Probing the Nature of Noncovalent Interactions in Dimers of Linear

Jan 10, 2019 - Search; Citation; Subject. Search in: .... In contrast, the lowest energy dimer of linear WY is stabilized by Ocarboxyl···H(N)trp an...
0 downloads 0 Views 3MB Size
Download PDF
This is an open access article published under an ACS AuthorChoice License, which permits copying and redistribution of the article or any adaptations for non-commercial purposes.

Article Cite This: ACS Omega 2019, 4, 911−919

http://pubs.acs.org/journal/acsodf

Probing the Nature of Noncovalent Interactions in Dimers of Linear Tyrosine-Based Dipeptides Maricris L. Mayes*,† and Lisa Perreault† Department of Chemistry and Biochemistry, University of Massachusetts Dartmouth, North Dartmouth, Massachusetts 02747, United States

ACS Omega 2019.4:911-919. Downloaded from pubs.acs.org by 109.94.221.15 on 01/10/19. For personal use only.

S Supporting Information *

ABSTRACT: Tyrosine-based dipeptides self-assemble to form higher order structures. To gain insights into the nature of intermolecular interactions contributing to the early stages of the self-assembly of aromatic dipeptides, we study the dimers of linear dityrosine (YY) and tryptophan−tyrosine (WY) using quantum-chemical methods with dispersion corrections and universal solvation model based on density in combination with energy decomposition and natural bond orbital (NBO) analyses. We find that hydrogen bonding is a dominant stabilizing force. The lowest energy structure for the linear YY dimer is characterized by Ocarboxyl···H(O)tyr. In contrast, the lowest energy dimer of linear WY is stabilized by Ocarboxyl···H(N)trp and πtyr···πtyr. The solvent plays a critical role as it can change the strength and nature of interactions. The lowest energy for linear WY dimer in acetone is stabilized by Ocarboxyl···H(O)tyr, πtrp···H(C), and πtrp···H(N). The ΔG of dimerization and stabilization energies of solvated dipeptides reveal that the dipeptide systems are more stable in the solvent phase than in gas phase. NBO confirms increased magnitudes for donor−acceptor interaction for the solvated dipeptides.

1. INTRODUCTION Noncovalent interactions (NCI), although not as strong as covalent bonds, are essential for some of the most basic biological and chemical phenomena. They are found in a wide variety of biological, chemical, organic crystal, and organic materials assemblies and are essential to their form and function.1−3 Perhaps most important to life, NCI hold together and stabilize the double strands of DNA and the secondary structures of proteins. They also play a crucial role in protein− protein interactions, protein−DNA interactions, and protein recognition and signaling.4−6 Aromatic peptides, in particular, can come together to form higher order discrete structures, such as micelles and vesicles, or continuous structures, such as fibers, ribbons, and tapes.7 Dipeptides self-assemble even more readily into three-dimensional structures,8 and the type of structure that forms is dependent on many environmental factors, including solvent choice, dipeptide concentration, and temperature, but one of the most common structures are nanotubes. Dipeptide nanotubes have an attractive set of physical and chemical properties useful for many potential applications such as biological grafts, drug delivery tools, vaccine carriers, biosensors, antibiotics, microelectronic devices, transmembrane ion channels, matrices for chromatography, and selfcleaning surfaces.7−15 They are readily available, easily customizable, and environmentally and biologically safe. They also tend to be robust in heat, pH, and enzymatic © 2019 American Chemical Society

degradation. For instance, diphenylalanine (FF) nanotubes, the most common and widely studied dipeptide nanotube, are stable in temperatures up to 150 °C.8,16,17 The FF nanotube structure has been well elucidated.9,14−18 The FF monomer has two aromatic rings, stabilized by NCI, and the FF nanotube is made up of symmetric six-membered rings stacked via NCI.14,17,18 The outer diameter of the tube ranges 10−100 nm, whereas the inner diameter is only around 10 Å.9,13 Within the nanotube, the monomers are uniform in their structure, which allows for uniform head to tail alignment.9,15 The FF molecules are in the trans position, and the L configuration of phenylalanine is used.15,16 An investigation of the isolated dipeptides showed that the peptide plane has less than a 20° variance from the 180° peptide plane.19 The most favorable backbone conformation was found to be folded up such that (NH)H···Ocarboxyl hydrogen bonding between the two exposed termini is formed.20,21 If this conformation was not assumed, a fully outstretched backbone was the next most stable form. Although the FF nanotube structure has been successfully elucidated, the mechanism of self-assembly remains elusive. Currently, there are three proposed mechanisms of selfassembly of dipeptides. First, the dipeptides come together to Received: October 25, 2018 Accepted: December 27, 2018 Published: January 10, 2019 911

DOI: 10.1021/acsomega.8b02934 ACS Omega 2019, 4, 911−919

ACS Omega

Article

form two-dimensional β-sheets, which then roll up into tubes. This “rolling up” process is proposed to be driven by side chain π−π interactions and hydrogen bonding.16,22 Second, hydrophobic interactions cause the dipeptides to form small vesicles to protect their hydrophobic motifs from the environment. Once a high enough concentration of these vesicles accumulates, they are thought to fuse into tubes.7,11,18 Finally, the nanotube comes together in a more piece-wise cooperative manner. It suggests that the side chains of adjacent FF dipeptides “zipper” together to form ringlike structures while stacking up into a tube.13,17 Each of these mechanisms demonstrates the well-known importance of NCI in the self-assembly of dipeptide nanotubes.7−10,18,22,23 Computational and laboratory studies of FF nanotubes have proven the presence of electrostatic interactions between charged residues or the carboxylic acid and amine termini, π−π stacking between aromatic side chains, hydrophobic interactions to protect the peptide backbones, and directional hydrogen bonding networks.7−11,18,23 A laboratory-based comparison of FF nanotubes made with modified monomers revealed that FF monomers with high hydrogen bonding capacity readily self-assembled into ordered nanostructures, whereas FF monomers with poor hydrogen bonding capacity only self-assembled into random β-sheets.11 Additionally, more extensive hydrogen bonding was found to allow for high-aspect-ratio packing, further stabilizing the nanotubes.11 A similar study of phenylalanine−alanine nanotubes confirmed the presence of both (NH)H···Ocarbonyl hydrogen bonding.16 There has been little research on nanotubes made from other aromatic peptides, such as tyrosine (Y) and tryptophan (W). These two dipeptides share many of the same characteristics of phenylalanine but have been suggested to involve stronger NCI.1,2,12 When comparing the binding of aromatic dipeptides with nucleosides, it was found that both tyrosine and tryptophan formed more energetically stable complexes than phenylalanine.1,2 Also, peptides containing a hydroxyl group form more ordered structures than peptides which do not, suggesting that tyrosine-based nanotubes would have an even higher aspect ratio than phenylalanine-based nanotubes.12 Our previous work on the conformational and vibrational analyses of the isolated linear dityrosine (YY) and tryptophan− tyrosine (WY), which we refer to as monomers in this paper, showed that the most stable conformers are characterized by hydrogen bonding between the carboxylic hydroxyl group and the carbonyl of the amide.25 In this paper, we present our computational study on the noncovalent dimers of linear YY and WY toward understanding the early stages of self-assembly of dipeptide nanotubes. We investigate the types and magnitudes of NCI, specifically the aromatic interactions and hydrogen bonding, that are implicated in providing the order and directionality needed for the self-assembly process. The orbital interactions were examined in detail. The understanding of such interactions would provide insights into the aggregation of peptide-based nanostructures.

intermolecular hydrogen bonds between the carboxylic group of one dipeptide and the tyrosine hydroxyl group of the other dipeptide (Ocarboxyl···H(O)tyr), each within 1.94 Å (Figure 1).

Figure 1. Five lowest energy noncovalent dimers of linear YY and their relative energies (in kcal/mol) at DFTB3-D level of theory.

YYd1 has a relatively low dipole moment of 1.50 D. Four other YY dimers are within 4.54 kcal/mol of YYd1; some of their properties are given in Table 1. YYd2 is 0.48 kcal/mol higher than YYd1 and is stabilized by two Ocarboxyl···H(O)tyr within 1.88 and 2.11 Å and a weak π−π stacking between two tyrosine rings (πtyr···πtyr) separated by ∼3.94 Å. The Ocarboxyl···H(O)tyr is a recurring structural motif, present in YYd1, YYd2, YYd3, and YYd5, whereas YYd4 has Ocarboxyl···H(O)carboxyl with a distance ∼1.86 Å. Also, YYd5 has NH···πtyr within 2.94 Å. These results imply that hydrogen bonding is an important NCI between dimers, in agreement with studies focused on the structure of FF nanotubes.7,9,10 In contrast, for linear WY dimers, the lowest energy WYd1 is stabilized in such a way that the carboxylic acid terminal of one monomer interacts with the other monomer’s tryptophan amine group, resulting in two Ocarboxyl···H(N)trp within ∼2.02 Å. In addition, WYd1 has πtyr···πtyr stacking interaction within ∼3.58 Å (Figure 2). WYd1 is quite polar with a dipole moment of 5.02 D. Eight more WY dimers are within 4.33 kcal/mol of WYd1 (Table 1). WYd2, which is 1.05 kcal/mol higher than WYd1, is stabilized by Ocarboxyl···H(O)tyr within 1.95 Å as well as πtrp···H(C) within 2.98 Å. WYd3 and WYd4 are both stabilized by one intermolecular Ocarboxyl···H(N)trp. WYd5 is mainly stabilized by intermolecular Ocarboxyl··· H(O)carboxyl. WYd6 is governed by Ocarboxyl···H(O)tyr and πtyr···πtyr. WYd7 and WYd8 are both stabilized by Ocarboxyl···H(N)trp. WYd7 and WYd8 are further stabilized by πtrp··· H(N) ∼2.88 Å apart and πtrp···H(C) in around 2.91 Å, respectively. Finally, WYd9 is mostly stabilized by intermolecular Ocarboxyl···H(O)tyr. Figure 3 shows the calculated IR spectra for the most stable YY monomer (YY1) and five lowest noncovalent dimer structures shown in Figure 1 at the DFTB3-D level of theory. For YY1, some of the representative peaks include free hydroxyl groups (OHtyr) on each of the phenolic rings of dityrosine (3657 and 3671 cm−1), peptide N−H stretching (NHpep, 3316 cm−1), the intense carboxyl O−H stretch (OHcarboxyl, 3244 cm−1), the carbonyl of the carboxyl group (COcarboxyl, 1745 cm−1), and carbonyl peptide stretching vibrations (COpep, 1628 cm−1), the peptide N−H in-plane bending vibration (NHipb, 1526 cm−1), and carboxyl O−H inplane bending vibration (OHipb, 1290 cm−1). The characteristic peaks that are present in the monomer are also present in the dimers. Some peaks occurred twice in the dimers since there are two monomers in each dimer, which are in slightly different chemical environments. The dimers consistently have more intense peaks than the monomers, especially for the functional groups which participated in hydrogen bonding such as OHtyr and COpep. In particular, for the lowest dimer

2. RESULTS AND DISCUSSION 2.1. Full Optimization Dimers. The lowest energy noncovalent dimer of linear YY, YYd1, as calculated by the dispersion-corrected third-order density functional-based tight binding method (DFTB3-D),26 is characterized by two 912

DOI: 10.1021/acsomega.8b02934 ACS Omega 2019, 4, 911−919

ACS Omega

Article

Table 1. Some Structural Properties of the Lowest Energy Dimers of Linear YY and WY at DFTB3-D Level of Theory

linear YY

linear WY

relative energy (kcal/mol)

binding energy (kcal/mol)

relative energy in acetone (kcal/mol)a

YYd1 YYd2 YYd3

0.00 0.48 1.22

−18.48 −18.00 −17.26

0.00 (sYYd1) 0.84 0.88

YYd4

4.24

−14.24

1.87

YYd5 WYd1 WYd2 WYd3 WYd4 WYd5 WYd6 WYd7 WYd8 WYd9

4.54 0.00 1.05 1.77 1.95 3.45 3.46 3.50 4.09 4.34

−13.95 −17.71 −16.65 −15.93 −15.76 −14.26 −14.25 −14.21 −13.62 −13.36

2.50 1.75 (sWYd1) 0.00 (sWYd2) 1.79 1.92 2.36 4.30 1.73 2.64 2.96

type and number of intermolecular hydrogen bonding 2 2 1 1 1 1 1 1 2 1 1 1 1 1 1 1 1

Ocarboxyl···H(O)tyr Ocarboxyl···H(O)tyr Ocarboxyl···H(O)tyr Ocarboxyl···H(NH)amine Ocarboxyl···H(N)pep Ocarboxyl···H(O)carboxyl HNpep···H(N)pep Ocarboxyl···H(O)tyr Ocarboxyl···H(N)trp Ocarboxyl···H(O)tyr Ocarboxyl···H(N)trp Ocarboxyl···H(N)trp Ocarboxyl···H(O)carboxyl Ocarboxyl···H(O)tyr Ocarboxyl···H(N)trp Ocarboxyl···H(N)trp Ocarboxyl···H(O)tyr

distance of intermolecular hydrogen bonding (Å) 1.94, 1.94 1.88, 2.11 1.99 2.56 2.56 1.86 2.16 1.92 2.02, 2.02 1.95 2.15 2.29 1.87 2.07 2.81 2.68 1.88

a

Solvation model based on density (SMD) is used.

cm−1), and OHipb (1266 cm−1). The lowest dimer WYd1 has very similar spectra compared to WY1, and the main differences are the presence of Ocarboxyl···H(N)trp at about 3299 cm−1 and the red-shift of intramolecular OHcarboxyl to about 3120 cm−1. For WYd2, the intermolecular Ocarboxyl··· OHtyr shows at 3549 cm−1 and one of the OHcarboxyl shifts to 3020 cm−1. For WYd3, the intermolecular Ocarboxyl···H(N)trp peaks at 3325 cm−1 and one of the OHcarboxyl shows at 3100 cm−1. For WYd4, both peaks of OHcarboxyl are red shifted to 2992 and 3094 cm−1 and the intermolecular Ocarboxyl···H(N)trp peaks at 3323 cm−1. For WYd5, the OHcarboxyl shows at 3138 and the intermolecular Ocarboxyl···H(O)carboxyl peaks at 3430 cm−1. For WYd6, the intermolecular Ocarboxyl···OHtyr shows at 3550 cm−1, and one of the OHcarboxyl shifts to 3130 cm−1. For WYd7 and WYd8, one of the OHcarboxyl shows at 3144 and 3129 cm−1, respectively. For WYd9, one of the OHcarboxyl shows at 3184 cm−1 and the intermolecular Ocarboxyl···H(O)tyr peaks at 3509 cm−1. The highest occupied molecular orbitals (HOMO) and lowest unoccupied molecular orbital (LUMO) of the most stable noncovalent dimers YYd1 and WYd1 consist of π and π* orbitals, respectively (see Figure S1). The YYd1 HOMO is localized around one of the tyrosine rings of each dipeptide, whereas the LUMO is localized around the other tyrosine ring on each dipeptide. In contrast, the WYd1 HOMO is localized around both tryptophan rings, whereas its LUMO is localized around both tyrosine rings. The difference in localization between the HOMO and LUMO demonstrates the importance of charge-transfer interactions in the stabilization of each dipeptide dimer. Table 1 also shows that the more negative the binding energy, the more stable the dimer is. YYd1 has a binding energy of −18.48 kcal/mol, whereas WYd1 has a slightly lower binding energy of −17.71 kcal/mol. Such a result implies that the stability of aromatic dipeptide dimers is dependent on the strength of intermolecular hydrogen bonding (Ocarboxyl··· H(O)tyr vs Ocarboxyl···H(N)trp). Table 2 shows the thermodynamic properties of YYd1 and WYd1 dimerization, which indicate that the dimerization of each dipeptide is thermodynamically favorable at room temperature.

Figure 2. Nine lowest energy noncovalent dimers of linear WY and their relative energies (in kcal/mol) at DFTB3-D level of theory.

YYd1, the OHtyr occurs at 3663 and 3470 cm−1, the latter a red-shift of about 200 cm−1 with respect to the monomer YY1, which is a characteristic of intermolecular Ocarboxyl···H(O)tyr. The intense OHcarboxyl peaks at around 3002 cm−1, also a redshift of about 200 cm−1. Other notable difference is the COcarboxyl, which appears at 1690 and 1684 cm−1, a red-shift of 50 cm−1. The rest of the YYd1 spectrum has similar characteristics as YY1. For YYd2, one of the OHtyr also shifts to about 3493 cm−1; COcarboxyl is also red shifted. The OHcarboxyl shows at around 3085 cm−1 but not as strong as YYd1, which has two of this type. YYd3 has distinct peaks at around 3533 and 3100 cm−1 for intermolecular and intramolecular OH bonding, respectively. Similarly, YYd4 shows intermolecular OHtyr and intramolecular OHcarboxyl around 3420 and 3150 cm−1, respectively. YYd5 shows intermolecular OHtyr at 3488 cm−1 and intramolecular OHcarboxyl around 3251 and 3161 cm−1. Figure 4 shows the calculated IR spectra for the most stable WY monomer (WY1) and the nine lowest WY noncovalent dimers, and the corresponding frequencies are summarized in Table S1. For WY1, the representative peaks include OHtyr (3658 cm−1), N−H stretching of the tryptophan group (NHtrp, 3361 cm−1), NHpep (3305 cm−1), OHcarboxyl (3223 cm−1), COcarboxyl (1742 cm−1), COpep (1626 cm−1), NHipb (1519 913

DOI: 10.1021/acsomega.8b02934 ACS Omega 2019, 4, 911−919

ACS Omega

Article

Figure 3. Calculated IR spectra of the lowest YY monomer and the five lowest YY dimers at DFTB3-D level of theory.

Figure 4. Calculated IR spectra of the lowest WY monomer and the nine lowest WY dimers at DFTB3-D level of theory.

intermolecular Ocarboxyl···H(O)tyr hydrogen bonds. In contrast, for WYd1, both electrostatic and dispersion are important because of the presence of two Ocarboxyl···H(N)trp and πtyr···πtyr stacking interactions with donor−acceptor (DA) stabilization energies of 6.55 and 0.47 kcal/mol, respectively. 2.2. Constrained Optimization of the Dimers. To deconstruct the nature of interaction for each specific functional group present, we performed constrained optimization of the dimers in such a way that the functional groups were oriented in either parallel displaced (PD) or T-shaped

The pair interaction energy decomposition analyses (PIEDA)27 using fragment molecular orbital (FMO) method28 at the second-order Møller−Plesset perturbation theory (FMO-MP2) and natural bond orbital (NBO)29 analysis were used to investigate the nature of the NCI present in each lowest energy dimer in more detail. Consistent with the binding energy trend, WYd1 has less total interaction energy compared to YYd1. As shown in Figure 5 and in agreement with the findings of Lee and co-workers30 and Karthikeyan and co-workers,3 electrostatic is dominant in YYd1 due to two 914

DOI: 10.1021/acsomega.8b02934 ACS Omega 2019, 4, 911−919

ACS Omega

Article

Table 2. ΔG for the Dimerization of the Lowest Energy GasPhase (YYd1 and WYd1) and Stabilization Energy of Solvated Structures (sYYd1 and sWYd2) at 25 °C at DFTB3-D and SMDa ΔG

gas-phase

in acetone sYYd1 sWYd2

YYd1 WYd1 ΔG −12.55 −12.32

ring−ring distances demonstrate that the ideal ring−ring distance between T-shaped aromatic rings is influenced heavily by the chemical environment. Figure 7 shows that for Tshaped orientations, the HOMO is mainly made up of π orbitals localized around the tryptophan ring of one dipeptide, whereas each LUMO is mainly made up of π* orbitals localized around the second dipeptide’s tryptophan ring. This suggests that for T-shaped orientations, charge transfer plays a role in stabilizing the linear WY dimer. As shown in Table 3, the W−W has the lowest HOMO−LUMO gap and is the most reactive. FMO-MP2 PIEDA results also show that W−W has the most negative interaction energy. Figure 8 shows that dispersion is dominant for both PD orientations of the constrained dimers. W−W has the most negative dispersion energy since its tryptophan has a larger area for delocalization. W−Y has more negative dispersion energy than Y−Y suggesting that having the tryptophan ring increases dispersion interactions. Furthermore, W−W also has the most positive exchange energy, exhibiting more π−π orbital repulsion. On the other hand, T-shaped constrained dimers showed that in addition to dispersion energy, electrostatic interactions are also significant which can be attributed to polarized (X)H···π interaction between the two monomers.2,5,6,34 Similar to the PD counterparts, W−W has the most negative dispersion energy and the most positive exchange energy; Y−Y has significantly less interaction energy of every type, likely because the tyrosine functional groups were farther away from one another. Table 4 shows that each interaction is stabilized by π from one monomer to π* of another monomer. This interaction occurs between the edges of the rings, which are closest to one another. W−W is stabilized by 0.51 and 0.34 kcal/mol for aPD and b-PD, respectively. For W−Y, aside from π to π*, which is stabilized by about 0.3 kcal/mol for PD, π to σ* is also a factor in its stabilization. The interactions in T-orientations are stronger, W−W is stabilized by π donation from one tryptophan ring to σ* C−H in the other tryptophan ring by about 0.78 kcal/mol. W−Y is stabilized by π of tyrosine ring to σ* C−H of the tryptophan by 0.95 kcal/mol. 2.3. Solvation Effects. In the presence of acetone, the range of the relative energies is much closer, about 2−3 kcal/ mol for five lowest energy YY and WY dimers compared to 3− 5 kcal/mol for the gas-phase dimers (Table 1). The ordering of

stabilization energy

−4.50 −3.04 solvation energyb

−13.08 −15.63

−11.42 −13.96

a

All values are in kcal/mol. bAt 1 atm standard state.

configurations instead of the full optimization employed in the previous section. The details are described in the Supporting Information. For linear WY, three functional group combinations were evaluated: W−W, Y−Y, and W−Y. The PD and Tshaped orientations were considered. For the PD orientation, two configurations for the carboxylic group are shown in this paper: on the opposite side (a-PD) and the same side (b-PD). In PD constrained functional group dimers, each functional group combination has a ring−ring distance within 3.45−3.67 Å, as shown in Table 3, consistent with the studies on peptidelike systems that the characteristic distance between PD aromatic rings ranges from 3.3 to 3.7 Å.2,4,30,31 The ring−ring distance is found to be closer to 3.7 Å if the monomers are small, as demonstrated by the Y−Y ring−ring distance of 3.67 Å for a-PD and 3.81 for b-PD, whereas closer to 3.3 Å for heterofunctional group, as supported the W−Y ring−ring distance of 3.45 Å for a-PD and 3.53 Å for b-PD.2,32 Figure 6 shows that for each functional group dimer, the HOMO and LUMO are made up of mainly π and π*, respectively. Y−Y has the largest HOMO−LUMO gap, whereas W−W has the smallest. W−W also has the most negative interaction energy for a-PD configuration, whereas W−Y for b-PD configuration. The T-shaped orientation constrained dimers have ring− ring distances within 6.3−7.8 Å (Table 3). Studies on peptidelike systems have shown that T-shaped orientations between aromatic rings typically range from 4.9 to 5.5 Å, a shorter range than observed for these WY dimers.6,30,33 Studies thus far have been on isolated functional rings. For example, to model tryptophan, a single indole ring would have been used. The uses of entire dipeptide molecules and the resulting longer

Figure 5. Interaction energy distribution for the lowest energy noncovalent dimers YYd1 and WYd1 at FMO-MP2/6-31G*. 915

DOI: 10.1021/acsomega.8b02934 ACS Omega 2019, 4, 911−919

ACS Omega

Article

Table 3. Some Energetic and Structural Parameters Describing the Interactions between Linear WY’s Functional Rings at MP2/6-31G*a a-PD

b-PD

T-shaped

W−W Y−Y W−Y W−W Y−Y W−Y W−W Y−Y W−Y

distance between rings (Å)

dipole moment (D)

HOMO−LUMO gap (eV)

interaction energy (kcal/mol)

3.59 3.67 3.45 3.76 3.81 3.53 6.30 7.38 6.41

2.60 4.86 6.82 10.58 13.01 13.69 11.69 12.38 15.47

10.60 10.78 10.64 10.40 10.68 10.57 10.39 10.57 10.49

−10.00 −7.33 −9.52 −7.97 −6.22 −8.14 −8.63 −3.88 −7.60

a

Interaction energy calculated at FMO-MP2 PIEDA.

Figure 6. HOMO (bottom) and LUMO (top) of the PD orientation constrained dimers of linear WY, describing the WW, Y−Y, and W−Y. The a-PD orientation is when the carboxylic groups are on the opposite sides of the dimer while b-PD is when they are on the same side.

Instead, the lowest is sWYd2, which showed closer Ocarboxyl··· H(O)tyr and πtrp···H(X) [X = C and N]. Solvation with acetone has little geometric effect on sWYd2, which has a rootmean-square deviation (RMSD) of 0.27, with respect to WYd2 but has more effect on the YYd1 dipeptide system, which has an RMSD of 0.69 (Figure S3). Similarly, the IR spectra show that when solvation is added, the peaks become typically more intense due to their greater polarity (Figure 9), and the peaks are also slightly red shifted, particularly for the intermolecular OHcarboxyl by about 100 cm−1 to about 2900 cm−1. Table 2 shows the ΔG of dimerization and stabilization energies of solvated dipeptides indicating that the dipeptide systems are more stable in the solvent phase. NBO confirms increased magnitudes for DA interaction for the solvated dipeptides.

3. SUMMARY AND CONCLUSIONS Analysis of the lowest energy noncovalent dimers of linear YY and WY dipeptides demonstrated that hydrogen bonding is a major stabilizing force. The most stable dimer of linear YY is stabilized by two intermolecular Ocarboxyl···H(O)tyr hydrogen bonds, whereas the linear WY dimer is stabilized by two intermolecular Ocarboxyl···H(N)trp and πtyr···πtyr stacking interactions. The interactions πtrp···H(C) and πtrp···H(N) are important in the solvated linear WY. The different types of intermolecular hydrogen bonding were all confirmed in the calculated IR spectra and NBO results. For the higher energy dimer configurations, when hydrogen bonding is present, electrostatic energy makes up a majority of the interaction energy within the dimer. The presence of tryptophan in linear WY indicates greater propensity for dispersion compared to

Figure 7. HOMO (bottom) and LUMO (top) of the T-shaped orientation constrained dimers of linear WY, describing (a) W−W, (b) Y−Y, and (c) W−Y.

the dimer conformers did not change for the solvated linear YY. The lowest energy structure for the solvated linear YY dimer is referred to as sYYd1. In general, the presence of acetone in the system makes the dipeptide monomers closer together. For example, the intermolecular Ocarboxyl···H(O)tyr for sYYd1 occurs at about 1.89 Å compared to 1.94 Å of the unsolvated YYd1. In contrast, the lowest energy structure for the solvated linear WY is not the same as that in the gas phase. 916

DOI: 10.1021/acsomega.8b02934 ACS Omega 2019, 4, 911−919

ACS Omega

Article

Figure 8. Interaction energy distribution for the W−W, Y−Y, and W−Y in a-PD, b-PD, and T orientation constrained dimers of linear WY as calculated by FMO-MP2/6-31G* PIEDA level of theory.

4. METHODS We use the structures of the lowest energy monomers (YY1 and WY1) as determined in ref 24. Three commonly reported dimer orientations were employed: (i) side by side, where the monomers are adjacent to one another, (ii) parallel displaced (PD), and (iii) T-shaped, where monomers are oriented perpendicularly.1,4,35,36 The monomers were also flipped around within these orientations to capture all possible functional group interactions. All dipeptides employed the LL configuration. Each dimer was then optimized using DFTB3 employing the parameter set, which is specifically designed for organic and biological systems (3OB).26,37 The dispersion effects were accounted for by using the universal force field dispersion model.38 This parameterized variation of density functional theory (DFT) has decent accuracy for estimating the geometry of noncovalent systems and reducing overbinding between monomers, and computational efficiency is two or three orders faster than conventional DFT methods. FMO-MP2 at 6-31G* basis set39 was used to perform PIEDA.38 NBO39 analysis was also employed to study the type of donor−acceptor (DA) relationship between dimers. The solvation model based on density (SMD)40 was used to study the influence of the solvent. Acetone was chosen as a solvent because it has a dielectric constant (ε) similar to hexafluoroisopropanol, which was used in typical experiments of self-assembly of dipeptides. The default values for the atomic radii were employed. Vibrational frequencies were computed to ensure that each stationary point is a local minimum, and reaction energetics were corrected for zero-

Table 4. Strongest Donor−Acceptor (DA) Interactions Present between Linear WY Functional Groups Calculated Using NBO at the MP2/6-31G* Level of Theory interaction a-PD

b-PD

T

Y−Y W−W W−Y W−Y Y−Y Y−Y W−W W−Y W−W W−Y

donor

acceptor

π π π π π π π π π π

π* π* π* σ* σ* π* π* π* σ* σ*

tyr1 trp1 trp trp tyr1 tyr1 trp1 trp trp1 tyr

tyr2 trp2 tyr tyr tyr2 tyr2 trp2 tyr trp2 trp

strength (kcal/mol) 0.43 0.51 0.32 0.42 0.21 0.15 0.34 0.33 0.78 0.95

linear YY. Exchange and charge-transfer interactions are consistently found to have slightly lower but important contributions to the total interaction energy. Analyses of PD and T-shaped constrained functional group dimers showed that the favored orientation is dependent on the functional group and chemical environment. The most favorable functional group interaction is Y−Y for linear YY and W−Y for linear WY. The addition of solvent stabilizes the dipeptide dimers and results in red-shift and higher IR intensities. The understanding of such interactions would provide insights into the aggregation of self-assembled structures from peptidebased molecular frameworks.

Figure 9. IR spectra of the (A) YYd1 and (B) WYd1 in gas phase using DFTB3-D and in acetone using SMD at DFTB3-D. 917

DOI: 10.1021/acsomega.8b02934 ACS Omega 2019, 4, 911−919

ACS Omega

Article

(8) Sedman, V. L.; Adler-Abramovich, L.; Allen, S.; Gazit, E.; Tendler, S. J. B. Direct Observation of the Release of Phenylalanine from Diphenylalanine Nanotubes. J. Am. Chem. Soc. 2006, 128, 6903− 6908. (9) Jeon, J.; Mills, C. E.; Shell, M. S. Molecular Insights into Diphenylalanine Nanotube Assembly: All-Atom Simulations of Oligomerization. J. Phys. Chem. B 2013, 117, 3935−3943. (10) Reches, M.; Gazit, E. Molecular Self-Assembly of Peptide Nanostructures: Mechanism of Association and Potential Uses. Curr. Nanosci. 2006, 2, 105−111. (11) Rajbhandary, A.; Brennessel, W. W.; Nilsson, B. L. Comparison of the Self-Assembly Behavior of Fmoc-Phenylalanine and Corresponding Peptoid Derivatives. Cryst. Growth Des. 2018, 18, 623−632. (12) Jeziorna, A.; Stopczyk, K.; Skorupska, E.; Luberda-Durnas, K.; Oszajca, M.; Lasocha, W.; Górecki, M.; Frelek, J.; Potrzebowski, M. J. Cyclic Dipeptides as Building Units of Nano- and Microdevices: Synthesis, Properties, and Structural Studies. Cryst. Growth Des. 2015, 15, 5138−5148. (13) Adler-Abramovich, L.; Gazit, E. The Physical Properties of Supramolecular Peptide Assemblies: From Building Block Association to Technological Applications. Chem. Soc. Rev. 2014, 43, 6881−6893. (14) Andrade-Filho, T.; Furlan Ferreira, F.; Andrade Alves, W.; Reily Rocha, A. The Effects of Water Molecules on the Electronic and Structural Properties of Peptide Nanotubes. Phys. Chem. Chem. Phys. 2013, 15, 7555−7559. (15) Oakley, M. T.; Johnston, R. L. Energy Landscapes and Global Optimization of Self-Assembling Cyclic Peptides. J. Chem. Theory Comput. 2014, 10, 1810−1816. (16) Guha, S.; Drew, M. G. B.; Banerjee, A. Dipeptide Nanotubes, with N-Terminally Located ω-Amino Acid Residues, That Are Stable Proteolytically, Thermally, and Over a Wide Range of PH. Chem. Mater. 2008, 20, 2282−2290. (17) Azuri, I.; Adler-Abramovich, L.; Gazit, E.; Hod, O.; Kronik, L. Why Are Diphenylalanine-Based Peptide Nanostructures so Rigid? Insights from First Principles Calculations. J. Am. Chem. Soc. 2014, 136, 963−969. (18) Amdursky, N.; Molotskii, M.; Gazit, E.; Rosenman, G. Elementary Building Blocks of Self-Assembled Peptide Nanotubes. J. Am. Chem. Soc. 2010, 132, 15632−15636. (19) Mondal, S.; Chowdhuri, D. S.; Ghosh, S.; Misra, A.; Dalai, S. Conformational Study on Dipeptides Containing Phenylalanine: A DFT Approach. J. Mol. Struct.: THEOCHEM 2007, 810, 81−89. (20) Celik, S.; Ozel, A. E.; Akyuz, S. Comparative Study of Antitumor Active Cyclo(Gly-Leu) Dipeptide: A Computational and Molecular Modeling Study. Vib. Spectrosc. 2016, 83, 57−69. (21) Domcke, W.; Sobolewski, A. L. Peptide Deactivation: Spectroscopy Meets Theory. Nat. Chem. 2013, 5, 257−258. (22) Lee, O.-S.; Stupp, S. I.; Schatz, G. C. Atomistic Molecular Dynamics Simulations of Peptide Amphiphile Self-Assembly into Cylindrical Nanofibers. J. Am. Chem. Soc. 2011, 133, 3677−3683. (23) Lampkins, A. J.; Abdul-Rahim, O.; Li, H.; Castellano, R. K. Enhanced Small-Molecule Assembly through Directional Intramolecular Forces. Org. Lett. 2005, 7, 4471−4474. (24) Mayes, M. L.; Perreault, L. Conformational Structures and Vibrational Spectroscopic Investigation of Isolated Dityrosine and Tryptophan−Tyrosine Dipeptides: A Theoretical Study. Comput. Theor. Chem. 2018, 1131, 99−109. (25) Macha, P.; Perreault, L.; Hamedani, Y.; Mayes, M. L.; Vasudev, M. C. Molecular Mechanisms of Tryptophan−Tyrosine Nanostructures Formation and Their Influence on PC-12 Cells. ACS Appl. Bio Mater. 2018, 1 (5), 1266−1275. (26) Gaus, M.; Cui, Q.; Elstner, M. DFTB3: Extension of the SelfConsistent-Charge Density-Functional Tight-Binding Method (SCCDFTB). J. Chem. Theory Comput. 2011, 7, 931−948. (27) Fedorov, D. G.; Kitaura, K. Pair Interaction Energy Decomposition Analysis. J. Comput. Chem. 2007, 28, 222−237. (28) Kitaura, K.; Ikeo, E.; Asada, T.; Nakano, T.; Uebayasi, M. Fragment Molecular Orbital Method: An Approximate Computa-

point energy differences. All quantum-chemical calculations were performed using GAMESS41 program.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsomega.8b02934. HOMO and LUMO of the lowest energy dimers; constrained geometry optimization information; structures of the lowest energy solvated dimers and their corresponding RMSD with respect to the unsolvated dimers; IR frequencies of WY dimers (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Maricris L. Mayes: 0000-0003-1184-7226 Author Contributions †

The manuscript was written through contributions of all authors. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work is supported by the faculty research start-up and Seed Funding program from the Office of the Provost at the University of Massachusetts Dartmouth. L.P. gratefully acknowledges the support from Office of Undergraduate Research. We thank Benjamin Burnett for his help with the figures. This research used the computing resources of UMass Dartmouth and the UMass Green High Performance Computing Cluster.

■

REFERENCES

(1) Rutledge, L. R.; Wetmore, S. D. Remarkably Strong T-Shaped Interactions between Aromatic Amino Acids and Adenine: Their Increase upon Nucleobase Methylation and a Comparison to Stacking. J. Chem. Theory Comput. 2008, 4, 1768−1780. (2) Mignon, P.; Loverix, S.; De Proft, F.; Geerlings, P. Influence of Stacking on Hydrogen Bonding: Quantum Chemical Study on Pyridine−Benzene Model Complexes. J. Phys. Chem. A 2004, 108, 6038−6044. (3) Karthikeyan, S.; Nagase, S. Origins of the Stability of Imidazole− Imidazole, Benzene−Imidazole, and Benzene−Indole Dimers: CCSD(T)/CBS and SAPT Calculations. J. Phys. Chem. A 2012, 116, 1694− 1700. (4) Rutledge, L. R.; Campbell-Verduyn, L. S.; Wetmore, S. D. Characterization of the Stacking Interactions between DNA or RNA Nucleobases and the Aromatic Amino Acids. Chem. Phys. Lett. 2007, 444, 167−175. (5) Morozov, A. V.; Misura, K. M. S.; Tsemekhman, K.; Baker, D. Comparison of Quantum Mechanics and Molecular Mechanics Dimerization Energy Landscapes for Pairs of Ring-Containing Amino Acids in Proteins. J. Phys. Chem. B 2004, 108, 8489−8496. (6) Gervasio, F. L.; Procacci, P.; Cardini, G.; Guarna, A.; Giolitti, A.; Schettino, V. Interaction between Aromatic Residues. Molecular Dynamics and Ab Initio Exploration of the Potential Energy Surface of the Tryptophan−Histidine Pair. J. Phys. Chem. B 2000, 104, 1108− 1114. (7) Rad-Malekshahi, M.; Lempsink, L.; Amidi, M.; Hennink, W. E.; Mastrobattista, E. Biomedical Applications of Self-Assembling Peptides. Bioconjugate Chem. 2016, 27, 3−18. 918

DOI: 10.1021/acsomega.8b02934 ACS Omega 2019, 4, 911−919

ACS Omega

Article

tional Method for Large Molecules. Chem. Phys. Lett. 1999, 313, 701− 706. (29) Glendening, E. D.; Badenhoop, J. K.; Reed, A. E.; Carpenter, J. E.; Bohmann, J. A.; Morales, C. M.; Landis, C. R.; Weinhold, F. NBO 6.0; Theoretical Chemistry Institute, University of Wisconsin: Madison. (30) Lee, E. C.; Kim, D.; Jurečka, P.; Tarakeshwar, P.; Hobza, P.; Kim, K. S. Understanding of Assembly Phenomena by Aromatic− Aromatic Interactions: Benzene Dimer and the Substituted Systems. J. Phys. Chem. A 2007, 111, 3446−3457. (31) Geng, H.; Jiang, F.; Wu, Y.-D. Accurate Structure Prediction and Conformational Analysis of Cyclic Peptides with Residue-Specific Force Fields. J. Phys. Chem. Lett. 2016, 7, 1805−1810. (32) Rutledge, L. R.; Durst, H. F.; Wetmore, S. D. Computational Comparison of the Stacking Interactions between the Aromatic Amino Acids and the Natural or (Cationic) Methylated Nucleobases. Phys. Chem. Chem. Phys. 2008, 10, 2801−2812. (33) Smith, Q. A.; Gordon, M. S.; Slipchenko, L. V. Benzene− Pyridine Interactions Predicted by the Effective Fragment Potential Method. J. Phys. Chem. A 2011, 115, 4598−4609. (34) Jurečka, P.; Š poner, J.; Hobza, P. Potential Energy Surface of the Cytosine Dimer: MP2 Complete Basis Set Limit Interaction Energies, CCSD(T) Correction Term, and Comparison with the AMBER Force Field. J. Phys. Chem. B 2004, 108, 5466−5471. (35) Sinnokrot, M. O.; Valeev, E. F.; Sherrill, C. D. Estimates of the Ab Initio Limit for Π−π Interactions: The Benzene Dimer. J. Am. Chem. Soc. 2002, 124, 10887−10893. (36) Sinnokrot, M. O.; Sherrill, C. D. Substituent Effects in π-π Interactions: Sandwich and T-Shaped Configurations. J. Am. Chem. Soc. 2004, 126, 7690−7697. (37) Gaus, M.; Goez, A.; Elstner, M. Parametrization and Benchmark of DFTB3 for Organic Molecules. J. Chem. Theory Comput. 2013, 9, 338−354. (38) Rappe, A. K.; Casewit, C. J.; Colwell, K. S.; Goddard, W. A.; Skiff, W. M. UFF, a Full Periodic Table Force Field for Molecular Mechanics and Molecular Dynamics Simulations. J. Am. Chem. Soc. 1992, 114, 10024−10035. (39) Hariharan, P. C.; Pople, J. A. The Influence of Polarization Functions on Molecular Orbital Hydrogenation Energies. Theor. Chim. Acta 1973, 28, 213−222. (40) Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. Universal Solvation Model Based on Solute Electron Density and on a Continuum Model of the Solvent Defined by the Bulk Dielectric Constant and Atomic Surface Tensions. J. Phys. Chem. B 2009, 113, 6378−6396. (41) Schmidt, M. W.; Baldridge, K. K.; Boatz, J. A.; Elbert, S. T.; Gordon, M. S.; Jensen, J. H.; Koseki, S.; Matsunaga, N.; Nguyen, K. A.; Su, S.; et al. General Atomic and Molecular Electronic Structure System. J. Comput. Chem. 1993, 14, 1347−1363.

919

DOI: 10.1021/acsomega.8b02934 ACS Omega 2019, 4, 911−919