Agrobacterium tumefaciens type II NADH dehydrogenase
Characterization and interactions with bacterial and thylakoid
membranes
Laetitia Bernard*
,‡
, Carine Desplats
‡
, Florence Mus
†
, Ste
´
phan Cuine
´
, Laurent Cournac and
Gilles Peltier
CEA Cadarache, Direction des Sciences du Vivant, De
´
partement d’Ecophysiologie Ve
´
ge
´
tale et Microbiologie des Bacte
´
ries et Microalgues,
UMR 6191 CNRS-CEA, Aix-Marseille II, Saint-Paul-lez-Durance, France
Electrons that enter the respiratory chain can originate
from three different types of NADH dehydrogenase.
Complex I (NDH-1), a multisubunit transmembrane
enzyme coupling quinone reduction to proton translo-
cation, is present in bacteria, as well as in plant, fungal
and mammal mitochondria [1]. Multisubunit sodium-

Note
‡These authors contributed equally to this
study
(Received 4 April 2006, revised 16 May
2006, accepted 9 June 2006)
doi:10.1111/j.1742-4658.2006.05370.x
Type II NADH dehydrogenases (NDH-2) are monomeric enzymes that cat-
alyse quinone reduction and allow electrons to enter the respiratory chain
in different organisms including higher plant mitochondria, bacteria and
yeasts. In this study, an Agrobacterium tumefaciens gene encoding a puta-
tive alternative NADH dehydrogenase (AtuNDH-2) was isolated and
expressed in Escherichia coli as a (His)
6
-tagged protein. The purified
46 kDa protein contains FAD as a prosthetic group and oxidizes both
NADH and NADPH with similar V
max
values, but with a much higher
affinity for NADH than for NADPH. AtuNDH-2 complements the growth
(on a minimal medium) of an E. coli mutant strain deficient in both
NDH-1 and NDH-2, and is shown to supply electrons to the respiratory
chain when incubated with bacterial membranes prepared from this
mutant. By measuring photosystem II chlorophyll fluorescence on thylak-
oid membranes prepared from the green alga Chlamydomonas reinhardtii,
we show that AtuNDH-2 is able to stimulate NADH-dependent reduction
of the plastoquinone pool. We discuss the possibility of using heterologous
expression of NDH-2 enzymes to improve nonphotochemical reduction of
plastoquinones and H
2
production in C. reinhardtii.

Escherichia coli-mutant deficient in NDH-1 and
NDH-2, they were proposed, because of their low activ-
ity, to have a sensor rather than a bioenergetic function
[20]. In other bacteria, although the physiological role of
NDH-2 is still unclear, the NDH-1 ⁄ NDH-2 ratio seems
to be regulated as a function of variations in the growth
conditions [21].
In chloroplasts of higher plants and algae, in addi-
tion to the photosynthetic electron transfer chain oxid-
izing water at photosystem II (PS II) and reducing
NADP
+
at photosystem I (PS I), the existence of a
respiratory chain including both nonphotochemical
reduction and oxidation of the plastoquinone (PQ)
pool has been shown [22]. In higher plant chloroplasts,
dark PQ reduction is mediated by a multisubunit com-
plex homologous to bacterial complex I [23,24]. In the
green unicellular alga Chlamydomonas reinhardtii such
a complex is absent from chloroplasts [22]. Based on
pharmacological studies, the involvement of a plasti-
dial NDH-2 enzyme has been proposed [25,26].
Whether bacterial or mitochondrial NDH-2s, which
normally reduce ubiquinones (UQs), are able to reduce
PQs and interact with the photosynthetic electron
transport chain is not established.
In this study, we report on the isolation of an Agro-
bacterium tumefaciens gene coding for a putative
NDH-2 (AtuNDH-2). Following expression in E. coli,
a His-tagged protein was purified by nickel-affinity

and cyanobacterial sequences. The last subgroup con-
tains plant NDA, NDB as well as yeast and C. rein-
hardtii sequences. AtuNDH-2 shares from 24 to 28%
identity and 40 to 48% similarity with rotenone-insensit-
ive NAD(P)H dehydrogenases of plant mitochondria.
Alignment of representative NDH-2 protein sequences
from the four different families (Fig. 2) revealed high
conservation in two domains showing most of the cri-
teria for dinucleotide binding, including a bab fold and
a GxGxxG motif [28]. Based on a comparison with the
lipoamide dehydrogenase sequence (an enzyme which
shares significant similarity with NDH-2s and the struc-
ture of which has been resolved), the first binding site
can be attributed to FAD and the second to NAD(P)H
[27]. The C-terminal domain of the protein, which con-
tains  20 hydrophobic residues (Fig. 2), has been sug-
gested to anchor the enzyme to the membrane [29].
Expression, purification and biochemical
characterization of AtuNDH-2
In order to express a His-tagged AtuNDH-2 protein
in E. coli, the AtuNDH-2 gene sequence was amplified
Agrobacterium tumefaciens NDH-2 L. Bernard et al.
3626 FEBS Journal 273 (2006) 3625–3637 ª 2006 The Authors Journal compilation ª 2006 FEBS
from genomic DNA, cloned into pSD80 with a C-ter-
minus (His)
6
-tag sequence. The resulting plasmid was
used to transform E. coli strain DH10b. The recom-
binant protein, mainly present in membrane fractions
(Fig. 3A), was purified by nickel-affinity chromatogra-

was analysed by HPLC and identified as FAD
(Fig. 3B). Figure 3C shows the absorption spectrum
of AtuNDH-2 protein. Typical peaks of FAD-con-
taining proteins were observed, confirming the nature
of the cofactor [19]. By using the extinction coeffi-
cient at 450 nm we calculated that the stoichiometric
ratio of flavinic cofactor per mole of protein reaches
1.02 mol, which is characteristic of NDH-2 proteins
[21]. The capacity of the purified enzyme to oxidize
NADH or NADPH was measured by monitoring
absorbance decay at 340 nm in the presence of var-
ious electron acceptors (Table 1). In the presence of
ferricyanide, the NADH-oxidizing activity saturated
around 50 lm NADH and reached 100 nmolÆmin
)1
Ælg
)1
protein. In the presence of NADPH, a significant
activity was measured, but it was not possible to
observe saturation within a concentration range suit-
able for spectrophotometric studies. An activity
of  50 nmolÆ min
)1
Ælg
)1
protein was measured at
200 lm NADPH. Various quinone acceptors were tes-
ted. In the presence of the soluble quinone Q
0
,a

MITOPROT II software (v. 1.0)
corresponding to the maximal local hydrophobicity indicated.
Agrobacterium tumefaciens NDH-2 L. Bernard et al.
3628 FEBS Journal 273 (2006) 3625–3637 ª 2006 The Authors Journal compilation ª 2006 FEBS
NADPH was about four times lower than for NADH.
Similar V
max
values were measured for NADPH and
NADH, but under these conditions, measured V
max
values were
~
50 times lower than in the presence of
ferricyanide or quinone acceptors.
In vitro interaction of AtuNDH-2 with bacterial
and thylakoid membranes
The ability of AtuNDH-2 to interact with bacterial
membranes was studied using membrane preparations
of an E. coli mutant lacking both NDH-1 and NDH-2.
The functionality of the respiratory chain of these
preparations was first checked in the presence of succi-
nate as substrate of complex II (data not shown). As
expected, no O
2
uptake was detected when mutant
membranes were supplemented with NADH (Fig. 4A).
Preincubation of mutant membranes with AtuNDH-2
resulted in an O
2
uptake upon NADH addition,

~
3.7 nmol O
2
Æmin
)1
Ælg
)1
protein) were obtained for
both NADH and NADPH oxidations (data not
shown). Diphenyleneiodonium (DPI), an inhibitor of
flavin enzymes, inhibited both NADH- and NADPH-
dependent reactions by
~
75%, half inhibition being
obtained at DPI concentrations of
~
13 lm (data not
shown).
A
B
C
Fig. 3. Purification of His-tagged AtuNDH-2 by nickel-affinity chro-
matography (A) and analysis of the flavinic cofactor (B). (A) Coo-
massie Brilliant Blue-stained SDS ⁄ PAGE and western blot analysis
using an anti-histidine IgG. T, total protein extract from bacterial
cells expressing AtuNDH-2; S, soluble proteins; M, membrane pro-
teins solubilized by dodecyl maltoside; E1 and E2, eluted fractions
from nickel-affinity chromatography using 200 m
M imidazole. MW,
molecular mass markers. (B) HPLC separation and fluorometric ana-

2
reduction by AtuNDH-2 measured
using an O
2
electrode in the presence of NADH and NADPH as
electron donors.
K
m
(lM) V
max
(nmolÆmin
)1
Ælg
)1
protein)
NADH 52.9 2.26
NADPH 201 2.70
L. Bernard et al. Agrobacterium tumefaciens NDH-2
FEBS Journal 273 (2006) 3625–3637 ª 2006 The Authors Journal compilation ª 2006 FEBS 3629
The ability of AtuNDH-2 to interact with PQs of
the photosynthetic electron transport chain was stud-
ied by performing chlorophyll fluorescence measure-
ments on thylakoid membranes of C. reinhardtii
(Fig. 5). When measured under nonactinic light, the
chlorophyll fluorescence level is an indicator of the
PQ pool redox state in the dark [31]. Under anaerobic
conditions to prevent dark reoxidation of the PQ
pool, addition of NADH (200 lm)toC. reinhardtii
thylakoid membranes provoked a slow increase in the
chlorophyll fluorescence level, indicating a reduction

with empty vector) could not grow on the minimal
medium, nor could untransformed cells (data not
shown). In contrast, the mutant strain expressing
AtuNDH-2 after induction with 0.1 mm isopropyl thio-
b-d-galactoside (IPTG) could grow under these condi-
tions. Note that partial complementation of the mutant
was observed in the absence of IPTG, likely indicating
significant IPTG-independent expression of the protein.
Using an antibody raised against the recombinant pro-
tein, AtuNDH-2 was detected in protein extracts
from the complemented ANN0222 strain (Fig. 6C).
Most of the protein was present in membrane fractions,
A
B
C
Fig. 5. Interaction of AtuNDH-2 with C. reinhardtii thylakoid mem-
branes. The increase in chlorophyll fluorescence was measured
under low light in response to NADH addition (final concentration
200 l
M) to a suspension of C. reinhardtii thylakoids (30 lg chloro-
phyllÆmL
)1
). (A) control; (B, C) thylakoid membranes were preincu-
bated with purified AtuNDH-2 at two protein concentrations (2.5
and 5 lgÆmL
)1
final protein concentration, respectively) for 30 min
before measurements.
A
B

sumption rates. Whereas only low O
2
uptake activity
was induced by NADH addition on membranes of the
control transformant strain, strong O
2
uptake was
observed in membranes containing AtuNDH-2
(Fig. 6B). Oxygen-uptake activities were measured in
membranes of the complemented ANN0222 strain to
determine apparent kinetic parameters of AtuNDH-2.
Under these conditions, AtuNDH-2 oxidized both
NADH and NADPH with similar maximal rates, and
with a much higher affinity for NADH (K
m
¼ 12.5 lm)
than for NADPH (K
m
¼ 1129 lm) (data not shown).
Discussion
We report here the cloning, expression and characteri-
zation of AtuNDH-2, the A. tumefaciens orthologue to
rotenone-insensitive NAD(P)H dehydrogenases. The
purified enzyme showed a NADH : Q
0
oxidoreductase
activity in the range of activities measured for other
purified enzymes, such as the C. glutamicum NDH-2
[9] or the Trypanosoma brucei NDH-2 [14]. AtuNDH-2
appears, however, 1000 less active than the purified

akoid membranes and PQs. Although NDH-2s are
membrane-bound enzymes, the nature of membrane ⁄
protein interactions has not been elucidated. Mem-
brane binding seems to rely on different mechanisms
depending on the enzyme, some involving transmem-
brane anchorage, whereas others involve electrostatic
interactions. A membrane association via amphipha-
tic helices has been suggested for E. coli and
Acidianus ambivalens NDH-2s [13,29]. 2D structure
analysis, performed on the 50 NDH-2 sequences used
to build the phylogenetic tree, predict the presence of
C-terminal transmembrane helices in the prokaryotic B
subgroup (including AtuNDH-2). AtuNDH-2 also
contains a hydrophobic domain enriched in aromatic
residues located between its two cofactor binding sites.
This domain has been proposed to be involved in the
interaction with hydrophobic quinones of respiratory
chains [11,13]. Respective contributions of both hydro-
phobic domains to membrane and quinone interactions
require further study.
In higher plants, addition of NAD(P)H to thylakoid
membrane preparations has been shown to stimulate
nonphotochemical reduction of the PQ pool [31].
Although higher plant chloroplasts contain a func-
tional NDH-1 complex [23,24], pharmacological stud-
ies concluded that a NDH-2-like enzyme is also
involved in this phenomenon [31]. C. reinhardtii chlo-
roplasts are recognized to lack NDH-1 complex [22].
A recent study conducted in our laboratory concluded
that, in C. reinhardtii, the enzyme involved in the dark

NDH-2 [26]. Because nonphotochemical reduction of
PQ may constitute a limiting step of hydrogen produc-
tion under certain experimental conditions, overexpres-
sion of a NDH-2 in C. reinhardtii plastid should be
considered as a valuable optimization strategy towards
improving the anaerobic phase of hydrogen produc-
tion. AtuNDH-2 appears as a suitable gene for such a
purpose because its high GC content is similar to that
of C. reinhardtii genomic DNA. The obtention of
transformants expressing AtuNDH-2 is currently in
progress in our laboratory.
AtuNDH-2 showed a much higher affinity for
NADH than for NADPH, although both substrates
were oxidized with comparable V
max
values (Table 2).
Several NDH-2s are strict NADH- [14,27] or strict
NADPH-dehydrogenases [15,38], whereas a few others,
mainly from plants, are able to oxidize both substrates
indifferently [5]. Michalecka et al. [38] suggested that
the presence of an acidic residue (E or D) at the end
of the second b sheet of the dinucleotide-binding site
would confer NADH specificity by providing hydrogen
bonding to the ribose moiety (Fig. 2). In enzymes
specifically oxidizing NADPH, this acidic residue is
replaced by a neutral counterpart (Q or N). The pres-
ence of a glutamic residue in AtuNDH-2 is in agree-
ment with the enzyme preference for NADH.
However, AtuNDH-2 was able to oxidize NADPH at
a similar maximal rate, but with a lower affinity.

Strains and media
A. tumefaciens C58 was grown in Luria–Bertani medium
(1% tryptone, 0.5% yeast extract, 0.5% NaCl, pH 7.0) in
the presence of rifampicine (100 lgÆmL
)1
)at30°C for 24 h
before genomic DNA extraction. Cloning and gene expres-
sion were performed in E. coli dH10b. The E. coli mutant
lacking both NDH-1 and NDH-2 ANN0222 was gener-
ously provided by T. Friedrich (Freiburg University,
Germany). Complementation assays were performed on a
minimal medium supplemented with mannitol (1· M9 salts,
2 · 10
)3
m MgSO
4
,10
)4
m CaCl
2
, 0.4% mannitol). Trans-
formed bacteria were selected on Luria–Bertani medium
containing ampicillin (100 lgÆmL
)1
). C. reinhardtii wild-
type (137c) was grown in a Tris-acetate phosphate medium
as described previously [42]. Algal culture was maintained
at 25 °C under continuous agitation and under an illumin-
ation of 100 lmolÆphotonÆm
)2

His-tagged NDH-2 was expressed in Luria–Bertani medium
in the presence of 50 lgÆmL
)1
ampicillin and 25 lgÆmL
)1
chloramphenicol and incubated at 37 °C under vigorous
shaking. Expression was induced for 2 h by the addition of
0.5 mm IPTG when the culture reached D ¼ 0.5. Cells were
then harvested by centrifugation and washed with a solu-
tion containing 500 mm KCl, 10 mm Tris ⁄ HCl, pH 7.5,
0.2 mm phenylmethylsulfonyl fluoride and stored at
)80 °C.
Protein purification
Membrane isolation and nickel-affinity purification of
A. tumefaciens His-tagged NDH-2 were performed follow-
ing the protocol developed by Bjo
¨
rklo
¨
f et al. [27] for purifi-
cation of an His-tagged E. coli NDH-2. Membranes were
solubilized using dodecyl maltoside and the enzyme was
purified by FPLC (A
˚
kta
˚
FPLC, Amersham Biosciences,
Uppsala, Sweden) using a HiTrap chelating nickel column
(Amersham Biosciences). The bound NDH-2 was eluted
using an imidazole gradient. Collected fractions were ana-

pSDN2ag6 h plasmid or empty vector as a control and
transformants were selected on Luria–Bertani agar contain-
ing ampicillin. Transformants were grown to mid-exponen-
tial phase at 37 °C in ampicillin-containing Luria–Bertani
medium. Cells were washed once with sterile M9 medium
supplemented with mannitol as the sole source of carbon.
Cells were then diluted in M9 ⁄ mannitol medium and spot-
ted on M9 ⁄ mannitol ⁄ agar plates containing ampicillin and
different IPTG concentrations. Plates were incubated at
37 °C for two days. As a control, cells were plated on a
Luria–Bertani agar medium containing ampicillin and
appropriate IPTG concentrations and incubated at 37 °C
overnight.
Flavin analysis
The nature of the flavinic cofactor was determined as des-
cribed previously [14]. The purified enzyme was boiled for
3–4 min followed by centrifugation. Supernatant was ana-
lysed by HPLC using a 150 · 4.6 mm internal diameter.
Supelcosil LC-DP column (Sigma-Aldrich). The composi-
tion of the mobile phase (flow rate of 1 mLÆmin
)1
) was
80% of 0.1% TFA in water and 20% of 0.1% TFA in
40% acetonitrile. Excitation and emission wavelengths of
the fluorescence detector were set at 450 and 525 nm,
respectively. FAD and FMN standards were used to iden-
tify the nature of the flavinic cofactor. UV-visible spectra
of native and boiled proteins were recorded using a Cary-
50 spectrophotometer (Varian, Palo Alto, CA, USA). Pro-
teins were diluted in 50 mm Tris ⁄ HCl buffer pH 7.5,

¨
rklo
¨
f et al. [27].
The same procedure was followed to prepare membranes
from ANN.0222 transformed cells either with pSD80 or
pSDN2Ag6H.
A 200 mL volume of C. reinhardtii culture grown on Tris-
acetate phosphate medium was harvested in exponential
growth phase, centrifuged, washed in 35 mm Hepes-NaOH
buffer, pH 7.2 and resuspended in 12 mL of buffer C
(50 mm Tricine-NaOH, 10 mm NaCl, 5 mm MgCl
2
;pH8)
supplemented with 1% BSA w ⁄ v, 1 m m benzamidine, 1 mm
phenylmethylsulfonyl fluoride. The following operations
were carried out in the dark at 4 °C. Cells were disrupted by
two cycles of a chilled French pressure cell (2000 p.s.i).
The homogenate was centrifuged at 500 g for 5 min using
an Eppendorf 8510R centrifuge (Eppendorf, Hamburg,
Germany). The pellet, containing unbroken cells, was discar-
ded. The supernatant was centrifuged at 3000 g for 15 min
using an Eppendorf 8510R centrifuge to collect thylakoid
membranes. Thylakoid membranes were resuspended in
250–500 lL of buffer C and stored on ice, in the dark. Chlo-
rophyll extraction was performed in 80% acetone ⁄ H
2
Ov⁄ v,
and chlorophyll concentration was calculated from absorp-
tion measurements at 663 and 646 nm [44].

consumption was
measured at 25 °C in the presence of NADH or NADPH.
O
2
uptake was also followed in absence of bacterial mem-
branes to measure direct oxidation by the enzyme. Catalase
(1000 unitsÆmL
)1
) and SOD (500 unitsÆmL
)1
) were added to
the assay medium. Inhibitory effects of DPI were quantified
after 10 min incubation with membrane samples before
measurements.
Chlorophyll fluorescence measurements
Measurements were achieved at 25 °C, in anaerobic condi-
tions a pulse modulated amplitude fluorometer (PAM
Agrobacterium tumefaciens NDH-2 L. Bernard et al.
3634 FEBS Journal 273 (2006) 3625–3637 ª 2006 The Authors Journal compilation ª 2006 FEBS
101-103, Walz, Effeltrich, Germany). The optic fibre of the
fluorometer was placed close to the glass tube of the O
2
electrode reaction chamber (1 mL reaction volume: 50 l L
of thylakoid membrane suspension and 950 lL of buffer C,
pH 7.2). Nonactinic modulated light (650 nm, 1.6 kHz) was
used to determine the chlorophyll fluorescence level F
0
. The
maximal chlorophyll fluorescence level (F
m

providing the E. coli double mutant strain ANN0222.
References
1 Friedrich T (2001) Complex I: a chimaera of redox and
conformation-driven proton pump? J Bioenerg Bio-
membr 33, 169–177.
2 Steuber J (2001) Na
+
translocation by bacterial NADH
: quinone oxidoreductases: an extension to the complex-
I family of primary redox pumps. Biochim Biophys Acta
1505, 45–56.
3 Bertsova YV & Bogachev AV (2004) The origin of the
sodium-dependent NADH oxidation by the respiratory
chain of Klebsiella pneumoniae. FEBS Lett 563, 207–212.
4 Menz RI & Day DA (1996) Purification and characteri-
zation of a 43-kDa rotenone-insensitive NADH dehy-
drogenase from plant mitochondria. J Biol Chem 271,
23117–23120.
5 Rasmusson AG, Svensson AS, Knoop V, Grohmann L
& Brennicke A (1999) Homologues of yeast and bacter-
ial rotenone-insensitive NADH dehydrogenases in
higher eukaryotes: two enzymes are present in potato
mitochondria. Plant J 20, 79–87.
6 Michalecka AM, Svensson AS, Johansson FI, Agius
SC, Johanson U, Brennicke A, Binder S & Rasmusson
AG (2003) Arabidopsis genes encoding mitochondrial
type II NAD(P)H dehydrogenases have different evolu-
tionary orgin and show distinct responses to light. Plant
Physiol 133, 642–652.
7 Bragg PD & Hou C (1967) Reduced nicotinamide ade-

14 Fang J & Beattie DS (2002) Novel FMN-containing
rotenone-insensitive NADH dehydrogenase from Trypa-
nosoma brucei mitochondria: isolation and characteriza-
tion. Biochemistry 41, 3065–3072.
15 Melo AM, Bandeiras TM & Teixeira M (2004) New
insights into type II NAD(P)H : quinone oxidoreduc-
tases. Microbiol Mol Biol Rev 68, 603–616.
16 Bertsova YV, Bogachev AV & Skulachev VP (2001)
Noncoupled NADH : ubiquinone oxidoreductase of
Azotobacter vinelandii is required for diazotrophic
growth at high oxygen concentrations. J Bacteriol 183,
6869–6874.
17 Bergsma J, Van Dongen MBM & Konings WN
(1982) Purification and characterization of NADH
L. Bernard et al. Agrobacterium tumefaciens NDH-2
FEBS Journal 273 (2006) 3625–3637 ª 2006 The Authors Journal compilation ª 2006 FEBS 3635
dehydrogenase from Bacillus subtilis. Eur J Biochem
128, 151–157.
18 Kim JY, Song K & Rhee S (1995) A novel aerobic
respiratory chain-linked NADH oxydase system in
Zymomonas mobilis. J Bacteriol 177, 5176–5178.
19 Cook SA & Shiemke AK (2002) Evidence that a type-2
NADH : quinone oxidoreductase mediates electron
transfer to particulate methane monooxygenase in
Methylococcus capsulatus. Arch Biochem Biophys 398,
32–40.
20 Howitt CA, Udall PK & Vermaas WFJ (1999) Type 2
NADH dehydrogenases in the cyanobacterium Synecho-
cystis sp. strain PCC 6803 are involved in regulation
rather than respiration. J Bacteriol 181, 3994–4003.

rklo
¨
f K, Zickermann V & Finel M (2000) Purifica-
tion of the 45 kDa, membrane bound NADH dehydro-
genase of Escherichia coli (NDH-2) and analysis of its
interaction with ubiquinone analogues. FEBS Lett 467,
105–110.
28 Wierenga RK, Terpstra P & Hol WGJ (1986) Prediction
of the occurrence of the ADP-binding bab-fold in pro-
teins, using an amino acid sequence fingerprint. J Mol
Biol 187, 101–107.
29 Rapisarda VA, Chehin RN, De las Rivas J, Rodriguez-
Montelongo L, Farias RN & Massa EM (2002) Evi-
dence for Cu(I)-thiolate ligation and prediction of a
putative copper-binding site in the Escherichia coli
NADH dehydrogenase-2. Arch Biochem Biophys 405,
87–94.
30 Messner KR & Imlay JA (1999) The identificationof
primary sites of superoxide and hydrogen peroxide for-
mation in the aerobic respiratory chain and sulfite
reductase complex of Escherichia coli. J Biol Chem 274,
10119–10128.
31 Corneille S, Cournac L, Guedeney G, Havaux M & Pel-
tier G (1998) Reduction of the plastoquinone pool by
exogenous NADH and NADPH in higher plant chloro-
plasts. Characterization of a NAD(P)H–plastoquinone
oxidoreductase activity. Biochim Biophys Acta 1363,
59–69.
32 Happe T & Naber JD (1993) Isolation, characterization
and N-terminal amino acid sequence of hydrogenase

NADH : ubiquinone oxidoreductases. Biochim Biophys
Acta 1459, 274–283.
40 Joseph-Horne T, Hollomon DW & Wood PM (2001)
Fungal respiration: a fusion of standard and alterna-
tive components. Biochim Biophys Acta 1504, 179–
195.
41 Luttik MAH, Overkamp KM, Kotter P, de Vries S,
van Dijken JP & Pronk JT (1998) The Saccharomyces
cerevisiae NDE1 and NDE2 genes encode separate
mitochondrial NADH dehydrogenases catalyzing the
oxidation of cytosolic NADH. J Biol Chem 273,
24529–24534.
Agrobacterium tumefaciens NDH-2 L. Bernard et al.
3636 FEBS Journal 273 (2006) 3625–3637 ª 2006 The Authors Journal compilation ª 2006 FEBS
42 Harris EH (1989) The Chlamydomonas Sourcebook. A
Comprehensive Guide to Biology and Laboratory Use.
Academic Press, San Diego, CA.
43 Smith SP, Barber KR, Dunn SD & Shaw GS (1996)
Structural influence of cation binding to recombinant
human brain S100b: Evidence for calcium-induced expo-
sure of a hydrophobic surface. Biochemistry 35, 8805–
8814.
44 Lichtenthaler HK & Wellburn AR (1983) Determina-
tion of total carotenoids and chlorophylls a and b of
leaf extracts in different solvents. Biochem Soc Trans 11,
591–592.
45 Altschul SF, Madden TL, Schaffer AA, Zhang J, Miller
W & Lipman DJ (1997) Gapped BLAST and PSI-
BLAST: a new generation of protein database search
programs. Nucleic Acids Res 25, 3389–3402.