Peptides at the Interface: Self-Assembly of Amphiphilic Designer Peptides and Their Membrane Interaction Propensity

Self-assembling amphiphilic designer peptides have been successfully applied as nanomaterials in biomedical applications. Understanding molecular interactions at the peptide–membrane interface is crucial, since interactions at this site often determine (in)compatibility. The present study aims to elucidate how model membrane systems of different complexity (in particular single-component phospholipid bilayers and lipoproteins) respond to the presence of amphiphilic designer peptides. We focused on two short anionic peptides, V4WD2 and A6YD, which are structurally similar but showed a different self-assembly behavior. A6YD self-assembled into high aspect ratio nanofibers at low peptide concentrations, as evidenced by synchrotron small-angle X-ray scattering and electron microscopy. These supramolecular assemblies coexisted with membranes without remarkable interference. In contrast, V4WD2 formed only loosely associated assemblies over a large concentration regime, and the peptide promoted concentration-dependent disorder on the membrane arrangement. Perturbation effects were observed on both membrane systems although most likely induced by different modes of action. These results suggest that membrane activity critically depends on the peptide’s inherent ability to form highly cohesive supramolecular structures.


Figure S9
Amino acid sequence of the peptide A₆YK S15 Figure S10 Determination of the CAC of A₆YK S15 Figure S11 A₆YK supramolecular structure formation S16 Figure S12 Biophysical characterization of DPPC-A₆YK and LDL-A₆YK interactions S17 Table S4 DSC parameters of DPPC + A₆YK S18 Figure S13 EPR spectra of LDL + peptides S19 Table S5 Derived EPR order parameters (S) and rotational correlation times (τ c ) for LDL + peptides S20 Table S6 Derived EPR isotropic splitting constants (a') for LDL + peptides S20 Figure S14 SAXS patterns of LDL + peptides S21

High-performance liquid chromatography (HPCL) and mass spectrometry (MS) analysis of A₆YD, V₄WD₂ and A₆YK
The peptides ac-A₆YD (A₆YD) and ac-V₄WD₂ (V₄WD₂) were custom synthesized and purified by Peptide 2.0 (Chantilly, VA, USA), ac-A₆YK-NH₂ (A₆YK) was derived from piCHEM GmbH (Graz, Austria). The peptide V₄WD₂ does not show ordered supramolecular structures for a large concentration regime. SAXS is extremely sensitive to aggregates, and an indicative rise in intensity at low q-values can already be seen at very low peptide concentrations (A). This suggests that self-assembly starts slowly and continuously and a distinct CAC value, such as it is conventionally determined (e.g. with fluorescence techniques), is probably not applicable. Although we have not observed ordered structures with SAXS, ATR-FTIR measurements suggest that the aggregates show internal β-sheet characteristics, seen as an amide I peak at 1630 cm⁻¹ (B). Most probably the electrostatic repulsion between charged aspartic acid residues is high and thus prevents the formation of large supramolecular structures on a long range order. Nevertheless, valine has a high propensity for hydrogen bonding and might to a certain extent outbalance electrostatic repulsion. Thus, assembly into small aggregates, held together by weak intermolecular hydrogen bonds is likely. For A₆YD, scattering patterns suggest that selfassembly into highly ordered structures starts between 1-5 mM, with characteristics of a cylindrical architecture. At very high A₆YD concentrations (> 30 mM) the SAXS patterns show the additional contribution of a structure factor at low q-values (q<0.4 nm⁻¹). This means that the individual superstructures are not in a dilute state any more, but distances of the cylinders relative to each other come into the same order of magnitude as the distances inside the individual cylinders. The distinct shape of this peak suggests that the cylinders are not only densely packed, but probably also arranged in an aligned orientation (C). ATR-FTIR measurements display a peak at 1636 cm⁻¹ in the amide I region, which can be attributed to βtype structures. Like for V₄WD₂, we observe a concentration-dependent scaling of the IR signal (D).
In conclusion, a combination of both techniques -SAXS and ATR-FTIR -shows that the morphology of the assemblies is rather sequence-dependent and does not change significantly with concentration. It means that the individual structural characteristics are developed already at low peptide concentrations, but are getting more pronounced the higher the concentration. Volume fraction (φ) 0.20 Interparticle potential, width of the square-well (λ) 1.9 Interparticle potential, depth of the square-well (ε) -0.07 Table S1 summarizes the parameters obtained by fitting the scattering pattern of 11 mM A₆YD with a cylinder-shell model, 1 combined with a sticky hard sphere interaction term. 2 This structure factor was used for simplicity. The structures show an outer radius of 3.3 nm, and an inner cylinder radius of 1.6 nm, resulting in a shell thickness of ~1.7 nm. The high volume fraction -in combination with the derived values for the interparticle potential -indicates an attractive interaction amongst the self-assembled structures. A hard sphere radius of 2.6 nm seems in contradiction to an outer cylinder radius of 3.3 nm. However, as the cylinders display either slight cross-sectional ellipticity or a slight distribution in size, both values are in agreement. Consequently, the cylindrical structures are tightly packed.

Negative-staining Transmission Electron Microscopy micrographs compared to Cryogenic Transmission Electron Microscopy (Cryo-TEM) micrographs
The procedure of negative-staining is often subject to a critical survey, since the sample solution is blotted several times and left to dry on the grid. Especially for peptide samples, where supramolecular structure development is often concentration-dependent, evaporation and blotting can lead to the formation of structures that are not present in the original solution. Cryo-TEM is known to be less invasive and should preserve vitrified structures. We were interested in differences and similarities, and investigated the peptide A₆YK with both techniques.
Negative-staining TEM sample preparation: Samples were adsorbed onto a glow-discharged carbon-coated copper grid and allowed to settle for 60 seconds. Excess fluid was carefully removed with filter paper and immediately replaced by 5 µl of a 2 % (w/v) uranyl acetate staining solution. It was allowed to settle for 30 seconds, blotted and replaced by another 5 µl of fresh solution. After 30 seconds the staining solution was removed and the sample was air-dried. Imaging was done by using a Fei Tecnai G² 20 transmission electron microscope (Eindhoven, The Netherlands) operating at an acceleration voltage of 120 kV.

Cryo-TEM sample preparation:
Sample solution was applied to a glow-discharged carbon-coated copper grid, blotted to create a thin film, plunged into liquid ethane and transferred to liquid nitrogen. Vitrified specimens were transferred onto a Gatan 626-DH cryo transfer specimen holder. Imaging was done at cryogenic temperatures in a Tecnai T12 (FEI, The Netherlands) microscope, operated at 120 kV. Figure S11 shows a direct comparison of a negatively-stained (C) and a vitrified A₆YK sample measured under cryogenic conditions (D). Negatively-stained samples exhibit a slightly denser network of fibers (C), which possibly originates from the drying procedure. In contrast, cryo-TEM samples seem to show a more aligned orientation of fibers (D). Nevertheless, both examples display the same structural morphology of A₆YK assemblies, which are characterized by almost the same sizes: lengths extend to several hundred of nanometers, whereas the fibers show diameters of 4-10 nm with negative-staining TEM, and 8-12 nm with cryo-TEM. Thus, we conclude that the micrographs of both techniques are of similar quality and yield the same information content.

Self-assembly and membrane interaction propensity of A₆YK
In order to test a positively charged analog of A₆YD, the peptide A₆YK was designed and compared to A₆YD and V₄WD₂ regarding self-assembling behavior and membrane activity. Figure S9 shows the amino acid sequence of the peptide. The critical aggregation concentration (CAC) was determined to be in the range of 1 mM (Fig. S10). Above this concentration the peptide self-assembles into cylindrical structures, which show an antiparallel internal organization of monomers, most probably stabilized by hydrogen bonding interactions. The supramolecular cylinders are several hundred nanometers in length and ~4-12 nm in diameter (Fig. S11).   With ATR-FTIR (B) we observed a major amide I peak at 1626 cm⁻¹, characteristic for either an asymmetric in plane bending vibration of lysine's NH 3 + group or intermolecular β-sheet interactions, most probably a combination of both. A minor peak was found at 1670 cm⁻¹, which we assigned to the antiparallel component of β-type structures. We assume that peptide monomers as the building blocks of the supramolecular cylinders are aligned in an antiparallel way. The presence of cylindrical structures was confirmed with TEM (C and D). Negative-staining TEM (C) shows a network of long fibers, characterized by diameters between 4-10 nm and lengths extending several micrometers. The same structures were also observed with cryo-TEM, although their diameters appeared slightly larger (8-12 nm). S17 A₆YK's propensity to interact with synthetic as well as biological membranes was assessed by a variety of biophysical techniques, namely differential scanning calorimetry (DSC), electron paramagnetic resonance spectroscopy (EPR), and small angle X-ray scattering (SAXS). Samples containing different lipid:peptide ratios (5:1, 10:1, 25:1, 50:1 mol/mol, orange curves) were prepared as described in the main article and compared to controls without peptide (grey curves). Single component phospholipid bilayers (Dipalmitoyl phosphatidylcholine multilamellar vesicles, DPPC MLVs) and low density lipoprotein (LDL) served as model systems. As can be seen in Figure S12A, B, and C the peptide A₆YK has almost no influence on DPPC MLVs. The thermotropic phase behavior of DPPC remained unaffected by the presence of A₆YK and displayed the well-characterized pre-and main-transition at 36 °C and 41.7 °C, respectively ( Fig.  S12A and Table S2). EPR spectra at 25 °C with a 5-DSA spin label showed typical characteristics of anisotropic lipid motion in membranes and did not differ whether peptides were present or not. This means that A₆YK did not affect the mobility of the spin label in both systems, DPPC and LDL ( Fig. S12B and D). SAXS patterns of pure DPPC at 25 °C showed three reflection orders with a characteristic d-spacing of 6.41 nm, which was maintained in the presence of A₆YK (Fig. S12C). Also LDL particles displayed the same characteristics of the scattering pattern as controls when A₆YK was added. The only difference could be observed in the low q-range where we noticed a slight increase in intensity. Since A₆YK has a very low CAC and all peptide concentrations are applied above the CAC we expect also a contribution of the supramolecular peptide assemblies in this particular region.   Figure S13. EPR spectra of LDL in the presence of the amphiphilic designer peptides A₆YD and V₄WD₂ at 10 °C (below the core lipid transition temperature), 25 °C (around the core lipid transition temperature) and 37 °C (above the core lipid transition temperature), with two different spin labels used (5-DSA, which probes lipid mobility in the surface lipid monolayer and 7-MeDSA, which probes lipid mobility in the hydrophobic core region). The spectra are shifted in the vertical axis for clearer visibility. Table S5. Derived EPR order parameters (S) and rotational correlation times (τ c ) for LDL + peptides The order parameters (S) were obtained as described in the main article. The EPR spectra of the 7-MeDSA spin label, which probes mobility in the core region, change from anisotropic motion to isotropic motion at temperatures above the core melting transition. As a result, the hyperfine tensors T ║ ' and T ┴ ' could not be determined without ambiguity. Therefore, the rotational correlation time (τ c ) was used as means to describe spin label mobility and was calculated as described in the main article.

S21
SAXS of LDL in the absence and presence of peptides A₆YD and V₄WD₂ Figure S14. SAXS patterns of LDL in the presence of A₆YD (A) and V₄WD₂ (B) only indicate a slight shift of the first order maximum at the highest peptide concentration. All spectra were shifted in the y-axes for clearer visibility. The interaction of A₆YK with LDL is shown in Fig.  S12E.