Self-Assembling Peptide Surfactants A6K and A6D Adopt α-Helical Structures Useful for Membrane Protein Stabilization

Elucidation of membrane protein structures have been greatly hampered by difficulties in producing adequately large quantities of the functional protein and stabilizing them. A6D and A6K are promising solutions to the problem and have recently been used for the rapid production of membrane-bound G protein-coupled receptors (GPCRs). We propose that despite their short lengths, these peptides can adopt α-helical structures through interactions with micelles formed by the peptides themselves. These α-helices are then able to stabilize α-helical motifs which many membrane proteins contain. We also show that A6D and A6K can form β-sheets and appear as weak hydrogels at sufficiently high concentrations. Furthermore, A6D and A6K together in sodium dodecyl sulfate (SDS) can form expected β-sheet structures via a surprising α-helical intermediate.


Chemicals
Peptides A 6 D and A 6 K of ≥98% purity were purchased from GL Biochem (Shanghai, China) and used without further purification. The concentration of peptide stock solutions in MilliQ water was estimated by weight due to the lack of aromatic residues that could be utilized for spectroscopic concentration determination. The pH of the peptide solutions was not adjusted. Sodium dodecyl sulfate (SDS) was purchased from Ameresco Solon Inc., Ohio, USA.

Circular Dichroism (CD) Measurements
Circular dichroism measurements were performed on an Aviv 410 spectropolarimeter using a rectangular 1 mM quartz cell with a fitted cap. Temperature melts were conducted by gradually increasing the temperature in a step-wise fashion and then inverting the process. The investigated temperatures ranged from 4-90 °C (in the following steps: 4 °C, 6 °C, 10 °C, 15 °C, 20 °C, 25 °C, 37 °C, 40 °C, 50 °C, 70 °C, 90 °C and then run again at 4 °C and 25 °C) or 20-90 °C (in steps of: 20 °C, 25 °C, 37 °C, 40 °C, 50 °C, 70 °C, 90 °C and then cooled while acquiring data at the same temperatures). Prior to data acquisition, samples were equilibrated for 1 min at each temperature. Data acquisition was performed in steps of 0.5 nm between 260-190 nm, except when kinetics was fast and a pitch of 1 nm between 270-185 nm was implemented instead. The averaging time for each data point was either 20 or 4 s-depending once more on the observed kinetics. The use of a long averaging time was only justified when no time-dependent effects could be observed during the full experimental time (ranging between 8 to 25 h). Between two and 100 scans were performed at each environmental state depending on the effects being studied. All spectra were baseline corrected with blanks corresponding to appropriate temperatures, but averaging was only performed when secondary structure did not change over time. The concentration of peptides in MilliQ water ranged between 0.25-10 mM. Ultimately, all spectra were converted into mean residue ellipticity (MREs) using Equation 1: where [θ] λ is the MRE at wavelength λ in deg cm 2 dmol −1 , l is the path length in cm, c is the concentration in M and (n − 1) designates the number of peptide bonds in the studied peptide.
Increases in the amount of helical content can then be judged from a deepening of the CD signal at 222 nm, while an increase in β-sheet content can be judged from a deepening of the CD signal at 217 nm. Quantification of the amount of α-helical and β-sheet contents was however not done, due to the absence of aromatic residues, which prevented the precise determination of peptide concentration needed for quantification.

Field Emission Scanning Electron Microscopy (FESEM) Studies
Samples were frozen at −80 °C or preferably shock frozen in liquid nitrogen and vacuum dried. Subsequently they were fixed onto a sample holder using conductive tape and sputtered with platinum from both the top and the sides in a JEOL JFC-1600 High Resolution Sputter Coater. The coating )) 1 Obs θ θ λ current used was 30 mA, and the process lasted for 60 s. The surface of interest was then examined with a JEOL JSM-7400F field-emission scanning electron microscopy system using an accelerating voltage of 5-10 kV.
Samples of A 6 D and A 6 K (0.4 mM) were heated from 4 °C to 90 °C and then cooled back to 4 °C, in the presence of 13 mM SDS. While A 6 D remained predominantly random coiled ( Figure 1A), A 6 K astonishingly formed well-defined α-helices as judged from the CD spectra ( Figure 1B). Typical α-helices have 3.6 residues per turn, and are usually formed by longer peptides. Being 7 residues long, A 6 K can hardly form a full turn of the α-helix and can only form 4 intra-molecular hydrogen bonds. Its surprising ability to form a stable α-helix makes it one of the shortest α-helices observed.
The length of an A 6 K molecule is about 1.1 nm, comparable to the length of an outstretched SDS molecule, which is 1.67 nm [23]. Given the structural similarity of A 6 K to amphiphilic SDS molecules, it is very likely that A 6 K interacted with SDS molecules to form micellar structures ( Figure 1C). This conformational change occurred instantly when SDS was added to A 6 K and the partial unfolding of the helix due to heating from 4 °C to 90 °C exhibited complete thermal reversibility when cooled to 4 °C.
A 6 D, however, did not show any observable secondary structural transitions when 13 mM SDS was added to it at 25 °C. This is probably because the positively-charged lysine headgroup of A 6 K interacts better with the anionic SDS molecules, as compared to the negatively-charged Asp headgroup of A 6 D.

B C
A major concern in the interpretation of the results was that some measurements were made below the Krafft point of SDS, 22 °C. Above this Krafft point, the solubility of SDS greatly increases in aqueous systems and is regarded as the melting temperature of the hydrated SDS solid. Hence below the Krafft point, SDS does not form micelles but hydrated crystals. In the experiment above, SDS was added to the peptides at 25 °C and then rapidly cooled to 4 °C. Micelles were formed and probably stabilized by the presence of the peptides, and hence remained intact even at 4 °C. Low temperatures were used in this case in an attempt to slow down the kinetics of the peptides and SDS interaction, to a suitably low rate for observation. The trend of the CD signal below and above the Krafft point agree with each other, indicating that the Krafft point probably did not affect the peptide-SDS interactions very significantly.
However, if SDS at 4 °C was added to the peptide, no structural induction was seen. This supports the hypothesis that SDS micelles were needed for the structural induction of A 6 K to α-helices.

Secondary Structures of A 6 D and A 6 K Peptide Surfactants in Water
Heating samples of each peptide (0.4 mM) from 4 °C to 90 °C and then cooling back to 4 °C in water revealed a mostly random coiled conformation with a small proportion of helical motifs which increased at higher temperatures. The secondary structural changes were fully reversible (Figure 2A and 2B), suggesting weak interactions.  At higher concentrations of 2.5 mM, A 6 K formed α-helices which were not fully reversible ( Figure 3A), but A 6 D remained predominantly random-coiled (data not shown), reflecting similar results to the previous experiments in 13 mM SDS. This is probably possible because the amphiphilic peptides at higher concentrations can form micelles in place of SDS, to induce α-helical structures on the peptides.
At even higher concentrations, both A 6 K and A 6 D can form β-sheets ( Figure 3B and 3C). Under FESEM, these are seen as spongy, cavity-containing membranous structures which allow large quantities of water to be held ( Figure 3D-G). These peptide solutions hence tend to appear as hydrogels. The process of assembly from α-helices to β-sheets is however not yet clear.

Secondary Structures of A 6 D and A 6 K in Combination, in Water
We were then interested in how the two peptides would interact when placed together, given their complementary structures. When 0.2 mM A 6 D and 0.2 mM A 6 K were combined and subjected to the same temperature treatment, the resultant CD spectra ( Figure 2C) appears to be a combination of the spectra observed in Figure 1A and 1B. This suggests that the combined A 6 D and A 6 K structures are similar to their individual ones.

Secondary Structural Changes of a Combination of A 6 D and A 6 K in SDS
However, when 20 mM SDS was added to a mixture of 0.2 mM A 6 D and 0.2 mM A 6 K at 50 °C and incubated for 8.5 h, anti-parallel β-sheets ( Figure 4A) were formed which were very stable and could not be reversed by heating to 90 °C, cooling to 4 °C, sonication (20 min) or subsequent SDS additions.  the two peptides are present in two different conformations and the CD spectra are simply average representations.
As temperature increases, the SDS micelles become destabilized, allowing A 6 D to interact with the less SDS-shielded A 6 K, probably initiated by electrostatic attraction between the oppositely-charged headgroups. The charge neutralization following the formation of an ionic bond allows the hydrophobic driven incorporation of peptide dimers into micelles, with a subsequent conversion into β-sheets. The presence of SDS provides the aprotic environment required to overcome the activation energy needed to induce assembly.

Conclusions
Our study of the above two peptides hypothesizes a probable mechanism in which short surfactant peptides such as A 6 D and A 6 K could aid in the cell-free production of membrane proteins. The amphiphilic property of these peptides allows the formation of micelles that induce peptides to adopt α-helical conformations and stabilize these membrane proteins.
Furthermore, we describe a mechanism by which A 6 D and A 6 K could assemble into β-sheets via an α-helical intermediate, requiring an aprotic condition provided by the presence of SDS. We believe these assembly processes are not only relevant to such simple systems, but may also be involved in the amyloid fibril formation of many diseases such as Alzheimer's and Parkinson's. By studying these relatively simple systems, the assembly processes of these diseases may also be unraveled.