In vitro RNA editing in plant mitochondria does not require added energy.

Abstract

RNA editing in flowering plant mitochondria is investigated by in vitro assays. These cauliflower mitochondrial lysates require added NTP or dNTP. We have now resolved the reason for this requirement to be the inhibition of the RNA binding activity of the glutamate dehydrogenases (GDH). Both GDH1 and GDH2 were identified in RNA-protein cross-links. The inhibition of in vitro RNA editing by GDH is confirmed by the ability of the GDH-specific herbicide phosphinothricin to substitute for NTP. NADH and NADPH, but not NAD or NADP, can also replace NTP, suggesting that the NAD(P)H-binding-pocket configuration of the GDH contacts the RNA. RNA editing in plant mitochondria is thus intrinsically independent of added energy in the form of NTP.

Full text

Page 1

In vitro RNA editing in plant mitochondria does not require added energy
Mizuki Takenakaa,*, Daniil Verbitskiya, Johannes A. van der Merwea, Anja Zehrmanna,
Uwe Plessmanna,b, Henning Urlauba,b, Axel Brennickea
aMolekulare Botanik, Universita ¨t Ulm, 89069 Ulm, Germany
bMax Planck Institute for Biophysical Chemistry, Bioanalytical Mass Spectrometry Group, Am Fassberg 11, 37077 Go ¨ttingen, Germany
Received 16 April 2007; revised 8 May 2007; accepted 8 May 2007
Available online 21 May 2007
Edited by Ulf-Ingo Flu ¨gge
Abstract
tigated by in vitro assays. These cauliflower mitochondrial ly-
sates require added NTP or dNTP. We have now resolved the
reason for this requirement to be the inhibition of the RNA bind-
ing activity of the glutamate dehydrogenases (GDH). Both
GDH1 and GDH2 were identified in RNA–protein cross-links.
The inhibition of in vitro RNA editing by GDH is confirmed
by the ability of the GDH-specific herbicide phosphinothricin
to substitute for NTP. NADH and NADPH, but not NAD or
NADP, can also replace NTP, suggesting that the NAD(P)H-
binding-pocket configuration of the GDH contacts the RNA.
RNA editing in plant mitochondria is thus intrinsically indepen-
dent of added energy in the form of NTP.
? ? 2007 Federation of European Biochemical Societies. Published
by Elsevier B.V. All rights reserved.
RNA editing in flowering plant mitochondria is inves-
Keywords: RNA editing; Plant mitochondria; Glutamate
dehydrogenase; RNA-binding; In vitro RNA editing;
Cauliflower
1. Introduction
In plants RNA molecules are in both organelles altered by
RNA editing. In plastids of flowering plants about 35 and in
mitochondria about 400 selected cytosines are changed to uri-
dines. The direct biochemical effect of RNA editing in these
plants is thus a site-specific deamination. Despite years of
investigation neither the reaction mechanism nor the enzymes
involved have been identified. Several clues on – or rather con-
ditions of – the biochemistry of the reaction have emerged
mostly from in vitro analyses of plastid and mitochondrial ly-
sates, respectively. The first in vitro assays of plant mitochon-
drial lysates suggested that the sugar-phosphate bonds of the
affected RNA molecule are not disrupted in the polynucleotide
chain [1]. This observation excludes insertional editing which
would excise and exchange either the nucleotide or the base.
Thus either direct deamination or transamination are the most
likely mechanisms. The first reaction would per se require no
added energy, while for the latter, the transamination reaction,
additional molecules of higher molecular energy would most
likely be involved.
For mitochondrial lysates a strict requirement for added
ATP has been observed [2]. Similarly in plastids additional
ATP has been reported to be required in most lysates [3–5],
while in some residual activity can be seen without any added
ATP [6,7]. Surprisingly the added ATP can be substituted to
full effect by some or all of the other NTPs and even dNTPs
[2,6,7]. While all of these molecules are virtually interchangable
in mitochondrial lysates, they vary in their effect in plastid
extracts [6,7]. In both organelles at least one of the dNTPs is
as effective as ATP, suggesting that one of the few enzymes
accepting either triphosphate is involved. One group of such
enzymes is a class of RNA helicases and their participation
has consequently been proposed [2].
We have now investigated the requirement for nucleotide tri-
phosphates in detail and come to the surprising conclusion
that this dependence is at least in mitochondria almost entirely
explained by the behaviour of the enzyme glutamate dehydro-
genase (GDH).
2. Materials and methods
2.1. Preparation of mitochondrial extracts
Cauliflower mitochondria were purified by differential centrifugation
steps and a Percoll gradient as described [2,8]. Isolated mitochondria
were lysed, the lysate was cleared and the supernatant was recovered
and dialyzed as detailed previously [2,8].
2.2. RNA substrates
DNA clones were constructed in an adapted pBluescript SK+to
allow run-off transcription of the editing substrate RNA as described
[2,8]. Substrate RNAs containing vector sequences at the 50- and 30-
ends of the mitochondrial insert sequences were synthesized from the
T7 RNA polymerase promoter in the linearized template DNA. The
bordering bacterial sequences were used for specific amplification of
the substrate RNAs by RT-PCR after the in vitro assay [9,10]. The
32P-labelled RNA templates for the gel shift assays were obtained by
direct incorporation of labelled ATP into the RNA during run-off
transcription from the DNA template.
2.3. In vitro RNA editing reactions
The in vitro RNA editing reactions were performed as described
[8,11]. After incubation, substrate sequences were amplified by RT-
PCR, the upstream primer being labelled with the Cy5 fluorophor.
RNA editing activity was detected by mismatch analysis employing
the TDG enzyme activity (thymine DNA glycosylase, Trevigen). The
TDG treated fragments were separated and the Cy5 fluorescence was
scanned and displayed using an ALF express DNA sequencer (GE
Healthcare).
*Corresponding author. Fax: +49 731 502 2626.
E-mail addresses: mizuki.takenaka@uni-ulm.de (M. Takenaka),
daniil.verbitskiy@uni-ulm.de (D. Verbitskiy), dre.van-der-merwe@
uni-ul.de (J.A. van der Merwe), anja.zehrmann@uni-ulm.de
(A. Zehrmann), uplessm@gwdg.de (U. Plessmann), henning.urlaub@
mpi-bpc.mpg.de (H. Urlaub), mo.bo@uni-ulm.de (A. Brennicke).
0014-5793/$32.00 ? 2007 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.febslet.2007.05.025
FEBS Letters 581 (2007) 2743–2747

Page 2

2.4. RNA binding proteins
Proteins in the mitochondrial lysate binding to RNA were analysed
by incubating 1 pmol RNA template for 30 min in the reaction mix-
ture. In the template RNA every uridine was labeled with
in vitro transcription without adding cold UTP. The proteins in con-
tact with the RNA were irradiated with UV light for 20 min (Stratalin-
ker 1800, stratagene). This step was followed by digestion of the RNA
with RNase A at 37 ?C for 30 min. Protein samples were dissolved in
buffer and separated by SDS–PAGE on a 10% gel. Labelled proteins
were visualized with a Bioimaging Analyzer BAS-3000 (Fuji Photo
Film Co).
32P by
2.5. Protein identification
RNA editing templates were labelled with biotin (biomers GmbH)
and 500 pmol were incubated with 100 ll of mitochondrial lysate in
a total volume of 200 ll under standard in vitro RNA editing condi-
tions in the presence of 1 mg total yeast RNA [2]. The proteins in con-
tact with the RNA were cross-linked by UV irradiation for 200(UV
Stratalinker 1800, Stratagene). The biotin-labelled RNA–protein com-
plexes were bound to streptavidine sepharose? High performance (GE
healthcare). Unbound proteins were washed off by six rinses in spin
columns with 400 ll wash buffer each (30 mM HEPES-KOH pH 7.7,
3 mM magnesium acetate, 45 mM potassium acetate, 30 mM ammo-
nium acetate, 10% glycerol). The sepharose beads were treated with
RNase A to release the RNA-bound proteins, which were collected
with wash buffer. Proteins with their bound residual RNA fragments
were analysed by PAGE and/or were (directly) identified by liquid
chromatography coupled (LC) electrospray ionization (ESI) tandem
mass spectrometry (MSMS). Proteins cross-linked to RNA were
digested within the gel using trypsine according to Shevchenko et al.
[12]. For digestion of proteins in solution after extraction from the
beads, the samples are denatured in 4 M urea and 2 M thiourea,
diluted to a final concentration of 61 M urea and incubated with
2 lg trypsine overnight at room temperature. Peptides are desalted
and separated by nanoLC equipped with a pre-column working in
backflush and directly analyzed by ESI MSMS in a Q-ToF (Q-ToF
ultima, Waters) or a hybrid triple quadrupole/linear iontrap mass spec-
trometer (4000 QTrap, ABI) under standard conditions. Fragment
spectra of peptides are searched against the NCBInr database using
Mascot as search engine. Matches with resulting protein identification
were usually obtained against the Arabidopsis thaliana sequences,
which is phylogenetically closely related to the here analysed cauli-
flower (Brassica oleracea). Since different species are compared, simi-
larities/differences in the respective orthologous protein sequences
additionally influence the generated scores.
3. Results and discussion
3.1. NTP as well as dNTP support the in vitro RNA editing
reaction
The strict requirement for added NTP in the in vitro RNA
editing reaction is documented in Fig. 1. No RNA editing
product with uridine in the relevant position of RNA editing
is detectable if no NTP is added. Addition of ATP at the opti-
mal concentration of 15 mM yields about 2–5% C to U con-
verted nucleotide, the amount depending upon the lysate, the
source of the mitochondria and the template used. The positive
effect of ATP can be fully recovered by substitution with CTP,
shown as example here (Fig. 1). Any of the NTPs or dNTPs
can replace ATP and will yield comparable editing rates [2].
Since in plastid as in mitochondrial in vitro systems NTPs
and dNTPs can substitute for ATP, analogous factors may
be responsible for the in vitro ATP requirement in both organ-
elles.
3.2. Influence of the added NTP on template RNA–protein
interactions
To investigate the protein moieties affected by the added
NTP/dNTP in their ability to contact and bind to the RNA
substrate molecules, we performed gel shift experiments with
radioactively labelled RNA editing substrates (data not
shown). Comparisons of the protein patterns cross-linked with
the template RNA in the presence or absence of ATP on
PAGE-gels and by more sensitive mass spectroscopy (MS)-
analyses did not reveal any qualitative differences. The overall
quantity of the proteins labelled with the cross-linked RNA,
however, changes dramatically with the presence or absence
of NTP: Without added ATP the same proteins are labelled
much stronger (Fig. 2). This suggests that addition of ATP dis-
turbs all protein–RNA interactions non-specifically.
3.3. Identification of proteins bound by the RNA editing
template
Since the gel shift and UV cross-linking experiments (Fig. 2)
did not yield detectable proteins differentially affected by NTP
in their affinity to the RNA editing template, we employed a
more sensitive affinity purification scheme to identify respec-
tive polypeptides. This procedure achieved an enrichment of
the proteins bound to the editing template RNA (Fig. 3).
The Coomasie-stained SDS–PAGE patterns of proteins
retained by the RNA template in the presence or absence of
ATP, however, again showed only little qualitative difference
and no clear candidate protein could be assigned by its ATP-
dependent affinity to the template in this gel analysis (data
not shown).
Because the ATP-dependence of the in vitro reaction may
not manifest in abundant specific RNA–protein interactions
and the PAGE analysis may not be sensitive enough, we pro-
ceeded to analyse the set of proteins interacting with the RNA
directly by the much more sensitive MS without prior gel sep-
aration. The proteins bound by the RNA editing template
were hydrolyzed and peptides were identified by liquid chro-
matography (LC) coupled tandem MS (MSMS) analysis. To
identify the proteins, the obtained fragment spectra were
searched against all entries in the NCBInr database. Since
cauliflower and A. thaliana, which latter is completely
sequenced, are closely related and the protein sequences of
many household enzymes are to a large extent identical, we
should be able to identify cauliflower proteins through their
Fig. 1. In vitro RNA editing of an atp4 mRNA template requires the
addition of ATP or CTP. Similar to the in vitro editing of an atp9
RNA [2], dNTP-nucleotides can substitute for the NTP in this reaction
(data not shown).
2744
M. Takenaka et al. / FEBS Letters 581 (2007) 2743–2747

Page 3

orthologues in A. thaliana. The proteins contained in the total
complement enriched by the affinity purification procedure
and identified by conserved peptide sequences indeed yielded
a list of mitochondrial proteins from A. thaliana (Table 1).
Within this list we focussed on proteins which might be in-
volved in a deaminase or transaminase reaction. Among the
general mitochondrial proteins such as heat shock proteins,
malate dehydrogenase and such, some of the more prominent
proteins were representatives of the glutamate dehydrogenase
group of proteins, i.e. GDH1 and GDH2 (Table 1) [13].
GDH can catalyze amination as well as deamination reactions
depending on the presence and concentration of various allo-
steric and isosteric regulators (Fig. 4). There is thus a (albeit
remote) possibility that one or more of these enzymes have
been recruited into the RNA editing reaction to catalyse the
deamination step from C to U.
Furthermore GDH-proteins have been previously identified
as RNA binding proteins [14] and have been investigated par-
ticularly with respect to their participation in RNA editing in
kinetoplasts of trypanosomes. Simpson and coworkers showed
that the preferential binding of GDH to guide RNA is fortu-
itous and not related to RNA editing [15]. However, in kine-
toplasts RNA editing involves the insertion and deletion of
specific uridines rather than the deamination events found in
mitochondria and plastids of flowering plants.
3.4. Is the GDH involved in the RNA editing reaction?
To investigate a potential participation of the GDH in the C
to U deamination reaction, we tested the influence of various
cofactors of this enzyme on the in vitro editing assay. The most
prominent and essential cofactors of the normal GDH cata-
lyzed reactions are NADH and NADP for the deamination
and amination reactions, respectively. NAD, NADP and their
respective reduced forms NADH and NADPH were individu-
ally added to the in vitro RNA editing reaction. The assay also
contained the suboptimal concentration of 10 mM ATP to be
able to observe either enhancing or detrimental effects. NADH
as well as NADPH stimulated the reaction while the oxidized
dinucleotides NAD and NADP did not (data not shown).
Since only enhancing effects were detected, the next series of
assays investigated the effect of NADH and the other dinucle-
otides in the absence of ATP to see whether they can substitute
at least partially for the mononucleotide. Both NADH and
NADPH were able to replace the NTP requirement, while
NAD and NADP did not support the in vitro RNA editing
reaction (Fig. 5).
3.5. Investigation of the role of the GDH in the in vitro RNA
editing assay
This result of NADH and NADPH being able to substitute
for NTP suggests that NTP may not be required for the RNA
editing reaction per se, but may be required to alleviate the
inhibitory effect of the GDH by releasing this protein from
the template RNA or by keeping it away a priori. To test this
hypothesis that the GDH might be a major inhibitor of the
in vitro RNA editing activity, we analysed the effect of a direct
blocker of the GDH proteins.
The compound phosphinothricin is such a specific inhibitor
of the GDH in plants and binds irreversibly to the active site of
the enzyme. Addition of phosphinothricin to the in vitro RNA
editing reaction indeed recovered the editing activity in the ab-
sence of any added NTP (Fig. 6). This result shows that NTP
solely serves to keep the GDH protein(s) away from the
Fig. 2. Similar proteins are cross-linked to radiolabelled RNA in the
presence or absence of ATP, but ATP reduces the overall amount of
the unspecifically bound proteins. Proteins contained in the mitochon-
drial lysate from cauliflower inflorescences used for in vitro RNA
editing assays were UV-cross-linked to32P labelled atp4 RNA in the
absence (?ATP) or presence (+ATP) of 15 mM added ATP. The RNA
template was digested with RNase A and proteins were separated by
SDS–PAGE in a 10% gel. Sizes of marker proteins are indicated in
kDa on the side of the autoradiogramme of the dried gel.
Fig. 3. Cauliflower proteins binding to the atp9 RNA editing template
were purified by cross-linking to the RNA. The atp9 RNA editing
template contains 40 nucleotides upstream and 10 nucleotides down-
stream of the first editing site. Proteins were spread on an analytical
PAGE-gel and silver-stained. The arrow indicates the position of the
GDH as determined from the molecular weight of the A. thaliana
protein(s). It is assumed that the cauliflower protein analysed here has
a comparable molecular weight.
M. Takenaka et al. / FEBS Letters 581 (2007) 2743–2747
2745

Page 4

template and that no added energy is required for the RNA
editing reaction per se.
The equivalent capability of NADH and the other reduced
dinucleotide NADPH suggests that the site of action of these
compounds is the GDH itself. Whether the added NTPs and
dNTPs also bind to the GDH remains open.
If the NTPs and dNTPs do not directly interact with the
GDH, the NTP/dNTP and the NADH/NADPH/phosphino-
thricin effects could be achieved by different modes of action:
While the first may activate an RNA helicase [2] which
removes the inhibiting bound GDH, the latter compounds
may have the same final effect by directly blocking the
GDH. The ultimate result observed with both sets of added
chemicals, namely active in vitro RNA editing, could be the
outcome from either mechanism. To decide this question, the
effect of added (d)NTPs and NAD(P)H on the RNA binding
of the GDH will have to be investigated with the purified
GDH enzyme. The observation that all unspecific RNA-
binding proteins in a mitochondrial extract are diminished
upon addition of ATP to the gel-shift or cross-linking reac-
tion (Fig. 2) suggests that the NTP/dNTP effect is indeed not
a direct action on the GDH, but could be rather a separate
general RNA-wiping effect, possibly mediated by an RNA
helicase.
Table 1
Typical protein identification results of an MS analysis after enrichment and UV-cross-linking of cauliflower mitochondrial proteins binding to the
atp9 RNA editing template
Gene identifierMass ScoreQueries matchedName
gi|4210330
gi|7076784
gi|2654210
gi|17939849
gi|15238762
gi|520478
gi|4210332
gi|415733
gi|15240793
gi|710400
gi|6435320
116581
114147
72330
63332
44496
39163
49908
62316
44671
43003
25248
839
575
489
435
324
258
226
210
191
134
123
19
12
9
7
7
4
7
4
4
5
3
2-oxoglutarate dehydrogenase, E1 subunit
2-oxoglutarate dehydrogenase, E1 subunit-like protein
Heat shock 70 protein
Mitochondrial F1 ATP synthase beta subunit
GDH1 (GLUTAMATE DEHYDROGENASE 1); oxidoreductase
Pyruvate dehydrogenase E1 beta subunit
2-oxoglutarate dehydrogenase E2 subunit
Mitochondrial chaperonin
GDH2 (GLUTAMATE DEHYDROGENASE 2); oxidoreductase
Pyruvate dehydrogenase E1 alpha subunit
Nucleoside diphosphate kinase
The orthologous proteins from A. thaliana with relevant scores listed here are those expected in plant mitochondria and previously identified in
proteome analyses [16].
Fig. 4. The reactions catalyzed by the glutamate dehydrogenase
include deamination and amination steps. The GDH enzymes bind
NAD(P) or NAD(P)H for the activities of amination and deamination,
respectively. The direction of the dominant reaction is influenced by
the levels of ATP. It is possible that one of these or a similar enzyme
has mutated to access also RNA molecules and deaminate specific
cytidines.
Fig. 5. NADH and NADPH can substitute the requirement for NTP
in the in vitro RNA editing reaction. These assays monitor the first
editing site in an atp4 template RNA in which no other site can be
edited since the adjacent downstream nucleotides have been altered [9].
The lane marked 0 contains no added (di)nucleotide in the control
in vitro incubation. The lanes marked CTP, NADH, NAD, NADP,
NADPH contain 15 mM of the respective compounds in the in vitro
reaction mix. The gel image shown is the relevant portion of the
fluorescent detection of the Cy5 labelled RT-PCR products after TDG
treatment, which recognizes the mismatches at the T moieties
introduced by in vitro editing.
Fig. 6. The GDH-specific inhibitor phosphinothricin can substitute
the requirement for NTP in the in vitro RNA editing reaction. All
assays monitor the first editing site in an atp4 template RNA. In the
control in vitro incubation no nucleotide was added (lane marked 0).
The lanes marked ‘‘phosphinothricin’’ contain this compound but no
added NTP or dNTP in the in vitro reaction mix. Only the relevant
portion of the gel image is shown.
2746
M. Takenaka et al. / FEBS Letters 581 (2007) 2743–2747

Page 5

Nevertheless, the substitution of NTP by phosphinothricin
shows that if another, general RNA-clearing enzyme such as
a helicase is stimulated by the added NTP/dNTP, the activity
of this enzyme is not essential for the actual RNA editing reac-
tion, but at least in vitro mainly serves to remove the inhibitory
GDH.
3.6. RNA editing does not require added NTP in vitro
The observed in vitro RNA editing in the absence of added
NTP suggests that the actual biochemical reaction does not
consume energy. The in vitro observed requirement of added
NTPs, dNTPs, NADH or phosphinothricin may not be rele-
vant in the in vivo situation: in the intact mitochondrion,
RNA editing as well as splicing and 50- and 30-processing most
likely occur compartimentalized and in safe distance from the
location of the GDH.
In chloroplast extracts in vitro RNA editing is similarly
stimulated by or depends on the addition of NTP or dNTP
moieties. Here the plastid located NADPH-dependent GDH
may play an analogous inhibitory role and corresponding spe-
cific inhibitors such as phosphinothricin should be tested to
resolve this question. It has been reported however that neither
NADH nor NADPH have any effect on the in vitro reaction,
which may argue against an analogous role [6]. On the other
hand the effects of individual NTP and dNTP identities vary
between chloroplast extracts from different plant species
[3,6,7].
Of course one cannot formally exclude the presence of low
amounts of ATP in the mitochondrial or chloroplast lysates
which escape the dialysis step by being bound to larger mole-
cules during the extract preparation. However, direct determi-
nation of the ATP content in a plastid lysate showed less than
10 nM ATP to be present, suggesting that the dialysis indeed
efficiently removes these small molecules [6].
The correct in vitro RNA editing in a plant mitochondrial
lysate in the presence of GDH-inhibiting phosphiniothricin
shows that the GDH is not involved in the deamination of C
to U although this enzyme regularly catalyzes an analogous
reaction and can bind RNA molecules. These conclusions
make a direct deamination process by a modified cytidine
deaminase or an analogous enzyme more probable than a
transaminase reaction, for which a requirement of some addi-
tional source of activation energy would be more likely –
although energetically not strictly necessary.
Acknowledgements: We thank Dagmar Pruchner for excellent experi-
mental help. M.T. was a fellow of the JSPS. J.A.v.d.M. was a fellow
of the DAAD. This work was supported by a grant from the DFG.
References
[1] Araya, A., Domec, C., Begu, D. and Litvak, S. (1992) An in vitro
system for the editing of ATP synthase subunit 9 mRNA using
wheat mitochondrial extracts. Proc. Natl. Acad. Sci. USA 89,
1040–1044.
[2] Takenaka, M. and Brennicke, A. (2003) In vitro RNA editing in
pea mitochondria requires NTP or dNTP, suggesting involvement
of an RNA helicase. J. Biol. Chem. 278, 47526–47533.
[3] Hirose, T. and Sugiura, M. (2001) Involvement of a site-specific
trans-acting factor and a common RNA-binding protein in the
editing of chloroplast RNA: development of an in vitro RNA
editing system. EMBO J. 20, 1144–1152.
[4] Miyamoto, T., Obokata, J. and Sugiura, M. (2002) Recognition
of RNA editing sites is directed by unique proteins in chloro-
plasts: biochemical identification of cis-acting elements and trans-
acting factors involved in RNA editing in tobacco and pea
chloroplasts. Mol. Cell. Biol. 22, 6726–6734.
[5] Miyamoto, T., Obokata, J. and Sugiura, M. (2003) A site-specific
factor interacts directly with its cognate RNA editing site in
chloroplast transcripts. Proc. Natl. Acad. Sci. USA 101, 48–52.
[6] Hegeman, C.E., Hayes, M.L. and Hanson, M.R. (2005) Substrate
and cofactor requirements for RNA editing of chloroplast
transcripts in Arabidopsis in vitro. Plant J. 42, 124–132.
[7] Nakajima, Y. and Mulligan, M. (2006) Nucleotide specificity of
the RNA editing reaction in pea chloroplasts. J. Plant Physiol.
162, 1347–1354.
[8] Neuwirt, J., Takenaka, M., van der Merwe, J.A. and Brennicke,
A. (2005) An in vitro RNA editing system from cauliflower
mitochondria: editing site recognition can vary between different
plant species. RNA 11, 1563–1570.
[9] Verbitskiy, D., Takenaka, M., Neuwirt, J., van der Merwe, J.A.
and Brennicke, A. (2006) Partially edited RNAs are intermediates
of RNA editing in plant mitochondria. Plant J. 47, 408–416.
[10] van der Merwe, J.A., Takenaka, M., Neuwirt, J., Verbitskiy, D.
and Brennicke, A. (2006) RNA editing sites in plant mitochondria
can share cis-elements. FEBS Lett. 580, 268–272.
[11] Takenaka, M., Neuwirt, J. and Brennicke, A. (2004) Complex cis-
elements determine an RNA editing site in pea mitochondria.
Nucleic Acids Res. 32, 4137–4144.
[12] Shevchenko, A., Wilm, M., Vorm, O. and Mann, M. (1996) Mass
spectrometric sequencing of proteins from silver-stained poly-
acrylamide gels. Anal. Chem. 68, 850–858.
[13] Melo-Oliveira, R., Oliveira, I.C. and Coruzzi, G.M. (1996)
Arabidopsis mutant analysis and gene regulation define a nonre-
dundant role for glutamate dehydrogenase in nitrogen assimila-
tion. Proc. Natl. Acad. Sci. USA 93, 4718–4723.
[14] Preiss, T., Sang, A.E., Chrzanowska-Lightowlers, Z.M.A. and
Lightowlers, R.N. (1995) The mRNA-binding protein COLBP is
glutamate dehydrogenase. FEBS Lett. 367, 291–296.
[15] Bringaud, F., Stripecke, R., Frech, G.C., Freedland, S., Turck,
C., Byrne, E.M. and Simpson, L. (1997) Mitochondrial glutamate
dehydrogenase from Leishmania tarantolae is a guide RNA-
binding protein. Mol. Cell. Biol. 17, 3915–3923.
[16] Millar, A.H., Heazlewood, L., Kristensen, B.K., Braun, H.-P. and
Møller, I.M. (2005) The plant mitochondrial proteome. Trends
Plant Sci. 10, 36–43.
M. Takenaka et al. / FEBS Letters 581 (2007) 2743–2747
2747