Interactions of the protein main chain are probed for their role in folding and self-assembly. The interactions are assessed with serine nonapeptide Ac-(Ser-Ala)4-Ser-NH2 in poly-L and alternating-L,D structure variations. Being a neutral molecule, Serine nonapeptide has been found to display not only folding-unfolding equilibrium, but also association-dissociation equilibrium as a function of solvent and concentration. Thus scrutiny of intra- and inter-molecular interactions have been undertaken in water, methanol, and DMSO solvents. In water, poly-L peptide displays a PPII-helix conformation which unfolds to extended β-conformation with increase of temperature, apparently in a two-state equilibrium. Poly-L peptide at high concentration and on transfer to the low polarity solvent, methanol, displays ordering as a β-hairpin. This implies folding of the peptide by self assembly. Self assembly and ordering possibly as double-stranded β-helix is also evidence for alternating-L,D peptide. Both isomers were observed to be unfolded in high polarity solvent DMSO. Dynamic light scattering suggests that assembly in both isomers may involve large size aggregates. The results have established that folding and self-assembly can be coupled equilibria dependent upon solute structure, concentration, and solvent. The interactions of the protein main chain involved in folding and self assembly of unfolded structure are illuminated and have been discussed.

A nascent protein on biosynthesis as unfolded structure may suffer diverse fates involving folding or self assembly. The physiologically relevant process apparently is folding to tertiary structure followed by its assembly into quaternary structure with bound stoichiometry.1 Alternatively, the polypeptide structure may assemble translationally in an open stoichiometry like a microtubule polymer.2 Other possible fates may involve misfolding of polypeptide structures followed by translational assembly to open stoichiometric complexes like amyloids.3 Amyloids as non-physiological aggregates can have pathological consequence. Thus the interplay of alternative fates i.e. folding to biologically functional state or misfolding followed by aggregation to pathological states remains one of the most vigorous pursuits in protein-folding research.4–6 

The problem of finding the physical basis of the interplay of folding and assembly is complicated by the size of proteins. Efforts to address the physical basis use molecular models that are structures simple enough to compute equilibria with rigor and yet realistic enough to observe equilibria with experiment. Using this approach, role of protein main chain has been assessed with oligoalanines models.7–10 The main chain has also been investigated by examining the effects by modifying the structure stereochemically.11–13 Poly-L structure is mutated to alternating-L,D stereochemistry and the isomers are compared in diverse solvents to test which interactions of the main chain are critical.11,14 Continuing the inquiry, this study focuses on a model suitable for studies of not only folding but also self assembly. The model is a serine variant of the previously examined lysine solubilized nonapeptide.11,14 Being a charged molecule, the peptide was observed as a monomer and thus allowed the focus to be entirely on folding-unfolding equilibrium. The serine solubilized nonapeptide, being a neutral molecule, was observed to have less solubility and thus manifested aggregation with increase of concentration. Thus equilibria of folding-unfolding was examined at low concentrations with increase of temperature, and association-dissociation was examined with concentration titration. Water, methanol, and DMSO were applied as test solvents. To probe main chain interactions, the peptide was examined in poly-L and alternating-L,D structure variations. Experiments were focused on characterization of both intra- and inter-molecular interactions. Conformational equilibria of polypeptide monomer were modeled with statistical mechanics followed by analysis of contributing microstates to understand the basis of folding-unfolding.

To investigate the competing interactions of main chain in folding and self association, we used adequately soluble serine solubilized nonapeptide. The stereochemical modification of polypeptide chain from Poly-L to alternating-L,D can impact not only interactions internal, but also external interaction of the structure. Comparison of isomers may illuminate interactions of main chain critical for not only folding but also self assembly. The triserine structure Ac-Ser-Ala3-Ser-Ala3-Ser-NH2 was made but turned out to be insoluble and unsuited for spectroscopic studies. The pentaserine structure Ac-(Ser-Ala)4-Ser-NH2 has been made and found adequately soluble for spectroscopic measurements. Though soluble, the structure displayed aggregation at high concentration with CD (Circular Dichroism) spectroscopy, DLS (Dynamic Light Scattering), and AFM (Atomic Force Microscopy). Thus the model has been applied to a study of not only folding but also self assembly. Requisite structures were made by solid phase synthesis using Fmoc chemistry. The peaks appear in ESI-MS spectra shown in Fig. S1 of Supplementary material15 and validate the structures.

CD spectra were recorded in water and methanol as a function of concentration and temperature. CD spectra of poly-L structure at 150 μm concentration recorded in water as a function of temperature are shown in Fig. 1. At low temperature the peptide displays a strong maximum of ellipticity at 198 nm and a weaker minimum at 218 nm. The spectrum has the appearance of β-hairpin signature. The peptide with multiple serines has side chain hydroxyls capable of interacting with NH and CO groups of the main chain. Such interactions within a residue are known to promote the interruption of regular helical secondary and promotion of local turns in which main chain CO and NH groups, in free access of solvent, can interact by hydrogen bonding with serine side chain OH. Consequently, serine is known as β-turns as well as β-sheet favouring residue. A β-turn in the middle of the chain could order oligopeptide chain structure as a β-hairpin allowing the main chain to hydrogen bond in an antiparallel β-sheet registry. The appearance of CD spectrum resembling that of the hairpin is thus consistent with the possibility that the poly-L structure may adopt a hairpin fold due to a serine-promoted β-turn conformation. As noted in Fig. 1, with the increase of temperature, possible β-hairpin spectrum changes gradually with the appearance of a minimum at ∼197 nm and disappearance of a minimum at 222 nm, display isodichoric point at ∼214 nm. The sigmoidal nature of the plot of ellipticity change at 197 nm with temperature suggests a two state equilibrium. On unfolding with temperature, the chain structure may reconfigure by maximizing solvation in peptides to semi-extended PPII conformation in the main chain or by maximizing electrostatics of peptides to fully extended-β conformation in the main chain. To test the transition, the temperature-unfolded peptide was allowed to anneal and the effect was monitored with CD, and results are presented in Fig. 2. Upon decrease of temperature, we expected the unfolded structure to fold as a hairpin. Surprisingly, the reversal of hairpin-like CD does not occur. Instead, ellipticity at 197 nm diminishes and at 222 nm increases. An isosbestic appears and the transition has the rough appearance of two-state equilibrium. The result is consistent with the possibility that the unfolded monomer structure is in fully extended-β conformation at high temperature. Upon cooling, the structure relaxes to semi-extended PPII conformation, possibly due to peptide groups being maximized in their solvation with water.

FIG. 1.

Temperature dependence of CD spectra of poly-L nonapeptide (150 μM concentration) in water during heating the sample. The inset shows the plot of intensity as a function of temperature at wavelength 197 nm.

FIG. 1.

Temperature dependence of CD spectra of poly-L nonapeptide (150 μM concentration) in water during heating the sample. The inset shows the plot of intensity as a function of temperature at wavelength 197 nm.

Close modal
FIG. 2.

Temperature dependence of CD spectra of poly-L nonapeptide (150 μM concentration) in water during cooling the sample. The inset shows the plot of intensity as a function of temperature at wavelength 197 nm.

FIG. 2.

Temperature dependence of CD spectra of poly-L nonapeptide (150 μM concentration) in water during cooling the sample. The inset shows the plot of intensity as a function of temperature at wavelength 197 nm.

Close modal

The possible conformational changes were also measured with concentration dependent CD, and the results are reported in Fig. 3. The appearance of isodichoric point ∼214 nm and sigmoidal trend of ellepticity at 197 nm as a function of concentration indicate a two state equilibrium, similar to observation with temperature. At low concentration, appearance of a minimum of ellipticity at 197 nm suggests the ordering of the peptide as a monomeric PPII - helix. This may be due to maximization of peptide groups for solvation with water, which is consistent with the reported observation.16,17 The result implies that with the increase of temperature, peptide oligomers that have a β-hairpin conformation are converted to monomers with an extended-β conformation. At high temperature, with an increase of solvent entropy, the main chain presumably desolvates and is maximized in electrostatics of peptides.

FIG. 3.

Concentration dependence of CD of poly-L nonapeptide in water. The inset shows the plot of intensity as a function of increasing peptide concentration at wavelength 197 nm.

FIG. 3.

Concentration dependence of CD of poly-L nonapeptide in water. The inset shows the plot of intensity as a function of increasing peptide concentration at wavelength 197 nm.

Close modal

The poly-L structure was also tested for the effects of solvent composition. The CD spectra of peptide in water-methanol progressive titration, and of temperature melting in methanol are presented in Fig. S2 of Supplementary material.15 With an increase in the proportion of methanol, the CD pattern does not change except for the intensity of the minimum and maximum ellipticity. Furthermore, we did not observe significant changes in CD pattern on increase of temperature in methanol. Presumably, the interactions ordering the peptide as an assembly of hairpins are strong in low polarity solvent methanol.18 

Assessment of solvent-dependent conformation of peptide at 250 μM concentrations was made with FTIR. FTIR spectra of peptide in water, methanol, and DMSO are reported in Fig. 3 of Supplementary material.15 The spectrum in DMSO is characterized with a single, strong band at wave number 1664 cm−1. A peak at 1664 cm−1, characterizes the free carbonyls that do not participate in hydrogen bonding,19,20 suggests the unfolded conformation of peptide in DMSO. In water, peptide manifests strongest bands at 1639 and 1649 cm−1, attributed to extended β structure and random coil respectively.19,21,22 In light of recent insights, we assume the random coil in short peptides as PPII conformation. Additional IR peak at 1628 and 1684 cm−1 have been attributed to a signature of aggregated β-hairpins.19,23 In summary, the IR results suggest the ordering of peptide as β-hairpins aggregates in water and methanol, whereas unfolding of peptide in DMSO.

An attempt was made to characterize aggregation of the peptide with AFM and DLS; results are presented in Fig. 4. The AFM experiments were performed on aqueous samples after air drying the solutions. Ordering of large aggregates of irregular morphology is observed in Fig. 4(a). The DLS experiments were performed at two concentrations 250 μM and 25 μM and results are shown in Fig. 4(b). The observed hydrodynamic radius at higher concentration is ∼400 nm and at low concentration is ∼50nm. The results suggest the ordering of peptide to large size aggregates at higher concentration, whereas to a smaller size at low concentration.

FIG. 4.

AFM images (panel a), and Particle size distribution by DLS for the poly-L nonapeptides at concentrations of 25 μM (panel b), and 250 μM (panel c), showing nonapeptide to populate in higher order assemblies.

FIG. 4.

AFM images (panel a), and Particle size distribution by DLS for the poly-L nonapeptides at concentrations of 25 μM (panel b), and 250 μM (panel c), showing nonapeptide to populate in higher order assemblies.

Close modal

The solvent dependent conformational transition of the alternating-L,D isomer was investigated by water-methanol titration with CD, and the results are presented in Fig. 5. We observed a sharp transition in the range of 70-90% methanol in water. The CD spectrum of the peptide in water appears to be low in ellipticity with a single broad minima at ∼220 nm. Contrastingly, in methanol, the spectra features a weak maximum at ∼220 nm and strong minimum at ∼198 nm, which is assigned to the β-helices of gramicidine-A.24,25 We tested the concentration dependency of observed conformation in methanol, result are shown in Fig. S5(a) of Supplementary material.15 The little variation over concentration suggests that that peptide adopt methanol-induced ordered conformation even at pretty low concentration (20 μm), which intensifies further with increase of concentration. The results of temperature dependent CD titration in methanol are shown in Fig. S4 of Supplementary material.15 According to result, the methanol induced ordered structures unfold at high temperature, display two state behavior with a midpoint at ∼50 °C; the high temperature CD is similar to that of the peptide in water. The CD spectra of the peptide on cooling reported in Fig. S5(b) of Supplementary material,15 indicate the irreversibility of temperature dependent order-disorder transition in methanol.

FIG. 5.

Solvent dependence of CD of alternating-L,D nonapeptide (60 μM concentration) in water and methanol progressive titration. The inset shows the plot of intensity as a function of increasing methanol concentration at wavelength 197 nm.

FIG. 5.

Solvent dependence of CD of alternating-L,D nonapeptide (60 μM concentration) in water and methanol progressive titration. The inset shows the plot of intensity as a function of increasing methanol concentration at wavelength 197 nm.

Close modal

Assessment of solvent-dependent conformation change were made with FTIR and the results are presented in Fig. S6 of Supplementary material.15 In DMSO, we observed a intense band at 1665 cm−1, which is believed to be a characteristic of non-hydrogen bonded free peptide-carbonyls groups. This suggest the unfolding of peptide due to strong solvation of peptide-NHs by DMSO. The spectra in water and methanol are broadly similar in appearance. The strongest peak in water at 1649 cm−1 has been attributed in literature to random coil structure.21 In methanol, the peak at 1649 cm−1 been assigned to β4.4 helix dimer.24,26 We observed strongest peak in methanol spectrum is at 1655 cm−1, that has been attributed to monomeric β4.4 helix.24,26 According to reported values, additional peakes at 1636, 1670, 1690 cm−1 were assigned to double stranded β5.6 helix.

An attempt was made to characterize aggregation with the AFM experiment of an air dried methanol solubilized peptide sample; results are presented in Fig. 6. It appears that alternating-L,D peptide assembled as smaller size aggregates than the observed aggregates of poly-L structure. To know the concentration dependency over the size of peptide aggregate, we performed DLS experiment and the results are presented in Fig. 6. The hydrodynamic radius of aggregated particles were observed to be ∼13 nm at 25 μm and ∼38 nm at 250 μm concentration. Contransting to poly-L peptide aggregates, alternating-L,D peptide aggregation appears to be smaller in size and independent of concentration.

FIG. 6.

AFM images (panel a), and Particle size distribution by DLS for the alternating-L,D nonapeptides at concentrations of 25 μM (panel b), and 250 μM (panel c), showing nonapeptide to populate in higher order assemblies.

FIG. 6.

AFM images (panel a), and Particle size distribution by DLS for the alternating-L,D nonapeptides at concentrations of 25 μM (panel b), and 250 μM (panel c), showing nonapeptide to populate in higher order assemblies.

Close modal

Monomers were modeled to equilibria and were analyzed in their contributing microstates of polypeptide structure. Conformational equilibria were computed with GROMACS at 298 K under NVT condition over 250 ns in water, methanol, and DMSO as solvents. A modified Gromos96 force field modified for treatment of D amino acids was applied. Evolution of discrete folds populating equilibria was assessed by clustering to a 0.15 nm RMSD cutoff over backbone atoms. The central member of each cluster was taken to model a discrete fold populating the equilibrium. Achievement of equilibrium was judged from the attainment of an asymptote in the time evolution of microstates. Results summarized in Fig. S7 of Supplementary material,15 indicate that asymptotes were attained or are reasonably well approximated in all ensembles. According to results reported in Table I, alternating-L,D isomers populate less number of microstates than poly-L peptide in water and methanol, whereas as to a higher density in DMSO. According to radius of gyration (Rg) distribution noted in Table I and Fig. 7, conformational ensemble of alternating-L,D peptide observed to be more compact than the poly-L structure in all solvents. Poly-L peptide has a bimodal distribution of Rg in water and a sharp maximum appears at 0.5 nm suggesting the population of mainly folded microstates. Rg distribution of poly-L peptide in methanol and DMSO manifests maxima around 0.85 nm, suggesting largely unfolded microstates.

Table I.

Statistics over macrostates of percentage of specific hydrogen bond, radius of gyration, and occupancies in ϕ,ψ basins.

  Rg (nm)% Occupency #% Hydrogen bonds$
 MTotalAvg.M1stαβPPIIAvg.SRMRLR
poly-L 
Water 847 0.62 ± 0.14 0.49 ± 0.03 24.2 36.9 31.6 1.4 19.8 57.0 23.2 
Methanol 653 0.77 ± 0.11 0.90 ± 0.02 10.6 48.5 35.1 0.6 63.2 26.7 10.1 
DMSO 259 0.82 ± 0.07 0.90 ± 0.03 8.1 45.9 39.2 0.2 93.1 6.9 0.0 
alternating-L,D 
Water 307 0.43 ± 0.03 0.42 ± 0.01 7.9 48.5 37.5 2.7 12.8 40.7 46.5 
Methanol 537 0.46 ± 0.05 0.44 ± 0.02 5.1 56.7 31.7 2.9 12.1 37.3 50.6 
DMSO 358 0.62 ± 0.07 0.55 ± 0.03 7.9 57.5 27.5 0.2 37.7 55.8 6.5 
  Rg (nm)% Occupency #% Hydrogen bonds$
 MTotalAvg.M1stαβPPIIAvg.SRMRLR
poly-L 
Water 847 0.62 ± 0.14 0.49 ± 0.03 24.2 36.9 31.6 1.4 19.8 57.0 23.2 
Methanol 653 0.77 ± 0.11 0.90 ± 0.02 10.6 48.5 35.1 0.6 63.2 26.7 10.1 
DMSO 259 0.82 ± 0.07 0.90 ± 0.03 8.1 45.9 39.2 0.2 93.1 6.9 0.0 
alternating-L,D 
Water 307 0.43 ± 0.03 0.42 ± 0.01 7.9 48.5 37.5 2.7 12.8 40.7 46.5 
Methanol 537 0.46 ± 0.05 0.44 ± 0.02 5.1 56.7 31.7 2.9 12.1 37.3 50.6 
DMSO 358 0.62 ± 0.07 0.55 ± 0.03 7.9 57.5 27.5 0.2 37.7 55.8 6.5 
a

*MTotal = Number of microstates in total ensemble; # Basin definitions are, α: L/Dϕ = −/+ 20 to −/+ 100, L/Dψ = −/+ 20 to −/+ 80; β: L/Dϕ = −/+ 30 to −/+ 170, L/Dψ = +/− 80 to +/− 180; ϕF: Lϕ = 0 to −180, Dϕ = 0 to 180; ϕU: Lϕ = 0 to 180, Dϕ = 0 to -180. $The hydrogen bonds are short (SR: i-i±2), medium (MR: i-i±3 + i-i±4), and long ranged (MR: i-i±5 + i-i±≥6).

FIG. 7.

Radius of gyration distribution of main chain atoms over all conformer populating macrostate of nonapeptide structures in specific solvents.

FIG. 7.

Radius of gyration distribution of main chain atoms over all conformer populating macrostate of nonapeptide structures in specific solvents.

Close modal

Conforming to the above results are statistics of macrostates in hydrogen bonds and in occupancies of specific basins of ϕ, ψ over entire ensemble. Poly-L peptide has its highest occupancy in α basin, ∼25% in water and lowest ∼8% in methanol. Contrastingly, alternating-L,D structure has comparable <8% occupancy in α basins in all three solvents. In spite of low occupancy in α basin, the macrostate in alternating-L,D structure has close to 3 main chain-main chain hydrogen bonds per molecule in water and methanol. Poly-L structure has an average of < 1.5 hydrogen bonds in water. In DMSO both the isomers are devoid of any intamolecular hydrogen bonding. The populated hydrogen bonds in alternating-L,D structure are mainly long ranged (LR) hydrogen bonds, whereas enrichment of medium ranged (MR) hydrogen bonds were observed in poly-L peptide.

Backbone trace representation of top 5 microstates of both the isomer, in all three solvent, are presented in Fig. S8-S10 of Supplementary material.15 Poly-L structure populates a greater diversity of folds in water compared to methanol and DMSO. In water, the ensembles appears to be largly populated by in hairpin-like folds. Though the most populous microstate is helical in simulation, but experimental observation indicate the presence of β-hairpin folds, which could be due to self association. CD spectra at low concentration suggest that peptide is in PPII conformation, which is the third most populous microstate in simulation. The microstates in methanol are ensembles of distorted PPII helices, which does not agree with our experimentally observed β-hairpin conformation. The β-hairpin conformation of poly-L peptide may be favored due to self association, which cannot be seen in simulation with monomeric species. Both simulation and experiment indicate the presence of unfolded structure in DMSO. In water and methanol, peptide hydrogen bonds are evident in ordering a relatively compact structure of alternating-L,D peptide. In methanol, the chain structure is practically excluded from the α basin and is folded over LR hydrogen bonds as right handed β4.4 helix in the first microstate, as β5.6 helix in the second microstates, as left handed β4.4 helix in the third microstate, and as left handed β5.6 helix in the fourth microstate. In DMSO, the microstates are largely alternating Lβ, Dβ structures, fully solvated in main chain and thus devoid of hydrogen bonds.

Interactions of solvent with polypeptide structures have been analyzed on the basis of solvent shell analysis. The calculations of RDF and SDF, over total ensemble, are presented in Fig. S11 of Supplementary material.15 The RDF peaks of the peptides are comparable in magnitude in water, and in DMSO, but are varied in methanol. The absence of the RDF maxima of the water oxygen atom against backbone-NHs and backbone-COs suggest the sequestration of backbone atoms and their involvement in intramolecular hydrogen bonding with other main chain atoms. The poly-L peptide manifests RDF maxima of the oxygen atom of methanol against backbone-NHs and backbone-CO at 0.28 nm which is the typical signature of hydrogen bonding. The alternating-L,D peptide lacks the first RDF maxima of methanol oxygen atoms against backbone atoms suggesting the sequestration of backbone atoms and their involvement in intra-molecular hydrogen bonds. The oxygen of DMSO manifests a peak in RDF against backbone NHs at 0.28 nm for both poly-L and alternating-L,D peptide, implies strong solvation of backbone-NHs with DMSO. The RDF results support the observed folded structure of both isomers in water and the unfolded structure in DMSO. Additionally, the RDF results support the experimentally observed folded structure of alternating-L,D peptide in methanol.

Cβ of alanines display a first RDF maxima at 0.36 nm against the oxygen atom of solvents, and another maxima at 0.4 nm against methyls of methanol and DMSO. These results suggest solvation of Cβ atoms of alanines by combination of C-H…O hydrogen bond and vdW interactions. Oγ of serines manifest the first RDF maxima at 0.28 nm against the oxygen atom of solvents, which suggests solvation of Oγ of serines through hydrogen bonding.

Interactions of the main chain are ubiquitous to protein structure in folding-unfolding equilibrium.27 It remains a tough challenge to find the physical basis and to quantify the exact contributions to energetics of the equilibrium. The interactions mediated by hydrogen bonds define specific folds of the protein main chain.28,29 The interactions of the peptide dipoles with solvent are now found to define the unfolded protein structure in energetics as well as in specificity of conformation.16,17,30–38 As per classical view of protein folding, conformational entropy of the unfolded polypeptide structure was considered to be a driving force of unfolding of protein as a random coil.39–41 The classical view of folding also considers the gain of solvent entropy, which is accompanied with burial of side chains into a solvent-sequestered core of folded protein, as the fundamental thermodynamic drive in protein folding.42 Probably entropy changes in polypeptide chain and in bulk solvent do contribute in thermodynamics of folding. However, interactions internal to the main chain and of the main chain with solvent are possibly the additional critical effects involved. The fact that the interactions of the main chain, rather than the entropy difference between the folded and unfolded structure, may be critical came to light largely from studies of oligoalanines as main chain models.16,17,27 Combination of spectroscopy and computational studies involving both quantum mechanics and statistical mechanics proved that the local segments of polypeptide structure are not completely random coils, but remarkably well ordered in water.7–10,43–47 It was established further that interactions of peptides with solvent water are strong enough to be an important thermodynamic drive in unfolding proteins.16,17

The intense studies of oligoalanines involving observation of specific folds and calculation of energetics have provided new insights. However, the complex effects of solvent and sequences are still not well addressed. Our lab pioneered and applied a unique approach to unraveling the complex energetics of main chain structure involved in folding-unfolding equilibrium. The mutual interactions among peptides of the main chain were probed by mutating the structure from natural poly-L to artificial alternating-L,D stereochemistry. This modification could alter the critical balance of forces involved, and the balance was probed with diverse solvents. An oligoalanine solubilized with lysine inserts was modeled to equilibria using water, methanol, and DMSO as test solvents and was targeted for observation mainly with CD and NMR.11,14 The studies have established that not only hydrogen bonds among peptides and their interaction with solvent, but also strong field effects of neighboring peptide dipoles, are critical in folding-unfolding equilibrium. The electrostatic interaction between the neighboring peptide dipoles could provide a specific basis to explain solvent and possibly sequence roles.

The present study has provided new insights into the roles of stereochemistry and solvent in intermolecular interactions of oligopeptides. The study is performed with oligoalanines involving the application of serines as solubilizing inserts. Serine nonapeptides being neutral manifest poorer solubility than lysine nonapeptides, which allowed the study of intermolecular interactions and their impacts on conformation. Therefore, the interactions of the models were examined as a function of concentration to probe association-dissociation equilibria, and as a function of temperature to probe folding-unfolding equilibria. The molecular dynamics studies were implemented for monomolecular species of the structures and thus are relevant only to draw relevant inferences about folding-unfolding equilibrium.

On the basis of FTIR and MD, the peptides appear to be so strongly solvated in DMSO that neither folding nor self assembly appears to be possible. According to CD, the PPII helix appears to be the minimum of energy in water at low temperature while extended-β structure may the minima of energy in this solvent at high temperature, specifically in poly-L structure. The equilibrium between these conformations structures has been reported in the literature.16 The critical effects appear to be maximization of peptides to solvation at low temperature, which necessitates residues to be in PPII conformation, and maximization of peptides in electrostatics at elevated temperature as an increase of solvent entropy promotes desolvation of peptides, which is possible when the structure is in extended-β conformation.

Poly-L peptide, on increase of concentration and on transfer to low polarity solvent methanol, displays ordering as a β-hairpin according to CD, implying folding on the basis of self assembly. This observation is again supported by DLS and AFM indicating the formation of large size aggregates. Possible assembly of the structure as a β-hairpin and dependence of folding on intermolecular interactions is a noteworthy observation with possible critical implications for ordering of unfolded proteins as an amyloid aggregates. Amyloids appear to essentially involve β-sheets as the organizing element of the structures. The self assembly observed in alternating-L,D structures presumably involves the ordering to gramicidin-like double stranded helical structures. The present study furnishes an interesting model to dissect the assembly in molecular principle.

The diverse observations are conformed to the model, according to which the interactions of the main chain are a function of at least two independent effects. One effect is the strength of solvation of peptides based on dipolar strength of solvent. As the strongest dipole, DMSO prevents peptide hydrogen bonds and thus prevents folding of the chain irrespective of stereochemistry. Probably, with this effect, the DMSO may also prevent oligopeptides from self assembly. Another effect is solvent's role as dielectric. As a poor screen of electrostatics of poly-L structure, methanol was found to promote fully-extended-β and semi-extended-PPII conformation in poly-L structure. The same conformational fold in residue level structures are promoted in alternating-L,D structure, and peptide hydrogen bonds are also ordered without occupancy of α basins. As a dielectric of intermediate strength, DMSO was found to promote ordering of isolated residues to α conformation in both poly-L and alternating-L,D diastereomer structures. Thus the ordering of residues to α conformation is an effect of a solvent dielectric, independent of promotion of occupancy of α basin due to mutual hydrogen bonding of peptides. Thus water contributes to the folding of poly-L structure as a strong dielectric, promoting excursion of residues to α conformation, as well as a relatively weak dipole, relative to DMSO, by allowing peptide hydrogen bonds. With water as the solvent, the oligomer assembles and folds as a hairpin. The poor dipole strength of water, relative to DMSO, may promote mutual hydrogen bonding interaction of peptides. Ordering of hydrogen bonds with concentration apparently promotes assembly, which may trigger folding. At low concentration, monomeric peptides are presumabely solvated in main chain and are thus promoted to a semi-extended structure as a PPII-helix structure. With the increase of temperature, we observed monomer-oligomer equilibrium in water contrasted to methanol where aggregates did not revert with temperature. Presumably, self assembly is strongest in low polarity solvent methanol and apparently of weaker strength in water.

Interactions of the polypeptide main chain were probed in a serine solubilized nonapeptide. The interactions were probed by examining the effect of mutating the structure from poly-L to alternating-L,D stereochemistry. The isomers were examined with water, methanol, and DMSO as test solvents. Poly-L peptide at low concentration in water displays semi-extended PPII-helix conformation which unfolds to fully extended β-conformation at higher temperature, apparently in two-state equilibrium. Poly-L peptide displays ordering as a β-hairpin at higher concentration and in low polarity solvent methanol, implying folding on basis of self assembly. Alternating-L,D peptide also manifest the self assembly and ordering possibly as double-stranded β-helix in methanol. In high polarity solvent DMSO,both isomers were observed to be fully unfolded. The results have established that folding and self-assembly can be coupled equilibria dependent upon solute structure, concentration, and solvent.

Synthesis was performed on Rink Amide AM resin using standard Fmoc chemistry and HOBt/DIC as coupling reagents.48 Each coupling, monitored with Kaiser and chloranil tests, typically required about 6 hrs. Deprotections were carried out with 30% (v/v) piperidine-DMF. N-terminus was acetylated (-NHCOCH3) with Ac2O: DIPEA:DMF in 1:2:20 ratio. The cleavage of the final polypeptide and deprotection of side chains were achieved together with reagent K (82.5% TFA/5% dry-phenol/5% thioanisole/2.5% ethandithiol/5% water). The product precipitated with anhydrous diethyl ether was lyophilized from 1:4 H2O:tBuOH solution as a white powder. Peptide purity was assessed with HPLC over RP-C18 (10 μM, 10 mm × 250 mm; Merck) eluting with CH3CN\H2O (0.1%TFA) 0-100% gradients. Mass spectra were recorded by QTOF-ESI MS.

Circular Dichroism (CD) was recorded on JASCO J-810 CD spectropolarimeter at 298 K in 0.2 cm path length quartz cell with 2 nm bandwidth in far-UV (190-250 nm) range. Scanning at 100 nm/min with 1.0 s time constant in 1 nm steps, five scans were averaged after baseline correction for solvent. Working solutions of specific concentration of peptides were prepared by optical measurements. The observations in millidegrees were converted to molar residue ellipticity [θMRW].

FTIR spectra were recorded on Perkin Elmer FTIR spectrometer. Data were collected in amide I region varying from 1600-1700 cm−1. Peptides were dissolved in specific solvents and then drop was put on KBr pellet.

The size-distributions of the peptide samples at different concentrations were assessed by dynamic light scattering instrument (Model: 90 Plus Particle Size Analyzer); Brookhaven Instrument Co., Holtsville, NY, USA) which is a technique to detect the fluctuations of the scattering intensity due to the Brownian motion of particles in suspension using a high power 35 mW diode laser source. The measurements were carried out at 90° scattering angle. The graphs were processed with the software program 90 Plus particle sizing software, Brookhaven Instrument Corporation. The samples were prepared in water or methanol and filtered using 0.22 μ filter followed by its bath sonication for one hour. DLS was performed on the peptide samples at two different concentrations namely 25 μM and 250 μM.

An aliquot of peptide solution (150 μL) incubated in water was transferred onto a freshly cleaved mica surface and the sample solution was uniformly spread throughout. The sample coated mica was air dried for 12 hours at room temperature and imaged by Vecco Digital Instrument Nanoscope-IV at a scanning rate of 1.507 Hertz.

Peptides were modeled either with the in-house software package CAPM (Computer Aided Peptide Modeling), capable of handling D-amino acid effectively. In-house program PDBmake was used for generation of PDB coordinates of CAPM modeled structure.

Molecular dynamics was performed with gromos-96 43A1 force field in GROningen MAchine for Chemical Simulations (GROMACS) 3.3.3 package49,50 in a periodic box of explicit solvent, using the water, methanol and DMSO. The simulation was performed under NVT condition, viz., fixed number of particles, constant volume, and constant temperature. Non-bonded list cutoff was 1.4 nm with shift at 0.8 nm. Integration step was 2 fs. Initial velocities were drawn from Maxwellian distribution. Temperature was coupled to an external bath with relaxation time constant of 0.1 ps. Bond lengths were constrained with SHAKE51 to geometric accuracy 10−4. Electrostatics was treated SHIFT method implementing a Coulomb cutoff of 1.4 nm, Fourier spacing of 0.12 nm, and an interpolation order of 4. Poly-L isomers were modeled in extended conformation with ϕ = -120o, ψ = 120o and alternating-L,D isomers were modeled in extended conformation with ϕL/D = −/+ 90o, ψL/D = −/+ 90o. Simulations were performed in five parallel trajectories by picking some random conformation sampled in one run to avoid any conformational biasing of starting conformation. Simulation was initialized and 3 ns trajectory was exempted from analysis as pre-equilibration period. The individual trajectories were then merged together. Peptides constrained to center of the periodic cubic box were soaked in specific solvents, which was added to 1 atm density at 298 K. First solute was energy minimized, and then solvent was energy minimized while restraining solute, and finally both were energy minimized after removing restraint. The trajectory was sampled thereafter at 10 ps intervals.

Peptide conformers were clustered in Cartesian space to ≤ 0.15 nm RMSD cutoff over backbone atoms using Daura et al. algorithm.52 This gives microstates in diminishing population, viz., diminishing thermodynamic stability. Helmholz free energy was calculated from relative probabilities pA and pB of finding the system in microstate A and B as: ΔFA-B = -RT ln pB/pA, with R as gas constant, T as temperature, and pA and pB the number of members in microstates A and B. Radius of gyration were calculated using g_gyrate utility in GROMACS.

Radial Distribution Function and Spatial Distribution Functions of specific solvent atoms were calculated over the most populous polypeptide microstate in each ensemble using g_rdf and g_spatial utility in GROMACS.

We acknowledge DST (09DST028), Government of India, for financial support and IIT Bombay for the computing facility “Corona”. KRS is recipients of fellowships from Council of Scientific and Industrial Research (CSIR).

1.
D. S.
Goodsell
and
A. J.
Olson
,
Annu. Rev. Biophys. Biomol. Struct.
29
,
105
(
2000
).
2.
E.
Pouget
 et al,
J. Am. Chem. Soc.
132
,
4230
(
2010
).
3.
F.
Chiti
and
C. M.
Dobson
,
Annu. Rev. Biochem.
75
,
333
(
2006
).
5.
C. M.
Dobson
,
Nat. Rev. Drug Discov.
2
,
154
(
2003
).
6.
A. W.
Fitzpatrick
,
T. P.
Knowles
,
C. A.
Waudby
,
M.
Vendruscolo
, and
C. M.
Dobson
,
PLoS Comput. Biol.
7
,
e1002169
(
2011
).
7.
A.
Chakrabartty
and
R. L.
Baldwin
,
Adv. Protein Chem.
46
,
141
(
1995
).
8.
V.
Ramakrishnan
,
R.
Ranbhor
, and
S.
Durani
,
J. Am. Chem. Soc.
126
,
16332
(
2004
).
9.
Z.
Shi
,
C. A.
Olson
,
G. D.
Rose
,
R. L.
Baldwin
, and
N. R.
Kallenbach
,
Proc. Natl. Acad. Sci. USA
99
,
9190
(
2002
).
10.
Z.
Shi
,
R. W.
Woody
, and
N. R.
Kallenbach
,
Adv. Protein Chem.
62
,
163
(
2002
).
11.
A.
Kumar
,
V.
Ramakrishnan
,
R.
Ranbhor
,
K.
Patel
, and
S.
Durani
,
J. Phys. Chem. B
113
,
16435
(
2009
).
12.
V.
Ramakrishnan
,
R.
Ranbhor
,
A.
Kumar
, and
S.
Durani
,
J. Phys. Chem. B
110
,
9314
(
2006
).
13.
R.
Ranbhor
,
V.
Ramakrishnan
,
A.
Kumar
, and
S.
Durani
,
Biopolymers
83
,
537
(
2006
).
14.
K. R.
Srivastava
,
A.
Kumar
,
B.
Goyal
, and
S.
Durani
,
J. Phys. Chem. B
115
,
6700
(
2011
).
15.
See supplementary material at http://dx.doi.org/10.1063/1.4885475 for mass spectra of peptides along with other spectral plots (CD & FTIR) illustratiing temperature, solvent and concentration dependence of peptides conformation. The material also summarise the analysis of MD results illustrating monomer conformational dependence on solvents.
16.
F.
Eker
,
K.
Griebenow
, and
R.
Schweitzer-Stenner
,
J. Am. Chem. Soc.
125
,
8178
(
2003
).
17.
M.
Mezei
,
P. J.
Fleming
,
R.
Srinivasan
, and
G. D.
Rose
,
Proteins
55
,
502
(
2004
).
18.
S.
Hwang
,
Q.
Shao
,
H.
Williams
,
C.
Hilty
, and
Y. Q.
Gao
,
J. Phys. Chem. B
115
,
6653
(
2011
).
19.
M.
Jackson
and
H. H.
Mantsch
,
Crit. Rev. Biochem. Mol. Biol.
30
,
95
(
1995
).
20.
C.
Liu
,
A.
Bo
,
G.
Cheng
,
X.
Lin
, and
S.
Dong
,
Biochim. Biophys. Acta
1385
,
53
(
1998
).
21.
A.
Barth
,
Biochim. Biophys. Acta
1767
,
1073
(
2007
).
22.
P. I.
Haris
and
F.
Severcan
,
Journal of Molecular Catalysis B: Enzymatic
7
,
207
(
1999
).
23.
E.
Saguer
,
P.
Alvarez
,
J.
Sedman
,
H. S.
Ramaswamy
, and
A. A.
Ismail
,
Food Hydrocolloids
23
,
874
(
2009
).
24.
S. V.
Sychev
,
L. I.
Barsukov
, and
V. T.
Ivanov
,
Eur. Biophys. J.
22
,
279
(
1993
).
25.
D. W.
Urry
,
J. D.
Glickson
,
D. F.
Mayers
, and
J.
Haider
,
Biochemistry
11
,
487
(
1972
).
26.
V. M.
Naik
and
S.
Krimm
,
Biochem. Biophys. Res. Commun.
125
,
919
(
1984
).
27.
G. D.
Rose
,
P. J.
Fleming
,
J. R.
Banavar
, and
A.
Maritan
,
Proc. Natl. Acad. Sci. USA
103
,
16623
(
2006
).
28.
L.
Pauling
and
R. B.
Corey
,
Proc. Natl. Acad. Sci. USA
37
,
729
(
1951
).
29.
L.
Pauling
,
R. B.
Corey
, and
H. R.
Branson
,
Proc. Natl. Acad. Sci. USA
37
,
205
(
1951
).
30.
J.
Apostolakis
,
P.
Ferrara
, and
A.
Caflisch
,
J. Chem. Phys.
110
,
2099
(
1999
).
31.
F.
Avbelj
and
R. L.
Baldwin
,
Proc. Natl. Acad. Sci. USA
99
,
1309
(
2002
).
32.
F.
Avbelj
and
R. L.
Baldwin
,
Proc. Natl. Acad. Sci. USA
100
,
5742
(
2003
).
33.
H. S.
Chan
and
K. A.
Dill
,
Annu. Rev. Biophys. Biomol. Struct.
26
,
425
(
1997
).
34.
Y.
Levy
and
J. N.
Onuchic
,
Annu. Rev. Biophys. Biomol. Struct.
35
,
389
(
2006
).
36.
M.
Sundaralingam
and
Y. C.
Sekharudu
,
Science
244
,
1333
(
1989
).
37.
C. S.
Poon
,
E. T.
Samulski
,
C. F.
Weise
, and
J. C.
Weisshaar
,
J. Am. Chem. Soc.
122
,
5642
(
2000
).
38.
H. T.
Tran
,
A.
Mao
, and
R. V.
Pappu
,
J. Am. Chem. Soc.
130
,
7380
(
2008
).
39.
J. D.
Bryngelson
,
J. N.
Onuchic
,
N. D.
Socci
, and
P. G.
Wolynes
,
Proteins
21
,
167
(
1995
).
40.
P. J.
Flory
, (InterScience Publishers, New York,
1969
).
41.
P. E.
Leopold
,
M.
Montal
, and
J. N.
Onuchic
,
Proc. Natl. Acad. Sci. USA
89
,
8721
(
1992
).
42.
K. A.
Dill
,
T. M.
Truskett
,
V.
Vlachy
, and
B.
Hribar-Lee
,
Annu. Rev. Biophys. Biomol. Struct.
34
,
173
(
2005
).
43.
T.
Head-Gordon
,
F. H.
Stillinger
,
M. H.
Wright
, and
D. M.
Gay
,
Proc. Natl. Acad. Sci. USA
89
,
11513
(
1992
).
44.
J.
Makowska
,
K.
Baginska
,
A.
Skwierawska
,
A.
Liwo
,
L.
Chmurzynski
, and
H. A.
Scheraga
,
Biopolymers
90
,
724
(
2008
).
45.
J.
Makowska
,
S.
Rodziewicz-Motowidło
,
K.
Baginska
,
M.
Makowski
,
J. A.
Vila
,
A.
Liwo
,
L.
Chmurzynski
, and
H. A.
Scheraga
,
Proc. Natl. Acad. Sci. USA.
103
,
1744
(
2006
).
46.
J.
Makowska
,
S.
Rodziewicz-Motowidło
,
K.
Bagińska
,
M.
Makowski
,
J. A.
Vila
,
A.
Liwo
,
L.
Chmurzyński
, and
H. A.
Scheraga
,
Biophys. J.
92
,
2904
(
2007
).
47.
B.
Zagrovic
,
G.
Jayachandran
,
I. S.
Millett
,
S.
Doniach
, and
V. S.
Pande
,
J. Mol. Biol.
353
,
232
(
2005
).
48.
W. C.
Chan
and
P. D.
White
, (IRS Press, Oxford, England,
1989
).
49.
E.
Lindahl
,
B.
Hess
, and
D.
Van Der Spoel
,
J. Mol. Mod.
7
,
306
(
2001
).
50.
W. F.
Van Gunsteren
,
S. R.
Billeter
,
A. A.
Eising
,
P. H.
Hünenberger
,
P.
Krüger
,
A. E.
Mark
,
W. R. P.
Scott
, and
I. G.
Tironi
, Hochschulverlag AG an der ETH Zürich, Zürich (
1996
).
51.
J. P.
Ryckaert
,
G.
Ciccotti
, and
H. J. C.
Berendsen
,
J. Comp. Phy.
23
,
327
(
1977
).
52.
X.
Daura
,
W. F.
van Gunsteren
, and
A. E.
Mark
,
Proteins
34
,
269
(
1999
).

Supplementary Material