Improved glucose-neopentyl glycol (GNG) amphiphiles for membrane protein solubilization and stabilization

Article abstract of DOI:10.1002/asia.201301303

Membrane proteins are inherently amphipathic and undergo dynamic conformational changes for proper function within native membranes. Maintaining the functional structures of these biomacromolecules in aqueous media is necessary for structural studies but difficult to achieve with currently available tools, thus necessitating the development of novel agents with favorable properties. This study introduces several new glucose-neopentyl glycol (GNG) amphiphiles and reveals some agents that display favorable behaviors for the solubilization and stabilization of a large, multi-subunit membrane protein assembly. Furthermore, a detergent structure-property relationship that could serve as a useful guideline for the design of novel amphiphiles is discussed. Copyright

Full text of DOI:10.1002/asia.201301303

FULL PAPER
DOI: 10.1002/asia.201301303
Improved Glucose-Neopentyl Glycol (GNG) Amphiphiles for Membrane
Protein Solubilization and Stabilization
Kyung Ho Cho,[a] Hyoung Eun Bae,[a] Manabendra Das,[a] Samuel H. Gellman,[b] and
Pil Seok Chae*[a]
Abstract: Membrane proteins are in-
herently amphipathic and undergo dy-
namic conformational changes for
proper function within native mem-
branes. Maintaining the functional
structures of these biomacromolecules
in aqueous media is necessary for
structural studies but difficult to ach-
ieve with currently available tools, thus
necessitating the development of novel
agents with favorable properties. This
study introduces several new glucose-
neopentyl glycol (GNG) amphiphiles
and reveals some agents that display
favorable behaviors for the solubiliza-
tion and stabilization of a large, multi-
subunit membrane protein assembly.
Keywords: amphiphiles
membrane proteins
design
protein stabilization
·
Furthermore,
a detergent structure–
·
molecular
property relationship that could serve
as a useful guideline for the design of
novel amphiphiles is discussed.
·
protein solubilization ·
Introduction
their instability in non-native environments, and their high
flexibility caused by dynamic conformation changes within
the lipid bilayer.[4] Maintaining native membrane protein
conformations is a particularly formidable task because
these macromolecules tend to aggregate and denature when
extracted from native membranes for study.[5] Thus, a mem-
brane-mimicking system must be used to prevent structural
degradation of these biomacromolecules.
Detergents tend to self-assemble to form micelles. Deter-
gent micelles are often illustrated having a globular or
eclipsed shape with a hydrophobic interior and hydrophilic
exterior. This architecture confers the micelles with an abili-
ty to interact with the hydrophobic portions of membrane
proteins in an aqueous environment. Thus, detergent mi-
celles serve as excellent media for membrane protein solubi-
lization and stabilization, and are indispensable tools for
membrane protein studies.[6] Membrane protein research
starts with protein extraction from the membrane by deter-
gent molecules. It is a prerequisite to keep the resulting pro-
tein–detergent complexes (PDCs) soluble in the subsequent
multistep processes, including protein purification and crys-
tallization.
Membrane proteins are amphipathic macromolecules that
reside in lipid bilayers called membranes. It is estimated
that a third of all open reading frames in the human
genome encode membrane proteins that are distributed
over several different types of membranes, including the
plasma membrane, nuclear membrane, and inner and outer
mitochondrial membranes.[1] Integral membrane proteins
play central roles in a variety of cellular processes such as
material transport, signal transduction, cell adhesion, and
cell-to-cell communication. The importance of these bio-
macromolecules in normal and disease states is reflected by
the fact that more than half of all pharmaceutical agents cur-
rently under development target membrane proteins.[2] De-
spite its biological and pharmaceutical importance, mem-
brane protein research lags far behind that of its soluble
counterpart, largely because of difficulty in handling these
proteins.[3] The structures of only several hundred mem-
brane proteins are known, in contrast to tens of thousands
of soluble proteins with known structure. This slow progress
is mainly due to a low natural abundance of these proteins,
A large number of conventional detergents with different
combinations of hydrophilic and hydrophobic groups are
available, but only a small number are widely used for mem-
brane protein manipulation. For instance, the five conven-
tional detergents, OG (n-octyl-b-d-glucopyranoside), NG
(n-nonyl-b-d-glucopyranoside), DM (n-decyl-b-d-maltopyra-
noside), DDM (n-dodecyl-b-d-maltopyranoside), and
LDAO (lauryldimethylamine-N-oxide), have facilitated crys-
tal structure determinations of about 70% of a-helical mem-
brane proteins.[7] It is interesting to note that these popular
agents share their behaviors for membrane protein manipu-
lation. They are not only efficient at extracting membrane
[a] K. H. Cho, H. E. Bae, Dr. M. Das, Prof. P. S. Chae
Department of Bionanotechnology
Hanyang University
Ansan, 426-791 (Korea)
Fax : (+81)31-436-8146
E-mail: pchae@hanyang.ac.kr
[b] Prof. S. H. Gellman
Department of Chemistry
University of Wisconsin-Madison
1101 University Avenue, Madison, WI 53706 (USA)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/asia.201301303.
Chem. Asian J. 2014, 9, 632 – 638
632
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemasianj.org
Pil Seok Chae et al.
proteins from membranes but also effective at stabilizing
the solubilized proteins. Despite their wide utility, mem-
brane proteins in detergent micelles often undergo structur-
al degradation through denaturation and/or aggregation.[4]
The absence of lateral pressure exerted by a membrane[8] is
regarded as the main reason for limited membrane protein
stability in detergent micelles. Protein extraction and purifi-
cation involves loss of lipid molecules specifically bound to
membrane protein surfaces.[9] Along with micelle inhomoge-
neity, this could contribute to the observed PDC instability.
Furthermore, most membrane proteins with unknown struc-
tures are more difficult to solubilize and stabilize than those
with known structures.[10] Thus, the development of novel
amphiphiles with enhanced protein solubilization and stabi-
lization efficacy is of great importance for the future success
of membrane protein research.
vorable solubilization and/or stabilization efficacy toward
a large, multi-subunit membrane protein assembly.
Results and Discussion
Design and Characterization of New GNG Amphiphiles
GNG agents were first designated according to their func-
tional groups in the central region (Scheme 1); GNG-1 and
GNG-5 derivatives contain ether and amide linkages, re-
spectively, while GNG-6 derivatives have thioether linkages.
Many conventional detergents have the common structur-
al features of a flexible tail group and polar head group
such as a glucoside, maltoside, or N-oxide. A number of
novel classes of amphiphiles with different architectures
have been devised to overcome the limited performance of
conventional detergents.[11] Representative examples include
amphipathic polymers (amphipols),[11a–c] tripod amphiphiles
(TPAs),[11d–f] hemifluorinated surfactants (HFSs),[11g,h] nano-
discs (NDs),[11c,i] rigid hydrophobic group-bearing amphiphi-
les,[11j–l] peptide-based amphipathic agents,[11m–o] facial amphi-
philes (FAs),[11p–r] and neopentyl glycol-based amphiphiles
(glucose-neopentyl glycol (GNG) and maltose-neopentyl
glycol (MNG).11s–u] Despite the structural diversity and inno-
vative designs of these novel agents, only a few classes are
successful in membrane protein crystallization, as exempli-
fied by FAs,[11r] GNGs and MNGs.[12] It is particularly note-
worthy that GNG and MNG agents have facilitated 10 new
X-ray crystal structure determinations of membrane pro-
teins over the last three years, including several G-protein
coupled receptors (GPCRs) such as b2 adrenergic recep-
tors,[12a–c] acetylcholine receptors,[12d,e] and opioid recep-
tors.[12f,g] This indicates that these class members are indis-
pensable tools for membrane protein science. GNG and
MNG class members share a quaternary carbon in the cen-
tral region from which two hydrophobic and hydrophilic
groups project. Our previous study showed that in addition
to favorable protein stabilization behavior, GNG agents
tend to form small PDCs relative to those formed by DDM,
presumably due to their short tail- and head-groups.[11u] This
attribute of GNG would provide a favorable environment
for PDC-based membrane protein crystallization by
facilitating nucleus and crystal lattice formation. The use of
a GNG led to the generation of high-quality crystals of
some membrane proteins, including sodium-pumping pyro-
phosphatase.[11u,12j]
Scheme 1. Chemical structures of previously described GNG amphiphiles
(GNG-1 and GNG-3) and newly prepared GNG analogues (GNG-1-1,
GNG-1-2, GNG-3-1, GNG-5, GNG-5-1, GNG-6, and GNG-6-1).
The hydrophobic groups of GNG-3 were directly attached
to the central quaternary carbon. GNG agents were further
designated according to their hydrophobic group variations.
The hexyl chains (C6) of the previously reported GNGs,
GNG-1 and GNG-3, were replaced with pentyl chains (C5)
to give GNG-1-1 and GNG-3-1, respectively. These short
alkyl chain GNGs were designed based on the idea that con-
ventional detergents with a short alkyl chain (e.g., OG) are
more favorable than their long alkyl chain counterparts
(e.g., DDM) in PDC-based protein crystallization, provided
that native protein structures can be effectively main-
tained.[13] Due to synthetic convenience, an amide coupling
reaction was utilized to generate GNG-5 with C6 alkyl
chains and GNG-5-1 with cyclohexane rings. Note that the
three new GNGs, GNG-1-2, GNG-5-1, and GNG-6, bear cy-
clohexane rings as hydrophobic groups. These GNGs were
designed based on the popular use of cyclohexane ring-con-
taining conventional detergents (e.g., CYGLUs and
CYMALs; Figure S1, Supporting Information). To investi-
gate potential differences in detergent efficacy between cy-
clohexane- and benzene-bearing GNGs, 2-phenylethanethiol
was used instead of cyclohexanethiol to produce GNG-6-1.
Another favorable aspect of this class is its high structural
flexibility and synthetic accessibility. Thus, detergent proper-
ties can be efficiently fine-tuned by preparing a number of
family members. In the current study, several new GNG an-
alogues were prepared based on previously reported
GNGs.[11u] Some of the new GNG amphiphiles showed fa-
Chem. Asian J. 2014, 9, 632 – 638
633
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemasianj.org
Pil Seok Chae et al.
Except for GNG-1-2, all new GNG agents were prepared
by straightforward chemical reactions (3 steps) in high over-
all yields (80–90%). The protocol allowed for the prepara-
tion of multi-gram quantities, thus supporting the availabili-
ty of new agents. One of main reasons why many novel am-
phiphiles have not become popular within the membrane
protein community is believed to be due to their impractica-
bility. Most GNG agents are highly water-soluble (>
10 wt%), but the solubility of the cyclohexane-bearing
GNGs in water was dependent on the functional group pres-
ent in the central region. An amide-functionalized GNG,
GNG-5-1, was water-insoluble and thus not studied further.
Ether- and thioether-functionalized GNGs (GNG-1-2 and
GNG-6, respectively) are water-soluble >10 wt%. Note that
the cyclohexyl version of GNG-3 (GNG-3-2; Figure S2, Sup-
porting Information) was not prepared because a previous
study showed that the constitutional isomer of this agent,
GNG-4, is barely water-soluble.[11u]
region hinders micelle formation.[15] The amide-containing
GNG, GNG-5, had a weak tendency to self-associate, there-
by giving a large CMC value (~18 mm; ~1.2 wt%). The thio-
ether-bearing GNGs, GNG-6 and GNG-6-1, showed high
propensity to aggregate, giving small CMC values (~2.1 mm
and ~0.82 mm, respectively). The relative polar and nonpo-
lar characteristics of amide and thioether groups, respective-
ly, are responsible for their different self-association behav-
iors. Micelles formed by all new GNGs and previously re-
ported GNGs are smaller than those formed by DDM,
which is likely related to their property of small PDC forma-
tion.[11u] Notably, short alkyl chain GNGs (GNG-1-1 and
GNG-3-1) formed even smaller micelles than their original
compounds, GNG-1 and GNG-3, respectively. This is be-
cause a decrease in the length of the aliphatic tail group
causes molecules to be more cone-shaped, thereby leading
to the formation of smaller micelles. This result is consistent
with the idea that the relative size of hydrophilic and hydro-
phobic groups (i.e., molecular geometry) is responsible for
the self-association behavior of detergents.[16] Micelles
formed by the amide-functionalized GNG (GNG-5; Rh = ~
2.40 nm) and two thioether-functionalized GNGs (GNG-6
and GNG-6-1; Rh = ~2.66 nm and ~2.98 nm, respectively)
were comparable to those formed by ether-functionalized
GNG-1 (Rh = ~2.64 nm). Thus, functional group differences
in the central region have only a minimal effect on the mi-
celle size. The micelle size range displayed by the new
GNGs (1.9–3.1 nm) overlaps with the reported range of OG
micelles (1.5–2.3 nm),[17] thus suggesting that these GNGs
resemble OG with respect to small PDC formation. GNGs
and DDM micelles were further characterized by their size
distribution. All agents except GNG-1-1 showed one micelle
distribution, as did DDM (Figure S3, Supporting Informa-
tion). The DLS diagram for GNG-1-1 indicates the presence
of two sets of micelles with hydrodynamic radii of about
2.3 nm and 59 nm, respectively. Scattering intensity analysis
indicates that the set of small micelles is an almost exclusive
entity in the solution (see the caption of Figure S3 for de-
tails, Supporting Information).
Micelles formed by all amphiphiles were characterized by
solubilization experiments using a hydrophobic fluorescent
dye, diphenylhexatriene (DPH),[14] and by dynamic light
scattering (DLS). Table 1 summarizes the critical micelle
Table 1. Molecular weight (MW), critical micelle concentration (CMC),
and hydrodynamic radius (Rh) of micelles (meanÆSD, n=4), and solubi-
lization yields (SYs) for GNGs (GNG-1, GNG-1-1, GNG-1-2, GNG-3,
GNG-3-1, GNG-5, GNG-6, and GNG-6-1) and conventional detergents
(DDM and OG).
Detergents
MW[a]
CMC [mM; wt%]
Rh [nm][b]
SY [%]
GNG-1
GNG-1-1
GNG-1-2
GNG-3
GNG-3-1
GNG-5
GNG-6
GNG-6-1
DDM
628.8
600.7
624.7
568.7
540.6
682.8
656.8
700.9
510.1
292.4
~1.6; ~0.10
~11; ~0.65
~17; ~1.1
~1.0; ~0.058
~6.9; ~0.37
~18; ~1.2
~2.1; ~0.13
~0.82; ~0.057
~0.17;~0.0087
~25;~0.73[e]
2.64Æ0.04
2.28Æ0.01[c,d]
2.30Æ0.02[c]
3.07Æ0.01
1.96Æ0.04[c]
2.40Æ0.03[c]
2.66Æ0.09
2.98Æ0.11
3.47Æ0.04
1.5–2.3[e]
~90
~40
~70
~80
~60
~80
~90
~95
~80
>95
OG
[a] Molecular weight of detergents. [b] Hydrodynamic radius of micelles
determined by dynamic light scattering at 0.5 wt%. [c] These agents were
tested at higher concentration (2.0 wt% for GNG-1-1 and GNG-3-1, or
4.0 wt% for GNG-1-2 and GNG-5) to obtain a strong signal. [d] Two
forms of aggregates were found with hydrodynamic radii of about 2.3 nm
and 59 nm (Figure S3, Supporting Information). [e] This value was ob-
tained from ref. [17].
Evaluation of GNGs for Membrane Protein Solubilization
and Stabilization
To evaluate the solubilization efficiency and stabilization ef-
ficacy of new GNG agents, the photosynthetic superassem-
bly of Rhodobacter (R.) capsulatus was employed. This
system comprises a resilient reaction center complex (RC)
and a labile light-harvesting complex I (LHI). The robust
light-harvesting complex II (LHII), which is present in the
natural source of the superassembly, was genetically deleted
for the purpose of detergent evaluation.[18] The LHI-RC su-
perassembly is a pigment–protein complex that contains 30–
40 subunits and various types of cofactors, including bacter-
iochlorophylls and carotenoids. The high susceptibility of
LHI to denaturation, which is due to its large size and multi-
ple quaternary structures, enabled us to discriminate be-
tween the efficacies of a set of mild detergents. LHI tends
concentrations (CMCs) and hydrodynamic radii (Rh) of mi-
celles formed by these new detergents. The table includes
data for two conventional detergents (DDM and OG) and
two previously described GNGs (GNG-1 and GNG-3) for
comparison. The CMC values of GNG-1-1 and GNG-3-
1 were around seven times larger than those of their parent
molecules, GNG-1 and GNG-3, respectively. Interestingly,
GNG-1-2 had a 10-fold larger CMC value than its constitu-
tional isomer, GNG-1. In spite of the similar hydrophobicity,
the large difference in the CMC value between GNG-1 and
GNG-1-2 is likely due to the bulkiness of the cyclohexyl
group. The presence of a bulky group in the lipophilic
Chem. Asian J. 2014, 9, 632 – 638
634
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemasianj.org
Pil Seok Chae et al.
to gradually denature even in mild conventional detergents
such as DDM and OG. Detergent efficacies with a medium
range of strength could be differentiated by virtue of the re-
silient character of RC. Only harsh detergents such as
sodium dodecyl sulfate (SDS) completely destroyed the RC
complex during solubilization. Accordingly, we unambigu-
ously assessed detergent efficacies in a graded way based on
the denaturation extent of the LHI-RC superassembly. This
complex is also a convenient test bed for detergent evalua-
tion because the multiple cofactors embedded in the com-
plex interior produce a characteristic UV/Vis spectrum. This
enabled us to utilize spectrophotometry to determine pro-
tein quantity and integrity. The native conformation of the
LHI-RC complex has an intense peak at 875 nm, while
intact RC with denatured LHI, and denatured LHI and RC
have characteristic peaks at around 800 nm and 760 nm, re-
spectively.
Figure 1. a,b) Absorbance spectra of R. capsulatus superassembly solubi-
lized (a) and purified (b) in representative GNGs (GNG-1-2 and GNG-
6-1; solid lines) and two conventional detergents (DDM and OG; dotted
lines). Detergents were used at CMC+1.0 wt% for superassembly solu-
bilization. For protein purification, Ni-NTA affinity chromatography was
performed and protein complexes were collected from the column by
using 1xCMC detergent concentration. Absorbance values were calculat-
ed from dilution factors (normally 1:5 or 1:10).
This study included two conventional detergents (DDM
and OG) and two previously described GNGs (GNG-1 and
GNG-3) for comparison. For the initial solubilization of
LHI-RC complexes, intracytoplasmic R. capsulatus mem-
branes were treated with CMC+1.0 wt% individual GNGs
or conventional detergents. The insolubilized parts of mem-
branes and cellular debris were then removed by ultracentri-
fugation. The resulting supernatant containing detergent-
solubilized superassembly was collected while the pellet
containing insolubilized parts was resuspended in an aque-
ous medium. UV/Vis spectra of these two portions were ob-
tained to assess detergent efficacy regarding protein solubili-
zation and stabilization. Protein solubilization yield (SY;
Table 1) for each detergent was calculated from the differ-
ence between the initial amount of superassembly and the
protein amount of homogenized pellets following solubiliza-
tion (Table 1 and Figure S4, Supporting Information). A
previous study showed that DDM is one of the best conven-
tional detergents for solubilization and stabilization of the
complexes, which is in agreement with the popular use of
this agent in membrane protein studies.[19] Consistent with
this observation, DDM efficiently extracted the complexes
from the membrane (~80%) without structural degradation
(Figure 1a and Figure S4, Supporting Information). Strong
absorbance at 875 nm indicates the intact nature of DDM-
solubilized complexes. The other non-ionic conventional de-
tergent, OG, extracted protein almost quantitatively (>
95%), but the resulting LHI structure was partially degrad-
ed in the course of protein solubilization. A characteristic
peak at 760 nm indicates this partial protein denaturation
(Figure 1a and Figure S5a, Supporting Information). Re-
markably, all GNG agents solubilized the superassembly
without denaturation, thus revealing their generally favora-
ble property with respect to membrane protein stabilization.
However, a large variation in detergent solubilization effi-
ciency was observed. The new GNG derivatives with short
alkyl chains (C5) (GNG-1-1 and GNG-3-1) were substantial-
ly inferior to their parent GNGs with C6 alkyl chains
(GNG-1 and GNG-3) (Figures S4 and S5a, Supporting In-
formation). A similar result was found for cyclohexane-con-
taining GNG-1-2. The high CMC values of these agents
(GNG-1-1, GNG-1-2, and GNG-3-1; 7–17 mm) relative to
those of other GNGs may be a good indication for such
poor protein solubilization efficiency (Table 1). Detergents
with high CMC values have a low tendency of self-associa-
tion mainly because of reduced lipophilicity. Thus, the tail
groups of high CMC detergents are likely to weakly interact
with the hydrophobic portion of membrane proteins. GNG-
5 with amide linkages in the central region was less efficient
than DDM in the complex solubilization (Figure S4, Sup-
porting Information). By contrast, thioether-functionalized
GNGs (GNG-6 and GNG-6-1) were superior to DDM and
other GNGs in this regard (Table 1 and Figure S4; ~90%
and ~95% vs. ~80%). Overall, the new GNG amphiphiles
showed intriguing behaviors in this initial evaluation with
LHI-RC complexes.
To investigate the utility of these new agents for mem-
brane protein purification, individual detergent-solubilized
LHI-RC complexes were subjected to immobilized nickel
affinity chromatography. The complex contains a nickel
binding moiety (i.e., hepta-histidine tag) on the C-terminus
of the M-subunit of RC. A highly pure LHI-RC in solution
was obtained by eluting the resin-bound complexes from the
affinity column using an elution buffer containing 1m imida-
zole and 1xCMC individual amphiphiles. As shown in Fig-
ure 1b and Figure S5b in the Supporting Information, the
spectra of GNGs-, DDM-, or OG-purified complexes resem-
ble those of the complexes solubilized with the individual
agents. This data indicates that all GNGs and DDM were
successful in retaining native complex conformations in the
course of protein purification. Conversely, the rather harsh
nature of OG further destroyed the complexes during purifi-
cation, as indicated by the small shoulders at 760 nm and
800 nm (Figure S5b, Supporting Information).
Chem. Asian J. 2014, 9, 632 – 638
635
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemasianj.org
Pil Seok Chae et al.
formation). The stabilization efficacies of new GNGs (short
alkyl chain GNGs, and amide- and thioether-functionalized
GNGs) were comparable to that of DDM. Notably, the cy-
clohexane-bearing GNG-1-2 was clearly superior to the con-
ventional detergents (OG and DDM) and other GNGs (Fig-
ure 2a). When the detergent concentration was increased to
CMC+1.0 wt%, the difference in detergent efficacy was
more prominent (Figure 2b and Figure S6b, Supporting In-
formation). Under these harsh conditions, OG-solubilized
proteins completely lost their integrity after a few days,
while DDM-solubilized complexes gradually underwent
structural degradation. Only 10% of the LHI-RC complexes
retained their native conformation after 20 days of storage.
Every GNG agent displayed a more favorable behavior
than DDM, with the best performance of GNG-1-2. In this
cyclohexane-containing GNG, approximately 50% of LHI-
RC complexes retained their native conformation after
20 days (vs. ~10% in DDM, Figure 2b). The difference in
the detergent stabilization efficacies of GNG-1-2 and DDM
is also well illustrated by time-course changes in the absorp-
tion spectra of the complexes solubilized in each detergent
(Figure 2c,d).
Long-Term Stabilization Efficacy of GNG Agents
The initial short-term solubilization and purification proto-
col was not satisfactory in discerning the membrane protein
stabilization efficacy of individual amphiphiles except for
OG. For this reason, these mild agents (DDM and GNGs)
were further evaluated for long-term stabilization efficacy.
LHI-RC complexes were solubilized with 1.0 wt% DDM
and purified in 1xCMC DDM through metal affinity chro-
matography according to the protocol described above. The
DDM-purified protein solution was then diluted into indi-
vidual amphiphile-containing solutions to reach a final
DDM concentration far below its CMC and individual am-
phiphile concentrations of CMC+0.04 wt% or CMC+
1.0 wt%. Protein integrity was monitored over a 20 day
period by measuring the absorbance at 875 nm during the
storage at room temperature (Figure 2 and Figure S6, Sup-
Conclusions
Hydrophobic variations of previously reported GNG amphi-
philes were prepared by introducing various lipophilic
groups and/or functional groups to the central region. Their
ability to solubilize and stabilize a challenging membrane
protein complex, R. capsulatus superassembly, was assessed
by spectrophotometry. The new GNG agents displayed
medium to excellent efficiency at solubilizing the superas-
sembly from the membrane, with the best performance of
a thioether-functionalized agent, GNG-6-1. We have devel-
oped a few novel classes of amphiphiles with satisfactory
protein stabilization efficacy, exemplified by TPAs, MNGs,
and TFAs. Among these agents, GNG-6-1 is one of the best
agents in terms of superassembly solubilization efficiency.
Amphiphiles with such high solubilization efficiency tend to
partially destroy the LHI-RC complexes, as observed for
conventional detergents (Triton X-100, OG, and
LDAO)[11u,20] and a tripod amphiphile (TPA-4).[11e] The fa-
vorable solubilization behavior likely results from the high
affinity of these agents to the hydrophobic surface of the su-
perassembly. This strong interaction could serve as a favora-
ble role in membrane protein crystallization by preventing
protein aggregation, which is one of the main mechanisms
of protein structural degradation. Based on the fact that the
three most widely used conventional detergents, DDM, OG,
and LDAO, are efficient in membrane protein solubilization,
we believe that the new GNG agents have potential utility
in membrane protein research. Membrane protein stability
is probably the most important determinant for the success
of membrane protein structural studies. In this context,
GNG-1-2 showed promising behavior as this agent was out-
standing at maintaining the native structure of the photosyn-
Figure 2. a,b) Time course stability of R. capsulatus superassembly at
CMC+0.04 wt% (a) and CMC+1.0 wt% (b). Absorption spectra were
taken over a 20 day incubation period at room temperature for com-
plexes solubilized in GNGs (GNG-1, GNG-1-1, GNG-1-2, GNG-3,
GNG-3-1, GNG-5, GNG-6, and GNG-6-1) and conventional detergents
(DDM and OG). Results are expressed as % absorbance at 875 nm
(A875) relative to measurements at day 0. c,d) Time course change in the
absorption spectra of GNG-1-2-solubilized (c) and DDM-solubilized (d)
protein samples. Detergents were used at CMC+1.0 wt%. A binding
buffer (10 mm Tris, pH 7.8, containing 100 mm NaCl) was used for long-
term protein storage.
porting Information). Consistent with a previous study,[11u] at
relatively low detergent concentration (i.e., CMC+
0.04 wt%), GNG-1- and GNG-3-solubilized protein com-
plexes preserved the native conformation as effectively as
DDM while OG-purified complexes gradually lost their in-
tegrity with time (Figure 2a and Figure S6a, Supporting In-
Chem. Asian J. 2014, 9, 632 – 638
636
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemasianj.org
Pil Seok Chae et al.
thetic superassembly at both low and high detergent concen-
trations.
Despite their wide use in membrane protein studies, the
Acknowledgements
The National Research Foundation of Korea (NRF) funded by the
Korean government (MSIP) (grant number 2008-0061891 and
2012R1A1A1040964 to P.S.C., K.H.C., H.E.B., M.D.) supported this
work. We thank Philip Laible (Argonne National Laboratory, IL, USA)
for supplying membrane preparation from R. capsulatus.
currently available conventional detergents have limitations.
DDM tends to form large PDCs while OG and LDAO are
often too harsh to stabilize delicate membrane proteins.[13]
New tools with favorable solubilization and stabilization ef-
ficacy and the ability to form small PDCs are needed. The
GNG amphiphiles introduced here are excellent candidates
for meeting these criteria because this class tends to form
small PDCs, as shown in a previous study,[11u] while display-
ing favorable solubilization efficiency and stabilization effi-
cacy, as described here. The short-chain detergents, GNG-1-
1 and GNG-3-1, may also find use in membrane protein
structural studies; these agents have alkyl chains as short as
that of OG, but they showed significantly enhanced stabili-
zation efficacy. Although it is unclear whether the results of
the current study will translate to other membrane protein
systems as well, this initial evaluation reveals that some
GNG members are promising alternatives to conventional
detergents for membrane protein study.
[1] a) E. Wallin, G. von Heijne, Protein Sci. 1998, 7, 1029–1038; b) A.
Krogh, B. Larsson, G. von Heijne, E. L. Sonnhammer, J. Mol. Biol.
2001, 305, 567–580.
[2] J. P. Overington, B. Al-Lazikani, A. L. Hopkins, Nat. Rev. Drug Dis-
covery 2006, 5, 993–996.
[3] http://blanco.biomol.uci.edu/Membrane_Proteins_xtal.html.
[4] J. J. Lacapꢁre, E. Pebay-Peyroula, J. M. Neumann, C. Etchebest,
Trends Biochem. Sci. 2007, 32, 259–270.
[5] a) S. H. White, W. C. Wimley, Annu. Rev. Biophys. Biomol. Struct.
1999, 28, 319–365; b) J. V. Møller, M. le Maire, J. Biol. Chem. 1993,
268, 18659–18672.
[6] a) R. M. Garavito, S. Ferguson-Miller, J. Biol. Chem. 2001, 276,
32403–32406; b) J. U. Bowie, Curr. Opin. Struct. Biol. 2001, 11, 397–
402.
[7] a) M. J. Serrano-Vega, F. Magnani, Y. Shibata, C. G. Tate, Proc.
Natl. Acad. Sci. USA 2008, 105, 877–882; b) S. Newstead, S. Ferran-
don, S. Iwata, Protein Sci. 2008, 17, 466–472.
The current study could provide insightful information on
which detergent moiety plays a crucial role in determining
protein solubilization efficiency and stabilization efficacy.
Here, we prepared several hydrophobic variations of GNG
agents with the same hydrophilic group of a branched diglu-
coside. Evaluation for the the superassembly showed that
the hydrophobic variations resulted in small changes in pro-
tein stabilization efficacy, but a wide range in protein solubi-
lization efficiency (Table 1; 40–95%). This result indicates
that detergent hydrophobic group dictates protein solubili-
zation efficiency rather than protein stabilization efficacy. In
previous studies,[11s,u] MNGs with a branched dimaltoside
outperformed GNGs at maintaining membrane protein sta-
bility. The MNG architecture bears a resemblance to that of
GNGs except for the hydrophilic group. Thus, these findings
suggest that the detergent hydrophilic group mainly deter-
mines protein stabilization efficacy. A similar trend was also
observed for the TPA architecture.[21] Therefore, these ob-
servations strongly support an important detergent struc-
ture–property relationship: protein stabilization efficacy is
most effectively altered by modifying the detergent hydro-
philic group, while protein solubilization efficiency is mainly
affected by the detergent hydrophobic group. Because both
detergent solubilization and stabilization properties play es-
sential roles in membrane protein crystallization, neither the
hydrophobic nor the hydrophilic group can be ignored when
designing novel amphiphiles.
[8] T. C. Anglin, J. C. Conboy, Biophys. J. 2008, 95, 186–193.
[9] D. Marsh, Biochem. Biophys. Acta 2008, 1778, 1545–1575.
[10] A. Alexandrov, M. Mileni, E. Y. Chien, M. A. Hanson, R. C. Ste-
vens, Structure 2008, 16, 351–359.
[11] a) C. Tribet, R. Audebert, J.-L. Popot, Proc. Natl. Acad. Sci. USA
1996, 93, 15047–15050; b) J.-L. Popot, Annu. Rev. Biochem. 2010,
79, 737–775; c) J.-L. Popot, T. Althoff, D. Bagnard, J.-L. Banꢁres, P.
Bazzacco, E. Billon-Denis, L. J. Catoire, P. Champeil, D. Charvolin,
M. J. Cocco, G. Crꢂmel, T. Dahmane, L. M. de la Maza, C. Ebel, F.
Gabel, F. Giusti, Y. Gohon, E. Goormaghtigh, E. Guittet, J. H.
Kleinschmidt, W. Kꢃhlbrandt, C. Le Bon, K. L. Martinez, M. Picard,
B. Pucci, J. N. Sachs, C. Tribet, C. van Heijenoort, F. Wien, F. Zito,
M. Zoonens, Annu. Rev. Biophys. 2011, 40, 379–408; d) D. T.
McQuade, M. A. Quinn, S. M. Yu, A. S. Polans, M. P. Krebs, S. H.
Gellman, Angew. Chem. 2000, 112, 774–777; Angew. Chem. Int. Ed.
2000, 39, 758–761; e) P. S. Chae, M. J. Wander, A. P. Bowling, P. D.
Laible, S. H. Gellman, ChemBioChem 2008, 9, 1706–1709; f) P. S.
Chae, P. D. Laible, S. H. Gellman, Mol. BioSyst. 2010, 6, 89–94;
g) C. Breyton, E. Chabaud, Y. Chaudier, B. Pucci, J.-L. Popot, FEBS
Lett. 2004, 564, 312–318; h) M. Abla, G. Durand, B. Pucci, J. Org.
Chem. 2011, 76, 2084–2093; i) A. Nath, W. M. Atkins, S. G. Sligar,
Biochemistry 2007, 46, 2059–2069; j) S. C. Howell, R. Mittal, L.
Huang, B. Travis, R. M. Breyer, C. R. Sanders, Biochemistry 2010,
49, 9572–9583; k) J. Hovers, M. Potschies, A. Polidori, B. Pucci, S.
Raynal, F. Bonete, M. J. Serrano-Vega, C. G. Tate, D. Picot, Y.
Pierre, J. L. Popot, R. Nehme, M. Bidet, I. Mus-Veteau, H. Bus-
skamp, K. H. Jung, A. Marx, P. A. Timmins, W. Welte, Mol. Membr.
Biol. 2011, 28, 171–181; l) P. S. Chae, S. G. F. Rasmussen, R. R.
Rana, K. Gotfryd, A. C. Kruse, S. Nurva, U. Gether, L. Guan, C. J.
Loland, B. Byrne, B. K. Kobilka, S. H. Gellman, Chem. Eur. J. 2012,
18, 9485–9490; m) C. E. Schafmeister, L. J. W. Miercke, R. M.
Stroud, Science 1993, 262, 734–738; n) C.-L. McGregor, L. Chen,
N. C. Pomroy, P. Hwang, S. Go, A. Chakrabartty, G. G. Privꢂ, Nat.
Biotechnol. 2003, 21, 171–176; o) X. Zhao, Y. Nagai, P. J. Reeves, P.
Kiley, H. G. Khorana, S. Zhang, Proc. Natl. Acad. Sci. USA 2006,
103, 17707–17712; p) Q. Zhang, X. Ma, A. Ward, W.-X. Hong, V.-P.
Jaakola, R. C. Stevens, M. G. Finn, G. Chang, Angew. Chem. 2007,
119, 7153–7155; Angew. Chem. Int. Ed. 2007, 46, 7023–7025; q) P. S.
Chae, K. Gotfryd, J. Pacyna, L. J. W. Miercke, S. G. F. Rasmussen,
R. A. Robbins, R. R. Rana, C. J. Loland, B. Kobilka, R. Stroud, B.
Byrne, U. Gether, S. H. Gellman, J. Am. Chem. Soc. 2010, 132,
16750–16752; r) S. C. Lee, B. C. Bennett, W.-X. Hong, Y. Fu, K. A.
Baker, J. Marcoux, C. V. Robinson, A. B. Ward, J. R. Halpert, R. C.
Experimental Section
Details regarding the synthesis and characterization of new GNG amphi-
philes and detergent screening protocols for membrane protein solubili-
zation and stabilization are given in the Supporting Information.
Chem. Asian J. 2014, 9, 632 – 638
637
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemasianj.org
Pil Seok Chae et al.
Stevens, C. D. Stout, M. J. Yeager, Q. Zhang, Proc. Natl. Acad. Sci.
USA 2013, 110, E1203–1211; s) P. S. Chae, S. G. F. Rasmussen, R. R.
Rana, K. Gotfryd, R. Chandra, M. A. Goren, A. C. Kruse, S. Nurva,
C. J. Loland, Y. Pierre, D. Drew, J.-L. Popot, D. Picot, B. G. Fox, L.
Guan, U. Gether, B. Byrne, B. Kobilka, S. H. Gellman, Nat. Methods
2010, 7, 1003–1008; t) K. H. Cho, B. Byrne, P. S. Chae, ChemBio-
Chem 2013, 14, 452–455; u) P. S. Chae, R. R. Rana, K. Gotfryd,
S. G. F. Rasmussen, A. C. Kruse, K. H. Cho, S. Capaldi, E. Carlsson,
B. K. Kobilka, C. J. Loland, U. Gether, S. Banerjee, B. Byrne, J. K.
Lee, S. H. Gellman, Chem. Commun. 2013, 49, 2287–2289.
Granier, Nature 2012, 485, 321–326; g) S. Granier, A. Manglik,
A. C. Kruse, T. S. Kobilka, F. S. Thian, W. I. Weis, B. K. Kobilka,
Nature 2012, 485, 400–404; h) J. F. White, N. Noinaj, Y. Shibata, J.
Love, B. Kloss, F. Xu, J. Gvozdenovic-Jeremic, P. Shah, J. Shiloach,
C. G. Tate, R. Grisshammer, Nature 2012, 490, 508–513; i) S. E.
Rollauer, M. J. Tarry, J. E. Graham, M. Jaaskelainen, F. Jager, S.
Johnson, M. Krehenbrink, S. M. Liu, M. J. Lukey, J. Marcoux, M. A.
McDowell, F. Rodriguez, P. Roversi, P. J. Stansfeld, C. V. Robinson,
M. S. Sansom, T. Palmer, M. Hcgbom, B. C. Berks, S. M. Lea,
Nature 2012, 492, 210–214; j) J. Kellosalo, T. Kajander, K. Kogan,
K. Pokharel, A. Goldman, Science 2012, 337, 473–476.
[12] a) S. G. F. Rasmussen, H.-J. Choi, J. J. Fung, E. Pardon, P. Casarosa,
P. S. Chae, B. T. DeVree, D. M. Rosenbaum, F. S. Thian, T. S. Kobil-
ka, A. Schnapp, I. Konetzki, R. K. Sunahara, S. H. Gellman, A.
Pautsch, J. Steyaert, W. I. Weis, B. K. Kobilka, Nature 2011, 469,
175–180; b) D. M. Rosenbaum, C. Zhang, J. Lyons, R. Holl, D.
Aragao, D. H. Arlow, S. G. F. Rasmussen, H.-J. Choi, B. T. DeVree,
R. K. Sunahara, P. S. Chae, S. H. Gellman, R. O. Dror, D. E. Shaw,
W. I. Weis, M. Caffrey, P. Gmeiner, B. K. Kobilka, Nature 2011, 469,
236–240; c) S. G. F. Rasmussen, B. T. DeVree, Y. Zou, A. C. Kruse,
K. Y. Chung, T. S. Kobilka, F. S. Thian, P. S. Chae, E. Pardon, D. Cal-
inski, J. M. Mathiesen, S. T. A. Shah, J. A. Lyons, M. Caffrey, S. H.
Gellman, J. Steyaert, G. Skiniotis, W. I. Weis, R. K. Sunahara, B. K.
Kobilka, Nature 2011, 477, 549–555; d) A. C. Kruse, J. Hu, A. C.
Pan, D. H. Arlow, D. M. Rosenbaum, E. Rosemond, H. F. Green, T.
Liu, P. S. Chae, R. O. Dror, D. E. Shaw, W. I. Weis, J. Wess, B. K. Ko-
bilka, Nature 2012, 482, 552–556; e) K. Haga, A. C. Kruse, H.
Asada, T. Y. Kobayashi, M. Shiroishi, C. Zhang, W. I. Weis, T.
Okada, B. K. Kobilka, T. Haga, T. Kobayashi, Nature 2012, 482,
547–551; f) A. Manglik, A. C. Kruse, T. S. Kobilka, F. S. Thian, J. M.
Mathiesen, R. K. Sunahara, L. Pardo, W. I. Weis, B. K. Kobilka, S.
[13] a) G. G. Privꢂ, Methods 2007, 41, 388–397; b) Q. Zhang, H. Tao, W.-
X. Hong, Methods 2011, 55, 318–323.
[14] A. Chattopadhyay, E. London, Anal. Biochem. 1984, 139, 408–412.
[15] Y. Li, J. Reeve, Y. Wang, R. K. Thomas, J. Wang, H. Yan, J. Phys.
Chem. B 2005, 109, 16070–16074.
[16] J. N. Israelachvili, D. J. Mitchell, B. W. Ninham, J. Chem. Soc. Fara-
day Trans. 2 1976, 72, 1525–1568.
[17] B. Lorber, J. B. Bishop, L. Delucas, Biochim. Biophys. Acta Bio-
membr. 1990, 1023, 254–265.
[18] P. D. Laible, C. Kirmaier, C. S. Udawatte, S. J. Hofman, D. Holten,
D. K. Hanson, Biochemistry 2003, 42, 1718–1730.
[19] M. Caffrey, D. Li, A. Dukkipati, Biochemistry 2012, 51, 6266–6288.
[20] P. S. Chae, M. J. Wander, K. H. Cho, P. D. Laible, S. H. Gellman,
Mol. BioSyst. 2013, 9, 626–629.
[21] P. S. Chae, K. H. Cho, M. J. Wander, H. E. Bae, S. H. Gellman, P. D.
Labile, Biochim. Biophys. Acta Biomembr. 2014, 1838, 278–286.
Received: September 25, 2013
Published online: November 28, 2013
Chem. Asian J. 2014, 9, 632 – 638
638
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim