An intermediate step in the evolution of ATPases ) the
F
1
F
0
-ATPase from Acetobacterium woodii contains F-type
and V-type rotor subunits and is capable of ATP synthesis
Michael Fritz and Volker Mu
¨
ller
Molecular Microbiology & Bioenergetics, Institute of Molecular Biosciences, Johann Wolfgang Goethe University, Frankfurt am Main, Germany
Membrane-bound, multisubunit, ion-translocating ATP
synthases ⁄ ATPases are present in every domain of life.
They arose from a common ancestor, but evolved into
three distinct classes of ATP synthases ⁄ ATPases: the
F
1
F
0
-ATP synthase present in bacteria, mitochondria
and chloroplasts, the A
1
A
0
-ATP synthase present in
archaea, and the V
1
V
0
-ATPase present in eukarya
[1,2]. A common feature of ATP synthases ⁄ ATPases

brane [13,14].
Subunit c of F
1
F
0
-ATP synthases has a molecular
mass of around 8 kDa, and folds in the membrane
like a hairpin, with two transmembrane helices that
are connected by a cytoplasmic loop [15]. Each
monomer contains an ion-binding site, and as 10–15
subunits constitute the rotor (depending on the spe-
cies), it has a total of 10–15 ion-binding sites [12,16–
19] This gives a H
+
(Na
+
) ⁄ ATP stoichiometry of
3.3–5, a value required for ATP synthesis, given a
Keywords
Acetobacterium woodii; ATP synthase;
F-type; rotor subunits; V-type
Correspondence
V. Mu
¨
ller, Molecular Microbiology &
Bioenergetics, Institute of Molecular
Biosciences, Johann Wolfgang Goethe
University Frankfurt ⁄ Main, Max-von-Laue-
Str. 9, 60438 Frankfurt, Germany
Fax: +49 69 79829306

1
F
0
-like subunits c
2
and c
3
. After incorporation into
liposomes, ATP synthesis could be driven by an electrochemical sodium
ion potential or a potassium ion diffusion potential, but not by a sodium
ion potential. This is the first demonstration that an ATPase with a V
0
–F
0
hybrid motor is capable of ATP synthesis.
Abbreviations
DY, membrane potential; DlNa
+
, electrochemical sodium ion potential; DpNa, sodium ion potential.
FEBS Journal 274 (2007) 3421–3428 ª 2007 The Authors Journal compilation ª 2007 FEBS 3421
transmembrane electrochemical ion gradient of
around  200 mV. The c subunit of V
1
V
0
-ATPases
arose by duplication and fusion of the bacterial c
subunit, giving rise to a  16 kDa protein with two
hairpins [20]. Most important, the ion-binding site is
not conserved in hairpin 1. If one assumes the same

from the a naerobic, a cetogenic bacterium Acetob acterium
woodii is unique and encodes nine F
1
F
0
-like subunits
along with one gene encoding a V
1
V
0
-like subunit. The
atp operon has one homolog each of a gene encoding
the F
1
F
0
subunits a, b, c, d, e, a, and b, but it has
three genes encoding differently sized c subunits [22].
Subunits c
2
and c
3
have a molecular mass of 8.18 kDa,
are identical at the amino acid level, and are similar to
c subunits from F
1
F
0
-ATP synthases. Like other F
1

2 ⁄ 3
ratio may
change the function of the enzyme from an ATP syn-
thase to an ATPase [23].
These questions have not so far been addressed, due
to the lack of a purified, intact ATP synthase. Despite
the clear genetic evidence for different c subunits, as
well as for the presence of subunits a and b, they were
not detected in previous preparations of the enzyme
[24,25]. Later, separation of gently solubilized mem-
brane protein complexes by blue native PAGE and
N-terminal sequencing of polypeptides present in the
gel revealed both types of c subunit, as well as sub-
units a and b, in the ATPase complex [26]. These stud-
ies prompted us to revise our purification scheme, with
the aim of obtaining an intact ATP synthase from
A. woodii. We here present a protocol yielding a com-
plete Na
+
F
1
F
0
-ATP synthase complex, including the
entire membrane motor. Most important, both types
of c subunits were present. This made it possible to
readdress the longstanding question of whether an
ATPase with F
0
- and V

gel filtration was 590 kDa. As can be seen from Fig. 1,
the enzyme preparation contained 12 polypeptides.
The identity of the peptides was established using
MALDI-TOF or western blot analyses. These studies
revealed that the 58 kDa fragment corresponds to sub-
unit a, the 54 kDa fragment to subunit b, the 35 kDa
fragment to subunit c, the 19 kDa fragment to subunit
d, the 18 kDa fragment to a mixture of subunits a and
b, the 16 kDa fragment to subunit e, the 14 kDa frag-
ment to subunit c
1
, and the 10 kDa fragment to sub-
unit c
2 ⁄ 3
(Fig. 1A). The 42 kDa fragment reacted with
antibobies against c
2
⁄ c
3
(which also recognize c
1
) and
with antibodies against c
1
(which do not recognize
c
2 ⁄ 3
; Fig. 1B). These data demonstrate that the 42 kDa
fragment represents the SDS-resistant, heterooligomeric
c ring of the Na

Step
Protein
(mg)
Volume
(mL)
Activity
(U)
Activity
(U ⁄ mg)
Purification
(fold)
Yield
(%)
Membranes 430 50 214 0.6 1 100
Solubilizate 41 47 188 4.6 8.3 87
Concentrated
solubilizate
16 15 98 6.2 11 45
Gel filtration 6.9 10 67 9.7 16 30
Na
+
F
1
F
0
-ATP synthase from A. woodii M. Fritz and V. Mu
¨
ller
3422 FEBS Journal 274 (2007) 3421–3428 ª 2007 The Authors Journal compilation ª 2007 FEBS
Characterization of the Na

could be substituted by Li
+
. Furthermore, the
K
m
for Na
+
or Li
+
(0.5 mm, 2.0 mm) was compar-
able to the values determined before. As described
before, the stimulation by Na
+
was less pronounced
at more acidic pH values, indicating competition of
Na
+
and H
+
for a common binding site (data not
shown). Furthermore, inhibition by N¢,N¢-dicyclo-
hexylcarbodiimide was abolished in the presence of
Na
+
(Fig. 3). In summary, the biochemical parame-
ters of the complete preparation were indistinguish-
able from those of the preparation described
previously [25].
α
β

30 -
20 -
14 -
45 -
94 -
-66
-45
-30
-20
-14
Fig. 1. Subunit composition of the Na
+
F
1
F
0
-ATP synthase from A. woodii. Proteins were separated by SDS ⁄ PAGE and stained with SERVA
Blue G (Serva GmbH, Heidelberg, Germany) (A) or blotted against specific antibodies (B). Lane 1: molecular mass marker. Lane 2: ATP syn-
thase preparation was denatured by incubation at 80 °C for 10 min. Lane 3: ATP synthase was heated for 5 min at 120 °C prior to
SDS ⁄ PAGE to disrupt the c oligomer and blotted against c
1
antibodies. Lane 4: ATP synthase was incubated for 10 min at 80 °C, and blotted
against c
1
antibodies. Lane 5: the sample was incubated for 10 min at 80 °C and hybridized with antibody specific for the a subunit. Lane 6:
ATP synthase was incubated for 5 min at 120 °C and hybridized with antibodies against subunit c
2 ⁄ 3
(which also detect subunit c
1
). Lane 7:

3
4
5
1.0
-1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1/ATPase activity
B
Fig. 2. Ion dependence of ATP hydrolysis by
the Na
+
F
1
F
0
-ATP synthase from A. woodii.
ATPase activity was measured at 30 °C
using the assay described in Experimental
procedures. NaCl (A) or LiCl (B) was added
from stock solutions to the concentrations
indicated. The insert shows the same data

ATP hydrolysis as catalyzed by these proteoliposomes
was 2.8 UÆmg
)1
. To analyze whether the enzyme
was reconstituted in a functionally coupled state that
allows for ATP synthesis, the following experiments
were performed. In a fully coupled system, ATP
hydrolysis is accompanied by ion transport into the
proteoliposomes, and the membrane potential estab-
lished slows down or even inhibits further ion trans-
port and thus ATPase activity. This thermodynamic
control can be overcome by addition of ionophores.
After addition of the Na
+
ionophore N¢,N¢,N¢,N-tetra-
cyclohexyl-1,2-phenylenedioxydiacetamide to the proteo-
liposomes, ATP hydrolysis was stimulated seven-fold.
Stimulation, but to a lower extent, was also observed
with the protonophore tetrachlorosalicylanilide. These
experiments demonstrated coupling of ATP hydrolysis
to the generation of a membrane potential in our pro-
teoliposome system.
Next, we applied artificial driving forces to the pro-
teoliposomes. The general strategy is shown in Fig. 4.
When the proteoliposomes were loaded with 200 mm
NaCl and incubated in the presence of 200 mm KCl,
thus creating a sodium ion potential (DpNa), there was
no ATP synthesis (Fig. 5). Upon addition of 2 lm vali-
nomycin, a membrane potential (DY, inside positive)
was created in addition by the influx of K

0
-ATP synthase from A. woodii.
Discussion
The Na
+
F
1
F
0
-ATP synthase from the anaerobic, ace-
togenic bacterium A. woodii purified here contained all
the subunits deduced from the operon sequence. Most
importantly, it contained the membrane motor sub-
units a and b, which were absent in previous prepara-
tions. Because the ATP synthase preparations from the
close relatives Moorella thermoautotrophicum and Moo-
rella thermoacetica were also devoid of subunits a and
b, it was suggested in the literature that the ATP
synthases of acetogenic bacteria may be simpler in
architecture than other F
1
F
0
-ATP synthases [27,28].
DCCD (µM)
ATPase activity (%)
0 25 50 75 100
0
25
50

¨
ller
3424 FEBS Journal 274 (2007) 3421–3428 ª 2007 The Authors Journal compilation ª 2007 FEBS
However, this argument was difficult to understand, as
the genes encoding subunits a and b were embedded
in the atp operons of M. thermoacetica and A. woodii
[22,29], and one would have to envision a mechanism
that post-trancriptionally or post-translationally specif-
ically removes these subunits. Furthermore, these sub-
units are essential for motor function. From the results
presented here, it is evident that the critical step in
purification is the solubilization procedure. The Triton
X-100 used in previous studies apparently did not solu-
bilize the stator subunits a and b. Whether or not our
previous preparation contained both types of c subunit
is difficult to retrace. However, Fig. 3 in Reidlinger &
Mu
¨
ller [25] shows a faint band at about 17 kDa, which
appeared when the enzyme was autoclaved. At that
time, c
1
was unknown, and an N-terminal sequence of
this fragment was not obtained. Therefore, it is not
only not excluded that it was the c
1
subunit but likely,
as the c ring is rather stable, and the use of Triton X-
100 should not lead to removal of c
1

1
F
0
-ATP
synthase from Propionigenium modestum, DlNa
+
as
well as DY but not DpNa were sufficient as driving
forces [31–33]. Most importantly, the Na
+
F
1
F
0
-ATP
synthase from A. woodii was able to catalyze ATP
synthesis despite its different c subunits. Although the c
subunit stoichiometry has not yet been solved, at least
one copy of c
1
must have been present. Further discus-
sion of the coupling efficiencies has to await the deter-
mination of the rotor subunit stoichiometry.
In summary, we have solved a longstanding problem,
the purification of the Na
+
F
1
F
0

, and 420 mm sucrose
(pH 7.5) (buffer A). After addition of 1 g of lysozyme
(Sigma-Aldrich Chemie GmbH, Steinheim, Germany), the
suspension was incubated at 37 °C for 60 min. All sub-
sequent steps were carried out at 4 °C unless otherwise
indicated. The resulting protoplasts were collected by cen-
trifugation (16 000 g, 10 min, Beckman Avanti J25, JA14
rotor) and suspended in one volume of buffer A containing
a few crystals of DNase (Sigma-Aldrich Chemie GmbH)
and phenylmethanesulfonyl fluoride (final concentration
0.5 mm). The protoplasts were disrupted by two passages
through a French pressure cell at 42 MPa. Cell debris and
unbroken cells were removed by two sequential centrifuga-
tion steps at 6000 g for 15 min (Beckman Avanti J25, JA14
rotor). The supernatant was diluted with one volume of
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7

-ATP synthase-containing proteo-
liposomes. The artificial driving forces DlNa
+
(j), DY (m)orDpNa
(r) were applied to the proteoliposomes as described in Experi-
mental procedures. In one assay (d), a DlNa
+
was applied but
ADP was omitted.
M. Fritz and V. Mu
¨
ller Na
+
F
1
F
0
-ATP synthase from A. woodii
FEBS Journal 274 (2007) 3421–3428 ª 2007 The Authors Journal compilation ª 2007 FEBS 3425
50 mm Tris ⁄ HCl (pH 7.5), 10 mm MgCl
2
, and 17% (v ⁄ v)
glycerol (buffer B), and centrifuged at 130 000 g for 60 min
(Beckman L100K, 50.2 Ti rotor) to collect the membranes.
The membranes were washed once with buffer B, and
resuspended in 60 mL of buffer B.
Purification of ATP synthase
The membranes were diluted to 10 mgÆmL
)1
with buffer B,

base, 100 mm maleic acid, 5 mm MgCl
2
). The pH was adjus-
ted to 7.5 with KOH. The characterization of the enzyme
was performed at 30 °C by a discontinuous assay following
the ATP-dependent formation of inorganic orthophosphate,
according to the method of Heinonen & Lahti [38] as des-
cribed previously [39]. The assay contained 5 mm MgCI
2
when carried out at pH 7.5, and 50 mm MgCl
2
at pH 5.3.
For inhibitor studies, the samples were incubated with the
inhibitor for 30 min before the reaction was started by addi-
tion of ATP. N¢,N¢-Dicyclohexylcarbodiimide (Sigma-Ald-
rich Chemie GmbH) was added as an ethanolic solution,
and controls received solvents only.
Western blot analysis
After separation by SDS ⁄ PAGE, the ATP synthase sub-
units were blotted onto a nitrocellulose membrane as des-
cribed previously [40]. Western blot ECL detection
reagents were either purchased from PerkinElmer Life Sci-
ences (Boston, MA, USA) or made in-house [solution A
(200 mL containing 0.1 m Tris ⁄ HCl, pH 6.8, 50 mg of
luminol), and solution B (10 mL of dimethylsulfoxide con-
taining 11 mg of p-hydroxycoumaric acid)]. Blot mem-
branes were incubated in a mixture of 4 mL of solution
A, 400 lL of solution B and 1.2 lLofH
2
O

)1
), indicating almost complete incor-
poration into the proteoliposomes. ATP synthesis was
determined via a standard luciferin ⁄ luciferase assay, monit-
oring the emitted light with a chemiluminometer (Lumac,
AC Landgraaf, The Netherlands).
To generate aDlNa
+
, the proteoliposomes were first
loaded with Na
+
to create a DpNa. The vesicles were incu-
bated in buffer D containing, in addition, 200 mm NaCl
for 12 h at 4 °C. After this, the Na
+
-loaded vesicles were
collected by gravity-driven gel filtration using a 10 mL pip-
ette filled with Sephadex 25 matrix and equilibrated with
buffer D containing, in addition, 200 mm KCl and 5 mm
KH
2
PO
4
. The synthesis reactions were carried out at 30 °C
with 2 mL of proteoliposome solution from gel filtration
and by adding 5 mm ADP. The synthesis reaction was star-
ted by addition of 2 mL of valinomycin (Sigma-Aldrich
Chemie GmbH) to induce a DY. Samples (10 mL) were
withdrawn every 30 s and immediately added to 250 mL of
an ATP determination buffer (5 mm NaHAsO

tion of 2 mL of valinomycin to induce a DY.
Furthermore, when only a DpNa was to be applied, the
vesicles were incubated in buffer D containing, in addition,
200 mm NaCl for 12 h at 4 °C. After this, the Na
+
-loaded
vesicles were collected by gravity-driven gel filtration using
a 10 mL pipette filled with Sephadex 25 matrix and equili-
brated with buffer D containing, in addition, 200 mm KCl
and 5 mm KH
2
PO
4
. The synthesis reactions were carried
out at 30 °C with 2 mL of proteoliposome solution from
gel filtration, and started by addition of 5 mm ADP.
MALDI-TOF analysis
Proteins were separated by SDS ⁄ PAGE, and all bands vis-
ible by Coomassie staining were entirely cut out and subjec-
ted to in-gel digestion protocols [42,43], which were adapted
for use on a Microlab Star digestion robot (Bonaduz,
Switzerland). Samples were reduced, alkylated and
subsequently digested overnight using bovine trypsin
(sequencing grade; Roche, Mannheim, Germany). The gel
pieces were extracted, and the extracts were dried in a
vacuum centrifuge and stored at ) 20 °C until use ⁄ analysis.
MALDI-TOF MS experiments were performed on an
Ultraflex TOF ⁄ TOF mass spectrometer (Bruker Daltonics
Inc., Billerica, MA, USA). The samples were prepared as
described previously [44]. Spectra were externally calibrated

3 Boekema EJ, van Breemen JF, Brisson A, Ubbink-Kok
T, Konings WN & Lolkema JS (1999) Connecting stalks
in V-type ATPase. Nature 401, 37–38.
4 Karrasch S & Walker JE (1999) Novel features in the
structure of bovine ATP synthase. J Mol Biol 290, 379–
384.
5 Coskun U
¨
, Chaban YL, Lingl A, Mu
¨
ller V, Keegstra
W, Boekema EJ & Gru
¨
ber G (2004) Structure and
subunit arrangement of the A-type ATP synthase com-
plex from the archaeon Methanococcus jannaschii visu-
alized by electron microscopy. J Biol Chem 279,
38644–38648.
6 Forgac M (2000) Structure, mechanism and regulation
of the clathrin-coated vesicle and yeast vacuolar H
+
-
ATPases. J Exp Biol 203, 71–80.
7 Cross RL (1981) The mechanism and regulation of ATP
synthesis by F
1
-ATPases. Annu Rev Biochem 50, 681–714.
8 Fillingame RH (1992) H
+
transport and coupling by

F
1
): direct observation. Science 286,
1722–1724.
14 Hirata T, Nakamura N, Omote H, Wada Y & Futai M
(2000) Regulation and reversibility of vacuolar H
+
-
ATPase. J Biol Chem 275, 386–389.
15 Fillingame RH (1992) Subunit c of F
1
F
0
ATP synthase )
structure and role in transmembrane energy transduc-
tion. Biochim Biophys Acta 1101, 240–243.
16 Seelert H, Dencher NA & Mu
¨
ller DJ (2003) Fourteen
protomers compose the oligomer III of the proton-rotor
in spinach chloroplast ATP synthase. J Mol Biol 333,
337–344.
M. Fritz and V. Mu
¨
ller Na
+
F
1
F
0

21 Forgac M (1999) Structure and properties of the vacuo-
lar H
+
-ATPases. J Biol Chem 274, 12951–12954.
22 Rahlfs S, Aufurth S & Mu
¨
ller V (1999) The Na
+
-F
1
F
0
-
ATPase operon from Acetobacterium woodii. Operon
structure and presence of multiple copies of atpE which
encode proteolipids of 8- and 18-kDa. J Biol Chem 274,
33999–34004.
23 Mu
¨
ller V, Aufurth S & Rahlfs S (2001) The Na
+
cycle
in Acetobacterium woodii: identification and characteri-
zation of a Na
+
translocating F
1
F
0
-ATPase with a

¨
ller V (2000) Identification
of subunits a, b, and c
1
from Acetobacterium woodii
Na
+
-F
1
F
0
-ATPase. Subunits c
1
, c
2
, and c
3
constitute a
mixed c-oligomer. J Biol Chem 275, 33297–33301.
27 Das A, Ivey DM & Ljungdahl LG (1997) Purification
and reconstitution into proteoliposomes of the F
1
F
0
ATP synthase from the obligately anaerobic gram-
positive bacterium Clostridium thermoautotrophicum.
J Bacteriol 179, 1714–1720.
28 Ivey DM & Ljungdahl LG (1986) Purification and char-
acterization of the F
1

membrane voltage. EMBO J 18, 4118–4127.
33 Dmitriev O, Deckershebestreit G & Altendorf K
(1993) ATP synthesis energized by DpNa and DY in
proteoliposomes containing the F
0
F
1
-ATPase from
Propionigenium modestum. J Biol Chem 268, 14776–
14780.
34 Heise R, Mu
¨
ller V & Gottschalk G (1993) Acetogenesis
and ATP synthesis in Acetobacterium woodii are coupled
via a transmembrane primary sodium ion gradient.
FEMS Microbiol Lett 112, 261–268.
35 Scha
¨
gger H & von Jagow G (1987) Tricine-sodium
dodecylsulfate-polyacrylamide gel electrophoresis for the
separation of proteins in the range from 1 to 100 kDa.
Anal Biochem 166, 369–379.
36 Weber K & Osborne M (1969) The reliability of the
molecular weight determination by dodecyl sulfate poly-
acrylamide gel electrophoresis. J Biol Chem 244, 4406–
4412.
37 Blum H, Beier H & Gross HJ (1987) Improved silver
staining of plant proteins, RNA, and DNA in poly-
acrylamide gels. Electrophoresis 8, 93–98.
38 Heinonen JE & Lahti RJ (1981) A new and convenient

Na
+
F
1
F
0
-ATP synthase from A. woodii M. Fritz and V. Mu
¨
ller
3428 FEBS Journal 274 (2007) 3421–3428 ª 2007 The Authors Journal compilation ª 2007 FEBS