Smith, D. W. Elastin covalent structure as determined by solidphase amino acid sequencing. Pathol. Biol. 1985, 33, 266-274. (14) Yeh, H.; Ornstein-Goldstein, N.; Indik, Z.; Sheppard, P.; Anderson, N.; Rosenbloom, J. C.; Cicila, G.; Yoon, K.; Rosenbloom, J. Sequence variation of bovine elastin mRNA due to alternative splicing. Collagen Relat. Res. 1987, 7, 235-247. (15) Urry, D. W. Physical Chemistry of Biological Free Energy Transduction As Demonstrated by Elastic Protein-Based Polymers. J. Phys. Chem. B 1997, 101, 11007-11028.

Ohgo et al. (16) McMillan, R. A.; Caran, K. L.; Apkarian, R. P.; Conticello, V. P. High-Resolution Topographic Imaging of Environmentally Responsive, Elastin-Mimetic Hydrogels. Macromolecules 1999, 32, 90679070. (17) Nicol, A.; Gowda, D. C.; Urry, D. W. Cell adhesion and growth on synthetic elastomeric matrixes containing Arg-Gly-Asp-Ser-. J. Biomed. Mater. Res. 1992, 26, 393-413. (18) Cook, W. J.; Einspahr, H.; Trapane, T. L.; Urry, D. W.; Bugg, C. E. Crystal structure and conformation of the cyclic trimer of a repeat pentapeptide of elastin, cyclo-(L-valyl-L-prolylglycyl-L-valylglycyl)3. J. Am. Chem. Soc. 1980, 102, 5502-5505. (19) Venkatachalam, C. M.; Urry, D. W. Development of a linear helical conformation from its cyclic correlate. β-Spiral model of the elastin poly(pentapeptide) (VPGVG)n. Macromolecules 1981, 14, 12251229. (20) Urry, D. W.; Trapane, T. L.; Sugano, H.; Prasad, K. U. Sequential polypeptides of elastin: cyclic conformational correlates of the linear polypentapeptide. J. Am. Chem. Soc. 1981, 103, 2080-2089. (21) Hong, M.; McMillan, R. A.; Conticello, V. P. Measurement of conformational constraints in an elastin-mimetic protein by residuepair selected solid-state NMR. J. Biomol. NMR 2002, 22, 175-179. (22) Rodriguez-Cabello, J. C.; Alonso, M.; Diez, M. I.; Caballero, M. I.; Herguedas, M. M. Structural investigation of the poly(pentapeptide) of elastin, poly(GVGVP), in the solid state. Macromol. Chem. Phys. 1999, 200, 1831-1838. (23) Asakura, T.; Ashida, J.; Ohgo, K. Conformational characterization of (val-pro-gly val-gly)6 with 13C solid-state NMR. Polym. J. 2003, 35, 293-296. (24) Kumashiro, K. K.; Kurano, T. L.; Niemczura, W. P.; Martino, M.; Tamburro, A. M. 13C CPMAS NMR Studies of the Elastin-Like Polypeptide (LGGVG)n. Biopolymers 2003, 70, 221-226. (25) Wu, X.; Burns, S. T.; Zilm, K. W. Spectral editing in CPMAS NMR. Generating subspectra based on proton multiplicities J. Magn. Reson., Ser. A 1994, 111, 29-36. (26) Kumashiro, K. K.; Niemczura, W. P.; Kim, M. S.; Sandberg, L. B. Selection of side-chain carbons in a high-molecular-weight, hydrophobic peptide using solid-state spectral editing methods. J. Biomol. NMR 2000, 18, 139-144. (27) Ashida, J.; Ohgo, K.; Komatsu, K.; Kubota, A.; Asakura, T. Determination of the torsion angles of alanine and glycine residues of model compounds of spider silk (AGG)10 using solid-state NMR methods. J. Biomol. NMR 2003, 25, 91-103. (28) Ashida, J.; Ohgo, K.; Asakura, T. Determination of the torsion angles of alanine and glycine residues of bombyx mori silk fibroin and the model peptides in the silk I and silk II forms using 2D spin diffusion solid-state NMR under off magic angle spinning. J. Phys. Chem. B 2002, 106, 9434-9439. (29) Peersen O. B.; Wu X. L.; Kustanovich I.; Smith S. O. VariableAmplitude Cross-Polarization MAS NMR J. Magn. Reson., Ser. A 1993, 104, 334-339. (30) Hutchinson, E. G.; Thornton, J. M. A revised set of potentials for β-turn formation in proteins. Protein Sci. 1994, 3, 2207-2216. (31) Asakura, T.; Sugino, R.; Okumura, T.; Nakazawa, Y. The role of irregular unit, GAAS, on the secondary structure of Bombyx mori silk fibroin studied with 13C CP/MAS NMR and wide-angle X-ray scattering. Protein Sci. 2002, 11, 1873-1877. (32) Asakura, T.; Iwadate, M.; Demura, M.; Williamson, M. P. Structural analysis of silk with 13C NMR chemical shift contour plots. Int. J. Biol. Macromol. 1991, 24, 167-171. (33) Iwadate, M.; Asakura, T.; Williamson, M. P. CR and Cβ carbon-13 chemical shifts in proteins from an empirical database J. Biomol. NMR 1999, 13, 199-211.

BM034355X