Download Prokaryotic orthologues of mitochondrial alternative oxidase and plastid terminal oxidase

Plant Molecular Biology 53: 865–876, 2003.
© 2004 Kluwer Academic Publishers. Printed in the Netherlands.
865
Prokaryotic orthologues of mitochondrial alternative oxidase and plastid
terminal oxidase
Allison E. McDonald, Sasan Amirsadeghi and Greg C. Vanlerberghe∗
Department of Life Sciences and Department of Botany, University of Toronto at Scarborough, 1265 Military Trail,
Scarborough, Ontario, M1C 1A4 Canada (∗ author for correspondence; e-mail [email protected])
Received 14 October 2003; accepted in revised form 4 December 2003
Key words: alternative oxidase, chlororespiration, endosymbiosis, orthologues, plastoquinol terminal oxidase,
prokaryotic
Abstract
The mitochondrial alternative oxidase (AOX) and the plastid terminal oxidase (PTOX) are two similar members
of the membrane-bound diiron carboxylate group of proteins. AOX is a ubiquinol oxidase present in all higher
plants, as well as some algae, fungi, and protists. It may serve to dampen reactive oxygen species generation by
the respiratory electron transport chain. PTOX is a plastoquinol oxidase in plants and some algae. It is required in
carotenoid biosynthesis and may represent the elusive oxidase in chlororespiration. Recently, prokaryotic orthologues of both AOX and PTOX proteins have appeared in sequence databases. These include PTOX orthologues
present in four different cyanobacteria as well as an AOX orthologue in an α-proteobacterium. We used PCR, RTPCR and northern analyses to confirm the presence and expression of the PTOX gene in Anabaena variabilis PCC
7120. An extensive phylogeny of newly found prokaryotic and eukaryotic AOX and PTOX proteins supports the
idea that AOX and PTOX represent two distinct groups of proteins that diverged prior to the endosymbiotic events
that gave rise to the eukaryotic organelles. Using multiple sequence alignment, we identified residues conserved
in all AOX and PTOX proteins. We also provide a scheme to readily distinguish PTOX from AOX proteins based
upon differences in amino acid sequence in motifs around the conserved iron-binding residues. Given the presence
of PTOX in cyanobacteria, we suggest that this acronym now stand for plastoquinol terminal oxidase. Our results
have implications for the photosynthetic and respiratory metabolism of these prokaryotes, as well as for the origin
and evolution of eukaryotic AOX and PTOX proteins.
Abbreviations: AOX, alternative oxidase; PTOX, plastoquinol (formerly plastid) terminal oxidase
Introduction
The mitochondrial electron transport chain in plants
as well as some algae, fungi and protists is branched
so that electrons at ubiquinol can be transferred to
oxygen via the familiar cytochrome path or they may
be transferred directly to oxygen via a ubiquinol oxidase called alternative oxidase (AOX; Berthold et al.,
2000; Siedow and Umbach, 2000). Potential metabolic roles for this oxidase in plants have been recently
reviewed and include a means to stabilize the reduction state of electron transport chain components,
thereby controlling the mitochondrial generation of re-
active oxygen species (Simons and Lambers, 1999;
Møller, 2001). The presence of AOX in plants may
have wide implications for growth, stress tolerance or
susceptibility to programmed cell death (Moore et al.,
2002; Robson and Vanlerberghe, 2002).
Analysis of an Arabidopsis thaliana mutant (IMMUTANS) defective in chloroplast biogenesis led to
the discovery of a plastid-localized protein with some
sequence similarity to AOX (Carol et al., 1999; Wu
et al., 1999). This protein, termed plastid terminal
oxidase (PTOX), is localized to the thylakoid membrane and functions as a plastoquinol oxidase (Carol
et al., 1999; Cournac et al, 2000; Joet et al., 2002;
866
Josse et al., 2003). The IMMUTANS mutation indicates that PTOX is required for carotenoid biosynthesis
(Carol et al., 1999; Wu et al., 1999) and, more generally, it may represent the elusive terminal oxidase in
chlororespiration (Peltier and Cournac, 2002). PTOX
has been identified in plants and some green algae.
AOX and PTOX are both members of the diiron
carboxylate group of proteins. Such proteins can be
distinguished by six conserved amino acids proposed
to represent iron-binding residues (Berthold and Stenmark, 2003). Electron paramagnetic resonance data
indicates that AOX does indeed contain a coupled binuclear iron center (Berthold et al., 2002). Both AOX
and PTOX are proposed to bind interfacially to the
membrane bilayer (Berthold and Stenmark, 2003).
Given the endosymbiotic origin of plastids, it
was previously suggested that an oxidase responsible
for chlororespiration may have originated from a cyanobacterial ancestor (Scherer, 1990). Interestingly,
heterologous expression experiments have shown that
both plant AOX and PTOX proteins are functional
in bacterial membranes (Berthold et al., 2002; Josse
et al., 2003). Here, we report on orthologues of
PTOX in the sequenced genomes of four different
cyanobacteria and of an AOX orthologue in an αproteobacterium. This work complements and extends
upon other recent reports of the appearance of these
proteins in sequence databases (Finnegan et al., 2003;
Stenmark and Nordlund, 2003).
Materials and methods
Organism and growth conditions
Anabaena variabilis sp. strain PCC 7120 was from
the University of Toronto Culture Collection. Cells
were grown at 28–30 ◦ C in BG-11 medium (Rippka
et al., 1979) and at a light intensity of ca. 50 µmol
photons m−2 s−1 PAR.
In silico analyses
PTOX protein sequences were retrieved by tBLASTn
searches at the National Centre for Biotechnology Information (NCBI) (http://www.ncbi.nlm.nih.gov), the
Institute for Genomic Research (TIGR) (http://www.
tigr.org/tdb/), the Department of Energy Joint Genome
Institute Microbial Genome Project (http://www.jgi.
doe.gov/JGI_microbial/html/index.html), and Cyano
Base (http://www.kazusa.or.jp/cyanobase/) with the A.
thaliana PTOX protein sequence (NM_118352). A
Chlamydomonas reinhardtii PTOX sequence identified in this search was then used to search the genomes
available on the NCBI and CyanoBase servers and
identified a PTOX protein in the genome of the cyanobacterium Anabaena variabilis PCC 7120. This
sequence was then used to recover other prokaryotic
PTOXs.
Sequence alignment and phylogenetic analyses
Nucleotide sequences were translated into amino acid
sequences with the Translate tool located on the ExPASy server (http://us.expasy.org/tools/dna.html). All
deduced amino acid sequences were aligned by means
of Clustal X Version 1.81 with default gap penalties
(Thompson et al., 1997). Only sequences complete
enough to include all four of the iron-binding motifs (Andersson and Nordlund, 1999; Berthold et al.,
2002) were used in multiple sequence alignments and
for the generation of phylogenies. The first and fourth
iron-binding motifs served as the boundaries for analysis. This region encompasses a total of ca. 146 amino
acids for AOX proteins, and ca. 165 amino acids for
PTOX proteins. In total, 67 proteins (52 AOX, 15
PTOX) in gene databases were complete enough for
these analyses. Species names and accession numbers
for these can be found in Figure 4.
Phylogenetic analyses were conducted with MEGA
version 2.1 (Kumar et al., 2001). The neighborjoining method was used with the p-distance model,
but phylogenies generated with the number of distances model or the Poisson model yielded identical
topologies. Bootstrap analyses consisted of 1000 replicates. Phylogenies were also generated using the
maximum parsimony method in order to verify topology. Phylogenies were prepared for publication with
Adobe Illustrator 9.0 (Adobe Systems, California).
DNA and RNA analyses
DNA and RNA were isolated from A. variabilis
with TRIzol reagent (Invitrogen Life Technologies,
Carlsbad, CA). DNA was eliminated from RNA
samples prior to RT-PCR with amplification grade
DNase I (Invitrogen Life Technologies). Oligonucleotide primers for PCR and RT-PCR were designed
with Omiga 2.0 software (Genetics Computer Group,
Madison, WI). Two primers were designed to amplify
a 511 bp PTOX fragment from A. variabilis (forward, 5 -TCCTCGCTTTTATGTACTAGAAACC-3;
reverse, 5 -GTGTTCCATTTCATCATCACG-3).
867
PCR was conducted in a reaction containing 5 µl
10× PCR buffer, 1 µl 10 mM dNTP mixture, 2 µl of
each primer (10 µM), 2 µl DNA (0.3 µg/µl) 0.25 µl
Taq DNA polymerase, and 38 µl dH2 O, all covered
by 2 drops of mineral oil. The thermal cycler program
consisted of 2 min at 94 ◦ C, followed by 36 cycles of
45 s at 94 ◦ C, 30 s at 49 ◦ C, and 2 min at 58 ◦ C. A
final 10 min extension at 58 ◦ C was performed before
cooling the tubes to 4 ◦ C. Aliquots of the PCR reaction
were then analyzed by restriction digests and agarose
gel electrophoresis.
RT-PCR was performed with the Access RT-PCR
System (Promega, Madison, WI). The reaction contained 1 µg RNA, 1 µl primers (50 pmol), 1 µl AMV
reverse transcriptase, 1 µl Tfl DNA polymerase, 10 µl
AMV/Tfl 5× buffer, 1 µl dNTP mix (10 mM), 2 µl of
25 mM MgSO4 and DEPC-treated water up to 50 µl.
The thermal conditions used were: first-strand cDNA
synthesis, 48 ◦ C for 45 min (1 cycle); reverse transcriptase inactivation and initial denaturation, 94 ◦ C
for 2 min (1 cycle); denaturation, 94 ◦ C for 30 s; annealing and extension, 60 ◦ C for 1 min followed by
68 ◦ C for 2 min (40 cycles); final extension, 68 ◦ C for
7 min (1 cycle). An aliquot of the RT-PCR reaction
was then analyzed by agarose gel electrophoresis.
For northern analyses, total cellular RNA (20 µg)
was separated on agarose gels containing formaldehyde and transferred onto Hybond-N membrane
(Amersham Biosciences) with standard protocols
(Maniatis et al., 1982). An RNA ladder was included on the gel. A 511 bp fragment of A. variabilis
PTOX cDNA amplified by RT-PCR (see above) and
radiolabeled with the T7 Quick Prime Kit (Amersham
Biosciences) was used as the probe. Hybridization and
wash conditions were done according to Church and
Gilbert (1984) followed by exposure to X-ray film.
Results
Identification of eukaryotic and prokaryotic AOX and
PTOX proteins
Using the strategies outlined in Materials and methods, we sought to identify all AOX and PTOX proteins
present in the public sequence databases. After distinguishing in our search results between PTOX and
AOX proteins (see below), 20 PTOX sequences were
found, including the five sequences previously annotated for the plants A. thaliana, Capsicum annuum,
Gossypium hirsutum, Lycopersicon esculentum and
Figure 1. Identification and expression of PTOX in the cyanobacterium Anabaena variabilis. A. Analysis by agarose gel electrophoresis and restriction digests of PCR reaction products generated
with DNA isolated from Anabaena variabilis PCC 7120 and the
designed primers. Lanes: 1, DNA ladder; 2, 511 bp PCR product; 3,
PCR product digested with ClaI (460 and 51 bp fragments); 4, PCR
product digested with PvuII (278 and 213 bp fragments); 5, PCR
product digested with ClaI and PvuII (227 and 213 bp fragments);
6, control PCR without template DNA; 7, control PCR reaction
without Taq DNA polymerase; 8, DNA ladder. B. Analysis by
agarose gel electrophoresis of RT-PCR reaction products generated
with RNA isolated from A. variabilis PCC 7120 and the designed
primers. Lanes: 1, DNA ladder; 2, RT-PCR kit control 323 bp
product; 3–7, RT-PCR reaction products (3 and 5, template RNA
was not DNase-treated prior to RT-PCR; 5 and 6, RT-PCR reaction
contained no reverse transcriptase; 7, RT-PCR reaction contained
no template RNA). The results show that a reaction product of the
expected size (511 bp) is generated with DNase-treated template
RNA in the presence (lane 4) but not the absence (lane 6) of reverse transcriptase, thus precluding the possibility that the product
is being amplified from contaminating DNA in the template RNA.
C. Northern blot analysis of total RNA showing the expression
of A. variabilis PTOX. The left panel shows the ethidium bromide-stained gel. The right panel shows the northern blot. An RNA
ladder predicted the transcript to be ca. 800 bp.
868
Table 1. Identification of PTOX proteins in cyanobacteria, algae and higher plants.
Group
Species
Accession number
or identifier
Previous annotation
Anabaena variabilis PCC 7120
Gloeobacter violaceus PCC 7421
Prochlorococcus marinus MED4
Synechococcus sp. WH 8102
NP_486136a
gll0601c
CAE18795a
NP_896980a
oxidase
oxidase
hypothetical protein
hypothetical protein
Bigelowiella natans
Chlamydomonas reinhardtii
AY267664a
AAM12876a
chloroplast quinol-to-oxygen oxidoreductase
chloroplast quinol-to-oxygen oxidoreductase
Hordeum vulgare
Oryza sativa
Secale cereale
Sorghum bicolor
Triticum aestivum
Zea mays
TC88984b
AAC35554a
BE704987a
TC41730b
TC68206b
TC163623b
EST of unknown function
oxidase
EST of unknown function
oxidase
oxidase
oxidase
Arabidopsis thaliana
Capsicum annuum
Glycine max
Gossypium hirsutum
Lotus japonicus
Lycopersicon esculentum
Medicago truncatula
Mesembryanthemum crystallinum
AF098072a
AF177981a
BG315967a
AI730398a
BI419452a
AF302932a
TC80949b
TC3799b
IMMUTANS / PTOX
PTOX
hypothetical 38.7 kDa protein
PTOX
EST of unknown function
GHOST / PTOX
EST of unknown function
PTOX
Cyanobacteria
Algae
Monocotyledons
Eudicotyledons
a at the GenBank database; b at the TIGR database; c at CyanoBase.
Mesembryanthemum crystallinum (Table 1). In the
Oryza sativa, Sorghum bicolor, Triticum aestivum and
Zea mays plants, proteins previously annotated simply
as oxidases can now also be classified as PTOXs.
In addition, several ESTs of unknown function and
several hypothetical proteins in the plants Glycine
max, Hordeum vulgare, Lotus japonicus, Medicago
truncatula and Secale cereale are also PTOXs. A
protein in the green alga Chlamydomonas reinhardtii
and one in the alga Bigelowiella natans characterized
as chloroplast quinol-to-oxygen oxidoreductases can
now also be classified as PTOXs. Our search also recovered many AOX proteins from a wide variety of
taxa including plants, algae, oomycetes, fungi, and
protists.
Significantly, our database searches also revealed
prokaryotic orthologues of AOX and PTOX proteins. PTOXs were found in the sequenced genomes of four cyanobacteria: Anabaena variabilis PCC
7120, Synechococcus sp. WH 8102, Prochlorococcus
marinus MED4, and Gloeobacter violaceus PCC 7421
(Table 1), while an AOX was retrieved from the genome of the α-proteobacterium Novosphigobium aromaticivorans. The presence of these prokaryotic proteins in the database (with the exception of that from
G. violaceus) has been recently reported (Finnegan
et al., 2003; Stenmark and Nordlund, 2003).
Expression of the PTOX orthologue in cyanobacteria
To confirm the presence and expression of a PTOX
orthologue in cyanobacteria, we designed primers to
amplify a PTOX fragment from A. variabilis PCC
7120. PCR of DNA was used to confirm the presence of PTOX in the prokaryotic genome. Analysis
of the PCR reaction indicated a single product of the
expected size (511 bp) and restriction digests of this
PCR product generated, in each case, fragments of the
expected size (Figure 1A).
RT-PCR of DNase-treated RNA extracted from
A. variabilis was used to confirm expression of the
PTOX orthologue. RT-PCR (with the same primers as
869
above) again generated the expected 511 bp product,
confirming the presence of PTOX transcript (Figure 1B). Expression was also confirmed by northern
analysis of total RNA, which generated a band of
ca. 800 bp (Figure 1C). The A. variabilis PTOX
sequence was analyzed in the upstream region of
the start codon with Softberry’s BPROM software
(http://www.softberry.com). This analysis predicted a
Pribnow box (GTTTATAAC) at position –60 to –52
(10 nucleotides from where the actual transcription
begins) and a –35 region (CTCACA) at position –83
to –78 relative to the PTOX start codon. Furthermore, secondary structure in the untranslated region
200 bp downstream of the stop codon was analyzed
with the RNA Folding application of the mfold server
(http://www.bioinfo.rpi.edu/applications/mfold/). This
analysis revealed a hairpin loop at position 50–67 relative to the PTOX stop codon with a closing pair A
and U at positions 54 and 63. The predicted hairpin
loop might serve as a transcription terminator structure. Considering these findings, a transcript size of
about 815 bp is predicted, which is in good agreement
with the size of transcript observed (Figure 1C).
Multiple sequence alignment of AOX and PTOX
proteins
We examined the level of sequence similarity between
AOX and PTOX proteins with a large set of sequences
that included many newly found eukaryotic proteins
and, most importantly, the recently found prokaryotic
proteins. Both AOX and PTOX are thought to contain a diiron carboxylate centre in their active site
that is coordinated by two conserved His residues and
four other carboxylate (Glu) residues, all comprising
four iron-binding motifs distributed within the central
region of the protein sequence (Andersson and Nordlund, 1999; Berthold et al., 2000, 2002). Hence, we
focused our analysis on the protein sequence bordered
by the first and fourth iron-binding motifs, hypothesizing that a high degree of sequence conservation
would exist in this central region to maintain protein
function but that there might nonetheless exist differences between the distantly related AOX and PTOX
groups (Carol and Kuntz, 2001) that could be used to
differentiate the two groups.
In all, 52 AOX and 15 PTOX sequences found
in public databases encompassed all four iron-binding
motifs and a multiple sequence alignment of these revealed consistent similarities and differences between
AOX and PTOX proteins. To illustrate this, a smaller
alignment including just eight AOX and six PTOX sequences is shown in Figure 2. This alignment includes
thirteen species from diverse taxa and is representative
of the diversity in AOX and PTOX protein sequences
seen in the larger alignment.
The alignment showed that all six putative ironbinding residues (E-183, E-222, H-225, E-273, E-324,
H-327) are conserved in all PTOX and AOX proteins
(Figure 2). Besides these, the following amino acids
were also completely conserved: L-182, A-186, H198, N-221, A-276, Y-280, and D-323 (numbering
according to Berthold et al., 2000). The alignment
also showed the presence of a ca. 20 amino acid insert between the third and fourth iron-binding motifs
of all eukaryotic PTOX proteins that all AOX proteins
lacked (Carol et al., 1999). Interestingly, the insert was
also present in the PTOX of the cyanobacteria A. variabilis and G. violaceus but was absent in P. marinus
and reduced to about half the length in Synechococcus
(Figure 2).
Toward the N-terminus, there are significant differences between the PTOX proteins of plants, algae and
cyanobacteria (data not shown). The N-terminus of
the C. reinhardtii PTOX is extremely long compared
to all others with 278 amino acid residues prior to
iron-binding motif 1. Conversely, the N-termini of the
cyanobacterial PTOXs are very short: 27 amino acids
in A. variabilis and G. violaceus, 30 in P. marinus and
36 in Synechococcus sp. The length of the A. thaliana PTOX N-terminus is between those of algae and
cyanobacteria at 134 amino acids, for which the first
51 amino acids comprise the plastid targeting peptide
(Carol and Kuntz, 2001). Similarly, the N-terminus of
the N. aromaticivorans AOX is extremely short compared to AOXs of plants, algae, and fungi (data not
shown). While the length of the N-terminus of the
C. reinhardti AOX1, A. thaliana AOX2, and N. crassa
AOX is 187, 179 and 150 amino acids, respectively,
that of N. aromaticivorans is only 47 amino acids prior
to iron-binding motif 1.
The multiple sequence alignment showed that consistent differences existed between the iron-binding
motifs of the 52 AOX proteins compared to the 15
PTOX proteins. For example, in iron-binding motif 1,
the sixth amino acid is an A or G in AOX, but an R
in PTOX, while the ninth amino acid is a G in AOX,
but a Y in PTOX (Figure 3). In a similar way, consistent differences can be established for each of the
other iron-binding motifs (Figure 3). Of special note
is the third amino acid in iron-binding motif 4. It is
usually an A in AOX and a D in PTOX. This residue
870
Figure 2. A multiple sequence alignment of eight AOX and six PTOX proteins. This alignment includes proteins from thirteen species representative of a wide range of prokaryotic and eukaryotic taxa. Stars above the alignment indicate identical amino acids in all sequences analyzed.
These include two His and four Glu resides that are putative iron-binding residues (see Figure 3). The colors and the other symbols above the
alignment follow the default conventions defined by the ClustalX program, regarding conservation of amino acids between sequences. The
PTOX insertion refers to a 19–20 amino acid insert that is present in most PTOX proteins but lacking in AOX proteins. The numbering of
important residues (see Discussion) is the same as in Berthold et al. (2000) and represents the A. thaliana Aox1a sequence (AF370166).
is critical for antibody recognition by a widely used
monoclonal antibody to AOX termed AOA (Elthon
et al., 1989). Substitution of the third amino acid (A)
abolishes the ability of AOA to recognize AOX (Finnegan et al., 1999). Clearly, this residue differs in
PTOX, providing an explanation for the apparent inability of the AOA antibody to recognize PTOX proteins.
Based on this logic, we hypothesize that the AOA antibody would also fail to recognize an AOX protein of
Zea mays (AAL27795) and the AOX of D. discoideum
(BAB82989).
The differences in iron-binding motifs (Figure 3)
hold true for AOX and PTOX proteins from a wide
range of prokaryotic and eukaryotic taxa and, together,
should facilitate the annotation of new proteins in
sequence databases.
members. The two algal PTOXs group together, as
do the PTOXs from the cyanobacteria A. variabilis
and G. violaceus. These two cyanobacterial PTOXs
do not group with the other cyanobacterial proteins,
but rather are placed between the plant and algal
groups. This placement of A. variabilis and G. vi-
Global phylogeny of prokaryotic and eukaryotic AOX
and PTOX proteins
The 52 AOX and 15 PTOX sequences were used
to generate an unrooted protein phylogeny by means
of distance methods. This analysis showed that the
prokaryotic and eukaryotic PTOXs grouped separate
from the eukaryotic and prokaryotic AOXs (Figure 4).
Within the PTOX group, the proteins from Synechococcus sp.and P. marinus are the most divergent
Figure 3. Comparison of the iron-binding motifs of AOX and
PTOX proteins. Black circles show the putative iron-binding
residues, all of which are completely conserved between AOX and
PTOX proteins. Black arrows indicate amino acids that can be reliably used to distinguish between AOX and PTOX proteins. The
star indicates a key residue required for recognition by the AOX
antibody AOA (Elthon et al., 1989; Finnegan et al., 1999). The first
residue in iron-binding motif 1 corresponds to the first amino acid
in the Figure 2 alignment. Results are based upon the analysis of 52
AOX and 15 PTOX sequences available in public databases.
871
olaceus was also confirmed by maximum parsimony
analysis. Amongst plants, the monocot PTOXs are
separate from the dicot PTOXs.
Within the AOX group, the proteins from the slime
mold D. discoideum and the animal parasite T. brucei
brucei are the most divergent members (100% bootstrap support). The remainder of AOX proteins fall
into two large groups, one containing plants and the
α-proteobacterial protein, the other containing fungi,
green algae, and oomycetes.
Several plant species contain a small AOX gene
family and so in some cases the phylogeny includes
multiple proteins from one species. Consistent with
previous results, the plant proteins fall into two subgroups (98% bootstrap support), one found exclusively in dicots, the other found in both monocots and
dicots (Considine et al., 2002).
Within the fungi, the AOX of Cryptococcus neoformans is the most divergent. The remaining fungal
AOXs fall into two subfamilies (81% bootstrap support). As with plants, there is evidence for the expansion of fungal AOXs, as Candida albicans has two
AOX proteins.
Discussion
This work explores the presence of prokaryotic orthologues of plastid PTOX and mitochondrial AOX proteins. PTOX proteins are found in the sequenced genomes of four different cyanobacteria (three of which
were recently reported by Finnegan et al., 2003) while
an AOX protein is present in the α-proteobacterium
N. aromaticivorans as recently reported by Stenmark
and Nordlund (2003). PCR confirmed the presence of
PTOX in the genome of the cyanobacterium A. variabilis PCC 7120 while RT-PCR and northern analyses
showed that the gene is indeed expressed.
AOX and PTOX are both members of the
membrane-bound diiron carboxylate oxidases (Berthold and Stenmark, 2003). A pairwise alignment
of the AOX1a (AF370166) and PTOX proteins of
A. thaliana indicates that the proteins share less than
50% sequence similarity (data not shown) and our
phylogeny (as well as those of Finnegan et al., 2003)
indicate that AOX and PTOX proteins represent two
distinct groups. These findings support the practice
that these proteins retain different common names.
However, with the presence of PTOX in cyanobacteria, the naming of these proteins as plastid terminal
oxidases becomes problematic. For simplicity, we sug-
gest retaining the acronym PTOX, but recommend it
stand for plastoquinol terminal oxidase. This designation is supported by the finding that the A. thaliana
PTOX does indeed oxidize plastoquinol (Joet et al.,
2002; Josse et al., 2003).
Sequence conservation amongst eukaryotic and
prokaryotic AOX and PTOX proteins
Our multiple sequence alignment identified several
residues that were conserved in all of the large set of
eukaryotic and prokaryotic AOX and PTOX proteins
analyzed. All the putative iron-binding residues were
conserved, strongly supporting the view that these
are the critical residues involved in the formation of
a diiron carboxylate center in both AOX and PTOX
(Berthold et al., 2000). Strict conservation amongst
other residues provides a starting point toward studying various aspects of structure and function (Berthold
et al., 2000).
In the third iron-binding motif, the only Glu conserved in all PTOX and AOX proteins is E-273, the
proposed iron-binding Glu in recent AOX models
(Berthold et al., 2000). However, mutation of E-275 to
N results in an inactive AOX (Albury et al., 2002). The
most likely explanation for this is that even disruption
of non-iron-binding residues in this region can disrupt the diiron center. Several PTOXs have an H-275,
while several AOXs and PTOXs possess multiple Glu
residues in the range of 273 to 275 (Figure 3). Hence,
species-specific flexibility in the ligation sphere in this
region is also possible (Finnegan et al., 2003).
Many other conserved residues are next to, or
in close proximity to, the proposed iron-binding
residues. This is the case for L-182, A-186, N-221,
A-276, and D-323. With the exception of A-186, the
conservation of these residues has been noted before
amongst a smaller group of non-prokaryotic AOX and
PTOX proteins (Berthold et al., 2000). It is not known
whether the conservation of these residues is simply
related to their proximity to iron-binding residues,
or whether they play a more direct role in catalysis,
regulation, or structure.
Other residues that are conserved in all of the AOX
and PTOX proteins include H-198 and Y-280. Y-280
has been hypothesized to play a role in Tyr radical
formation in the oxidation reaction, electron transfer,
or quinol substrate interactions (Berthold et al., 2000)
and recent site-directed mutagenesis of this residue
to F yielded an inactive AOX (Albury et al., 2002).
Alternatively, Y-258 is not conserved in all AOX and
872
Figure 4. Phylogeny of eukaryotic and prokaryotic AOX and PTOX proteins, generated by the Neighbour-Joining method with the p-distance
model. Bootstrap values are based on 1000 trials. Branch lengths are indicated by the scale at the bottom left.
873
PTOX proteins (it is a T in Synechococcus sp.) suggesting it is not a critical Tyr for function. Consistent
with this hypothesis, the mutation of this residue has
no effect on AOX activity (Albury et al., 2002).
Including the iron-binding residues, a total of 13
amino acids are completely conserved in the region
bordered by the first and fourth iron-binding motifs of
all AOX and PTOX proteins. To our knowledge, the
following conserved residues have not yet been investigated: L-182, E-183, A-186, H-198, N-221, A-276,
and D-323.
The endosymbiotic origin of eukaryotic AOX and
PTOX proteins
The phylogeny of eukaryotic and prokaryotic AOX
and PTOX proteins indicates that AOX and PTOX
proteins represent two distinct groups. Nonetheless,
given their sequence similarity, we assume that these
proteins would have at one time arisen from a single
bacterial ancestral protein, as recently proposed by
Finnegan et al. (2003). Below, we offer two scenarios
for the origin of eukaryotic AOX and PTOX proteins.
The prokaryotic AOX sequence is found in an
α-proteobacterium while the prokaryotic PTOX sequence is found in several cyanobacteria. This is
particularly intriguing given that an α-proteobacteriallike organism is thought to have given rise to mitochondria (Gray et al., 2001) while the chloroplast is
thought to have arisen from a cyanobacterial symbiont
(Cavalier-Smith, 2002). The most straightforward proposal, then, is that modern-day mitochondrial AOXs
arose from the endosymbiotic event that gave rise to
mitochondria while modern-day chloroplast PTOXs
arose from the endosymbiotic event that gave rise to
chloroplasts (Finnegan et al., 2003). Given this, the
divergence of AOX and PTOX proteins would have
begun prior to the endosymbiotic events that gave rise
to eukaryotes. Our phylogeny supports this in that the
prokaryotic AOX is more similar to the eukaryotic
AOX than to the prokaryotic PTOX and conversely,
the prokaryotic PTOX is more similar to the eukaryotic PTOX than to the prokaryotic AOX. Assuming
an endosymbiotic origin for these two proteins, the
genes encoding these proteins would have later moved
from the organelle genomes to the nucleus or in some
cases simply been lost from the organism.
An alternative possibility is that an ancestral AOX
or PTOX protein was introduced to the eukaryotic line
through a single endosymbiotic event, upon which the
ancestral gene was relocated to the nuclear genome
and then duplicated. Over time, these copies would
diverge to give rise to AOX and PTOX proteins being targeted to different organelles. Recently, Martin
et al. (2002) compared 24 990 proteins encoded in
the A. thaliana nuclear genome to the proteins from
three cyanobacterial genomes and estimated that ca.
18% of the protein-coding nuclear genes (ca. 4500
genes) were acquired from the cyanobacterial ancestor
of plastids. Significantly, more than half of these proteins are predicted to now be non-plastid, indicating
that gene origin is not a good predictor of protein
compartmentation. Based on such findings, it seems
reasonable to suggest that AOX and PTOX may have
both arisen from a single gene in the cyanobacterial or
proto-mitochondrial endosymbiont that subsequently
diverged into the two proteins after endosymbiosis.
However, if this were the case, we would expect AOX
and PTOX proteins to group more closely together
than our protein phylogeny indicates. This model
would also fail to readily explain the presence of AOX
and PTOX sequences in extant organisms that are
the closest living relatives to proto-mitochondria and
proto-chloroplasts, respectively.
The AOX proteins
It is proposed that the α-protobacteria are the closest
representation of the proto-mitochondrion. It is interesting that the AOX found in the α-proteobacterium
is most closely related to higher-plant AOXs. For instance, one might have expected it to group more
closely with the trypanosomal AOX given that organisms in this group are known to be amongst the
earliest emerging mitochondrial protists (Emelyanov,
2003). This seemingly odd result may be explained if
one considers that the genes encoding the plant AOXs
would have originally been present in mitochondrial
DNA. In extant organisms, mitochondrial DNA is
classified as either ancestral or derived (Gray et al.,
1999). Ancestral groups (which include higher plants)
possess a gene content that approximates that of ancestral protist mitochondrial DNA, suggesting that such
DNA has been buffered against extensive loss. Derived groups (which includes fungi and algae) have
undergone a great deal of reductionism in their mitochondrial genomes. When viewed in this light, the
clustering of the α-proteobacterial AOX with the plant
AOX group is less surprising. This may also explain
the clustering of the algal, oomycete and fungal AOXs.
Another possibility is that the similarity of the bacterial AOX to that of plants is due to horizontal gene
874
transfer from plants to bacteria as previously suggested (Finnegan et al., 2003; Stenmark and Nordlund,
2003). However, our BLAST searches found no evidence for such transfer of AOX to other less ancient
bacteria. Also, there is still much debate over the range
and frequency of horizontal gene transfer from plants
to bacteria (Kurland et al., 2003). It is considered
extremely rare and is constrained by the bacterium
needing to contain DNA of significant homology to
the transfer DNA and by the recombination event surviving the action of mismatch repairing enzymes (de
Vries et al., 2001; Tepfer et al., 2003).
The PTOX proteins
The phylogeny indicated two groups of PTOX proteins within the cyanobacteria. Synechococcus and
P. marinus grouped separately from A. variabilis and
G. violaceus, with the algal PTOXs between them. Interestingly, Rubisco phylogenies classify Synechococcus and P. marinus as α-cyanobacteria, while A. variabilis is a β-cyanobacterium (Badger and Price, 2003).
It is believed that the divergence of these two groups
was a very ancient event that may have pre-dated
primary endosymbiosis. Studies also suggest that βcyanobacterial Rubisco Form 1B, such as that found in
A. variabilis, is more closely related to that of higher
plants than is the α-cyanobacterial Rubisco Form 1A
(Badger and Price, 2003). Similarly, the PTOX phylogeny shows that plant/algal PTOXs are more closely
related to the PTOXs of β- than α-cyanobacteria.
PTOXs are thus far only found in photosynthetic
organisms. Despite rigorous searching, no PTOX protein could be identified in fungi or non-photosynthetic
heterotrophs. This is most likely the result of chloroplast loss (Cavalier-Smith, 2002). The clustering of the
PTOX of the chlorarachniophyte Bigelowiella natans
with that of C. reinhardtii is consistent with the theory that B. natans acquired photosynthesis secondarily
from a green alga (Figure 4; Archibald et al., 2003).
Interestingly, no PTOX was found in the sequenced
genome of the commonly studied Synechocystis sp.
PCC 6803, indicating that not all photosynthetic organisms have PTOX.
Peltier and Cournac, 2002). A chlororespiration-like
path is also hypothesized to exist in cyanobacteria
(Berry et al., 2002; Schreiber et al., 2002) and so
one working hypothesis is that PTOX plays a role in
such a path. To our knowledge, terminal oxidases potentially involved in chlororespiration-like pathways
in cyanobacteria have only been extensively studied
in Synechocystis sp. PCC 6803, a species for which
we found no evidence of a PTOX. Indeed, previous
analysis of the genome sequence indicated that, in
addition to the aa3 -type cyt c oxidase (CtaI), Synechocystis harbored two additional terminal respiratory
oxidases (Howitt and Vermass, 1998). These were a
quinol oxidase of the cyt bd type (Cyd) and an oxidase
resembling another cyt aa3 -type cyt c oxidase (CtaII).
Cyd (but not CtaII) can participate in plastoquinol oxidation in the thylakoids of this species (Berry et al.,
2002; Berry, 2003), suggesting it to be the most functionally similar to a PTOX. Interestingly, A. variabilis
contains both Cyd genes and a PTOX (Kaneko et al.,
2001).
The complete genomes of two different isolates of
P. marinus (MIT9313 and MED4) are available but we
only find PTOX in P. marinus MED4 (Table 1). Interestingly, MIT9313 is adapted to low light levels and,
despite having a much larger genome than MED4,
it lacks many of the high-light-inducible proteins
found in MED4 (Ting et al., 2002; Rocap et al.,
2003). This may indicate that PTOX is required for
high-light acclimation. It has been suggested that the
chlororespiration-like pathways in cyanobacteria may
act to prevent over-reduction of the plastoquinone pool
under high light (Berry et al., 2002). It has also been
shown that PTOX is induced by high-light stress in
A. thaliana (Rizhsky et al., 2002).
Acknowledgments
This work was supported by research grants from
the Natural Sciences and Engineering Research Council of Canada and the Premiers Research Excellence
Award of Ontario (both to G.C.V.) and by a Natural Sciences and Engineering Research Council of
Canada Postgraduate Scholarship (to A.E.M.).
Metabolic implications for a cyanobacterial PTOX
Chlororespiration in plastids represents a possible ‘alternate path’ in the photosynthetic electron transport
chain whereby electrons at plastoquinol can be passed
to oxygen, and it is now hypothesized that PTOX is the
oxidase responsible for this activity (Joet et al., 2002;
References
Albury, M.S., Affourtit, C., Crichton, P.G. and Moore, A.L. 2002.
Structure of the plant alternative oxidase: site-directed mutagenesis provides new information on the active site and membrane
topology. J. Biol. Chem. 277: 1190–1194.
875
Andersson, M.E. and Nordlund, P. 1999. A revised model of the
active site of alternative oxidase. FEBS Lett. 449: 17–22.
Archibald, J.M., Rogers, M.B, Toop, M., Ishida, K. and Keeling, P.J.
2003. Lateral gene transfer and the evolution of plastid-targeted
proteins in the secondary plastid-containing alga Bigelowiella
natans. Proc. Natl. Acad. Sci. USA 100: 7678–7683.
Badger, M.R. and Price, G.D. 2003. CO2 concentrating mechanisms in cyanobacteria: molecular components, their diversity and
evolution. J. Exp. Bot. 54: 609–622.
Berry, S. 2003. Endosymbiosis and the design of eukaryotic electron
transport. Biochim. Biophys. Acta 1606: 57–72.
Berry, S., Schneider, D., Vermass, W.F.J. and Rogner, M. 2002.
Electron transport routes in whole cells of Synechocystis sp.
strain PCC 6803: the role of the cytochrome bd-type oxidase.
Biochemistry 41: 3422–3429.
Berthold, D.A. and Stenmark, P. 2003. Membrane-bound diiron
carboxylate proteins. Annu. Rev. Plant Biol. 54: 497–517.
Berthold, D.A., Andersson, M.E. and Nordlund, P. 2000. New insight into the structure and function of the alternative oxidase.
Biochim. Biophys. Acta 1460: 241–254.
Berthold, D.A., Voevodskaya, N., Stenmark, P., Graslund, A. and
Nordlund, P. 2002. EPR studies of the mitochondrial alternative
oxidase. Evidence for a diiron carboxylate center. J. Biol. Chem.
277: 43608–43614.
Carol, P. and Kuntz, M. 2001. A plastid terminal oxidase comes to
light: implications for carotenoid biosynthesis and chlororespiration. Trends Plant Sci. 6: 31–36.
Carol, P., Stevenson, D., Bisanz, C., Breitenbach, J., Sandmann,
G., Mache, R., Coupland, G. and Kuntz, M. 1999. Mutations in
the Arabidopsis gene IMMUTANS cause a variegated phenotype
by inactivating a chloroplast terminal oxidase associated with
phytoene desaturation. Plant Cell 11: 57–68.
Cavalier-Smith, T. 2002. Chloroplast evolution: secondary symbiogenesis and multiple losses. Curr. Biol. 12: R62–R64.
Church, G.M. and Gilbert, W. 1984. Genomic sequencing. Proc.
Natl. Acad. Sci. USA 81: 1991–1995.
Considine, M.J., Holtzapffel, R.C., Day, D.A., Whelan, J. and Millar, A.H. 2002. Molecular distinction between alternative oxidase
from monocots and dicots. Plant Physiol. 129: 949–953.
Cournac, L., Redding, K., Ravenel, J., Rumeau, D., Josse, E.M.,
Kuntz, M. and Peltier, G. 2000. Electron flow between photosystem II and oxygen in chloroplasts of photosystem I-deficient
algae is mediated by a quinol oxidase involved in chlororespiration. J. Biol. Chem. 275: 17256–17262.
de Vries, J., Meier, P. and Wackernagel, W. 2001. The natural
transformation of the soil bacteria Pseudomonas stutzeri and
Acinetobacter sp. by transgenic plant DNA strictly depends on
homologous sequences in the recipient cells. FEMS Microbiol.
Lett. 195: 211–215.
Elthon, T.E., Nickels, R.L., and McIntosh, L. 1989. Monoclonal antibodies to the alternative oxidase of higher plant mitochondria.
Plant Physiol. 89: 1311–1317.
Emelyanov, V.V. 2003. Mitochondrial connection to the origin of
the eukaryote cell. Eur. J. Biochem. 270: 1599–1618.
Finnegan, P.M., Umbach, A.L. and Wilce, J.A. 2003. Prokaryotic
origins for the mitochondrial alternative oxidase and plastid
terminal oxidase nuclear genes. FEBS Lett. 555: 425–430.
Finnegan, P.M., Wooding, A.R. and Day, D.A. 1999. An alternative oxidase monoclonal antibody recognizes a highly conserved
sequence among alternative oxidase subunits. FEBS Lett. 447:
21–24.
Gray, M.W., Burger, G. and Lang, B.F. 1999. Mitochondrial
evolution. Science 283: 1476–1481.
Gray, M.W., Burger, G. and Lang, B.F. 2001. The origin and early
evolution of mitochondria. Genome Biol. 2: 1018.1–1018.5.
Howitt, C.A. and Vermaas, W.F.J. 1998. Quinol and cytochrome
oxidases in the cyanobacterium Synechocystis sp. PCC6803.
Biochemistry 37: 17944–17951.
Joet, T., Genty, B., Josse, E.M., Kuntz, M., Cournac, L. and
Peltier, G. 2002. Involvement of a plastid terminal oxidase
in plastoquinone oxidation as evidenced by expression of the
Arabidopsis thaliana enzyme in tobacco. J. Biol. Chem. 277:
31623–31630.
Josse E.-M., Alcaraz J.-P., Labouré A.-M. and Kuntz, M. 2003. In
vitro characterization of a plastid terminal oxidase (PTOX). Eur.
J. Biochem. 270: 3787–3794.
Kaneko, T., Nakamura, Y., Wolk, C.P., Kuritz, T., Sasamoto,
S., Watanabe, A., Iriguchi, M., Ishikawa, A., Kawashima, K.,
Kimura, T., Kishida, Y., Kohara, M., Matsumoto, M., Matsuno, A., Muraki, A., Nakazaki, N., Shimpo, S., Sugimoto, M.,
Takazawa, M., Yamada, M., Yasuda, M. and Tabata, S. 2001.
Complete genomic sequence of the filamentous nitrogen-fixing
cyanobacterium Anabaena sp. strain PCC 7120. DNA Res. 8:
205–213.
Kumar, S., Tamura, K., Jakobsen, I.B. and Nei, M. 2001. MEGA2:
Molecular Evolutionary Genetics Analysis Software, Arizona
State University, USA.
Kurland, C.G., Canback, B. and Berg O.G. 2003. Horizontal gene
transfer: a critical view. Proc. Natl. Acad. Sci. USA 100: 9658–
9662.
Maniatis, T., Fritsch, E.F. and Sambrook, J. 1982. Molecular
Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory
Press, Plainview, NY.
Martin, W., Rujan, T., Richly, E., Hansen, A., Cornelsen, S.,
Lins, T., Leister, D., Stoebe, B., Hasegawa, M. and Penny, D.
2002. Evolutionary analysis of Arabidopsis, cyanobacterial, and
chloroplast genomes reveals plastid phylogeny and thousands of
cyanobacterial genes in the nucleus. Proc. Natl. Acad. Sci. USA
99: 12246–12251.
Møller, I.M. 2001. Plant mitochondria and oxidative stress: Electron
transport, NADPH turnover, and metabolism of reactive oxygen
species. Annu. Rev. Plant Physiol. Plant Mol. Biol. 52: 561–591.
Moore, A.L., Albury, M.S., Crichton, P.G. and Affourtit, C. 2002.
Function of the alternative oxidase: is it still a scavenger? Trends
Plant Sci. 7: 478–481.
Peltier, G. and Cournac, L. 2002. Chlororespiration. Annu. Rev.
Plant Biol. 53: 523–550.
Rippka, R., Deruelles, J., Waterbury, J.B., Herdman, M. and Stanier,
R.Y. 1979. Generic assignments, strain histories and properties
of pure cultures of cyanobacteria. J. Gen. Microbiol. 111: 1–61.
Rizhsky, L., Hallak-Herr, E., Van Breusegem, F., Rachmilevitch, S.,
Barr, J.E., Rodermel, S., Inzé, D. and Mittler, R. 2002. Double
antisense plants lacking ascorbate peroxidase and catalase are
less sensitive to oxidative stress than single antisense plants
lacking ascorbate peroxidase or catalase. Plant J. 32: 329–342.
Robson, C.A. and Vanlerberghe, G.C. 2002. Transgenic plant
cells lacking mitochondrial alternative oxidase have increased
susceptibility to mitochondria-dependent and -independent pathways of programmed cell death. Plant Physiol. 129: 1908–1920.
Rocap, G., Larimer, F.W., Lamerdin, J., Malfatti, S., Chain, P., Ahlgren, N.A., Arellano, A., Coleman, M., Hauser, L., Hess, W.R.,
Johnson, Z.I., Land, M., Lindell, D., Post, A.F., Regala, W.,
Shah, M., Shaw, S.L., Steglich, C., Sullivan, M.B., Ting, C.S.,
Tolonen, A., Webb, E.A., Zinser, E.R. and Chisholm, S.W. 2003.
Genome divergence in two Prochlorococcus ecotypes reflects
oceanic niche differentiation. Nature 424: 1042–1047.
876
Scherer, S. 1990. Do photosynthetic and respiratory electron transport chains share redox proteins? Trends Biochem. Sci. 15:
458–462.
Schreiber, U., Gademann, R., Bird, P., Ralph, P., Larkum, A.W.D.
and Kuhl, M. 2002. Apparent light requirement for activation of photosynthesis upon rehydration of desiccated beachrock
microbial mats. J. Phycol. 38: 125–134.
Siedow, J.N. and Umbach, A.L. 2000. The mitochondrial cyanideresistant oxidase: structural conservation amid regulatory diversity. Biochim. Biophys. Acta 1459: 432–439.
Simons, B.H. and Lambers, H. 1999. The alternative oxidase: is it a
respiratory pathway allowing a plant to cope with stress? In: H.R.
Lerner (Ed.) Plant Responses to Environmental Stresses: From
Phytohormones to Genome Reorganization, Marcel Dekker,
New York, pp. 265–286.
Stenmark, P. and Nordlund, P. 2003. A prokaryotic alternative oxidase present in the bacterium Novosphingobium aromaticivorans. FEBS Lett. 552: 189–192.
Tepfer, D., Garcia-Gonzales, R., Mansouri, H., Seruga, M., Message, B., Leach, F., and Perica, M.C. 2003. Homology-dependent
DNA transfer from plants to a soil bacterium under laboratory conditions: implications in evolution and horizontal gene
transfer. Transgenic Res. 12: 425–437.
Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F. and
Higgins, D.G. 1997. The ClustalX windows interface: flexible strategies for multiple sequence alignment aided by quality
analysis tools. Nucl. Acids Res. 24: 4876–4882.
Ting, C.S., Rocap, G., King, J., and Chisholm, S.W. 2002.
Cyanobacterial photosynthesis in the oceans: the origins and
significance of divergent light-harvesting strategies. Trends Microbiol. 10: 134-142.
Wu, D.Y., Wright, D.A., Wetzel, C., Voytas, D.F. and Rodermel, S.
1999. The IMMUTANS variegation locus of Arabidopsis defines
a mitochondrial alternative oxidase homolog that functions during early chloroplast biogenesis. Plant Cell 11: 43–56.