The proximity between C-termini of dimeric vacuolar
H
+
-pyrophosphatase determined using atomic force
microscopy and a gold nanoparticle technique
Tseng-Huang Liu
1
, Shen-Hsing Hsu
1
, Yun-Tzu Huang
1
, Shih-Ming Lin
1
, Tsu-Wei Huang
2
,
Tzu-Han Chuang
3
, Shih-Kang Fan
4
, Chien-Chung Fu
3
, Fan-Gang Tseng
2,
* and Rong-Long Pan
1,
*
1 Department of Life Sciences and Institute of Bioinformatics and Structural Biology, College of Life Sciences, National Tsing Hua Univer-
sity, Hsin Chu, Taiwan, ROC
2 Department of Engineering and System Science, National Tsing Hua University, Hsin Chu, Taiwan, ROC
3 Institute of NanoEngineering and MicroSystems, National Tsing Hua University, Hsin Chu, Taiwan, ROC

Taiwan, ROC
Fax: +886 3 5742688
Tel: +886 3 5742685
E-mail: [email protected]
*These authors contributed equally to this
work
(Received 1 March 2009, revised 17 May
2009, accepted 10 June 2009)
doi:10.1111/j.1742-4658.2009.07146.x
Vacuolar H
+
-translocating inorganic pyrophosphatase [vacuolar H
+
-pyro-
phosphatase (V-PPase); EC 3.6.1.1] is a homodimeric proton translocase; it
plays a pivotal role in electrogenic translocation of protons from the cyto-
sol to the vacuolar lumen, at the expense of PP
i
hydrolysis, for the storage
of ions, sugars, and other metabolites. Dimerization of V-PPase is neces-
sary for full proton translocation function, although the structural details
of V-PPase within the vacuolar membrane remain uncertain. The C-termi-
nus presumably plays a crucial role in sustaining enzymatic and proton-
translocating reactions. We used atomic force microscopy to visualize
V-PPases embedded in an artificial lipid bilayer under physiological condi-
tions. V-PPases were randomly distributed in reconstituted lipid bilayers;
approximately 43.3% of the V-PPase protrusions faced the cytosol, and
56.7% faced the vacuolar lumen. The mean height and width of the cyto-
solic V-PPase protrusions were 2.8 ± 0.3 nm and 26.3 ± 4.7 nm, whereas
those of the luminal protrusions were 1.2 ± 0.1 nm and 21.7 ± 3.6 nm,

above-mentioned ions, as well as proton translocation.
Truncation of the C-terminus induces a dramatic
decline in V-PPase enzymatic activity, proton translo-
cation, and coupling efficiency [9]. In addition, deletion
of the C-terminus of V-PPase increases its susceptibil-
ity to heat stress and substantially increases the appar-
ent K
+
binding constant. It is thus likely that the
C-terminus plays an essential role in sustaining the
physiological functions of V-PPase.
Interactions between the subunits of the V-PPase
dimer have been studied [1,2,10–12]. Radiation inac-
tivation analysis demonstrated that the proper
dimeric structure of V-PPase on tonoplastic mem-
branes is a prerequisite for both enzymatic activity
and PP
i
-supported proton translocation [2,11,12].
Further target size measurements revealed that only
one subunit of the purified dimeric complex was suf-
ficient for the enzymatic reaction of V-PPase,
although proton translocation requires the presence
of both subunits [2]. Moreover, high hydrostatic
pressure was employed to inhibit V-PPase through
subunit dissociation of the enzyme, resulting in inac-
tive forms [10]. The physiological substrate and sub-
strate analogs enhance the high-pressure inhibition of
V-PPase, indicating the vulnerability of the subunit–
subunit interaction [10]. The above lines of evidence

located at the interface of subunits.
Results and discussion
AFM analysis of purified V-PPase adsorbed onto
mica
Recombinant DNAs for overexpression of V-PPases
containing a 6·His tag at either the C-terminus or
N-terminus were prepared and transformed into a
yeast host. Recombinant V-PPase containing a 6·His
tag at the C-terminus (Fig. 1C) was overexpressed in
yeast and successfully purified from microsomes.
Unfortunately, V-PPase containing a 6·His tag at
the N-terminus was poorly expressed in yeast and
was therefore excluded from the study (data not
shown). SDS/PAGE analysis of the purified C-termi-
nal 6·His-tagged V-PPase followed by Coomassie
Blue staining or western blotting showed that it was
highly purified, comprising a single major band with
a molecular mass of 73 kDa (Fig. 1A), as expected
from the known structure of the V-PPase monomer
[1,2,10]. During size exclusion chromatography,
V-PPase was eluted with an apparent molecular mass
of  145 kDa, similar to its native form and in
agreement with previous studies suggesting a dimeric
conformation [2,10–12].
The purified V-PPase was then reconstituted into
liposomes by a detergent removal method using Bio-
Rad SM-2 beads combined with freeze–thaw sonica-
tion [13]. On addition of PP
i
to the proteoliposome

between the polar surface of the protein molecules
and the charged surface of the mica [24]. These
images represent the first direct nanoscale observation
of V-PPase.
Determination of molecular volume provides the
stoichiometry of subunit components for functional
enzymes [24]. In this study, the volume of V-PPase
was calculated using Eqn (1) and determined to be
302.4 ± 40.6 nm
3
(V
s
)(n = 21), which was slightly
larger than the theoretical value (V
prot
; 274.5 nm
3
)of
the protein (Table 1). This slight overestimation in
volume probably arose from the broadening effect of
the AFM tip [24]. It is also likely that variations in
volume measurements might arise from distinct inter-
actions of the tip with the individual purified
V-PPase particles [25]. Nonetheless, these results
unambiguously demonstrate the feasibility of this
technique for nanoscale investigation of purified
V-PPase molecules.
AFM analysis of V-PPase reconstituted into
liposomes
The homodimeric structure of V-PPase in a planar

tuted V-PPase. Molecular mass (kDa) mark-
ers are indicated on the left. (B) PP
i
-
associated proton translocation of reconsti-
tuted V-PPase. Proton transport was initi-
ated by adding 1.0 m
M PP
i
. At the end of
each reaction, 5 lgÆmL
)1
gramicidin D was
added to stop the fluorescence quenching
of acridine orange. (C) Topological model of
V-PPase. Cylinders 1–16 indicate mem-
brane-spanning domains.
T H. Liu et al. Adjacent C-termini of dimeric H
+
-pyrophosphatase
FEBS Journal 276 (2009) 4381–4394 ª 2009 The Authors Journal compilation ª 2009 FEBS 4383
from a cross-section of the lipid bilayer. The bilayer
thickness was consistent with previous AFM mea-
surements of a lipid bilayer composed of a phospha-
tidylcholine/cholesterol mixture and prepared in a
similar aqueous environment [26]. V-PPases reconsti-
tuted into the lipid bilayer protruded from the
bilayer surface in a diffuse pattern with a random
distribution. The V-PPase images fell within two cat-
egories according to the extramembranous protrusion

standing this, current AFM techniques suffice to
provide unambiguous images of the dimeric structure
of V-PPase. Four representative examples exhibiting
minor variations are shown in Fig. 4A. The small
differences in topography of the individual V-PPases in
A
B
D
C
Fig. 2. AFM analysis of purified V-PPase. (A) Three-dimensional
AFM image of purified V-PPase adsorbed onto mica. (B) Profile of
peak heights along the cross-section shown in (A). Purified V-PPase
protrudes 1.6 ± 0.4 nm (n = 21) from the mica surface. (C) Histo-
gram of V-PPase height determined using the AFM image in (A).
(D) Histogram of V-PPase width determined using the AFM image
in (A).
Adjacent C-termini of dimeric H
+
-pyrophosphatase T H. Liu et al.
4384 FEBS Journal 276 (2009) 4381–4394 ª 2009 The Authors Journal compilation ª 2009 FEBS
the reconstituted lipid bilayer have probably resulted
from contact with the AFM tip during scanning.
Nevertheless, these AFM images are adequate for
nanoscale resolution of the structural details of
V-PPase [27,28]. Moreover, the resolution of the
images from V-PPases in reconstituted membranes was
Table 1. Dimensions of free and membrane-bound recombinant V-PPase determined by AFM. Values represent means ± SD. n = number
of observations. Observed and predicted volumes were determined from AFM analysis using Eqn (1) and from theoretical analysis using
Eqn (2).
Protein (145 kDa) Height (nm) Width (nm)

measured as 332.9 ± 46.9 nm
3
(n = 17). The V
prot
of
a V-PPase homodimer with a molecular mass of
145 kDa, calculated on the basis of the amino acid
composition, was determined to be 274.5 nm
3
[29].
This theoretical volume correlates very well with that
measured from the AFM images. Note that these
images were obtained by AFM scanning in a fluid,
and therefore probably provide an authentic illustra-
tion of V-PPase structure under physiological condi-
tions. The AFM images indicate the dimeric structure
of V-PPase reconstituted in a lipid bilayer. This study
provides the first 3D representation of individual
V-PPases protruding from the cytosolic and luminal
sides of a membrane in aqueous solution.
Proximity of V-PPase C-termini in reconstituted
membranes
Topology studies examining heterologous V-PPase
expression in yeast suggested that both the C-termini
and the N-termini of each subunit are located on the
lumen side and are opposite the catalytic domain on
the cytosolic side of the vesicular membrane [1].
Because V-PPase is homodimeric, there are two possi-
ble configurations for association of the two subunits;
the C-termini of both subunits may protrude from the

antibody could bind to V-PPase on either one or two
molecules (Fig. 5B). Clearly, Fig. 5B2 depicts that two
antibodies bind respectively to a single V-PPase mole-
cule in close vicinity. AFM image analysis using spip
software was used to generate histograms delineating
the distribution of protrusion heights and widths
(Fig. 5C,D), and this revealed three major groups of
protrusions: (a) lower peaks (peak 1; 1.4 ± 0.2 nm
mean height, n = 10) for structures of V-PPase on the
lumen side of the membrane lacking bound antibody;
(b) intermediate peaks (peak 2; 2.9 ± 0.2 nm mean
height, n = 20) for those on the cytosolic side of the
membrane; and (c) higher peaks (peak 3; 4.2 ± 0.3 nm
mean height, n = 10) for antibodies bound presumably
to the lumen side. The ratio of the sum of integrals for
peak 1 and peak 3 (free lumen side and antibodies bind-
ing to the lumen side) to peak 2 (cytosolic side) is con-
sistent with our prior results (approximately 5.6 : 4.4).
Fig. 5. AFM analysis of V-PPase in a reconstituted lipid bilayer immunolabeled with an antibody against His to detect the C-terminal 6·His
tag of V-PPase. (A) Image of a large section of immunolabeled lipid bilayer reconstituted in the presence of V-PPase. (B) High-resolution
images of immunolabeled protrusions in (A). (1) Protrusion showing a single antibody bound to V-PPase. (2) Protrusion showing two
antibodies bound to V-PPase. (C) Histogram of protrusion height determined using the AFM image in (A). (D) Histogram of protrusion width
determined using the AFM image in (A).
T H. Liu et al. Adjacent C-termini of dimeric H
+
-pyrophosphatase
FEBS Journal 276 (2009) 4381–4394 ª 2009 The Authors Journal compilation ª 2009 FEBS 4387
Previous AFM imaging studies have demonstrated
that the height of a single IgG molecule is
2.4 ± 0.1 nm [24]. Taking this value into account, the

2+
–nitrilotriacetic acid GNPs were
bound to the 6·His tags of V-PPase C-termini recon-
stituted in lipid bilayers in aqueous solution, resulting
in two major types of protrusion as observed with
AFM: the cytosolic side of V-PPase, and the particles
bound to the lumen side of V-PPase, respectively
(Fig. 6). The solid circle in Fig. 6B indicates GNP
bound to V-PPase C-terminus protruding from the
surface of the lipid bilayer, whereas the dotted circle,
V-PPase protrusion at the lumen side lacking bound
GNP (Fig. 6B). More than 70% of V-PPases were
covered by GNPs on the luminal side (data not
shown). The height distribution histogram indicated
that the heights of the lower V-PPase protrusions
(peak 1) were consistent with those of its cytosolic por-
tions, whereas the heights of the higher ones (peak 2)
represented those of the GNPs bound to the C-termini
of the enzyme (Fig. 6C). The height of the latter
protrusions (4.9 ± 0.1 nm) reflects the sum of Ni
2+
–
nitrilotriacetic acid GNP heights (mean height =
2.0 ± 0.1 nm, n = 16) and V-PPase heights on the
luminal side (mean height = 1.2 ± 0.1 nm, n = 20).
In contrast, the lower V-PPase protrusions represent
those of its cytosolic sides alone. Moreover, in the
width distribution histogram, the higher peaks (peak
1) represent either cytosolic protrusions of V-PPase
lacking GNPs (mean width = 26.3 ± 4.7 nm, n = 17)

mini of purified V-PPase. The TEM image displays the
bound GNPs as solid spheres with a diameter of
2.0 ± 0.2 nm (n = 18) (Fig. 7A). In addition, GNPs
bound to V-PPase C-termini occurred in pairs (Fig. 7B),
indicating the dimeric structure of the enzyme. The his-
togram showing the distribution of distances between
GNP pairs observed from the TEM image yields a mean
distance of 1.9 ± 0.8 nm (Fig. 7C), concurring with the
result generated by AFM analysis of GNP-labeled
V-PPase (Fig. 6E; distance = 2.2 ± 1.4 nm). The slight
fluctuation in distances between GNP pairs most likely
arose from the flexibility of the V-PPase C-termini. For
instance, the shorter distance observed indicates two
closed GNPs on the C-termini of the enzyme. In con-
trast, the longer distance indicates a probable extension
of the C-termini of V-PPase. Verification of these possi-
bilities requires further investigation.
The C-terminus of V-PPase has been determined to
be relatively conserved among various plant V-PPases,
Adjacent C-termini of dimeric H
+
-pyrophosphatase T H. Liu et al.
4388 FEBS Journal 276 (2009) 4381–4394 ª 2009 The Authors Journal compilation ª 2009 FEBS
and is presumed to be proximal to the catalytic site
[32]. In addition, the importance of the V-PPase C-
terminus in sustaining enzymatic and proton-translo-
cating function and for indirect regulation of K
+
binding has been demonstrated [9]. Moreover, inter-
subunit interactions of V-PPase are critical for proper

the protrusion heights determined using the
AFM image in (A). (D) Histogram of the pro-
trusion widths determined using the AFM
image in (A). (E) Simulation of potential
V-PPase protrusion widths based on dis-
tances between GNP pairs bound to the
V-PPase C-termini. Solid rhombus, predicted
protrusion width based on a single GNP
molecule bound to the C-terminus of
V-PPase; solid circle, predicted protrusion
widths based on the distance between two
GNP molecules bound to V-PPase C-termini;
solid triangle, actual protrusion width of
GNP-bound V-PPase determined by AFM.
Data represent the mean ± SD.
T H. Liu et al. Adjacent C-termini of dimeric H
+
-pyrophosphatase
FEBS Journal 276 (2009) 4381–4394 ª 2009 The Authors Journal compilation ª 2009 FEBS 4389
Experimental procedures
Cloning, expression, and purification
The mung bean (Vigna radiata L.) V-PPase cDNA (VPP;
accession number P21616) [33] was cloned into the yeast
expression vector pYES2 (Invitrogen, Carlsbad, CA, USA),
and the two synthesized oligonucleotides P
his
(5¢-CCTCG
AGCCATCATCATCATCATCATTAGGGCCGCATCAT
GTAATTAGTTATGT-3¢) and P
MluI

supernatant was incubated with Ni
2+
–nitrilotriacetic acid
beads prewashed with the extraction buffer for 1 h. The
Ni
2+
–nitrilotriacetic acid beads were injected into the
empty column and eluted at a flow rate of 0.5 mLÆmin
)1
with the elution medium [10 mm Tris/HCl (pH 8.0), 15%
(w/v) glycerol, 10 mm b-mercaptoethanol, 1 mm phen-
ylmethanesulfonyl fluoride, 0.1% (w/v) DDM] with a step
gradient of 20, 40, 60 and 250 mm imidazole, respectively.
The fractions with highest PP
i
hydrolysis activity at
250 mm imidazole were pooled and dialyzed against med-
ium containing 10 mm Tris/HCl (pH 8.0), 15% (w/v) glyc-
erol, and 0.1% (w/v) Triton X-100, and then stored at
)70 °C for further studies.
Fig. 7. TEM analysis of Ni
2+
–nitrilotriacetic acid GNP-labeled
V-PPase. (A) TEM image of Ni
2+
–nitrilotriacetic acid GNP-labeled
purified V-PPase. (B) A gallery of zoomed images for the GNP pairs
labelled on purified V-PPases. (C) Histogram of the distances
between GNP pairs determined using the TEM image in (A).
Fig. 8. A working model of V-PPase. The distance between

was determined spectrophotometrically as described previ-
ously [36,37]. The protein concentration was calculated by a
modified Bradford method with BSA as the standard [38].
Measurement of proton translocation
Proton translocation was measured as the initial rate of flu-
orescence quenching of acridine orange (excitation wave-
length, 495 nm, emission wavelength, 530 nm) as described
previously [34,39–41]. The reaction mixture for proton
translocation contained 5 mm Tris/HCl (pH 8.0), 1 mm
EGTA/Tris (pH 7.6), 400 mm glycerol, 100 mm KCl,
1.3 mm MgSO
4
,5lm acridine orange, and 100 lgÆ mL
)1
microsomes. The reaction was initiated by adding 1 mm
sodium pyrophosphate (pH 7.6). The initial rate of fluores-
cence quenching was calculated as the proton transport
activity [34,39–41]. The ionophore, gramicidin D
(5 lgÆmL
)1
), was then included at the end of each assay to
confirm the integrity of the membrane.
SDS/PAGE and western analysis
SDS/PAGE was performed according to Laemmli [42].
Denatured proteins were subjected to SDS/PAGE on a
Phast System (Pharmacia, Uppsala, Sweden) with a 12.5%
(w/v) polyacrylamide PhastGel. The gels were stained with
Coomassie Blue or electrotransferred to a poly(vinylidene
difluoride) membrane by using semidry electrotransblotting
apparatus (Nova Blot, Amersham Pharmacia Biotech,

filtration through a paper filter. The mixture was applied to a
Sephadex G-50 column equilibrated with 0.25 m sorbitol,
10 mm Tricine-Na (pH 7.5), 1 mm MgSO
4
, 0.1 mm EGTA
and 2 mm dithiothreitol to remove NaCl and glycerol. The
proteoliposome fraction was then frozen in liquid nitrogen,
thawed on ice, and sonicated for 20 s at 4 °C in a bath-type
sonicator. The proteoliposomes thus obtained were immedi-
ately used for measurement of proton-translocating activity.
Immunofluorescence microscopy
V-PPase proteoliposomes were prepared on glass coverslips
and then fused into the lipid bilayer. After being washed with
NaCl/P
i
, the coverslips were placed in blocking solution
[NaCl/P
i
containing 3% (w/v) BSA] for 30 min at room tem-
perature. Samples were then incubated with mouse monoclo-
nal antibody against the C-terminal 6·His tag of V-PPase in
NaCl/P
i
(1 lgÆmL
)1
) for 2 h at room temperature. After
being rinsed with NaCl/P
i
, samples were subsequently incu-
bated with carboxymethylindocyanine 3-coupled goat anti-

T H. Liu et al. Adjacent C-termini of dimeric H
+
-pyrophosphatase
FEBS Journal 276 (2009) 4381–4394 ª 2009 The Authors Journal compilation ª 2009 FEBS 4391
Nanogold solution (Nanoprobes, NY, USA), washed twice
with NaCl/P
i
, and then subjected to AFM imaging.
AFM
AFM was performed in the contact mode using a Nano-
scope IIIa Multimode atomic force microscope equipped
with an E-type scanner (Digital Instruments, Santa Bar-
bara, CA, USA) and a Picoplus instrument (Molecular
Imaging, MI, USA). Muscovite mica (Electron Microscopy
Sciences, Hatfield, PA, USA) was freshly cleaved and
immediately covered with adsorption buffer [10 mm Tris/
HCl (pH 7.8), 300 mm KCl, 25 mm MgCl
2
]. Subsequently,
10 lgÆmL
)1
of protein solution was dropped onto the mica
surface. After 1 h, the sample was rinsed with imaging buf-
fer [10 mm Tris/HCl (pH 7.8), 150 mm KCl]. V-shaped sili-
con nitride (Si
3
N
4
) cantilevers with spring constants of
0.08 NÆm

[29]:
V
s
¼ðph=6Þð3r
2
þ h
2
Þð1Þ
where V
s
is the molecular volume, and h and r are the
height and the radius (half of the measured width) of the
protein, respectively.
In addition, molecular volume based on molecular mass
was calculated using the Eqn (2):
M
o
¼ðN
A
V
prot
Þ=ðV
1
þ dV
2
Þð2Þ
where N
A
is the Avogadro constant (6.022 · 10
23

tion transmission electron microscope (Fa. Philips,
Eindhoven, the Netherlands) operating at 200 keV.
Acknowledgements
This work was supported by grants from National
Science Council, Republic of China: NSC 96-2627-
M-007-003 and NSC 97-2627-M-007-003 to R. L. Pan,
NSC 96-2627-M-009-001 to S. K. Fan, NSC 96-2627-
M-007-004 to C. C. Fu, and NSC 96-2627-M-007-003
to F. G. Tseng.
References
1 Maeshima M (2000) Vacuolar H
+
-pyrophosphatase.
Biochim Biophys Acta 1465, 37–51.
2 Tzeng CM, Yang CY, Yang SJ, Jiang SS, Kuo SY,
Hung SS, Ma JT & Pan RL (1996) Subunit
structure of vacuolar proton pyrophosphatase as
determined by radiation inactivation. Biochem J 316,
143–147.
3 Drozdowicz YM & Rea PA (2001) Vacuolar proton-
pyrophosphatases: from the evolutionary backwaters
into the mainstream. Trends Plant Sci 6, 206–211.
4 Rea PA, Kim Y, Sarafian V, Poole RJ, Davies JM &
Sanders D (1992) Vacuolar H
+
-translocating pyrophos-
phatases: a new category of ion translocase. Trends
Biochem Sci 17, 348–353.
5 Gordon-Weeks R, Steele SH & Leigh RA (1996) The
role of magnesium, pyrophosphate, and their complexes

H
+
-pyrophosphatase. A simple three-state model.
FEBS Lett 327, 199–202.
9 Lin HH, Pan YJ, Hsu SH, Van RC, Hsiao YY, Chen
JH & Pan RL (2005) Deletion mutation analysis on
C-terminal domain of plant vacuolar H
+
-pyrophospha-
tase. Arch Biochem Biophys 442, 206–213.
10 Yang SJ, Ko SJ, Tsai YR, Jiang SS, Kuo SY, Hung
SH & Pan RL (1998) Subunit interaction of vacuolar
H
+
-pyrophosphatase as determined by high hydrostatic
pressure. Biochem J 331, 395–402.
11 Maeshima M (1990) Oligomeric structure of H
+
-trans-
locating inorganic pyrophosphatase of plant vacuoles.
Biochem Biophys Res Commun 168, 1157–1162.
12 Sarafian V, Potier M & Poole RJ (1992) Radiation-
inactivation analysis of vacuolar H
+
-ATPase and H
+
-
pyrophosphatase from Beta vulgaris L. Functional sizes
for substrate hydrolysis and for H+ transport. Biochem
J 283, 493–497.

choline and phosphatidylcholine/cholesterol bilayers.
Biophys J 81, 1059–1069.
18 Slade A, Luh J, Ho S & Yip CM (2002) Single molecule
imaging of supported planar lipid bilayer-reconstituted
human insulin receptors by in situ scanning probe
microscopy. J Struct Biol 137, 283–291.
19 Saslowsky DE, Lawrence J, Ren X, Brown DA,
Henderson RM & Edwardson JM (2002) Placental
alkaline phosphatase is efficiently targeted to rafts in
supported lipid bilayers. J Biol Chem 277, 26966–26970.
20 Milhiet PE, Giocondi MC, Baghdadi O, Ronzon F,
Roux B & LeGrimellec C (2002) Spontaneous insertion
and partitioning of alkaline phosphatase into model
lipid rafts. EMBO Rep 3, 485–490.
21 Liu LN, Aartsma TJ & Frese RN (2008) Dimers of
light-harvesting complex 2 from Rhodobacter sphaero-
ides characterized in reconstituted 2D crystals with
atomic force microscopy. FEBS J 275, 3157–3166.
22 Bahatyrova S, Frese RN, Siebert CA, Olsen JD, van
Der Werf KO, van Grondelle R, Niederman RA,
Bullough PA, Otto C & Hunter CN (2004) The native
architecture of a photosynthetic membrane. Nature 430,
1058–1062.
23 Fotiadis D, Liang Y, Filipek S, Saperstein DA, Engel
A & Palczewski K (2003) Atomic-force microscopy:
rhodopsin dimers in native disc membranes. Nature
421, 127–128.
24 Schneider SW, La
¨
rmer J, Henderson RM & Oberleith-

(1995) Immunolocalization of lamins and nuclear pore
complex proteins by atomic force microscopy. Pflu
¨
gers
Arch 430, 795–801.
30 Rapp M, Seppa
¨
la
¨
S, Granseth E & von Heijne G (2007)
Emulating membrane protein evolution by rational
design. Science 315, 1282–1284.
31 Brekke OH, Michaelsen TE & Sandlie I (1995) The
structural requirements for complement activation by
IgG: does it hinge on the hinge? Immunol Today 16,
85–90.
32 Takasu A, Nakanishi Y, Yamauchi T & Maeshima M
(1997) Analysis of the substrate binding site and car-
boxyl terminal region of vacuolar H
+
-pyrophosphatase
of mung bean with peptide antibodies. J Biochem 122,
883–889.
33 Hung SH, Chiu SJ, Lin LY & Pan RL (1995) A cDNA
clone encoding the A subunit of the vacuolar H
+
-AT-
Pase from etiolated mung bean seedlings. Plant Physiol
109, 1125–1127.
34 Hsiao YY, Van RC, Hung SH, Lin HH & Pan RL

In Methods in Enzymology, vol. 55 (Fleischer S &
Packer L eds), pp 547–569. Academic Press, New York.
41 Clerc S & Barenholz Y (1998) A quantitative
model for using acridine orange as a
transmembrane pH gradient probe. Anal Biochem
259, 104–111.
42 Laemmli UK (1970) Cleavage of structural proteins
during the assembly of the head of bacteriophage T4.
Nature (Lond) 222 , 680–685.
43 Lin H, Zhu YJ & Lal R (1999) Amyloid beta protein
(1–40) forms calcium-permeable Zn
2+
-sensitive
channel in reconstituted lipid vesicles. Biochemistry 38,
11189–11196.
44 Edstrom RD, Meinke MH, Yang R, Yang X, Elings V
& Evans DF (1990) Direct visualization of phosphory-
lase–phosphorylase kinase complexes by scanning
tunneling and atomic force microscopy. Biophys J 58,
1437–1448.
Adjacent C-termini of dimeric H
+
-pyrophosphatase T H. Liu et al.
4394 FEBS Journal 276 (2009) 4381–4394 ª 2009 The Authors Journal compilation ª 2009 FEBS