Nucleation Effects in Peptide Foldamers

Abstract

Oligomers composed of β³-amino acid residues and a mixture of α- and β³-residues have emerged as proteolytically stable structural mimics of α-helices. An attractive feature of these oligomers is that they adopt defined conformations in short sequences. In this manuscript, we evaluate the impact of β³-residues as compared to their α-amino acid analogs in prenucleated helices. Our hydrogen-deuterium exchange results suggest that heterogeneous sequences composed of “αααβ” repeats are conformationally more rigid than the corresponding homogeneous α-peptide helices, with the macrocycle templating the helical conformation having a significant influence.

INTRODUCTION

Conformationally defined synthetic oligomers, termed foldamers, have emerged as attractive molecular scaffolds for the discovery of new materials, catalysts and ligands for protein receptors.^1–5 Oligomers composed of β³- and mixtures of α- and β³-residues represent well-studied classes of foldamers, specifically as mimics of α-helices and inhibitors of previously intractable protein-protein interactions.^6–12 One surprising aspect of β-peptide oligomers is that they assume defined helical structures in very short sequences despite the presence of the extra methylene units which would be expected to endow the backbone with an increased freedom of rotation.^2,6,13 In contrast, α-peptides composed of less than fifteen α-amino acids generally do not adopt defined helical conformations, in the absence of structural constraints. The helix-coil theory suggests that long sequences and multiple intramolecular hydrogen bonds are needed in order to overcome the energetically demanding nucleation parameter.^14–18 Attainment of helical configurations in β-peptides composed of as little as six residues suggests that the nucleation or the propagation parameters are distinct in helices containing β³-residues.¹³ In this manuscript, we compare the helical propensities of α- and β³-amino acids in prenucleated helices. Such comparisons have been challenging because the circular dichroism spectra of α-peptides and peptides containing β³-residues feature different sets of maxima and minima.^9,19,20 Although a number of β-peptide foldamers have been crystalized, the contribution of crystal structure packing to the observed conformation is difficult to quantify. Our results with oligomers composed of “αααβ” repeats are more stable than their α-peptide analogs.^9,10,13 We also investigated the protein-binding properties of α3β oligomers and find that the heterogeneous oligomer can bind the target receptor with a similar affinity to the α-peptide. A combination of the biophysical and protein binding results provide valuable insights for designing α3β foldamers. Importantly, the results suggest that the underlying strategy provides a unique approach for comparing properties of nonnatural residues to known parameters of α-amino acids.

RESULTS AND DISCUSSION

We designed a host helix in which the propensity of α- and β^3-guest residues could be evaluated.^21–29 Helices were nucleated using the hydrogen bond surrogate (HBS) approach, in which an N-terminal i → i+4 intramolecular hydrogen bond is replaced with a covalent bond (Figure 1a).³⁰ The HBS approach leads to defined α-helices in short peptides. Solution and solid-state conformations of HBS helices have previously been characterized with 2D NMR and circular dichroism spectroscopies, and X-ray crystallography.^31,32 Comparisons of experimental thermal denaturation curves with simulations of the Zimm–Bragg model^14–17 suggest that HBS helices are nucleated with the constant, σ, close to unity.^32–35 We have also demonstrated that HBS α-helices can target their cognate protein receptors with high affinity and specificity.^36–40 Significantly, the stabilized α-helices can modulate chosen intracellular protein-protein interactions while their unconstrained counterparts remain ineffective.^36,37

Figure 1.

(a) HBS α-helices feature a carbon-carbon bond in place of an N-terminal main-chain (i, i+4) hydrogen-bond. (b) An α3β sequence contains 13- and 14-membered hydrogen bonds. (c) Design of HBS α- and HBS α3β-peptides with different nucleation macrocycles; hydrogen bonds spanning 13- and 14-atoms are shown as hashed and bold bonds, respectively.

The key advantage of the HBS approach is its ability to provide conformational rigidity in short peptides without utilizing side chain functionality. Oligomers composed of β³- and mixtures of α- and β³-residues are typically preorganized using cyclic amino acid analogs with predefined φ, ψ-dihedral angles,^8,41,42 or through side chain to side chain contacts,^7,12,43–45 for example where one side of the helix has a hydrophobic face and a second features ionic bridges. A drawback of the later strategy is that it limits the diversity of side chains that may be placed on a helix for specific biomolecular recognition. An attractive feature of the HBS method is that it provides conformational control over sequences composed of acyclic residues without sacrificing side chain functionalities. Herein we show that conformationally defined chimeric helices can be accessed from acyclic residues with the HBS strategy.

Prenucleated HBS helices allow us to monitor the subtle effects of α→β³ substitution on helix propagation while controlling the nucleation parameters. We inserted one β³-residue per helical turn in the host HBS α-helix such that the heterogeneous sequence contains three α-residues followed by one β³-residue. Such α3β sequences have previously been characterized as suitable mimics of α-helices.⁹ We based the oligomer design on a short segment from the p53 activation domain⁴⁶ – whose design and protein binding properties have been previously reported (Table 1).^38,47 This sequence was chosen to allow us to investigate the potential of HBS α3β analogs to target MDM2 in comparison to the parent HBS α-helix. The p53 sequence was also deemed appropriate because it lacks multiple side chain contacts such as ionic bridges, which may bias the results. The parent constrained α-peptide is roughly 50% helical according to circular dichroism spectroscopy.^38,47 We conjectured that this range was ideal for examining potential increase or decrease in conformational rigidity of the heterogeneous oligomers.

Table 1.

Design of α and α3β-peptides.

Peptide	Sequencea
HBS α-helix 1	XQEG*FSDLWKLLS-NH2
HBS α3β-helix 2	XQEGβ*FSDLβWKLLβS-NH2
HBS α3β-helix 3	XQEβG*FSDβLWKLβLS-NH2
α3β peptide 4	AcQEβGFSDβLWKLβLS-NH2

^aX, G*, and G_β* denote 4-pentenoic acid, N-allyl glycine, and N-allyl β-glycine residues, respectively. Blue letters denote β³-residues.

Design of the α3β helices raises an interesting question regarding the nucleation of heterogeneous sequences. A canonical α3β sequence would be expected to feature hydrogen bonds spanning 13-atoms within turns comprised of α-residues but 14-membered hydrogen bonds in turns that contain a β³-residue (Figure 1b). This alternating pattern of putative hydrogen bonds within folded oligomers is unique to heterogeneous sequences, although 3(10) helices with 10-membered intramolecular hydrogen bonds are known to initiate α-helical regions within proteins.⁴⁸ We conjectured that the stability of the α3β sequences would fluctuate with the size of the nucleation macrocycle. A 13-membered HBS macrocycle is a mimic of a tripeptide α-helical turn and promotes intramolecular hydrogen bonds that span 13-atoms in canonical α-helices. HBS 1 is a mimic of an α-helix with a 13-membered HBS macrocycle. However, a 13-membered macrocycle would not be expected to be optimal if it is followed by a β³-residue and three successive 14-atom intramolecular hydrogen bonds as in HBS 2. The 13- and 14-membered hydrogen bonds are shown as hashed and bold bonds, respectively, in Figure 1c. A 14-memberd HBS macrocycle, as in HBS 3, with an embedded β³-residue should be a better nucleator of α3β sequences. In support of this hypothesis, we have previously shown that stability of α-helices is optimal with a 13-membered HBS macrocycle replacing the 13-membered intramolecular hydrogen bond rather than a 14-membered HBS macrocycle.³³

Design and Synthesis

We began the design of HBS α3β helices by examining the potential of a 13- and a 14-membered macrocycle to control the desired helical conformation in the attached peptide (Figure 1c and Table 1). HBS helices contain a carbon-carbon bond in place of a main chain i → i+4 hydrogen bond. The hydrocarbon bridge is inserted using a ring-closing metathesis reaction between two appropriately-placed alkene groups (Supporting Information).⁴⁹ Detailed protocols for the synthesis of the HBS helices have been reported previously.^50–53

Structural Characterization by Circular Dichroism

The helicity of the peptides was examined by circular dichroism spectroscopy. CD studies were performed in 10 % trifluoroethanol (TFE) in phosphate buffered saline (PBS). The organic co-solvent was included because the p53-derived sequence aggregates in pure aqueous solutions. In 10% TFE, molar ellipticity remained consistent between ^25–120 μM concentration range indicating reduced aggregation effects (Supporting Information, Figure S6). Figure 2a shows the CD spectra of 1-3. HBS α-helix 1 affords a CD signature typical of a canonical α-helix, with double minima near 206 and 222 nm and a maximum at 190 nm.³⁸ The traces obtained for HBS α3β 2-3 are similar to those observed for α-helices except with a weaker 222 nm band. It is difficult to compare the conformational stability of these three different oligomers using CD since they all contain different structural topographies and feature different minima.⁵⁴ The unconstrained α3β peptide analog 4 provides weaker signal as compared to the constrained derivative consistent with the idea that conformational rigidity is endowed by the HBS constraint (Figure 2a). The CD spectrum of the α3β sequences are consistent with the previously reported spectrum of β- and chimeric α/β-peptides in shape and intensity.^20,55

Figure 2.

(a) Circular dichroism spectra of peptides 1-4 at 25 °C. (b) Effect of temperature on the stability of HBS helices. The CD spectra were obtained in 10% TFE/PBS. Denaturation of 1, 2, and 3 was monitored at 222, 205, and 202 nm wavelengths, respectively. (c) Theoretical denaturation curves as a function of different nucleation constant, σ. The theoretical curves were obtained by simulating the Zimm–Bragg model for a helix that can only denature in one direction, as described.³³

Thermal Stability of HBS helices 1-3

We next investigated the thermal stability of HBS helices by monitoring the temperature-dependent change in its CD spectrum (Figure 2b). Previous thermal denaturation studies with HBS helices have shown that the conformation of these nucleated peptides stays remarkably consistent at high temperatures.^32–34 The broad melting transition in HBS helices is consistent with σ ≥ 1.⁵⁶ This σ value (and the broad transition) implies a noncooperative case in which each unit behaves independently, as shown by the simulation of the Zimm-Bragg model for a helix that can only denature in one direction (Figure 2c).^14–17,33 The thermal denaturation curves show that both HBS α3β 2 and 3 retain the broad melting characteristics of HBS helices with σ ≥ 1, and highlight the potential of the HBS strategy to nucleate heterogeneous sequences composed of acyclic residues. The results also indicate that both the 13- and 14-membered macrocycles in 2 and 3 are capable of initiating defined conformations.

Structural Characterization by NMR

Circular dichroism spectroscopy provides compelling evidence that the hydrogen bond surrogate approach can stabilize heterogeneous sequences. However, CD spectroscopy does not allow a detailed analysis of the peptide structure at the atomic level. For instance, we wanted to determine if individual β³-residues propagate with similar effectiveness as the α-residues. For answers to these pertinent questions, we fully characterized the compounds by NMR spectroscopy. The NMR studies were performed in 20% CF₃CD₂OD in PBS (pH 3.5) rather than in lower amounts of d3-TFE; we used this solvent mixture for two reasons: (1) in purely aqueous solutions or 10% TFE solutions, these sequences showed observable aggregation at concentrations needed for NMR studies while aggregation effects were negligible in 20% TFE, and (2) this solvent system provided minimal peak overlap allowing unambiguous assignment of a larger number of resonances. To evaluate the conformational stability and dynamics of 1-3, we obtained rates of amide proton H/D exchange, which provide a convincing measure of the extent to which a particular main chain proton is involved in intramolecular hydrogen bonding.^57,58

We utilized a combination of 1D and 2D NMR experiments to further define the conformation of 1-3. A set of 2D TOCSY and NOESY spectroscopies were used to assign ¹H NMR resonances for 1-3. Sequential NH-NH (i and i + 1) NOESY cross-peaks, a signature of helical structure, were observed for 1-3 as shown in the NOE correlation charts (Figure 3 and Supporting Information). The NOESY spectra further reveal several medium to weak (i, i+3) and (i, i+4) NH-CHα cross peaks that support an α-helix like conformation in these peptides. A larger number of contiguous medium range NOEs, suggestive of a more stable helical conformation, are seen with 3 than the other two peptides. Analysis suggests predominance of a single helical conformation in both HBS α3β sequences. This result is important because earlier solution studies have indicated a mixture of two conformations is observed in oligomers composed of acyclic β³-residues.^59,60

Figure 3.

(a–c) NOESY correlation charts for 1-3, respectively. The NMR spectra, included in the Supporting Information, were obtained in 20% TFE/PBS. Blue letters denote β³-residues. α refers to CH protons adjacent to the amide NH in α- and β³-amino acid residues.

Amide H/D Exchange Rates

Main-chain amide hydrogen-deuterium exchange rates offer a sensitive measure of structural stability and dynamics of proteins.^57,61,62 Structured protein amide protons are involved in backbone hydrogen bonding and are shielded from solvents resulting in their slow H/D exchange kinetics compared to unstructured protein amide protons. Figure 4 shows spectra for 1-3 at different time intervals following the addition of D₂O. The tabulated exchange values for residues outside the HBS macrocycle for oligomers 1-3 are shown in Tables 2–4. The individual hydrogen-exchange rates in these helices can be determined precisely which is typically not possible for short peptides, indicating the conformational stability of these oligomers. The measured exchange rates, k_ex, were compared to the predicted intrinsic chemical exchange rate, k_ch, for an unstructured α-peptide of the same sequence, to assess individual protection factors (log k_ch/k_ex) and the corresponding free energies of protection (−ΔG).⁶³ The predicted intrinsic chemical exchange rates, protection factors, and the free energy of protection were calculated using the spreadsheet at http://hx2.med.upenn.edu and are shown in Table 2. (This worksheet was developed for α-peptides and not for heterogeneous sequences; however, we think its use offers critical insights.)

Figure 4.

(a–c) Hydrogen-deuterium exchange spectra for backbone amide protons in 1-3, respectively. (d) Exchange curves for 3. The H/D exchange experiments were performed in duplicate. The NMR spectra were obtained in 20% TFE/PBS.

Table 2.

Summary of amide proton temperature coefficients and deuterium exchange data for 1.

Peptide 1 Residues:	F4	S5	D6	L7	W8	K9	L10	L11	S12
H/D rate constant × 10−5 (h−1)	38.08	38.08	38.08	1.69	0.18	0.51	0.20	0.17	0.78
Protection factor (log kch/kex)a	−0.81	−0.32	0.14	0.54	0.96	0.97	1.08	0.84	1.09
Stabilization, −ΔG (kcal/mol)	N/A	N/A	−0.58	0.52	1.22	1.23	1.40	1.04	1.42

^acalculated using the spreadsheet at http://hx2.med.upenn.edu.download.html. k_ex: measured exchange rates, k_ch: intrinsic chemical exchange rate.

Table 4.

Summary of amide proton temperature coefficients and deuterium exchange data for 3. β³-Residues are shown in blue font.

Peptide 3 Residues:	F4	S5	Dβ6	L7	W8	K9	Lβ10	L11	S12
H/D rate constant × 10−5 (h−1)	1.33	1.07	0.32	0.16	0.14	0.08	0.19	0.23	0.26
Protection factor (log kch/kex)	0.65	1.23	2.22	1.56	1.07	1.75	1.1	0.70	1.58
Stabilization, −ΔG (kcal/mol)	0.72	1.61	2.97	2.07	1.39	2.34	1.43	0.8	2.10

The H/D spectra suggest a striking degree of stability in 3 – more of the amides remain partially protected after 10 hours in 3 than 1 and 2. Nearly all NH protons have exchanged after 20 h of incubation in 1 and 2 but some residues retain their proton label in 3 even after 24 h. Several residues in 3 are protected from exchange with free energy values of 1 kcal/mol or greater than in 1 and 2. Overall, the data indicate that peptide 3 with the HBS macrocycle of 14-atoms templating two successive 14-atom hydrogen-bonds is more stable than peptide 2 with a 13-membered macrocycle nucleating three consecutive 14-atom hydrogen bonds (Figure 1c). This result is in line with our previous observation that α-helices containing a 13-membered HBS macrocycle are more conformationally stable than those containing an extra atom in the template.³³ Significantly, the exchange rates in oligomer 3 are much slower than in 1 suggesting that β³-residues have a higher helical propensity than α-residues. α3β-Peptide 3 contains a highly stable hydrogen-bonded network with significant protection factors and associated free energies of protection. Such a degree of stabilization is typically observed for buried amide protons in proteins but not in short peptides.^32,58 H/D exchange rates of oligomers 1 and 2, both of which contain 13-membered nucleation cycles, are similar in magnitude. This interesting result highlights the conflicting factors contributing to conformational stability of 2 and supports results seen with 3 – the nucleation in 2 is not optimal; however, the higher helical propensity of β³-residues stabilizes the conformation. Comparison of exchange rates for F4 and S5, the two residues that follow the macrocycle, in the three sequences is particularly revealing: these amides in 3 exchange much more slowly than in 1 and 2 indicating that the 14-membered macrocycle endows a high level of conformational rigidity. This result has significant implications for the future design of HBS helices as inhibitors of biomolecular interactions.

Solution Structure of 3

The solution structure of the HBS 3 was determined from NOESY cross-peaks and ³J_NHCHα coupling constants (Supporting Information, Table S6) using Monte Carlo conformational search in Macromodel 2011.^64,65 A total of 200 NOE restraints (45 medium and long range, 51 sequential, and 104 intraresidue) and 11 φ angle restraints were used during the dynamics. No explicit hydrogen bond restraints were used in the calculations. The final 20 lowest energy structures had no significant distance violations (Figure 5). The 20 conformer ensemble obtained for the peptide shows a backbone root mean squared deviation (RMSD) of 0.47 ± 0.06 Å. From the top down view, it can be seen that the macrocycle does not protrude from the helix (Figure 5b). Overall, the NMR structure of 3 confirms that a well-defined conformation is accessed in this α3β helix.

Figure 5.

NMR-derived structures of HBS α3β-helix 3. Side (a,c) and top (b) views of 20 lowest energy structures. Carbon, nitrogen and oxygen atoms are shown in grey, blue and red, respectively except in (c) where the hydrocarbon bridge is shown in gold color. The β-residues are highlighted in green.

Potential to Target Protein Receptors that Recognize α-Helices

The circular dichroism and NMR studies provide compelling evidence that HBS α3β helices 2 and 3 adopt configurations similar to that of an α-helix. To evaluate the potential of HBS α3β helices to target proteins that recognize α-helices, we measured the affinity of 2-3 for MDM2. We performed fluorescence polarization-based competition binding experiments and found that 3 binds to MDM2 with high affinity (K_D = 80 ± 21 nM) comparable to that previously reported for the optimized HBS p53 α-helix analog 1 (Figure 5 and Supporting Information).³⁸ Surprisingly, HBS α3β helix 2 targets MDM2 (K_D = 12.6 ± 1.8 μM) with a 150-fold lower affinity than 3. This large change may be attributed to two factors: (1) the differential conformational stability between peptides 2 and 3, and (2) the precise placement of the β-residues between peptide 2 and 3. In 3, all three of the p53 residues (Phe19, Trp23, and Leu26 in protein data bank numbering) which make important contacts with MDM2 are retained as α-amino acids;⁴⁶ whereas in 2, Leu11, or Leu26 in PDB numbering, has been converted to the β-analog. It is unclear if minor distortions caused by βLeu may be leading to decreased affinity.⁹

The high affinity of 3 for MDM2 demonstrates that judicious substitution of α-residues with β³-residues in HBS helices does not introduce structural perturbations that compromise their binding. To evaluate the specificity of HBS 3 for MDM2 we designed a negative control (HBS 5: XQE_βG*ASD_βLWKL_βAS-NH₂) by mutating two of the residues important for binding in 3 (Phe 19 and Leu 26) to alanines. As expected, compound 5 does not bind to MDM2 with measurable affinity (Figure 6). MDM2 has previously been targeted with β-peptide oligomers.^11,66–71 α3β 3 competes favorably with these oligomers with regards to its K_D for MDM2. Studies to evaluate the potential of this p53 mimetic to reactivate the p53 pathway are underway.⁷²

Figure 6.

Determination of peptide binding to His₆-tagged MDM2 by a fluorescence-polarization assay.

Conclusions

The studies described herein were designed to investigate three broad questions pertaining to the design of nonnatural peptide oligomers: (1) Can the HBS nucleation strategy effectively stabilize heterogeneous sequences composed of α- and β³-residues? (2) Since α3β oligomers may contain 13 or 14 atom intramolecular hydrogen bonds, which size nucleus would best stabilize the resulting foldamers? And (3), in optimally prenucleated systems, are α3β or all α sequences more stable?

This study shows that the HBS method compares favorably with previous approaches in stabilizing oligomers composed of α- and β³-residues.^8,12,41–45 The results also illustrate that the optimum α3β oligomer 3 is more conformationally defined than HBS α-helix 1, in line with the stable β-conformations observed in crystal and NMR structures.^6,9,13,73,74 Importantly, the studies provide a host scaffold for evaluating the propensities of β-amino acids. The Cα-Cβ torsion angles of the β-residues in the calculated lowest energy NMR structure of 3 (Figure 5) are 64.6° and 56.9° in D6 and L10, respectively. These values suggest that the preferred angle of 60° for the Cα-Cβ torsion in homogeneous 3₁₄-helices may be accessible in α3β helices;^74,75 although, in α3β crystal structures this angle is found to range between 70–80°.^9,13,76

Oligomer 3 with a 14-membered HBS macrocycle provides a highly stable α3β foldamer as compared to the 13-membered macrocycle in 2. This result speaks to the effective volume and helical pitch of α- versus β³-residues. Crystal structure overlays by Gellman et al. suggest that ααβαααβ and α3β repeats are suitable mimics of α-helices.^9,10 If the β³-residues have a similar pitch as the α-residues in these scaffolds, a 13- versus 14-atom nucleus should stabilize the foldamer to a similar extent. The CD results indicate that the two oligomers are equally stable, supporting the solid-state structures;⁹ however, subtle differences emerge in the NMR H/D exchange data. Although we designed the study with a biological sequence to obtain binding information, it was carefully chosen to avoid any apparent sequence dependent effects biasing the results. We are currently examining other biological sequences to confirm the finding that conformational stability of nucleated α-helices can be significantly enhanced by judicious incorporation of β³-residues. Combined, the analyses illustrate the potential of prenucleated helices to evaluate fundamental properties of nonnatural residues while controlling for any differences in nucleation.³⁵ In ongoing studies we are using the lessons learned in the present work to design next generations of protein-protein interaction inhibitors.

Experimental Section

General

Commercial-grade reagents and solvents were used without further purification except as indicated. Dichloroethane was distilled before use in the metathesis reactions. All reactions were stirred magnetically or mechanically shaken; moisture-sensitive reactions were performed under nitrogen or argon atmosphere. Reverse-phase HPLC experiments were conducted with 0.1% aqueous trifluoroacetic acid and 0.1% trifluoroacetic acid in acetonitrile buffers as eluents on C₁₈ reversed-phase columns using a Beckman Coulter HPLC equipped with a System Gold 168 Diode array detector. ESIMS data was obtained on an Agilent 1100 series LC/MSD (XCT) electrospray trap. The microwave reactions were performed in the CEM Discover single-mode reactor with controlled power, temperature, and time settings. Proton NMR spectra of HBS peptides were recorded on a Bruker AVANCE 600 or 900 MHz spectrometer.

Peptide Synthesis

Peptides were synthesized on a CEM Liberty microwave peptide synthesizer using Fmoc solid-phase chemistry on Rink amide resin, and purified by reversed-phase HPLC.⁵² The identity and the purity of the peptides were confirmed by ESI-MS.

Synthesis of HBS helices

HBS 1, HBS 2, HBS 3, and HBS 5 were synthesized as previously described.^50–53 Briefly, peptide sequences up to the i+5^th residue of the putative helix were synthesized using Fmoc solid-phase chemistry on Rink amide resin on a CEM Liberty Series microwave peptide synthesizer. N-allylation of the i+4^th residue was achieved over two steps by coupling of bromoacetic acid followed by an allylamine displacement reaction (Supporting Information, Figure S1). Coupling of the next two Fmoc-amino acid residues, followed by coupling of 4-pentenoic acid afforded the bis-olefin peptide.⁵² Ring-closing metathesis of the bis-olefin peptide was performed with Hoveyda-Grubbs II catalyst in dichloroethane under microwave irradiation as described.^50,52 Metathesized peptides were cleaved from the resin using TFA/TIS/water (95:2.5:2.5), purified by reversed-phase HPLC (C₁₈ column) and characterized by ESI-MS.

Circular dichroism spectroscopy

CD spectra were recorded on AVIV 202SF CD spectrometer equipped with a temperature controller using 1 mm length cells and a scan speed of 0.5 nm/min. The spectra were averaged over 10 scans with the baseline subtracted from analogous conditions as that for the samples. The samples were prepared in 0.1X phosphate buffered saline (13.7 mM NaCl, 1 mM phosphate, 0.27 mM KCl, pH 7.4), containing 10% trifluoroethanol, with the final peptide concentration of 100 μM. The concentrations of peptides were determined by the UV absorption of tryptophan residue at 280 nm. The helix content of the α-peptide was determined from the mean residue CD at 222 nm, [θ]₂₂₂ (deg cm² dmol⁻¹) corrected for the number of amino acids. Percent helicity was calculated from the ratio [θ]222/[θ]max, where [θ]max = (−44000 + 250T)(1 − k/n), with k = 4.0 and n = number of residues.³²

2D NMR spectroscopy

Experiments were performed on a Bruker AVANCE 600 or 900 MHz spectrometer equipped with a TXI probe (500 and 600) or a cryoprobe (900) and 3D gradient control. Samples were prepared by dissolving 2 mg of peptide in 450 μL of PBS buffer (137 mM NaCl, 10 mM phosphate, 2.7 mM KCl, pH 7.4) and 120 μL of TFE-d3. The 1D proton spectra or 2D TOCSY spectra (when overlapping is severe) were employed to discern the chemical shifts of the amide protons. Solvent suppression was achieved with a 3919 Watergate pulse sequence. All 2D spectra were recorded at 20 °C by collecting 4092 complex data points in the t2 domain by averaging 64 scans and 128 increments in the t1 domain with the States-TPPI mode. All TOCSY experiments are performed with a mixing time of 80 ms, and NOESY with the mixing time of 200 ms. The data were processed and analyzed using the Bruker TOPSPIN program. The original free induction decays (FIDs) were zero-filled to give a final matrix of 2048 by 2048 real data points. A 90° sine-square window function was applied in both dimensions.

Amide hydrogen-deuterium exchange experiments

Lyophilized samples of 1-3 from the above experiments were dissolved in 300 μL of a D2O/TFE-d3 mixture (80/20) to initiate the H/D exchange. The pH of the solution was confirmed. Spectra were recorded on a pre-shimmed Bruker AVANCE 600 or 900 MHz spectrometer. The recorded temperature was 20 °C both inside and outside the probe. The dead time was ca. 2 min. The intensity changes for each amide proton were determined by monitoring either the HN peaks on 1D spectra or the cross-peaks between HN and HR on 2D TOCSY spectra when overlapping was severe. The peak height data were fit into one phase exponential equation to get the exchange rate constants using GraphPad Prism 4.0 program.

Structure calculations

The solution structure of the peptide was computed using Monte Carlo conformational search in Macromodel 2011.^64,65 The macromodel force field was applied to a random starting conformation. A total of 74 conformers were obtained using 45 medium and long range, 51 sequential, and 104 intraresidue constraints. The 20 lowest energy structures from different starting conformations show minimal overall deviation. The NOE restraints were categorized into three groups: strong (2.5 Å upper limit), medium (4.0 Å upper limit), and weak (5.5 Å upper limit) (Supporting Information, Table S5). The ³J_NH-CHa coupling constants for all residues except G3 (due to lack of amide hydrogen) were used to calculate φ angles by application of the Pardi parameterized Karplus equation.^77,78

Description of protein binding studies

The relative affinities of peptides for N-terminal His₆-tagged MDM2 (25–117) were determined using fluorescence polarization based competitive binding assay with fluorescein labeled p53 peptide, Flu-p53. The polarization experiments were performed with a DTX 880 Multimode Detector (Beckman) at 25° C, with excitation and emission wavelengths at 485 nm and 535 nm, respectively. All samples were prepared in 96 well plates in 0.1% pluronic F-68 (Sigma). Prior to the competition experiments, the affinity of the Flu-p53 for MDM2 was determined by monitoring polarization of the fluorescent probe upon binding MDM2 (Figure S4). For competition binding experiments, appropriate concentrations of the peptides (1 nM–100 μM) were added to the MDM2-Flu-p53 mixture and the resulting solution was incubated at 25 °C for 1h before measuring the degree of dissociation of Flu-p53 by polarization. The binding affinity (K_D) values reported for each peptide are the averages of 3–5 individual experiments, and were determined by fitting the experimental data to a sigmoidal dose-response nonlinear regression model on GraphPad Prism 4.0.⁷⁹

Table 3.

Summary of amide proton temperature coefficients and deuterium exchange data for 2. β³-Residues are shown in blue font.

Peptide 2 Residues:	F4	S5	D6	Lβ7	W8	K9	L10	Lβ11	S12
H/D rate constant × 10−5 (h−1)	38.09	38.09	4.17	0.20	0.69	0.28	0.32	0.25	5.85
Protection factor (log kch/kex)	−0.59	−0.10	1.32	1.70	0.61	1.45	1.07	0.89	0.44
Stabilization, −ΔG (kcal/mol)	N/A	N/A	1.76	2.30	0.66	1.95	1.40	1.12	0.33