Expedient total synthesis of small to medium-sized membrane proteins via fmoc chemistry

Article abstract of DOI:10.1021/ja500222u

Total chemical synthesis provides a unique approach for the access to uncontaminated, monodisperse, and more importantly, post-translationally modified membrane proteins. In the present study we report a practical procedure for expedient and cost-effective synthesis of small to medium-sized membrane proteins in multimilligram scale through the use of automated Fmoc chemistry. The key finding of our study is that after the attachment of a removable arginine-tagged backbone modification group, the membrane protein segments behave almost the same as ordinary water-soluble peptides in terms of Fmoc solid-phase synthesis, ligation, purification, and mass spectrometry characterization. The efficiency and practicality of the new method is demonstrated by the successful preparation of Ser64-phosphorylated M2 proton channel from influenza A virus and the membrane-embedded domain of an inward rectifier K⁺ channel protein Kir5.1. Functional characterizations of these chemically synthesized membrane proteins indicate that they provide useful and otherwise-difficult-to-access materials for biochemistry and biophysics studies.

Full text of DOI:10.1021/ja500222u

Article
pubs.acs.org/JACS
Expedient Total Synthesis of Small to Medium-Sized Membrane
Proteins via Fmoc Chemistry
Ji-Shen Zheng,†,‡ Mu Yu,‡ Yun-Kun Qi,† Shan Tang,† Fei Shen,† Zhi-Peng Wang,† Liang Xiao,‡
Longhua Zhang,‡ Chang-Lin Tian,*,‡ and Lei Liu*,†
†Tsinghua-Peking Center for Life Sciences, Key Laboratory of Bioorganic Phosphorus Chemistry & Chemical Biology (Ministry of
Education), Department of Chemistry, Tsinghua University, Beijing 100084, China
‡High Magnetic Field Laboratory, Chinese Academy of Sciences and School of Life Sciences, University of Science and Technology of
China, Hefei 230026, China
S
* Supporting Information
ABSTRACT: Total chemical synthesis provides a unique approach for
the access to uncontaminated, monodisperse, and more importantly,
post-translationally modified membrane proteins. In the present study
we report a practical procedure for expedient and cost-effective
synthesis of small to medium-sized membrane proteins in multimilli-
gram scale through the use of automated Fmoc chemistry. The key
finding of our study is that after the attachment of a removable
arginine-tagged backbone modification group, the membrane protein
segments behave almost the same as ordinary water-soluble peptides in
terms of Fmoc solid-phase synthesis, ligation, purification, and mass
spectrometry characterization. The efficiency and practicality of the
new method is demonstrated by the successful preparation of Ser64-
phosphorylated M2 proton channel from influenza A virus and the membrane-embedded domain of an inward rectifier K+
channel protein Kir5.1. Functional characterizations of these chemically synthesized membrane proteins indicate that they
provide useful and otherwise-difficult-to-access materials for biochemistry and biophysics studies.
then connected using native chemical ligation.11 Extraordinary
difficulty in preparing and handling the transmembrane
segments was encountered, because such peptides are very
INTRODUCTION
■
Membrane proteins (MPs) comprise 20−30% of the
proteomes of most organisms.1 They perform many essential
poorly soluble in almost any solvent. Although strategies such
physiological functions such as signal transduction and
as attachment of a solubilizing tag20,21 or addition of various
molecule transportation.2,3 They are also the drug targets of
organic cosolvents or detergents12,13 have been tested, every
over 50% of modern pharmaceuticals.1,4 Therefore MPs have
new preparation of MPs requires tedious and case-specific
been intensively studied in structural, functional proteomics,
optimizations.22 Moreover, due to the difficulty of using Fmoc
and pharmaceutical research.5−8 A bottleneck in these studies is
SPPS to prepare either hydrophobic peptides23 or peptide
that recombinant production of MPs often suffers from low
thioesters,24−26 most of the previous MP syntheses relied on
yield, cell toxicity, and incomplete post-translational modifica-
tions (PTMs).9,10 Chemical protein synthesis11 offers an
the use of Boc SPPS, which unfortunately is not suitable for
making MPs carrying acid-sensitive PTMs (e.g., phosphor-
important alternative route to the production of MPs. This
ylation and glycosylation).17,27
method may be particularly suitable for the studies on small to
medium-sized MPs or MP domains.12,13 Because chemical
Here we present the first Fmoc and potentially general
method for total chemical synthesis of small to medium-sized
synthesis can produce proteins with diverse predesigned
changes with atomic precision,14 it provides a unique tool for
MPs at multimilligram scale (Figure 1a). A critical feature of
this method is that we develop a practical removable Arg4-
the structural and functional studies of MPs bearing PTMs such
as phosphorylation, lipidation or glycosylation.15 Moreover,
tagged backbone modification group for MP peptides.
Remarkably, with such backbone modification the MP peptides
behave as if they are ordinary water-soluble peptides during
purification, ligation, as well as for mass spectrum character-
chemical synthesis allows precise incorporation of specific labels
for advanced biophysical studies.16−18
Despite the promising value of chemical synthesis, only a
small number of MPs have been synthesized.12,13,15,16,18
A
izations. This new method, combined with the use of peptide
hydrazides in native chemical ligation,28,29 allows expedient and
representative example showing the state of the art is the total
synthesis of membrane-embedded diacylglycerol kinase DAGK
(121 residues).13 In the synthesis the segments of MPs are
prepared through solid-phase peptide synthesis (SPPS)19 and
Received: January 10, 2014
Published: February 21, 2014
© 2014 American Chemical Society
3695
dx.doi.org/10.1021/ja500222u | J. Am. Chem. Soc. 2014, 136, 3695−3704

Journal of the American Chemical Society
Article
Figure 1. Removable Arg-tagged backbone modification for chemical synthesis of membrane proteins. (a) General concept for the new Fmoc-based
method for total chemical synthesis of MPs. (b) The procedure for Fmoc-SPPS synthesis of peptides with removable Arg-tagged backbone
modification.
scalable synthesis, purification, and ligation of transmembrane
peptides to produce target MPs. Its practicality and usefulness
are demonstrated by the successful total synthesis and
functional characterizations of Ser64-phosphorylated M2
proton channel from influenza A virus and the membrane-
embedded domain of an inward rectifier K+ channel protein
Kir5.1.
To solve the dilemma we noticed a remarkable property of
Johnson’s N-(2-hydroxy-4-methoxy benzyl) (Hmb) group.32
When attached to peptides, this group can be readily cleaved
using TFA. However, if the 2-OH group is acylated, the masked
Hmb group is stable toward TFA.33 We hypothesized that such
a subtle change of activity may be used for removable backbone
modification (Figure 1a). The Boc-N-methyl-N-[2-(methyl-
amino)-ethyl]-carbamoyl group can be attached to the 2-OH
group to prevent the cleavage by TFA during Fmoc SPPS.34 In
the process of peptide ligations at neutral pH, the N-methyl-N-
[2-(methylamino)ethyl]carbamoyl group will be autocleaved
via intramolecular cyclization.35 Finally, with 2-OH group
unprotected, the ligated full-length peptide can be dissolved in
TFA to remove the tag and generate the final MP ready for in
vitro folding.
RESULTS AND DISCUSSION
■
Design of a Backbone Modification Group That Is
Removable by TFA. The difficulty revealed by the previous
strategies (e.g., C-terminal solubilizing tag20 or addition of
organic cosolvents or detergents12,13) led us to hypothesize that
a general approach to dissolve MP peptides may require the
breaking of their tendency to aggregate via formation of helices
or β-sheets. Indeed Kent et al. showed recently that backbone
N-methylation of an MP peptide can increase its solubility by
disrupting secondary structure formation.30 Unfortunately N-
methylation is irreversible, so that the current challenge is to
develop a practical approach for removable backbone
modification. Considering that a fully assembled MP would
be homogeneously soluble only in strong organic acids such as
trifluoroacetic acid (TFA),23,31 we hypothesized that TFA is the
best choice for backbone modification removal in the final step
of MP synthesis. This creates a demanding problem because
treatment with TFA is already conducted during Fmoc SPPS.
To test the hypothesis we synthesized the corresponding N-
tagged Gly residue (Scheme 1 in Supporting Information [SI]).
This amino acid can be readily used in automated Fmoc SPPS
to synthesize the MP segment (Figure 1b). Note that Gly is
one of the six most frequently occurring residues in
transmembrane sequences.34,36 Thus our strategy of backbone
modification at Gly residues should be general for MP
synthesis.
We next synthesized a model peptide Ac-AQFRGRBMSLA-
NH2 (1, Figure 2a and Scheme 2 in SI). GRBM denotes the Gly
with the removable backbone modification. An Arg4 tag was
3696
dx.doi.org/10.1021/ja500222u | J. Am. Chem. Soc. 2014, 136, 3695−3704

Journal of the American Chemical Society
Article
Figure 2. Two-step conversion of a peptide with removable Arg-tagged backbone modification to the native peptide. (a) Conversion of peptide 1 to
the native peptide. (b) Analytical RP-HPLC trace for the autocyclization of 1 in the ligation buffer (pH 7.2) at room temperature. (c) Analytical RP-
HPLC trace for the removal of the modification group in TFA cocktails at room temperature.
attached to the modification group during Fmoc SPPS (Figure
1b). This peptide was obtained after the treatment with
standard TFA cocktail (TFA/TIPS/H2O). ESI-MS analysis
confirmed that 2-OH was capped by the N-methyl-N-[2-
(methyl amino)ethyl]carbamoyl group (Figure 1 in SI).
We dissolved 1 in the neutral aqueous phosphate solution
commonly used for native chemical ligation (pH 7.2). A clean
conversion of 1 to 2 was observed on analytical RP-HPLC.
Peptide 2 possessed a free 2-OH group according to ESI-MS.
Over 95% of 1 was converted to 2 after 5 h at room
temperature, while the conversion was almost quantitative after
12 h. Moreover, we did not observe any decomposition of 2
after it stayed in the ligation buffer for one week (Figure 2b and
Figure 2 in SI). Finally, we dissolved 2 in the TFA/TIPS/H2O
cocktail. RP-HPLC analysis showed that 2 was quantitatively
converted to 3 in 3 h.
Performance with Transmembrane Peptides. We
chose the fourth transmembrane domain (23-mer) of signal
peptide peptidase (SPP, an intramembrane-cleaving protease)30
to examine the solubilizing effect of the removable backbone
modification. Seven peptides (SPP4-1 to SPP4-7) were
synthesized and labeled by fluorescein isothiocyanate (FITC)
to measure the solubility (Figure 3a). The native sequence
(SPP4-1) was poorly soluble in 50% aqueous CH3CN
containing 0.1% TFA (Figure 3). Addition of an N-terminal
Arg4 tag increased the solubility only by 2-fold (SPP4-2),
whereas the removable backbone modification group (without
any Arg tag yet) increased the solubility by 5-fold (SPP4-3).
When an Arg2−6 tag was added to the backbone modification
group, the solubility (SPP4-4, SPP4-5, and SPP4-6) was
further increased. Arg4 provided the best compromise between
the solubilizing effect and synthetic cost, leading to an overall
increase of solubility by 45-fold (SPP4-5). Finally, when the
removable backbone modification group was placed on a
different Gly residue (SPP4-7), we observed a similar
magnitude of solubilizing effect.
We next measured the solubility of the above peptides in
neutral ligation buffer (pH 7.4, 6 M guanidinium·HCl (Gn·
HCl), no detergent) after they were dissolved overnight. The N-
methyl-N-[2-(methyl amino)ethyl]carbamoyl group was re-
moved under these conditions, and the resulting peptides were
denoted as SPP4-1′ to SPP4-7′. The solubility of SPP4-1′ was
measured to be 0.05 mM, which is too low for the ligation
reaction (which usually requires 1 mM). N-terminal Arg4-
tagged SPP4-2′ showed improved solubility (0.37 mM),
whereas backbone-modified SPP4-3′ was soluble to 0.93
mM. Addition of an Argn tag to the backbone modification
group again proved critical (SPP4-4′ to SPP4-7′). With Arg4
the solubility of SPP4-5′ or SPP4-7′ was about 5 mM,
corresponding to an increase by ∼100 fold over SPP4-1′
(Figure 3a).
The above data showed that with the Arg4-tagged backbone
modification, the MP peptides behaved as if they are regular
water-soluble protein segments with regard to elution on RP-
HPLC, solubility under the ligation conditions, and ionization
in mass spectrometry (Figures 5−7 in SI). Backbone
modification was necessary for disrupting the secondary
structure formation as revealed by the circular dichroism
(CD) analysis (Figure 8 in SI). In the CD experiment, SPP4-2
showed a secondary structure similar to that of SPP4-1,
whereas SPP4-5 and SPP4-7 exhibited different spectra.
Furthermore, during the experiment we observed the
aggregation of SPP4-2 when its solution was allowed to
stand for 15−30 min. On the other hand, we did not observe
any aggregation of SPP4-5 and SPP4-7 after 24 h.
Finally, the ability of the backbone modification group to
disrupt secondary structure formation was also advantageous
for Fmoc SPPS. As shown in Figure 3c, in the crude SPPS
products SPP4-5 was identified as the major component
(isolated yield = 15%), whereas SPP4-2 was generated in a
much lower yield (isolated yield <5%).
3697
dx.doi.org/10.1021/ja500222u | J. Am. Chem. Soc. 2014, 136, 3695−3704

Journal of the American Chemical Society
Article
Figure 3. The solubility of FITC-labeled peptides SPP4-1 to SPP4-7. (a) The structures of SPP4 derivatives. (b) The photograph for solubility of
crude peptides SPP4-1, SPP4-2 and SPP4-5. (c) RP-HPLC elution profile of crude SPP4-2 and SPP4-5 analogues with a C4 column with a gradient
of 30−50% buffer B in buffer A over 40 min.
Total Synthesis of Wild-Type and Ser64-Phosphory-
lated M2. The M2 proton channel of influenza A virus is a 97-
residue homotetrameric integral membrane protein essential for
mediating transport of protons across the viral envelope. It has
been the subject of numerous structural and functional
studies,37−39 but some details regarding the role of M2 in
virus replication still remain unclear. For instance, M2 carries a
number of modifications including palmitoylation, fatty
acylation, and phosphorylation,40 whose roles in the function,
membrane localization, or assembly of M2 remain to be
elucidated.41 In this context chemical synthesis of modified M2
provides useful research reagents.
Kochendoerfer et al. previously used the Boc method to
prepare M2.42 Here we used the new strategy to synthesize
both wild-type and Ser64-phosphorylated M2 through Fmoc
chemistry (Figure 4a). The 2-Cl-Trt-NHNH2 resin was used to
prepare M2[1−49, 4Arg-Tag]-NHNH2 (5) by using automated
Fmoc SPPS. In this segment Gly34 was modified with the Arg4-
tagged removable modification group. We then used the Nova-
PEG Wang resin to prepare M2[50−97] (6) also by automated
Fmoc SPPS, in which Fmoc-Ser(HPO3Bzl)−OH was used to
install pSer64. Both M2[1−49, 4Arg-Tag]-NHNH2 and
M2[50-97] (bearing either Ser64 (6) or pSer64 (p6)) were
produced as readily soluble peptides (isolated yields = 12%,
20%, 15%) amenable to routine RP-HPLC purification and
ESI-MS characterizations (Figure 9 in SI).
The native chemical ligation of M2[1−49, 4Arg-Tag]-
NHNH2 5 and M2[50−97] 6 was conducted using the in
situ NaNO2 activation/thiolysis protocol.28,29 The reaction
proceeded smoothly in aqueous Gn·HCl (6 M) without any
organic cosolvent or detergent to yield the full-length polypeptide
in 39% (for Ser64, 7) or 36% (for pSer64, p7) yield (Figure
4a). 7 or p7 was purified through RP-HPLC and treated with
TFA/TIPS/H2O to remove the backbone modification group
in 5 h. The final wild-type (8) or Ser64-phosphorylated M2
(p8) was obtained through a further RP-HPLC separation in
26% (for Ser64) and 17% (for pSer64) yield. Characterizations
of 7 and 8 by ESI-MS and SDS-PAGE are shown in Figure 4b,c.
Both wild-type and Ser64-phosphorylated M2’s were incorpo-
rated into dodecylphosphocholine (DPC) vesicles. The CD
spectra of the M2-containing vesicles exhibited evidence of
significant α-helical structures with characteristic negative bands
at 208 and 222 nm (Figure 4d), suggesting the successful
generation of correctly folded M2.
Chemically synthesized full-length M2 proteins were then
incorporated into POPC/POPG (POPC: 1-palmitoyl-2-oleoyl-
sn-glycero-3-phosphocholine; POPG: 1-palmitoyl-2-oleoyl-sn-
glycero-3-phosphor-(1′-rac-glycerol)) bilayers in the presence
3698
dx.doi.org/10.1021/ja500222u | J. Am. Chem. Soc. 2014, 136, 3695−3704

Journal of the American Chemical Society
Article
Figure 4. Chemical synthesis and characterizations of wild-type and Ser64-phosphorylated M2. (a) Left: Strategy for chemical synthesis. Right:
Analytical data. (b) Analytical data for ligation product Arg-tagged M2-pS64 p7 and the final M2-pS64 p8. Up: RP-HPLC of the purified ligation
product Arg-tagged M2-pS64 p7 with a C4 column (4.6 mm × 150 mm) at 56 °C. According to the previous work, the HPLC analysis should be
operated at elevated temperatures to facilitate the elution of a membrane peptide from the HPLC column.13,30 Inset shows MALDI-TOF-MS of p7.
Down: RP-HPLC of the purified native M2-pS64 p8 with a C4 column (4.6 mm × 150 mm) at 56 °C. Inset shows MALDI-TOF-MS of p8. (c)
SDS-PAGE/Ag staining analysis: lane left, the ligation product Arg-tagged M2-pS64 p7; lane right, M2-pS64 p8. (d) CD analysis of reconstituted
M2-pS64 p8 (5 μM, in a 1 mm quartz cell) in DPC micelles. (e) Single-channel currents of chemical synthetic native and Ser64-phosphorylated M2
protein channels after reconstitution in POPE/POPG (3:1) lipid vesicles. The traces were recorded at various membrane potentials in 5 mM Tris-
MOPS, 150 mM NaCl, pH 5.0 both inside and outside. Solid line: opening (O); dotted line: closure (C).
of symmetrical solutions under acidic conditions (5 mM
tris(hydroxymethyl)aminomethane (Tris) and 3-morpholino-
propanesulfonic acid (MOPS), pH 5.0). Single-channel
currents were measured in response to different applied
voltages (−60, −40, −20, 0, 20, 40, and 60 mV) (Figure 4e).
Represented traces in Figure 4e show that the channel opens at
positive voltages as well as negative ones. Nonetheless,
relatively low signal-to-noise ratio was observed in the M2
channel conductance traces, similar to previous reports of
reconstituted M2 channels in lipid bilayers, probably due to
3699
dx.doi.org/10.1021/ja500222u | J. Am. Chem. Soc. 2014, 136, 3695−3704

Journal of the American Chemical Society
Article
Figure 5. Synthesis of inward rectifier K+ channel Kir5.1[64−179]. (a) Kir5.1[64−179] monomer (116 amino acids) is divided into four segments:
Kir5.1[64−89]-NHNH2 9, Kir5.1[90−116]-NHNH2 10, Kir5.1[117−137]-NHNH2 11, and Kir5.1[138−179]−OH 12. These four segments were
synthesized by Fmoc SPPS and ligated as described in the text. (b) RP-HPLC of the purified product Kir5.1[64−179] 15 with a C4 column (4.6 mm
× 150 mm) at 56 °C. Inset shows MALDI-TOF-MS of 15. (c) ESI-MS of 15. The observed baseline shift always occurs (also with blank injections).
(d) SDS-PAGE/Ag staining analysis: lane i, the final product Kir5.1[64−179] 16 by TFA cocktails; lane ii, reconstituted tetrameric Kir5.1[64−179]
in POPC/POPG liposomes.
disturbance of lipid bilayer integrity under acidic conditions or
in nonuniform oligomeric states.43 Then, the channel activity at
relatively low potentials (−20, 0, and 20 mV) was not obvious.
Meanwhile, control experiments without addition of M2 did
not produce any channel activity under the same conditions. To
determine the effect of amantadine (an anti-influenza drug
through blocking the M2 channel) on chemically synthesized
M2, the channel activity was recorded in the absence of
amantadine at first. Then aliquots of amantadine solution were
added to both chambers to a final concentration of 20 μM.
After stirring for about 10 min, the channel activity was
measured again. As shown in Figure 4e, the almost total
channel blockage in the presence of 20 μM amantadine was
observed at a holding potential of +60 mV.
phosphorylation caused little influence on both the channel
activity and the amantadine inhibition effect.
Total Synthesis of Core Transmembrane Domain of
Kir5.1. The inward rectifier family of potassium (Kir) channels
plays pivotal roles in controlling the excitability of various
cells.44−46 Kir5.1, a typical Kir channel that is widely expressed
in brain and kidney, has been implicated in many physiological
and pathological processes such as the modulation of brainstem
neurons and the SeSAME/EAST syndrome.47 It has been long
believed that Kir5.1 formed no functional channel when
expressed alone. Yet a functional homomeric Kir5.1 was
reported in a rare case, when it was expressed in HEK293T
cells with the help of PSD-95, a member of the membrane-
associated guanylate kinase (MAGUK) family which can help
ion channels to locate on plasma membrane.48 This largely
overlooked finding raised a hypothesis that the failure to detect
functional homomeric Kir5.1 may be due to the lack of ability
to locate on the plasma membrane.
Collectively our results demonstrated that chemically
synthesized M2, when incorporated into planar lipid bilayers,
can produce single-channel activity after spontaneous fusion at
low pH. Comparing the current results of wild-type and Ser64-
phosphorylated M2 (Figure 4e), it seemed likely that Ser64-
To investigate whether pure Kir5.1 can form a functional
channel, we synthesized its core transmembrane domain (64−
3700
dx.doi.org/10.1021/ja500222u | J. Am. Chem. Soc. 2014, 136, 3695−3704

Journal of the American Chemical Society
Article
Figure 6. Biophysical characterization of functional Kir5.1[64−179] channels made by total chemical synthesis. (a) CD analysis of functional
Kir5.1[64−179] channels (∼10 μM, in a 1 mm quartz cell). (b) Representative channel activity of the Kir5.1[64−179] channels after reconstitution
in POPE/POPG (3:1) lipid vesciles. The traces were recorded at −100 mV in 5 mM HEPES/150 mM KCl (pH 8.0) both inside and outside. A
zoom-in of the gating behavior is shown beneath. C: level of channel closure; O1: level 1 of channel opening; O2: level 2 of channel opening. (c)
Single-channel conductance of Kir5.1[64−179] in planar lipid bilayers. Left: Single-channel currents were recorded at various membrane potentials.
Right: Single-channel conductance was calculated from the current−voltage plot, with the slope conductance of 45 pS. The traces were recorded at
−100 mV in 5 mM HEPES/150 mM KCl (pH 8.0) both inside and outside.
179, 116 amino acids) containing two transmembrane helices
and a membrane-embedded P-loop. The target protein was
divided into four segments, Kir5.1[64−89]-NHNH2,
Kir5.1[90−116]-NHNH2, Kir5.1[117−137, 4Arg-Tag]-
NHNH2, and Kir5.1[138−179] (Scheme 4 in SI). Three
segments (i.e., Kir5.1[64−89]-NHNH2, Kir5.1[90−116]-
NHNH2, and Kir5.1[138−179]) could be directly prepared in
15%, 20%, and 12% isolated yields. Nonetheless, the P-loop
segment Kir5.1[117−137]-NHNH2 could not be prepared by
Fmoc SPPS as the crude peptide was a complex mixture of
poorly soluble materials. Only after the Arg4-tagged removable
backbone modification group was added at Gly133, could this
segment be obtained through automated Fmoc SPPS in 18%
isolated yield. The purified peptide was well soluble (Figure 11
in SI).
The four segments were condensed by N-to-C sequential
ligation in aqueous Gn·HCl solutions (isolated yields = 57%,
29%, 45%), producing Kir5.1[64−179] with the Arg4-tagged
backbone modification group in 7.5% overall isolated yield
(Figures 12−14 in SI). This molecule was well soluble in
aqueous CH3CN and could be readily isolated by RP-HPLC to
>95% purity (Figure 5b,c). The purified, backbone-modified
Kir5.1[64−179] powder was then dissolved into TFA/TIS/
H2O for 5 h at room temperature. The solution was
concentrated by N2 blowing, and the final target protein was
obtained through precipitation from cold diethyl ether. The
final Kir5.1[64−179] could not be eluted from RP-HPLC.
Nonetheless, SDS-PAGE analysis indicated that the final
product was the desired protein (Figure 5d).
Synthetic Kir5.1[64−179] was used in the channel
reconstitution in unilamellar liposome vesicles with POPC/
POPG lipids. Kir5.1[64−179] successfully formed a homoge-
neous tetramer according to SDS-PAGE (Figure 5d). The CD
spectrum of the Kir5.1[64−179]-containing lipids (Figure 6a)
showed two minima at 208 and 222 nm, indicating the
formation of a well-structured α-helical protein. Channel
conductance in the planar lipid bilayer was examined using
the proteoliposomes (Kir5.1[64−179] in POPC/POPG), and
the single-channel activity was measured at various voltages.
Shown in Figure 6b is a 4-s train of recorded current amplitudes
with −100 mV voltage applied on the cis side. Uniform current
amplitudes of 20 pA or 0 pA were observed with the alternating
channel opening or closing. Fast switching between channel
opening and closing was also demonstrated in the zoom-in in
Figure 6c.
3701
dx.doi.org/10.1021/ja500222u | J. Am. Chem. Soc. 2014, 136, 3695−3704

Journal of the American Chemical Society
Article
capped with Ac2O/DIEA/DMF (1/1/8, v/v/v). After deprotection of
Fmoc, the peptide was treated with a double or triple coupling step for
the next amino acid and an Ac2O capping step. After completion of the
peptides assembly, Pd(PPh3)4/PhSiH3 in CH2Cl2/DMF (1/1, v/v)
was added to remove the Alloc group. The deprotection step was
repeated once. The Arg-Tag was then coupled by using standard
Fmoc-SPPS protocols. The peptide was cleaved from the resin with
TFA/PhOH/TIPS/H2O cocktails.
Chemical Ligation Using Peptide Hydrazides As Thioester
Surrogates. Peptide hydrazides were dissolved using 0.2 M
phosphate solution containing 6 M Gn·HCl (pH 3.0−3.1) in a
Eppendorf reaction tube. The reaction tube was placed in a −15 °C
ice−salt bath and the solution was gently agitated by magnetic stirring.
0.5 M NaNO2 (10−15 equiv) was added, and the solution was gently
agitated for 15 min at −15 °C to oxidize the peptide hydrazide to the
azide. Subsequently, the phosphate solution containing MPAA (4-
mercaptophenylacetic acid, ∼100 equiv) and N-terminal Cys-peptide
(0.7−1.6 equiv) was added to the reaction mixture to convert the
peptide azide to the thioester for native chemical ligation. The tube
was removed from the ice−salt bath and warmed to room
temperature. The pH of the ligation reaction at the pH to 6.5−7.4
was monitored using a micro pH probe. The ligation was monitored
by analytical HPLC and ESI-MS. Note that the autocyclization
procedure was performed for 3−5 h under pH 7.4 at ∼25−30 °C to
completely liberate the phenol group. After completion of ligation, 0.1
M neutral TCEP solution (∼100 equiv) was added for 20 min to
reduce the ligation system.
To measure the channel conductance, current amplitudes of
Kir5.1[64−179] in the planar lipid bilayer were recorded in
duration of 1 s at various voltages (−60, −40, −20, 0, 20, 40,
and 60 mV) (Figure 6c). More than one channel was identified
in the preparation of Kir5.1[64−179] in the planar lipid bilayer
of POPC/POPG. Therefore, the measured current amplitudes
exhibited stepwise addition of multiple channels during their
openings. Different steps of current amplitudes were estimated
from the stepwise sets of data, referring to the predefined
baseline. The lowest amplitude level was extracted as the single-
channel current amplitude value. For example, two sets of
current amplitudes (1.1 and 2.2 pA) were recorded at a voltage
of 20 mV, demonstrating a 1.1 pA single-channel current
amplitude at 20 mV. The extracted single-channel current
amplitudes were plotted against different voltages. The
current−voltage was in linear relationship. Slope of the
regressed line was 45 pS, representing the single-channel
conductance of Kir5.1[64−179] in the planar lipid bilayer of
POPC/POPG.
Taken together, the successful channel conductance
measurement of synthetic Kir5.1[64−179] indicated that
Kir5.1 can form functional homomeric K+ channels. This
finding supported the previous hypothesis that trafficking and
membrane insertion are important limiting steps for channel
conductance function, especially using traditionally in situ
patch-clamp techniques. Our work also indicated that chemical
synthesis and reconstitutive refolding provide an effective way
to analyze the functions of channel proteins, especially those
that cannot be obtained by traditional biological methods or
those that exhibit no function in situ.
Removal of the Backbone Modification by TFA Cocktails.
Arg-tagged protein was dissolved in TFA cocktails (TFA/TIPS/H2O,
95/2.5/2.5, v/v/v). The native protein can be obtained quantitatively
for 5 h with TFA. After the cleavage, the reaction mixture was
concentrated and precipitated with Et2O to yield the final product.
Reconstitution in Unilamellar Liposome Vesicles. All lipids
used were obtained from Avanti Polar Lipids (Alabaster, AL, USA).
Chemically synthesized protein was centrifuged at 14,000 rpm for 10
min to recover the sample that sticks to the tube wall. Into the tube
was added 200 μL reconstitution buffer (0.1% SDS, 20 mM Tris, 200
mM NaCl, pH 8.0). The sample was completely dissolved and
uniformly dispersed using pipettes. POPC (1.5 mg/mL) and POPG
(0.5 mg/mL) were mixed and well dispersed in reconstitution buffer
(0.1% SDS, 20 mM Tris, 200 mM NaCl, pH 8.0) through five cycles of
liquid nitrogen freezing and thawing at rt. The protein solution was
added into the liposome solution to a final molar ratio of 1/600
(protein/lipid). The mixture was rotated at 4 °C for 1 h, and dialyzed
for 3 days at 25 °C in Tris buffer (20 mM Tris, 200 mM NaCl, pH
8.0) to completely remove the detergent. To prepare unilamellar
liposome vesicles, samples were extruded using a 400 nm
polycarbonate membrane by the Avanti Mini-Extruder (Alabaster,
AL). Final concentration of protein sample was ∼40 μg/mL.
CONCLUSION
■
The growing research interest in membrane proteins demands
a toolbox of approaches for their production. Total chemical
synthesis provides a unique technique to the arsenal for the
generation of pure, monodisperse, and modified MP samples
required for structural and biochemical studies. We showed
that, with the removable backbone modification group, the MP
peptides behaved almost the same as the peptide segments of
water-soluble proteins in terms of Fmoc SPPS, solubility,
HPLC elution, and mass characterization. With this method
and the technology of using peptide hydrazides as thioester
surrogates, small- to middle-sized MPs comprising more than
one transmembrane domain can be prepared readily through
the use of automated Fmoc chemistry. The efficiency and
practicality of the new method was demonstrated by the
multimilligram-scale synthesis of the M2 proton channel from
influenza A virus and the core transmembrane domain of
Kir5.1. These synthetic proteins were free of contamination by
other channel-forming proteins derived from cell membranes in
recombinant expressions. As a result they provided unique
materials for in vitro study of MP’s biophysical and biochemical
properties.
Single-Channel Conductance Measurement in Planar Lipid
Bilayer. Channel conductance measurements in planar lipid bilayer
were conducted using Ionovation Compact (Osnabruck, Germany).
̈
Two polycarbonate compartments in volume of 1.2 mL were separated
by a TEFLON foil with 25 μm thickness and 50−100 μm aperture
diameter. The premixed liposome (POPC/POPG = 3/1) was
employed to paint the aperture. Planar lipid bilayer formation was
monitored optically or by capacitance measurements. After successful
formation of a stable bilayer in the aperture, proteoliposomes (protein
sample in POPC/POPG) were added to the cis chamber next to the
bilayer. Fusion of protein and planar lipid bilayer was detected through
observation of channel conductance. Working solution (5 mM 4-(2-
hydroxyethyl)-1-piperazine-ethanesulfonic acid (HEPES), 150 mM
KCl, pH 8.0) was present in both cis- and trans chambers.
EXPERIMENTAL SECTION
■
Detailed experimental procedures for the synthesis of compounds,
peptides and proteins, their characterization data, and the electro-
physiology experiments are given in the Supporting Information.
Synthesis of Gly RMB-Containing Peptides. The peptide-chain
assembly was performed using standard Fmoc-SPPS protocols except
for GlyRMB0 and the amino acids immediately after GlyRMB0. GlyRMB0
residue was coupled using Fmoc-GlyRMB0−OH (2 equiv), HATU (2
equiv), HOAt (2 equiv) and DIEA (4 equiv) for 1.5 h, and then
Different DC voltage was used to measure channel conductance
using an EPC-10 amplifier (HEKA Elektronik). Currents were
measured with a 2 kHz low-pass filter at 10 kHz sampling rate. Data
was analyzed using the pCLAMP 10.0 software (Axon Instruments).
In each measurement at various voltages, current amplitudes were
recorded referring to a predefined baseline, based on 50% threshold
crossing methods. Current−voltage (I−V) relationship was plotted
3702
dx.doi.org/10.1021/ja500222u | J. Am. Chem. Soc. 2014, 136, 3695−3704

Journal of the American Chemical Society
Article
with the measured current amplitudes against different applied
voltages. The I−V plot was regressed to a linear line with the slope
representing channel conductance of the membrane protein in the
planar lipid bilayer.
(15) (a) Tolbert, T. J.; Wong, C. H. J. Am. Chem. Soc. 2000, 122,
5421. (b) Huse, M.; Holford, M. N.; Kuriyan, J.; Muir, T. W. J. Am.
Chem. Soc. 2000, 122, 8337. (c) Asahina, Y.; Kamitori, S.; Takao, T.;
Nishi, N.; Hojo, H. Angew. Chem., Int. Ed. 2013, 52, 9733−9737.
(16) Kochendoerfer, G. G.; Jones, D. H.; Lee, S.; Oblatt-Montal, M.;
Opella, S. J.; Montal, M. J. Am. Chem. Soc. 2004, 126, 2439−2446.
(17) Muralidharan, V.; Muir, T. W. Nat. Methods 2006, 3, 429−438.
(18) Valiyaveetil, F. I.; Leonetti, M.; Muir, T. W.; Mackinnon, R.
Science 2006, 314, 1004−1007.
(19) Kent, S. B. H. Annu. Rev. Biochem. 1988, 57, 957−989.
(20) (a) Sato, T.; Saito, Y.; Aimoto, S. J. Pept. Sci. 2005, 11, 410−416.
(b) Johnson, E. C.; Kent, S. B. H. Tetrahedron Lett. 2007, 48, 1795−
1799. (c) Johnson, E. C.; Malito, E.; Shen, Y.; Rich, D.; Tang, W. J.;
Kent, S. B. H. J. Am. Chem. Soc. 2007, 129, 11480−11490.
(21) (a) Tan, Z.; Shang, S.; Danishefsky, S. J. Proc. Natl. Acad. Sci.
U.S.A. 2011, 108, 4297−4302. (b) Huang, Y. C.; Li, Y. M.; Chen, Y.;
Pan, M.; Li, Y. T.; Yu, L.; Guo, Q. X.; Liu, L. Angew. Chem., Int. Ed.
2013, 52, 4858−4862.
(22) Olschewski, D.; Becker, C. F. Mol. Biosyst. 2008, 4, 733−740.
(23) Coin, I.; Beyermann, M.; Bienert, M. Nat. Protoc. 2007, 2,
3247−3256.
(24) Pattabiraman, V. R.; Bode, J. W. Nature 2011, 480, 471−479.
(25) Zheng, J. S.; Chang, H. N.; Wang, F. L.; Liu, L. J. Am. Chem. Soc.
2011, 133, 11080−11083.
(26) Zheng, J. S.; Tang, S.; Huang, Y. C.; Liu, L. Acc. Chem. Res.
2013, 46, 2475−2484.
ASSOCIATED CONTENT
■
S
* Supporting Information
Experimental details and compound characterizations. This
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Authors
lliu@mail.tsinghua.edu.cn
cltian@ustc.edu.cn
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This study was supported by the National Basic Research
Program of China (973 program, No. 2013CB932800),
Chinese Ministry of Science and Technology with the Protein
Science Key Projects (2011CB911104), NSFC Grants
(91313301, 31100847, and 21225207), the Specialized
Research Fund for the Doctoral Program of Higher Education
(Grant No. 20120002130004), and China Postdoctoral Science
Foundation (Nos. 2012M520009, 2013T60098).
(27) Shin, Y.; Winans, K. A.; Backes, B. J.; Kent, S. B. H.; Ellman, J.
A.; Bertozzi, C. R. J. Am. Chem. Soc. 1999, 121, 11684−11689.
(28) Fang, G. M.; Li, Y. M.; Shen, F.; Huang, Y. C.; Li, J. B.; Lin, Y.;
Cui, H. K.; Liu, L. Angew. Chem., Int. Ed. 2011, 50, 7645−7649.
(29) Zheng, J. S.; Tang, S.; Qi, Y. K.; Wang, Z. P.; Liu, L. Nat. Protoc.
2013, 8, 2483−2495.
REFERENCES
■
(30) Johnson, E. C.; Kent, S. B. H. J. Am. Chem. Soc. 2006, 128,
7140−7141.
(1) Bill, R. M.; Henderson, P. J.; Iwata, S.; Kunji, E. R.; Michel, H.;
Neutze, R.; Newstead, S.; Poolman, B.; Tate, C. G.; Vogel, H. Nat.
Biotechnol. 2011, 29, 335−340.
(31) Speers, A. E.; Wu, C. C. Chem. Rev. 2007, 107, 3687−3714.
(32) Johnson, T.; Quibell, M.; Owen, D.; Sheppard, R. C. J. Chem.
Soc., Chem. Commun. 1993, 369−372.
(2) Gouaux, E.; Mackinnon, R. Science 2005, 310, 1461−1465.
(3) Venkatakrishnan, A. J.; Deupi, X.; Lebon, G.; Tate, C. G.;
Schertler, G. F.; Babu, M. M. Nature 2013, 494, 185−194.
(33) Quibell, M.; Turnell, W. G.; Johnson, T. J. Org. Chem. 1994, 59,
1745−1750.
́
(4) Yildirim, M. A.; Goh, K. I.; Cusick, M. E.; Barabasi, A. L.; Vidal,
(34) (a) Wahlstrom, K.; Planstedt, O.; Unden
́
, A. Tetrahedron Lett.
, A.
M. Nat. Biotechnol. 2007, 25, 1119−1126.
(5) White, S. H. Nature 2009, 459, 344−346.
2008, 49, 3779−3781. (b) Wahlstrom, K.; Planstedt, O.; Unden
́
Tetrahedron Lett. 2008, 49, 3921−3924.
(6) Rosenbaum, D. M.; Rasmussen, S. G.; Kobilka, B. K. Nature
(35) Saari, W. S.; Schwering, J. E.; Lyle, P. A.; Smith, S. J.;
Engelhardt, E. L. J. Med. Chem. 1990, 33, 97−101.
2009, 459, 356−363.
(7) Lappano, R.; Maggiolini, M. Nat. Rev. Drug Discovery 2011, 10,
47−60.
(36) Eilers, M.; Shekar, S. C.; Shieh, T.; Smith, S. O.; Fleming, P. J.
Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 5796−5801.
(8) Neuzi, P.; Giselbrecht, S.; Lange, K.; Huang, T. J.; Manz, A. Nat.
̌
̈
(37) Schnell, J. R.; Chou, J. J. Nature 2008, 451, 591−596.
(38) Stouffer, A. L.; Acharya, R.; Salom, D.; Levine, A. S.; Di
Costanzo, L.; Soto, C. S.; Tereshko, V.; Nanda, V.; Stayrook, S.;
DeGrado, W. F. Nature 2008, 451, 596−599.
(39) Rossman, J. S.; Jing, X.; Leser, G. P.; Lamb, R. A. Cell 2010, 142,
902−913.
(40) Holsinger, L. J.; Shaughnessy, M. A.; Micko, A.; Pinto, L. H.;
Lamb, R. A. J. Virol. 1995, 69, 1219−1225.
(41) Grantham, M. L.; Wu, W. H.; Lalime, E. N.; Lorenzo, M. E.;
Klein, S. L.; Pekosz, A. J. Virol. 2009, 83, 8655−8661.
(42) Kochendoerfer, G. G.; Salom, D.; Lear, J. D.; Wilk-Orescan, R.;
Kent, S. B. H.; DeGrado, W. F. Biochemistry 1999, 38, 11905−11913.
(43) (a) Duff, K. C.; Ashley, R. H. Virology 1992, 190, 485−489.
(b) Tosteson, M. T.; Pinto, L. H.; Holsinger, L. J.; Lamb, R. A. J.
Membr. Biol. 1994, 142, 117−126.
(44) Hibino, H.; Inanobe, A.; Furutani, K.; Murakami, S.; Findlay, I.;
Kurachi, Y. Physiol. Rev. 2010, 90, 291−366.
(45) Bavro, V. N.; De Zorzi, R.; Schmidt, M. R.; Muniz, J. R.;
Rev. Drug Discovery 2012, 11, 620−632.
(9) Schwarz, D.; Junge, F.; Durst, F.; Frolich, N.; Schneider, B.;
Reckel, S.; Sobhanifar, S.; Dotsch, V.; Bernhard, F. Nat. Protoc. 2007,
̈
̈
2, 2945−2957.
(10) Junge, F.; Schneider, B.; Reckel, S.; Schwarz, D.; Dotsch, V.;
̈
Bernhard, F. Cell. Mol. Life Sci. 2008, 65, 1729−1755.
(11) (a) Dawson, P. E.; Muir, T. W.; Clark-Lewis, I.; Kent, S. B. H.
Science 1994, 266, 776−779. (b) Nilsson, B. L.; Soellner, M. B.; Raines,
R. T. Annu. Rev. Biophys. Biomol. Struct. 2005, 34, 91−118. (c) Haase,
C.; Seitz, O. Angew. Chem., Int. Ed. 2008, 47, 1553−1556.
(12) (a) Bianchi, E.; Ingenito, R.; Simon, R. J.; Pessi, A. J. Am. Chem.
Soc. 1999, 121, 7698−7699. (b) Clayton, D.; Shapovalov, G.; Maurer,
J. A.; Dougherty, D. A.; Lester, H. A.; Kochendoerfer, G. G. Proc. Natl.
Acad. Sci. U.S.A. 2004, 101, 4764−4769.
(13) Lahiri, S.; Brehs, M.; Olschewski, D.; Becker, C. F. Angew.
Chem., Int. Ed. 2011, 50, 3988−3992.
(14) (a) Kent, S. B. H. Chem. Soc. Rev. 2009, 38, 338−351.
(b) Devaraneni, P. K.; Komarov, A. G.; Costantino, C. A.; Devereaux,
J. J.; Matulef, K.; Valiyaveetil, F. I. Proc. Natl. Acad. Sci. U.S.A. 2013,
110, 15698−15703. (c) Matulef, K.; Komarov, A. G.; Costantino, C.
A.; Valiyaveetil, F. I. Proc. Natl. Acad. Sci. U.S.A. 2013, 110, 17886−
17891.
Zubcevic, L.; Sansom, M. S.; Ven
Struct. Mol. Biol. 2012, 19, 158−163.
(46) Tao, X.; Avalos, J. L.; Chen, J.; MacKinnon, R. Science 2009,
326, 1668−1674.
́
ien-Bryan, C.; Tucker, S. J. Nat.
3703
dx.doi.org/10.1021/ja500222u | J. Am. Chem. Soc. 2014, 136, 3695−3704

Journal of the American Chemical Society
Article
(47) Paulais, M.; Bloch-Faure, M.; Picard, N.; Jacques, T.;
Ramakrishnan, S. K.; Keck, M.; Sohet, F.; Eladari, D.; Houillier, P.;
Lourdel, S.; Teulon, J.; Tucker, S. J. Proc. Natl. Acad. Sci. U.S.A. 2011,
108, 10361−10366.
(48) Tanemoto, M.; Fujita, A.; Higashi, K.; Kurachi, Y. Neuron 2002,
34, 387−97.
3704
dx.doi.org/10.1021/ja500222u | J. Am. Chem. Soc. 2014, 136, 3695−3704