Elsevier

Tetrahedron

Volume 72, Issue 43, 27 October 2016, Pages 6898-6908
Tetrahedron

Rational design of novel high molecular weight solubilization surfactants for membrane proteins from the peptide gemini surfactants (PG-surfactants)

https://doi.org/10.1016/j.tet.2016.09.024Get rights and content

Abstract

We successfully developed novel high molecular weight (Mw) solubilization surfactants for membrane proteins, Bis-D3-DKDKC12 (Mw ∼3 kDa), Bis-K3-DKDKC12 (Mw ∼3 kDa), and Tris-D3-DKDKC12 (Mw ∼4.3 kDa), which were rationally designed from our previously established peptide gemini surfactants (PG-surfactants) DKDKC12K and DKDKC12D (Mw ∼1.3 kDa). An increase in Mw enhanced affinity between membrane proteins and surfactants, thereby allowing effective solubilization, even for lower concentration ranges (<0.0005 wt %).

Introduction

For most membrane protein studies, the use of solubilization surfactants for membrane proteins is indispensable, and historically, an interesting correlation exists between the development of new surfactants and progress in membrane protein researches.1, 2 Solubilization surfactants are one of the categories of surfactants that can solubilize membrane proteins in an aqueous buffer by binding to the hydrophobic transmembrane regions of the proteins originally embedded in lipid bilayers. Unlike most ionic surfactants such as sodium dodecyl sulfate (SDS), solubilization surfactants do not cause protein denaturation at either the extracellular domain or the transmembrane domain upon solubilization in a buffer containing these surfactants. Some nonionic surfactants having sugars or polyols as a hydrophilic head group, such as n-dodecyl-β-d-maltoside (β-DDM)3 and n-octyl-β-d-glucoside (β-OG),4 and zwitterionic surfactants with zwitterion as a head group, such as lauryl dimethylamine-N-oxide (LDAO),5 and 3-[(3-cholamidopropyl) dimethylammonio]-1-propanesulfonate (CHAPS),6 are well-known for their ability to solubilize membrane proteins without protein denaturation (Fig. 1). These surfactants are now commercially available and utilized to carry out both the isolation of membrane proteins from biological membranes and their structural analyses using X-ray4 and NMR.3, 7

If the research targets were only limited to the isolation of target membrane proteins from biological membranes, the evaluation of their biological activates, and the structural analyses before and after interactions with ligand molecules in a buffer, the use of current commercially available surfactants that can solubilize membrane proteins in a native-state with maintenance of single supramolecular dispersion might be sufficient in most cases. However, to improve resolution of structural analyses of membrane proteins via high-quality single crystallization, decrease in surfactants' usage, i.e., decrease in surfactants' concentration in a buffer, is effective both upon isolation of membrane proteins with maintenance of supramolecular structures on a native-state and crystallization of membrane proteins in a single crystal with less structural defects. On the other hand, recently the construction of sophisticated composite bio-nanodevices by combining membrane proteins with other enzymes or inorganic catalysts has gained a considerable attention.8 Therefore, development of novel surfactants that can solubilize membrane proteins even at much lower concentrations should be effective to organize various components in and/or onto various substrates.

The necessary number of surfactant molecules to solubilize one membrane protein molecule is generally defined from the critical micelle concentrations (CMCs) of surfactants, since solubilization of membrane proteins is achieved by including the hydrophobic transmembrane domains in a surfactants' micelle; upon less than CMCs, membrane proteins cannot be solubilize. However, the CMCs of most commercially available surfactants are over 100 μM1 and therefore, the number of free surfactant molecules that do not directly interact with membrane protein surfaces is relatively high. Hence, in order to decrease necessary number of surfactant molecules; i.e., decrease in surfactant concentration, we need to circumvent this CMC limitation. In order to solubilize each membrane protein molecule with a small number of surfactant molecules, increasing the molecular weight (Mw) of surfactants seems to be one of the reasonable approaches as an increase in molecular size of surfactants could facilitate multi-point non-covalent interactions with membrane protein surfaces, thereby reinforcing affinity. However, commercially available surfactants having higher Mw such as Triton X-100 (Mw ∼647, CMC=0.24 mM),1 Tween 20 (Mw 1227.5, CMC=0.06 mM),1 and C12E8 (Mw 538.75, CMC=0.09 mM)9 do not have adequate low CMC values. Furthermore, we cannot judge which surfactants are suitable to circumvent undesirable interactions causing protein denaturation, since it is hard to predict state of interactions between membrane proteins with surfactants (Fig. 1). Therefore, in order to construct novel high-Mw solubilizing surfactants, we need to establish a novel design concept.

Meanwhile, we recently reported that two lipopeptide-based peptide gemini surfactants (PG-surfactants), DKDKC12K and DKDKC12D, that can function as a novel class of solubilization surfactants for membrane proteins (Fig. 2).10 Generally, ‘gemini surfactants’ have two or more hydrophobic alkyl chains in a single molecule and therefore, have a superior assembling capability compared to single-chain surfactants.11 The aptitude of surfactants to form micelle assemblies from lower concentration ranges could be an advantage in terms of decreasing the necessary number of surfactant molecules required to solubilize membrane proteins in an aqueous buffer. Therefore, gemini surfactants could be one of the promising molecular frameworks to design new solubilization surfactants for membrane proteins. The basic molecular scaffold of PG-surfactants consists of three constituents,10 the linker peptide from 3 to 5 residues (X), two alkylamidomethyl-modified Cys residues at both sides of the linker peptide, and the peripheral peptide (Y) at the N-terminal side of the alkylamidomethyl-modified Cys residue (Fig. 2). From a set of screenings done on hydrophilic peptide sequences at X and Y to function the PG-surfactants as solubilization surfactants, we found that two PG-surfactants, DKDKC12K and DKDKC12D,10 were able to function as solubilization surfactants for membrane proteins. DKDKC12K and DKDKC12D have the same linker peptide (X) consisting of alternatively repeated Asp and Lys, -Asp-Lys-Asp-Lys-, and at the N-terminal, acetylated Asp or Lys as the peripheral peptide (Y), respectively. More recently, we reported the PG-surfactants having a β-turn forming peptide Asn-Pro-Asp-Gly for X and Ac-(Lys)2- for Y, which showed superior extraction efficiency for photosystem I (PSI) and photosystem II (PSII) from thylakoid membranes.12 These PG-surfactants have relatively low CMCs (10∼30 μM),10 but in this study, we aimed at designing new high-Mw surfactants, which can function with much lower concentration ranges.

The molecular structures of surfactants used in this study are summarized in Fig. 3. Upon solubilizing membrane proteins with DKDKC12K and DKDKC12D, the U-shaped [-Cys(C12)-Asp-Lys-Asp-Lys-Cys(C12)-] units might bind to the surface of the transmembrane domains of membrane proteins, orienting their C12 alkyl chains towards the protein side. Therefore, if we could tether several of these units through a flexible linker peptide without interfering their interactions, we could rationally design high Mw surfactants with reinforced interaction via multi-point interactions with protein surfaces, without losing the excellent solubilization ability of DKDKC12K and DKDKC12D. Since such a design concept, multimerizing the functional structural units for stable solubilization, is impossible to apply for other commercially available solubilization surfactants, this strategy is unique and should also provide many insights on general surfactant chemistry. In this study, we chose PSI and PSII of Thermosynecoccus vulcanus as representative membrane proteins, which involved in photosynthesis of higher plants and cyanobacteria, and evaluated the solubilization abilities of new high-Mw surfactants, from detailed examinations using fluorescence measurements, dynamic light scattering (DLS), photoinduced electron transfer activities estimated by an oxygen electrode, and transient absorption measurements.

Section snippets

Design of multimerized PG-surfactants and study of their assembly behaviors in a buffer solution

The multimerized PG-surfactants were synthesized on a Rink-Amide-AM resin by conventional Fmoc peptide synthesis, using commercially available Fmoc-protected amino acids and Fmoc-Cys(C12)single bondOH with a side chain modified with a dodecylamidomethyl group.13 The surfactants containing two [-Cys(C12)-Asp-Lys-Asp-Lys-Cys(C12)-] units linked via (Gly)4 linker and an acetylated (Asp)3 attached at the N-terminal side were named Bis-D3-DKDKC12 and those having an acetylated (Lys)3 at the N-terminal side,

Conclusion

In this report, we successfully designed novel high-Mw solubilization surfactants for membrane proteins, Bis-D3-DKDKC12 (Mw ∼3 kDa), Bis-K3-DKDKC12 (Mw ∼3 kDa), and Tris-D3-DKDKC12 (Mw ∼4.3 kDa), by simple multimerization of our previously reported monomeric DKDKC12D and DKDKC12K with the flexible (Gly)4 linker peptide. Using PSI and PSII as representative membrane proteins, we proved that neither the membrane integral domain nor the extracellular domain of PSI and PSII suffered any damage upon

Materials

N-(9-Fluorenyloxycarbonyl) (Fmoc)-protected l-amino acids, 1-hydroxybenzotriazole (HOBT), 2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate (HBTU), Rink-amide AM resin (200–400 mesh), N,N-diisopropylethylamine (DIEA), piperidine, trifluoroacetic acid (TFA), and N-methylpyrrolidone (NMP) were purchased from Merck Biosciences (Darmstadt, Germany), Novabiochem (Läufelfingen, Switzerland), and Watanabe Chemical Industries (Hiroshima, Japan). Dichloromethane (DCM) and methanol

Acknowledgements

This work was supported by JSPS KAKENHI (grant numbers 26410177, 15J07454), the Tatematsu Foundation, and in part by the Program for Advancing Strategic International Networks to Accelerate the Circulation of Talented Researchers.

References and notes (38)

  • A. Helenius et al.

    BBA Rev. Biomembr.

    (1975)
  • C. Tanford et al.

    BBA Rev. Biomembr.

    (1976)
  • A. Calcutta et al.

    BBA Biomembr.

    (2012)
  • J.U. Bowie

    Curr. Opin. Struct. Biol.

    (2000)
  • R. Zana et al.

    J. Colloid Interface Sci.

    (1998)
  • S. Itoh et al.

    Biochim. Biophys. Acta Bioenerg.

    (2001)
  • B.T. Arachea et al.

    Protein Expr. Purif.

    (2012)
  • D. Guschin et al.

    Anal. Biochem.

    (1997)
  • R.W. Egan

    J. Biol. Chem.

    (1976)
  • P. Fromme et al.

    Biochim. Biophys. Acta Bioenerg.

    (1998)
  • K. Takasaka et al.

    Biochim. Biophys. Acta Bioenerg.

    (2010)
  • I. Wittig et al.

    Mol. Cell. Proteomics

    (2007)
  • K. Kawakami et al.

    Biochim. Biophys. Acta Bioenerg.

    (2011)
  • K. Kawakami et al.

    FEBS Lett.

    (2007)
  • S. Hoshina et al.

    Biochim. Biophys. Acta Bioenerg.

    (1990)
  • R.J. Porra et al.

    Biochim. Biophys. Acta Bioenerg.

    (1989)
  • R.M. Garavito et al.

    J. Cell Biol.

    (1980)
  • S. Hiller et al.

    Science

    (2008)
  • J. Cladera et al.

    Eur. J. Biochem.

    (1997)
  • Cited by (0)

    View full text