Biochem. J. (2012) 442, 631–638 (Printed in Great Britain)
631
doi:10.1042/BJ20111284
Functional analysis of membranous Fo -a subunit of F1 Fo -ATP synthase
by in vitro protein synthesis
Yutetsu KURUMA*1 , Toshiharu SUZUKI†1 , Sakurako ONO†, Masasuke YOSHIDA†‡ and Takuya UEDA*2
*Department of Medical Genome Sciences, Graduate School of Frontier Sciences, The University of Tokyo, Kashiwanoha 5-1-5, Kashiwa-shi, Chiba 277-0882, Japan, †ATP Synthesis
Regulation Project, ICORP (International Co-operative Research Project), JST (Japan Science and Technology Corporation), Aomi 2-41, Tokyo 135-0064, Japan, and ‡Faculty of
Engineering, Kyoto Sangyo University, Kamigamo Motoyama, Kyoto 603-8555, Japan
The a subunit of F1 Fo (F1 Fo -ATP synthase) is a highly hydrophobic
protein with five putative transmembrane helices which plays
a central role in H +-translocation coupled with ATP synthesis/hydrolysis. In the present paper, we show that the a subunit
produced by the in vitro protease-free protein synthesis system
(the PURE system) is integrated into a preformed Fo a-less F1 Fo
complex in Escherichia coli membrane vesicles and liposomes.
The resulting F1 Fo has a H +-coupled ATP synthesis/hydrolysis
activity that is approximately half that of the native F1 Fo . By
using this procedure, we analysed five mutations of F1 Fo , where
the conserved residues in the a subunit (Asn90 , Asp112 , Arg169 ,
Asn173 and Gln217 ) were individually replaced with alanine. All of
the mutant Fo a subunits were successfully incorporated into F1 Fo ,
showing the advantage over conventional expression in E. coli by
which three (N90A, D112A, and Q217A) mutant a subunits were
not found in F1 Fo . The N173A mutant retained full activity and the
mutants D112A and Q217A had weak, but detectable, activity. No
activity was observed for the R169A and N90A mutants. Asn90 is
located in the middle of putative second transmembrane helix and
likely to play an important role in H +-translocation. The present
study exemplifies that the PURE system provides an alternative
approach when in vivo expression of membranous components in
protein complexes turns out to be difficult.
INTRODUCTION
Recent progress in cell-free protein synthesis methods enables
us to quickly prepare the target membrane proteins in a
combination with liposomes. Several active membrane proteins,
such as bacteriorhodopsin [27] and GP91phox which is the catalytic
core protein of the NADH oxidase complex [28], have been
prepared by cell-free systems previously [29–38]. The method
(hereafter called in vitro expression) is simple when compared
with the recombinant cell methods (hereafter called in vivo
expression). In vitro expression normally requires only addition
of template DNA to the reaction mixture and incubation for
several hours to produce the target proteins. With available mutant
template DNA, mutant proteins can be obtained within just one
day. This advantage becomes evident when a large number of
proteins have to be tested, e.g. analyses of proteomic [39] and
alanine-scanning [40] studies. On the other hand, it is also true that
the in vitro expression method contains a basic weakness that unrecognized elements in the cell-extract solution sometimes bring
unexpected side reactions, e.g. proteolytic degradation, action of
chaperones and protein modification. Such effects make the result
complicated, in the worst cases leading to misinterpretation of
the assay results. To eliminate such risks, we had established
a novel in vitro cell-free protein synthesis system (the PURE
system) in which all of the components necessary for protein synthesis are highly purified and assembled [41]. This means that
the components in the solution are all defined and unwanted side
reactions do not occur during protein synthesis reactions.
Our previous study showed that F1 Fo from thermophilic
Bacillus PS3 has an ability to form the complex without the a
subunit, and furthermore that the complete F1 Fo complex can
F1 Fo (F1 Fo -ATP synthase) is ubiquitously found in the membranes
of bacteria, chloroplasts and mitochondria, and synthesizes
most of the cellular ATP coupled with translocation of H + (and
in some cases Na + ) [1–3]. F1 Fo is composed of two portions:
a membrane integral Fo (subunit composition, ab2 cn(10–15) in
most bacteria) and a water-soluble catalytic portion F1 (subunit
composition, α3 β3 γδε). Using the energy of downhill H +translocation, Fo drives rotation of the cn ring and its connecting
γ ε subunits, which induces a sequence of reactions of ATP
synthesis in F1 . Of the three kinds of Fo subunit, the a subunit is a
26 kDa protein containing five TMHs (transmembrane helices)
[4–6]. Since H + channels are present inside the a subunit it
plays a central role in H +-translocation. When the a subunit was
genetically eliminated, it resulted in formation of F1 Fo which had
lost all of its Fo -mediated functions [7], i.e. neither ATPase-driven
H +-translocation nor ATP synthesis proceeded. Mutation studies
have identified the key residues responsible for H +-translocation
[8–11], such as Arg169 (Arg210 in Escherichia coli numbering) [12–
15], and insertion-scanning studies have suggested the membrane
topology of the five TMHs [4,6,16]. Fillingame and colleagues
have extensively surveyed the a subunit of E. coli Fo by cysteine
scanning combined with the reactivity of Ag + [17–19]. Further
mutation analyses are still necessary, but these studies often
encounter the difficulty that the mutant a subunit is found only
scarcely [4,6,9,17,20] or not at all [8,15] in membranes. The same
difficulty has widely limited the versatility of mutagenesis studies
of other membrane proteins [21–26].
Key words: a subunit, cell-free protein synthesis, F1 Fo -ATP
synthase (F1 Fo ), H +-pump, membrane protein, proteoliposome.
Abbreviations used: ACMA, 9-amino-6-chloro-2-methoxyacridine; Ap5a, P1 ,P5 -di(adenosine-5 )pentaphosphate; CBB, Coomassie Brilliant Blue; DDM,
n-dodecyl-β-D-maltoside; F1 Fo , F1 Fo -ATP synthase; FCCP, carbonyl cyanide p -trifluoromethoxyphenylhydrazone; IMV, inverted inner-membrane vesicle;
PL, proteoliposome; TMH, transmembrane helix; wt, wild-type.
1
These authors equally contributed to this work.
2
To whom correspondence should be addressed (email [email protected]).
c The Authors Journal compilation c 2012 Biochemical Society
632
Y. Kuruma and others
be formed in vitro by supplementing the purified a subunit
[7]. In the present study, the a subunit was synthesized by
the PURE system in the presence of a subunit-less F1 Fo in
the form of PL (proteoliposome). Remarkably, F1 Fo recovered
from the reaction mixture gained H +-translocation activity driven
by ATP hydrolysis and ATP synthesis activity driven by H +translocation. This indicates that the synthesized a subunit was
inserted into membranes and integrated into a subunit-less F1 Fo to
form active F1 Fo . F1 Fo s with mutant a subunits were also obtained
by the PURE system in the presence of a subunit-less F1 Fo .
In comparison with in vivo expression, where some mutant a
subunits were lost and not integrated into F1 Fo , almost the same
amount of a subunit, irrespective of mutation, was produced and
integrated into F1 Fo , enabling analysis of these mutations.
1.4 M sucrose layers, was carefully isolated. The fraction was
supplemented with 5 mM MgCl2 , diluted 4-fold with 10 mM
Hepes/KOH buffer (pH 7.5) containing 5 mM MgCl2 , and then
ultracentrifuged at 55 000 rev./min for 20 min at 4 ◦ C using a S55A
rotor (Hitachi). The pellets were suspended in PA3 buffer and used
as IMVs for the following analyses.
Purification of F1 Fo and quantification of the F1 Fo preparation
was performed as described previously [46]. In the present study,
the F1 Fo preparation was further purified by anion-exchange
column chromatography (ResourceQ, GE Healthcare) in 10 mM
Hepes/NaOH buffer (pH 7.5) containing 0.2 mM EDTA. F1 Fo was
eluted from the column with a linear gradient of NaCl (0–0.5 M),
concentrated by a centrifugal concentrator (50 kDa, Amicon),
frozen by liquid N2 and then stored at − 80 ◦ C until use.
EXPERIMENTAL
Cell-free protein synthesis by the PURE system
Construction of expression vectors
In vitro synthesis of the a subunit was performed using a
reconstructed cell-free translation system, the PURE system [41],
in the presence of either 0.9 mg of membrane proteins/ml IMVs
or 0.12 mg/ml PLs containing F1 Fo a. Briefly, the PURE system
consists of aminoacyl-tRNA synthetases, translational factors,
ATP-regenerating systems, T7 RNA polymerase, 70S ribosome
and low molecular mass compounds. As the template DNA,
either 20 μg/ml pET-a or 5 μg/ml uncB DNA with a fragmentattaching T7 promotor sequence at the 5 terminus, amplified by
PCR, was supplemented into the reaction mixture. The reaction
mixture was incubated for 30 min at 30 ◦ C and then terminated
by placing the solution on ice as described previously [41].
Preliminary experiments showed that a reaction time of 30 min
was sufficient to obtain an active F1 Fo complex, and that no
further increase in H +-translocating and ATP synthesis activities
was observed even after prolonged reaction time. To analyse
the H +-translocating activity, a further purification step is not
necessary and therefore the reaction mixtures were tested directly.
When measurement of ATP synthesis activity was required, the
following purification step was added to eliminate the components
required for ATP regeneration (adenylate kinase and creatine
kinase) from the reaction mixture: the reaction mixture was
diluted 4-fold with 10 mM Hepes/KOH (pH 7.5) containing 5 mM
MgCl2 . The IMVs or PLs in the mixture were then collected
by ultracentrifugation [58 000 rev./min for 20 min at 4 ◦ C using
a S80AT3 rotor (Hitachi)], washed twice with the buffer and
suspended in PA3 buffer.
In order to assess the incorporation of the synthesized a
subunit into the F1 Fo a complex, after the synthesis reaction
F1 Fo was purified from the reaction mixture. As F1 Fo has a His
tag (10 histidine residues) at N-terminus of its β subunit, the
purification was performed by using Ni-charged magnetic beads
(MagneHis Protein Purification system, Promega), according to
the manufacturer’s protocol with some modifications. To the
reaction mixture (50 μl), 100 μl of 50 mM Tris/HCl (pH 8.0)
buffer containing 300 mM NaCl, 5 mM MgCl2 , 20 mM imidazole
and 1 % DDM (n-dodecyl-β-D-maltoside) was added. The
resulting mixtures were incubated at room temperature (25 ◦ C)
for 1 h with gentle mixing prior to the purification step. After
the addition of the magnetic beads, the beads were collected
and washed twice with 150 μl of the Tris/HCl buffer containing
0.05 % DDM. F1 Fo was then eluted with the Tris/HCl buffer
containing 500 mM imidazole and 0.05 % DDM. The eluted
fraction was subjected to acetone precipitation to concentrate
F1 Fo . The purifications of the wt (wild-type) and mutants were
performed under the same conditions at the same time. The F1 Fo
preparations obtained were analysed by SDS/PAGE (10–20 %
gel) and immunoblotting using anti-a antibody.
The expression vector for Bacillus PS3 F1 Fo , pTR19-ASDS [42],
expresses an active F1 Fo complex in E. coli cells, but the unc
operon encoding the F1 Fo subunits on the vector lacks the
uncI gene. Therefore a DNA fragment containing uncIBE was
amplified by PCR from the plasmid pUC-F1 Fo [43] and introduced
into the vector in order to restore uncI to its original position on
the Bacillus PS3 unc operon. The resulting plasmid was named
pTR-ISBS2. A mutation, genetically lacking the whole of uncB
(encoding the a subunit), was transferred from pTR19-ASDSa
[7] to pTR-ISBS2. The resultant plasmid, pTR-ISBS2a, was
used for IMV (inverted inner-membrane vesicle) analyses. For PL
analysis, a derivative of pTR19-ASDSεc was used instead of
pTR-ISBS2a because the measuring system for ATP synthesis
activity in PLs was optimized for the plasmid system. There is no
detectable difference between the two plasmid systems when their
Fo functions were compared with each other. The plasmid pTR19ASDSεc has a mutation at the ε subunit by which the C-terminal
domain of ε was truncated to eliminate the inhibitory effect (longlag phase) of the ε subunit [44]. This results in rapid appearance
of ATP synthesis activity. The mutation can be assumed to have
no effect on Fo function itself because the mutated region, the
C-terminal helices, has no direct interaction with the Fo part [45].
For a subunit synthesis, uncB was amplified by PCR solely from
pTR-ISBS2 and introduced into pET-29 at the NdeI/EcoRI sites.
The resulting plasmid, pET-a, was used as a template DNA. In
all plasmids, the regions prepared by PCR amplification were
verified by nucleotide sequencing. Mutations at the a subunit
were introduced using synthetic oligonucleotides produced by
PCR into pTR-ISBS2 by the Megaprimer method as described
previously [42,46].
Preparation of membrane vesicles and purification of F1 Fo
Expression of F1 Fo in E. coli cells and preparation of membrane
vesicles were performed as described previously [42]. In the
present study, the vesicles were further purified by the sucrosedensity gradient method to isolate IMVs as follows: membrane
vesicles dissolved in PA3 buffer [10 mM Hepes/KOH (pH 7.5)
containing 5 mM MgCl2 and 10 % glycerol] were supplemented
with 10 mM EDTA and subjected to brief sonication (10 s). The
resulting vesicles were carefully put on the top of a stepwise
gradient of sucrose (0.8, 1.4, 2.0 M) in 10 mM Hepes/KOH
(pH 7.5) containing 10 % glycerol. Bacillus PS3 F1 Fo did not
disassemble even after this procedure because of its high
structural stability. The resulting solution was ultracentrifuged
at 32 000 rev./min for 2 h at 4 ◦ C using a P40ST rotor (Hitachi).
After centrifugation, a brown layer, located between 0.8 M and
c The Authors Journal compilation c 2012 Biochemical Society
In vitro synthesis of F1 Fo -a subunit
Measurements of ATP synthesis
ATP synthesis activity in IMVs was measured at 35 ◦ C in 100 mM
Hepes/KOH (pH 7.5) containing 5 mM MgCl2 , 0.5 mM ADP,
10 μM Ap5a [P1 ,P5 -di(adenosine-5 )pentaphosphate], 10 mM
potassium phosphate and 1:10 vol. of luciferin/luciferase mixture
CLSII (ATP Bioluminescence Kit, Roche) as described previously
[46]. The synthesis reaction was initiated by adding 3.5 mM
sodium succinate. Production of ATP was measured in real-time
by monitoring luminescence (light intensity) from the reaction
mixture at 560 nm by using a spectrofluorometer (FP6500,
JASCO) without excitation light. To convert the light intensity
into the ATP concentration value, defined amounts of ATP were
added to the assay mixture at the end of each measurement. The
activity that synthesized 1 μM ATP per min was defined as 1 unit.
Measurement of the ATP synthesis activity of PLs was done
by the acid/base-transition method, combined with valinomycininduced diffusion of potassium ion. This procedure was
established by modifying the method of Fischer et al. [47] and
was published recently [48]. Briefly, F1 Fo PLs were prepared from
a solution containing 32 mg/ml soybean L-α-phosphatidylcholine
(TypeII-S, Sigma), 0.6 mg/ml F1 Fo preparation and 6 % n-octylβ-D-glucopyranoside (Anatrace) by eliminating the detergent
using Bio-Beads (SM-2, Bio-Rad Laboratories). The resulting
PLs (final F1 Fo concentration was 0.3 mg/ml) were acidified by
soaking in acidification buffer (20 mM succinate, 0.6 mM KOH,
2.5 mM MgCl2 , 10 mM NaH2 PO4 , 0.7 mM ADP, 1 μM Ap5a and
280 nM valinomycin, adjusted to pH 4.5 by NaOH) for 4 min at
room temperature. The acidified PLs (30 μl) were then rapidly
injected into Base buffer (1 ml) containing 200 mM Tricine,
130 mM KOH, 2.5 mM MgCl2 , 10 mM NaH2 PO4 , 0.5 mM ADP,
1 μM Ap5a and 0.1 ml ATP luminescence assay kit CLS II
reagent (Roche), adjusted to pH 8.8 by NaOH. Changes in the
luminescence from luciferin/luciferase were monitored at 37 ◦ C
using a luminometer (Luminescencer-PSN AB-2200, ATTO).
To calibrate the luminescence intensity as ATP concentration,
200 nM ATP was added into the reaction mixture at the end of
each measurement.
Other analytical methods
ATPase-driven H +-translocation activity was analysed by
monitoring fluorescence quenching of ACMA (9-amino-6-chloro2-methoxyacridine) (excitation at 410 nm and emission at
480 nm) as described previously [42] at 42 ◦ C in 10 mM
Hepes/KOH buffer (pH 7.5) containing 100 mM KCl, 5 mM
MgCl2 and 0.3 μg/ml ACMA. Additionally, 2.5 mM creatine
phosphate and 25 μg/ml creatine kinase were added to the assay
mixture to regenerate ATP during measurement. The reaction was
initiated by adding 1 mM ATP, and terminated with 0.25 μg/ml
FCCP (carbonyl cyanide p-trifluoromethoxyphenylhydrazone).
All experiments were performed at least three times and the
averages are shown. Protein concentrations of membrane vesicles
were determined using a BCA (bicinchoninic acid) protein assay
kit from Pierce, with bovine serum albumin as a standard. All
SDS/PAGE in the present study used a gradient polyacrylamide
gel (10–20 %).
RESULTS
Expression of an F1 Fo mutant lacking the a subunit, F1 Fo a , in
E. coli cells
The a subunit-less F1 Fo (F1 Fo a) was expressed from the
expression plasmid, pTR-ISBS2a, in an F1 Fo -deficient E. coli
strain, DK8 [42]. Western blotting analysis of IMVs, prepared
from the recombinant cells, confirmed that F1 Fo a was stably
633
expressed into the membrane fraction of the E. coli host cells
(Figure 1A). The band intensity of α and β subunits of the
gel (Figure 1A, lanes 2 and 3) suggests that the expression
level of F1 Fo a was comparable with that of the wt-F1 Fo .
Immunoblot analysis using anti-a antibody showed no detectable
a subunit in the IMVs of F1 Fo a (Figure 1B, lane 3), but a
clear band for wt-F1 Fo (Figure 1B, lane 2). F1 Fo complexes were
purified from respective IMVs for evaluation of their subunit
stoichiometry (Figure 1C). With the exception of the absence
of the a subunit, other subunits were similarly observed on
SDS/PAGE gels (Figure 1C), which indicate the stable nature
of F1 Fo a. This result is consistent with an earlier finding that
F1 Fo still remains stable even when the a subunit is absent [7].
Formation of active F1 Fo on IMVs by in vitro expression of the
a subunit
The a subunit was synthesized by using the PURE system
in the presence of the IMVs. The reaction was performed
with or without the plasmid pET-a which contains the uncB
gene (encoding the a subunit). After the synthesis reaction,
IMVs were isolated by ultracentrifugation, washed with buffer
and then analysed. Figures 1(A) and 1(B) show SDS/PAGE and
immunoblot analyses of the IMVs reconstituted with in vitrosynthesized a subunit (lanes 4–6). Because ribosome components
were also simultaneously precipitated by the centrifugation step,
the band intensity of some proteins increased by overlaying of
these components (Figure 1A). Immunoblot analysis using anti-a
antibody (Figure 1B) clearly showed integration of the a subunit
into F1 Fo a IMVs (lane 6), the amount of which was comparable
with that of wt-F1 Fo (lane 4).
ATPase-driven H+ -translocating activity in the IMVs
reconstituted with in vitro-synthesized a subunit was analysed by
monitoring fluorescence quenching of ACMA, which is a pH
indicator of membrane pH. In the absence of uncB DNA, the
IMVs showed no quenching of fluorescence after the addition of
ATP (Figure 1D, uncB − trace). The rapid quenching observed just
after the addition of ATP is due to a direct interaction between ATP
and the ACMA fluorophore. In contrast, significant quenching
was observed when uncB was present during the synthesis reaction
(Figure 1D, uncB + trace). As a control, the synthesis reaction was
performed with wt IMVs in the presence of uncB (Figure 1D,
WT-uncB + trace). The initial quenching rate of the F1 Fo a IMVs
when reacted with uncB DNA showed 32 % of the quenching
observed for wt IMV, when fluorescence changes were assessed
2–20 s after the addition of ATP. The ATP synthesis activity of
the IMVs reconstituted with in vitro-synthesized a subunit was
assessed by a real-time monitoring method using luciferin/
luciferase (Figure 1E). As shown in the WT-uncB + trace, a linear
increase in light intensity was observed in authentic wt-F1 Fo IMVs
that were prepared through the synthesis reaction in the presence
of uncB. Although no increase in luminescence was observed
in F1 Fo a IMVs without uncB (uncB − trace), a moderate, but
obvious, increase in luminescence was observed in F1 Fo a IMVs
reacted with uncB (uncB + trace). ATP synthesis activities were
determined as 13.4 m-units/mg of membrane proteins in wt-F1 Fo
with uncB, and 2.6 or 0.01 m-units/mg of membrane proteins in
F1 Fo a with or without uncB. The result showed that the ATP
synthesis function was acquired due to synthesis of the a subunit,
although the activity was relatively lower than that of wt-F1 Fo .
In vitro expression of the a subunit with F1 Fo a PLs
Recently, we have established a new assay system for ATP
synthesis activity using PLs [48]. The expression vector
c The Authors Journal compilation c 2012 Biochemical Society
634
Figure 1
Y. Kuruma and others
In vitro expression of a subunit and formation of active F1 Fo in IMVs
SDS/PAGE (A) and immunoblot (B) analyses of IMVs. Lanes 2 and 3 show IMVs carrying wt F1 Fo and F1 Fo a respectively. wt F1 Fo IMVs (lane 4) and mutant F1 Fo a IMVs (lanes 5 and 6) were
respectively supplemented to the reaction mixture. The reactions were performed with (lanes 4 and 6) or without (lane 5) template DNA (uncB ). (C) SDS/PAGE analysis of F1 Fo preparations. In
(A–C), lanes 1 and 7 correspond to F1 Fo preparations of wt and F1 Fo a respectively. The position of a subunit is indicated by an arrowhead. (D) H +-translocating activity of the IMVs reconstituted
with in vitro -synthesized a subunit by monitoring fluorescence of ACMA. H +-translocation was initiated by adding ATP and terminated by FCCP, as indicated by the arrows.%ACMA indicates the
fluorescence intensity of ACMA which before adding ATP was calibrated as 100 %. (E) ATP synthesis activity of the IMVs reconstituted with in vitro -synthesized a subunit. The activity was analysed
by monitoring luminescence of luciferin/luciferase. The reaction was initiated by adding sodium succinate as indicated by the arrow. A defined amount of ATP was added to calibrate the light intensity
at the end of reaction.
(pTR19-ASDSεc) necessary for the measurement has a
mutation in the C-terminal domain of the ε subunit to eliminate
an inhibitory effect of the ε subunit. A mutation lacking uncB
was introduced into the vector to obtain pTR19-ASDSεc-a.
The a subunit-less F1 Fo mutant, named F1 Fo (εc)a, was stably
expressed in E. coli DK8 cells and purified by the same method
as wt-F1 Fo . The purified F1 Fo was reconstituted into soybean
phosphatidylcholine to obtain F1 Fo (εc)a PLs. An in vitro
expression reaction of the a subunit was then performed in the
presence of the PLs. Figures 2(A) and 2(B) show SDS/PAGE
analyses of the reacted PLs, which were reacted without (lane 2)
or with (lane 3) the pET-a plasmid. In immunoblot analysis of the
reacted PLs (Figure 2B), a clear band was detected at the position
of the a subunit (lane 3), although a few faint bands were also
observed above the a subunit band.
The resultant PLs were further analysed for ATPase-driven H +translocating activity (Figure 2C). PLs in the presence of uncB
DNA (uncB + trace) exhibited significant fluorescence quenching
of ACMA upon the addition of ATP, whereas PLs without uncB
DNA (uncB − trace) showed no quenching activity at all. This
indicates that the appearance of H +-translocating activity resulted
from a subunit synthesis. Quenching velocity in F1 Fo (εc)a
PLs with de novo-synthesized a subunit was approximately
52 % of the wt PLs (WT-uncB + trace). Subsequently, ATP
synthesis activity of the reacted PLs was also analysed by using
c The Authors Journal compilation c 2012 Biochemical Society
an acid/base-transition method combined with a K +-induced
membrane potential () in combination with valinomycin. The
amount of ATP synthesized by F1 Fo was assessed by monitoring
the luminescence of the assay mixture. F1 Fo (εc)a PLs reacted
with uncB DNA were isolated by ultracentrifugation, acidified and
then added to the assay mixture. The addition of PLs (indicated
with an arrow) resulted in a rapid increase in luminescence
light intensity (Figure 2D, uncB + trace). When the synthesis
reaction was performed without uncB, addition of PLs induced
no increase in the light intensity (Figure 2D, uncB − trace).
When μH + was dissipated by supplementing with the K + /H +exchanging ionophore nigericin, the increase in luminescence
almost disappeared (Nig + uncB + trace). These results suggest
that F1 Fo (εc)a on PLs acquired the ATP synthetic function by
synthesizing the a subunit. When compared with the activity of wtF1 Fo (Figure 2D, WT-uncB + trace), the activity of F1 Fo (εc)a
PLs that reacted with uncB was 51 %.
Mutation analysis of a subunits by the in vitro protein expression
system
Figure 3 shows a sequence alignment for a subunits of E.
coli, Propionigenium modestum and Bacillus PS3. The sequence
identities of a subunit of Bacillus PS3 are moderately similar to
those of E. coli (31 %) and P. modestum (25 %). In the regions of
In vitro synthesis of F1 Fo -a subunit
Figure 2
635
In vitro expression of a subunit and formation of active F1 Fo in PLs
(A and B) SDS/PAGE analysis of F1 Fo a PLs after a subunit synthesis. Synthesis of the a subunit was performed in the presence of F1 Fo a PLs with (lane 3) or without (lane 2) uncB . Lane 1
shows purified wt-F1 Fo . Gels were stained by CBB (A) or analysed by anti-a antibody immunoblotting (B). An arrowhead indicates the position of the a subunit. (C) H +-translocating activity of the
reacted F1 Fo a PLs. (D) ATP synthesis activity of the reacted F1 Fo a PLs. ATP synthesis was initiated by adding of the acidified PLs as shown with an arrow. Representations of samples were
same as (C) except for Nig + uncB + , which indicates F1 Fo a PLs reacted with uncB , but treated with nigericine before the measurement in order to dissipate μH + across membranes. A defined
amount of ATP was added for calibration as indicated with an arrow.
Figure 3
Amino acid sequence alignment of a subunits of bacterial ATP synthases
The alignment was obtained by using ClustalX [49]. Five putative TMHs in the a subunit are underlined, according to the previous proposal for the E. coli a subunit [19]. Conserved residues among
the three a subunits were highlighted with grey. Of these, five residues substituted with alanine in the present study are shown boxed. PS3, Bacillus PS3 H +-translocating F1 Fo ; E. coli , Escherichia
coli H +-translocating F1 Fo ; P.mod., Propionigenium modestum Na +-translocating F1 Fo .
the five TMHs [19], eight polar residues are conserved among
the three a subunits. Of the eight, five polar residues, Asn90
(TMH2), Asp112 (TMH3), Arg169 (TMH4), Asn173 (TMH4) and
Gln217 (TMH5), were selected (Figure 3, marked by boxes),
and were each replaced by alanine (N90A, D112A, etc.). The
five mutants were then subjected to analyses using the in vitro
or in vivo expression methods. First, the in vitro expression
method was performed. DNA fragments of uncB including the
T7 promoter and Shine–Dalgarno sequences were prepared by a
two-step PCR (Figure 4A). Next, the variants of the a subunit
were synthesized from the amplified uncB fragment in the
presence of F1 Fo (εc)a PLs. Immunoblot analysis by anti-a
antibody of the reaction mixture indicated that similar amounts
of a subunits were synthesized from the wt and mutated uncB
DNAs (Figure 4B). H + -translocating activity in the reacted
F1 Fo (εc)a PLs was directly analysed after the synthesis
reaction without any purification step of PLs (Figure 4C). The
result from this series of experiments was that the N173A mutant
showed strong H + -translocating activity, which was comparable
with the wt. Two other mutants, D112A and Q217A, demonstrated
weak activity, which indicated that side chains of the two residues,
though important, are not obligatory for H + -translocation. On
the other hand, the N90A and R169A mutants demonstrated no
ACMA quenching by ATP, which suggests that these residues
are required by the enzyme to function. To clarify whether these
differences in the activities are faithfully reflecting the effect of
the mutation, F1 Fo was purified from the reaction mixture of the
a subunit synthesis by using Ni-charged magnet beads (His-tags
are present at the β subunit). The yields of the isolated F1 Fo
were almost same as judged from band intensities of α and β
on the CBB (Coomassie Brilliant Blue)-stained SDS/PAGE gel
(Figure 4D). An anti-a antibody immunoblot analysis showed that
F1 Fo of the mutants contains a similar amount of a subunit when
compared with the wt (Figure 4E). These data suggest that the
mutated a subunits were incorporated into F1 Fo a complex at a
similar level to the wt a subunit.
Secondly, the in vivo expression method was carried out
in a conventional manner. Expression plasmids for the five
c The Authors Journal compilation c 2012 Biochemical Society
636
Figure 4
Y. Kuruma and others
Mutation study of the a subunit by the in vitro expression method
(A) Agarose gel electrophoresis of uncB DNA fragments prepared by PCR. (B) Immunoblot analysis of synthesized a subunits by anti-a antibody. The synthesis reaction of the a subunits
was performed with F1 Fo a PLs in the presence of either wt or mutated uncB . After the synthesis reaction, the samples were subjected to immunoblot analysis by anti-a antibody.
(C) H + -translocating activity of the reacted PLs. The PL samples after the synthesis reaction were directly subjected to the measurement. The analytical conditions were same as those in Figure 1(D).
(D) SDS/PAGE analysis of F1 Fo preparations. F1 Fo was purified from the reaction mixture of the a subunit synthesis by Ni-charged magnetic beads and applied to an SDS/PAGE gel. The gel was
stained with CBB. (E) Detection of a subunit in the F1 Fo preparations. The F1 Fo used in (D) were subjected to anti-a antibody immunoblot analysis. In (B), (D) and (E), the lanes represented with
TF1 Fo show purified wt-F1 Fo as a size marker.
Figure 5
Mutation analysis of the a subunit by the ‘normal’ in vivo expression method and its comparison with the in vitro expression method
(A) SDS/PAGE analysis of IMVs containing F1 Fo mutations prepared by the ‘normal’ in vivo expression method. (B) H + -translocation activity of the IMVs. Analytical conditions were the same as
those in Figure 1(D). (C) Immunoblot analysis of the IMVs by anti-a antibody. The arrow indicates the position of the a subunit. The lane headings are the same as in (A). (D) Comparison of the
results from the in vivo and in vitro expression methods. Initial quenching rates of ACMA fluorescence after the addition of ATP were determined. Values of the rates relative to the wt are shown.
Open and closed bars indicate the in vivo and in vitro expression methods respectively.
mutated F1 Fo and wt-F1 Fo were introduced into E. coli DK8,
and F1 Fo was expressed by cultivating the recombinant strains.
SDS/PAGE analysis of prepared IMVs showed that all of the
mutants more or less decreased the expression levels of F1 Fo ,
especially R169A, judging from the band intensities of the α and
β subunits (Figure 5A). Figure 5(B) shows that of the five, only
N173A showed H + -translocating activity, even though substantial
amounts of the mutated F1 Fo was detected in IMVs in the four
mutants (Figure 5A). The apparent inconsistency can be explained
by results from anti-a antibody immunoblot analysis where three
c The Authors Journal compilation c 2012 Biochemical Society
mutants (N90A, D112A and Q217) had no detectable a subunit
in the IMVs fraction (Figure 5C).
Figure 5(D) summarizes the results of H + -translocating
activities obtained from the in vitro expression (Figure 4C) and
in vivo expression (Figure 5B) methods. In order to compare
the results of both methods, each initial fluorescence-quenching
rate was determined from the respective traces just after ATP
addition, and their values relative to the wt moiety were presented.
This comparison demonstrates that there are critical differences
between the results from the in vitro and in vivo expression
In vitro synthesis of F1 Fo -a subunit
637
methods with regard to the D112A and Q217A mutants. The
activities of these two mutants were detected only in the in vitro
expression method.
conditions of F1 Fo . Sakurako Ono prepared cell membranes containing a subunit variants.
Masasuke Yoshida supervised the research and edited the paper before submission.
Takuya Ueda obtained grant support and edited the paper before submission.
DISCUSSION
ACKNOWLEDGEMENTS
In the present study, the a subunit of Bacillus PS3 F1 Fo was
successfully synthesized by the PURE system. Remarkably, when
an a subunit-less F1 Fo was added to the reaction mixture, it
acquired Fo functions after synthesizing the a subunit without
any additional agent. This result indicates that the synthesized
a subunit was spontaneously integrated into membranes, folded
correctly and complexed with the a subunit-less F1 Fo to form
active F1 Fo . The results from mutation analysis of a subunits
produced by the in vitro expression method were compared with
results from a subunits produced by the in vivo expression method.
The two methods produced analogous results for some mutants;
however, differences were found for mutants of D112A and
Q217A. As shown in Figure 5(D), the in vitro expression method
provided a small, but substantial, amount of activity for these
mutants (see also Figure 4C), in contrast to the ‘normal’ in vivo
expression results which showed no detectable H + -translocating
activity (Figure 5C). The reason for this difference is explained by
immunoblot analysis using anti-a antibody, which revealed that
the a subunit was absent in the IMVs of the mutants. The results
from the two mutants might have been affected by proteolytic
digestion of a subunits in the IMVs. In fact, when the Bacillus
PS3 a subunit was expressed in E. coli cells without other F1 Fo
subunits, a short expression time (2 h) was optimal to accumulate
the a subunit on the membrane fraction [7], which implies that the
a subunit is sensitive to protease. Such decreased accumulation
of the a subunit has been frequently observed in E. coli a subunit
studies [4,6,8,9,15,17,20].
The in vitro expression procedure clarified the necessity of
the five conserved residues. Of the five, Asp112 and Gln217 were
important for ATPase-driven H + -translocating activity, but were
not absolutely required (Figure 4C). This result is in good
agreement with that from previous E. coli studies [17,18] where
E. coli residues corresponding to Asp112 and Gln217 of Bacillus
PS3 were suggested to lie on H + -translocating channels [17,18].
Additionally, the obligate role of the side chain of Asn90 and
Arg169 was suggested. Arg169 (Arg210 and Arg226 in E. coli and
P. modestum respectively) has been widely known as one of the
most important residues in the a subunit [12–15]. Asn90 is located
in the middle of the second putative TMH and also appears to
play an indispensable role.
The PURE system represents a different environment from that
of intact cells or cell extract, since, in the PURE system, protease
is absent and therefore abnormal proteins are not destroyed by
proteolytic digestion. This observation suggests that the PURE
system conditions allow the a subunit to form the correct
conformations on a lipid bilayer without susceptibility to protease
attack during expression (Figure 4E). This advantage of the PURE
system enables us to analyse such ‘difficult-to-produce proteins’
that cannot be prepared by the ‘normal’ in vivo expression method
using cells. Simultaneous use of the in vitro expression procedure
with the normal in vivo method would greatly facilitate studies
that require such difficult proteins.
AUTHOR CONTRIBUTION
Yutetsu Kuruma and Toshiharu Suzuki equally contributed as primary researchers
responsible for experimental design, data acquisition, data analysis and the preparation
of the paper. Yutetsu Kuruma generated experimental data using the PURE system and
activity assay of F1 Fo ; Toshiharu Suzuki prepared samples of F1 Fo and optimized the assay
We thank Dr Boris Feniouk, Mr Chiaki Wakabayashi, and Mr Kazumi Tanaka (ICORP, JST)
for assistance with Fo F1 purification and valuable discussions.
FUNDING
This work was supported in part by the Targeted Proteins Research Program and Grantin-Aid for Scientific Research (A) [grant number 18201040 to (T.U.)] from the Ministry
of Education, Culture, Sports, Science and Technology (MEXT), Japan, and the Human
Frontier Science Program Organization (HFSPO).
REFERENCES
1 Boyer, P. D. (2002) A research journey with ATP synthase. J. Biol. Chem. 277,
39045–39061
2 Junge, W., Lill, H. and Engelbrecht, S. (1997) ATP synthase: an electrochemical
transducer with rotatory mechanics. Trends Biochem. Sci. 22, 420–423
3 Yoshida, M., Muneyuki, E. and Hisabori, T. (2001) ATP synthase: a marvelous rotary
engine of the cell. Nat. Rev. Mol. Cell Biol. 2, 669–677
4 Long, J. C., Wang, S. and Vik, S. B. (1998) Membrane topology of subunit a of the F1 F0
ATP synthase as determined by labeling of unique cysteine residues. J. Biol. Chem. 273,
16235–16240
5 Valiyaveetil, F. I. and Fillingame, R. H. (1998) Transmembrane topography of subunit a in
the Escherichia coli F1 F0 ATP synthase. J. Biol. Chem. 273, 16241–16247
6 Wada, T., Long, J. C., Zhang, D. and Vik, S. B. (1999) A novel labeling approach supports
the five-transmembrane model of subunit a of the Escherichia coli ATP synthase. J. Biol.
Chem. 274, 17353–17357
7 Ono, S., Sone, N., Yoshida, M. and Suzuki, T. (2004) ATP synthase that lacks
F0 a -subunit: isolation, properties, and indication of F0 b2 -subunits as an anchor rail of a
rotating c-ring. J. Biol. Chem. 279, 33409–33412
8 Vik, S. B., Patterson, A. R. and Antonio, B. J. (1998) Insertion scanning mutagenesis of
subunit a of the F1 F0 ATP synthase near His245 and implications on gating of the proton
channel. J. Biol. Chem. 273, 16229–16234
9 Wang, S. and Vik, S. B. (1994) Single amino acid insertions probe the α subunit of the
Escherichia coli F1 F0 -ATP synthase. J. Biol. Chem. 269, 3095–3099
10 Cain, B. D. and Simoni, R. D. (1989) Proton translocation by the F1 FO ATPase of
Escherichia coli . Mutagenic analysis of the a subunit. J. Biol. Chem. 264, 3292–3300
11 Vik, S. B., Cain, B. D., Chun, K. T. and Simoni, R. D. (1988) Mutagenesis of the α subunit
of the F1 FO ATPase from Escherichia coli . Mutations at Glu196 , Pro190 , and Ser199 . J. Biol.
Chem. 263, 6599–6605
12 Lightowlers, R. N., Howitt, S. M., Hatch, L., Gibson, F. and Cox, G. B. (1987) The proton
pore in the Escherichia coli F0 F1 -ATPase: a requirement for arginine at position 210 of the
a -subunit. Biochim. Biophys. Acta 894, 399–406
13 Valiyaveetil, F. I. and Fillingame, R. H. (1997) On the role of Arg210 and Glu219 of subunit a
in proton translocation by the Escherichia coli F0 F1 -ATP synthase. J. Biol. Chem. 272,
32635–32641
14 Vorburger, T., Ebneter, J. Z., Wiedenmann, A., Morger, D., Weber, G., Diederichs, K.,
Dimroth, P. and von Ballmoos, C. (2008) Arginine-induced conformational change in the
c -ring/a -subunit interface of ATP synthase. FEBS J. 275, 2137–2150
15 Mitome, N., Ono, S., Sato, H., Suzuki, T., Sone, N. and Yoshida, M. (2010) Essential
arginine residue of the Fo-a subunit in Fo F1 -ATP synthase has a role to prevent the proton
shortcut without c -ring rotation in the Fo proton channel. Biochem. J. 430, 171–177
16 Zhang, D. and Vik, S. B. (2003) Close proximity of a cytoplasmic loop of subunit a with c
subunits of the ATP synthase from Escherichia coli . J. Biol. Chem. 278, 12319–12324
17 Angevine, C. M. and Fillingame, R. H. (2003) Aqueous access channels in subunit a of
rotary ATP synthase. J. Biol. Chem. 278, 6066–6074
18 Angevine, C. M., Herold, K. A., Vincent, O. D. and Fillingame, R. H. (2007) Aqueous
access pathways in ATP synthase subunit a . Reactivity of cysteine substituted into
transmembrane helices 1, 3, and 5. J. Biol. Chem. 282, 9001–9007
19 Moore, K. J., Angevine, C. M., Vincent, O. D., Schwem, B. E. and Fillingame, R. H. (2008)
The cytoplasmic loops of subunit a of Escherichia coli ATP synthase may participate in
the proton translocating mechanism. J. Biol. Chem. 283, 13044–13052
20 Ishmukhametov, R. R., Pond, J. B., Al-Huqail, A., Galkin, M. A. and Vik, S. B. (2008) ATP
synthesis without R210 of subunit a in the Escherichia coli ATP synthase. Biochim.
Biophys. Acta 1777, 32–38
c The Authors Journal compilation c 2012 Biochemical Society
638
Y. Kuruma and others
21 Lee, G. F., Dutton, D. P. and Hazelbauer, G. L. (1995) Identification of functionally
important helical faces in transmembrane segments by scanning mutagenesis. Proc. Natl.
Acad. Sci. U.S.A. 92, 5416–5420
22 Ward, C. L., Omura, S. and Kopito, R. R. (1995) Degradation of CFTR by the
ubiquitin-proteasome pathway. Cell 83, 121–127
23 Akiyama, Y., Kihara, A., Tokuda, H. and Ito, K. (1996) FtsH (HflB) is an ATP-dependent
protease selectively acting on SecY and some other membrane proteins. J. Biol. Chem.
271, 31196–31201
24 Blott, E. J., Higgins, C. F. and Linton, K. J. (1999) Cysteine-scanning mutagenesis
provides no evidence for the extracellular accessibility of the nucleotide-binding
domains of the multidrug resistance transporter P-glycoprotein. EMBO J. 18,
6800–6808
25 Mimura, H., Nakanishi, Y., Hirono, M. and Maeshima, M. (2004) Membrane topology of
the H +-pyrophosphatase of Streptomyces coelicolor determined by cysteine-scanning
mutagenesis. J. Biol. Chem. 279, 35106–35112
26 Mitchell, S. M., Lee, E., Garcia, M. L. and Stephan, M. M. (2004) Structure and function
of extracellular loop 4 of the serotonin transporter as revealed by cysteine-scanning
mutagenesis. J. Biol. Chem. 279, 24089–24099
27 Kalmbach, R., Chizhov, I., Schumacher, M. C., Friedrich, T., Bamberg, E. and Engelhard,
M. (2007) Functional cell-free synthesis of a seven helix membrane protein: in situ
insertion of bacteriorhodopsin into liposomes. J. Mol. Biol. 371, 639–648
28 Marques, B., Liguori, L., Paclet, M. H., Villegas-Mendéz, A., Rothe, R., Morel, F. and
Lenormand, J. L. (2007) Liposome-mediated cellular delivery of active gp91phox . PLoS
ONE 9, e856
29 Klammt, C., Schwarz, D., Eifler, N., Engel, A., Piehler, J., Haase, W., Hahn, S., Dötsch, V.
and Bernhard, F. (2007) Cell-free production of G protein-coupled receptors for functional
and structural studies. J. Struct. Biol. 158, 482–493
30 Goren, M. A. and Fox, B. G. (2008) Wheat germ cell-free translation, purification, and
assembly of a functional human stearoyl-CoA desaturase complex. Protein Expr. Purif.
62, 171–178
31 Liguori, L., Marques, B., Villegas-Mendez, A., Rothe, R. and Lenormand, J. L. (2008)
Liposome-mediated delivery of pro-apoptotic therapeutic membrane proteins. J. Control.
Release 126, 217–227
32 Cappuccio, J. A., Blanchette, C. D., Sulchek, T. A., Arroyo, E. S., Kralj, J. M., Hinz, A. K.,
Kuhn, E. A., Chromy, B. A., Segelke, B. W., Rothschild, K. J. et al. (2008) Cell-free
co-expression of functional membrane proteins and apolipoprotein, forming soluble
nanolipoprotein particles. Mol. Cell. Proteomics 7, 2246–2253
33 Hovijitra, N. T., Wuu, J. J., Peaker, B. and Swartz, J. R. (2009) Cell-free synthesis of
functional aquaporin Z in synthetic liposomes. Biotechnol. Bioeng. 104, 40–49
34 Kai, L., Kaldenhoff, R., Lian, J., Zhu, X., Dötsch, V., Bernhard, F., Cen, P. and Xu, Z. (2010)
Preparative scale production of functional mouse aquaporin 4 using different cell-free
expression modes. PLoS ONE 5, e12972
Received 18 July 2011/21 November 2011; accepted 14 December 2011
Published as BJ Immediate Publication 14 December 2011, doi:10.1042/BJ20111284
c The Authors Journal compilation c 2012 Biochemical Society
35 Stiegler, N., Dalbey, R. E. and Kuhn, A. (2011) M13 procoat protein insertion into YidC
and SecYEG proteoliposomes and liposomes. J. Mol. Biol. 406, 362–370
36 Nishiyama, K., Maeda, M., Abe, M., Kanamori, T., Shimamoto, K., Kusumoto, S., Ueda, T.
and Tokuda, H. (2010) A novel complete reconstitution system for membrane integration
of the simplest membrane protein. Biochem. Biophys. Res. Commun. 394, 733–736
37 Shin, J. and Noireaux, V. (2010) Efficient cell-free expression with the endogenous E. coli
RNA polymerase and sigma factor 70. J. Biol. Eng. 4, 8
38 Kuruma, Y., Stano, P., Ueda, T. and Luisi, P. L. (2009) A synthetic biology approach to the
construction of membrane proteins in semi-synthetic minimal cells. Biochim. Biophys.
Acta 1788, 567–574
39 Niwa, T., Ying, B. W., Saito, K., Jin, W., Takada, S., Ueda, T. and Taguchi, H. (2009)
Bimodal protein solubility distribution revealed by an aggregation analysis of the entire
ensemble of Escherichia coli proteins. Proc. Natl. Acad. Sci. U.S.A. 106, 4201–4206
40 Osada, E., Shimizu, Y., Akbar, B. K., Kanamori, T. and Ueda, T. (2009) Epitope mapping
using ribosome display in a reconstituted cell-free protein synthesis system. J. Biochem.
145, 693–700
41 Shimizu, Y., Inoue, A., Tomari, Y., Suzuki, T., Yokogawa, T., Nishikawa, K. and Ueda, T.
(2001) Cell-free translation reconstituted with purified components. Nat. Biotechnol. 19,
751–755
42 Suzuki, T., Ueno, H., Mitome, N., Suzuki, J. and Yoshida, M. (2002) Fo of ATP synthase is
a rotary proton channel. Obligatory coupling of proton translocation with rotation of
c -subunit ring. J. Biol. Chem. 277, 13281–13285
43 Ohta, S., Yohda, M., Ishizuka, M., Hirata, H., Hamamoto, T., Otawara-Hamamoto, Y.,
Matsuda, K. and Kagawa, Y. (1988) Sequence and over-expression of subunits of
adenosine triphosphate synthase in thermophilic bacterium PS3. Biochim. Biophys. Acta
933, 141–155
44 Feniouk, B. A., Kato-Yamada, Y., Yoshida, M. and Suzuki, T. (2010) Conformational
transitions of subunit epsilon in ATP synthase from thermophilic Bacillus PS3.
Biophys. J. 98, 434–442
45 Stock, D., Leslie, A. G. and Walker, J. E. (1999) Molecular architecture of the rotary motor
in ATP synthase. Science 286, 1700–1705
46 Suzuki, T., Ozaki, Y., Sone, N., Feniouk, B. A. and Yoshida, M. (2007) The product of uncI
gene in F1 Fo -ATP synthase operon plays a chaperone-like role to assist c -ring assembly.
Proc. Natl. Acad. Sci. U.S.A. 104, 20776–20781
47 Fischer, S., Etzold, C., Turina, P., Deckers-Hebestreit, G., Altendorf, K. and Gräber, P.
(1994) ATP synthesis catalyzed by the ATP synthase of Escherichia coli reconstituted into
liposomes. Eur. J. Biochem. 225, 167–172
48 Soga, N., Kinoshita, Jr, K., Yoshida, M. and Suzuki, T. (2011) Efficient ATP synthesis by
thermophilic Bacillus Fo F1 -ATP synthase. FEBS J. 278, 2647–2654
49 Larkin, M. A., Blackshields, G., Brown, N. P., Chenna, R., McGettigan, P. A., McWilliam,
H., Valentin, F., Wallace, I. M., Wilm, A., Lopez, R. et al. (2007) ClustalW and ClustalX
version 2.0. Bioinformatics 23, 2947–2948