Location and Conformation of the LKα14 Peptide in Water/Ethanol Mixtures

It is widely recognized that solvation is one of the major factors determining structure and functionality of proteins and long peptides, however it is a formidable challenge to address it both experimentally and computationally. For this reason, simple peptides are used to study fundamental aspects of solvation. It is well established that alcohols can change the peptide conformation and tuning of the alcohol content in solution can dramatically affect folding and, as a consequence, the function of the peptide. In this work, we focus on the leucine and lysine based LKα14 peptide designed to adopt an α-helical conformation at an apolar–polar interface. We investigate LKα14 peptide’s bulk and interfacial behavior in water/ethanol mixtures combining a suite of experimental techniques (namely, circular dichroism and nuclear magnetic resonance spectroscopy for the bulk solution, surface pressure measurements and vibrational sum frequency generation spectroscopy for the air–solution interface) with molecular dynamics simulations. We observe that ethanol highly affects both the peptide location and conformation. At low ethanol content LKα14 lacks a clear secondary structure in bulk and shows a clear preference to reside at the air–solution interface. When the ethanol content in solution increases, the peptide’s interfacial affinity is markedly reduced and the peptide approaches a stable α-helical conformation in bulk facilitated by the amphiphilic nature of the ethanol molecules.


S1. SFG spectrum for LKα14 in D 2 O in the CH-stretching region
. SFG spectrum in the CH-stretching region acquired after equilibration (~30 minutes) following the peptide injection in D2O.
The acquired SFG spectrum for LKα14 peptide in pure D2O in the CH-stretching region ( Figure   S1) is similar to that obtained for leucine molecules at the air-water interface. S1

S2. SFG spectra fitting for the amide I spectral region
The SFG spectra were fitted according to the following formula: where A $% and ϕ $% are the amplitude and the phase of the non-resonant contribution and A ( , ω )%,( , and Γ ( are amplitude, center frequency, and half-width at half maximum (HWHM) of the j-th resonant contribution, respectively.
The fits of the amide I region SFG spectra for LKα14 in fEtOH=0 (pure H2O) and fEtOH=0.1solutions are shown in Figure S2 and the corresponding fitting parameters are listed in Table S1.
As shown in the Figure S2, on the high frequency side of the spectrum there is a very broad feature extending above 1800 cm -1 . Its nature is still a debated point in the SFG community and, to the best of our knowledge, this contribution has not yet been assigned, even though it has been observed in previous SFG studies, for example, for LKα14 peptides at the air/water interface. S2-S4 In our fit it appears as a very broad peak centered at ~1655 cm -1 . Figure S2. SFG spectra in the amide I region (empty circles) acquired after equilibration (~30 minutes) following the peptide injection in H2O/EtOH mixtures with ethanol volume fraction fEtOH=0 (pure H2O, red) and fEtOH=0.1 (orange). The solid lines present the corresponding fitting curves. Based on the fitting results we can conclude that ~5 cm -1 center frequency shift which can be observed in the SFG spectra ( Figure S2) actually originates from the difference in the contribution of the higher-frequency shoulder. Figure S3. The probability of the N-terminus pointing away from the interface and towards the bulk phase for the interfacial LKα14 molecules in H2O/EtOH mixtures with various EtOH volume fractions fEtOH.

S3. N-terminus orientation of the interfacial LKα14 molecules from interfacial MD simulations
Whether the N-terminus of the peptides points away or towards the bulk was determined by comparing the center of mass of the whole backbone with the center of mass of the backbone of the seven residues closest to the N-terminus. If the center of mass of the backbone of the seven residues was closer to the center of the simulation box than the center of mass of the whole peptide was, the N-terminus was considered to be pointing away from the interface towards the bulk.

S7
The probability that the N-terminus of the peptides at the interface points away from the interface is 0.77 in pure H2O, which matches previously reported results. S2 For fEtOH in the range from 0.10 to 0.68 the probability is lower, between 0.55 and 0.60. For higher EtOH volume fraction the probability shows less of a clear pattern, likely in part due to less prevalence of the peptide at the interface. Still, the probability of the N-terminus pointing towards the bulk is always higher than the probability of it pointing towards the interface, except in pure EtOH where the probability of the N-terminus pointing towards the bulk is 0.47.