BMC Evolutionary Biology
BioMed Central
Open Access
Research article
Phylogenomics of the archaeal flagellum: rare horizontal gene
transfer in a unique motility structure
Elie Desmond1, Celine Brochier-Armanet2,3 and Simonetta Gribaldo*1
Address: 1Unite Biologie Moléculaire du Gène chez les Extremophiles, Institut Pasteur, 25 rue du Dr. Roux, 75724 Paris Cedex 15, France,
2Université de Provence Aix-Marseille I, Marseille, France and 3Laboratoire de chimie bactérienne, Institut de Biologie Structurale et de
Microbiologie (CNRS), Marseille, France
Email: Elie Desmond - [email protected]; Celine Brochier-Armanet - [email protected];
Simonetta Gribaldo* - [email protected]
* Corresponding author
Published: 2 July 2007
BMC Evolutionary Biology 2007, 7:106
doi:10.1186/1471-2148-7-106
Received: 28 February 2007
Accepted: 2 July 2007
This article is available from: http://www.biomedcentral.com/1471-2148/7/106
© 2007 Desmond et al; licensee BioMed Central Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abstract
Background: As bacteria, motile archaeal species swim by means of rotating flagellum structures
driven by a proton gradient force. Interestingly, experimental data have shown that the archaeal
flagellum is non-homologous to the bacterial flagellum either in terms of overall structure,
components and assembly. The growing number of complete archaeal genomes now permits to
investigate the evolution of this unique motility system.
Results: We report here an exhaustive phylogenomic analysis of the components of the archaeal
flagellum. In all complete archaeal genomes, the genes coding for flagellum components are colocalized in one or two well-conserved genomic clusters showing two different types of
organizations. Despite their small size, these genes harbor a good phylogenetic signal that allows
reconstruction of their evolutionary histories. These support a history of mainly vertical
inheritance for the components of this unique motility system, and an interesting possible ancient
horizontal gene transfer event (HGT) of a whole flagellum-coding gene cluster between
Euryarchaeota and Crenarchaeota.
Conclusion: Our study is one of the few exhaustive phylogenomics analyses of a noninformational cell machinery from the third domain of life. We propose an evolutionary scenario
for the evolution of the components of the archaeal flagellum. Moreover, we show that the
components of the archaeal flagellar system have not been frequently transferred among archaeal
species, indicating that gene fixation following HGT can also be rare for genes encoding
components of large macromolecular complexes with a structural role.
Background
Motile archaeal species swim by means of rotating flagellum structures driven by a proton gradient force [1,2], as
in bacteria [3]. Interestingly, although they are both
responsible for swimming, archaeal and bacterial flagella
are not homologous, either in terms of overall structure,
components and assembly (for a recent review see [4,5]).
The bacterial flagellum is a complex rotary structure made
up of as much as 20 proteins and composed of three
major parts -the basal body, the hook, and the filament.
Rotation is provided by an ATPase exploiting a proton gradient force, and can be switched by specific proteins in
Page 1 of 13
(page number not for citation purposes)
BMC Evolutionary Biology 2007, 7:106
response to attractants or repellents in the environment
through the chemotaxis system. The filament is a hollow
structure about 20 nm in diameter and is composed of a
single type of protein called flagellin. Bacterial flagellins
are assembled by a complex type III secretion system
located in the basal body and are added to the distal tip of
the flagellum after passing through the hollow cavity [4].
Much less is known about the archaeal flagellum. It has
been extensively studied in terms of components, assembly, and mutation experiments, in Halobacteria and
Methanococcales (for recent reviews [4-6]). The archaeal
flagellum is a structure thinner than its bacterial counterpart, where at least a filament and a hook are evident [79]. The archaeal flagellum has been shown to have a
unique symmetry in Halobacterium salinarium. In fact, it
has 3.3 subunits/turn of a 1.9 nm pitch left-handed helix
compared to 5.5 subunits/turn of a 2.6 nm pitch righthanded helix for plain bacterial flagellum filaments
[10,11]. The archaeal filament can be made up of different
types of homologous flagellin proteins (called FlaA or
FlaB). The filament is ~10 nm in diameter and is not hollow, resembling more to bacterial type IV pili in this
respect [10]. A few other characteristics of archaeal flagella
make them more alike bacterial pili than flagella: as bacterial pilins, archaeal flagellins (i) are made as preproteins
with short signal peptides that are processed by a recently
identified archaeal-specific signal peptidase (called FlaK)
[12-14] that shows weak sequence similarity with the bacterial pili leader peptidase PilD, (ii) are likely added at the
base of the filament as in bacterial pili, and (iii) undergo
glycosylation as post-translational modification [15] (see
[5] for a recent review). Moreover, one component of
archaeal flagella (FlaI) is homologous to bacterial PilT, an
ATPase involved in bacterial pilin export (a type II/IV
secretion system) and pilus retraction during twitching
motility [16]. However, none of the remaining archaeal
flagellum components are homologous to those of bacterial pili [5]. Moreover, bacterial pili are not rotating structures, and no specific anchoring structures have ever been
observed, indicating substantial differences between these
two cellular structures.
A number of putative flagellum accessory genes lie close
to flagellin genes in archaeal genomes (called flaC, flaD/
flaE, flaF, flaG, flaH, flaI, and flaJ) [5]. Their putative role
in flagellum structure and assembly was tentatively
deduced based on their sequences, cellular location, and
mutation experiments [4-6].
FlaC, FlaD, FlaH and FlaI are associated with the membrane fraction in Methanococcus voltae and may thus be
peripheral components of the archaeal flagellum [5,6].
FlaH, FlaI and FlaJ may be important for the assembly of
archaeal flagella and possibly form a secretion complex
http://www.biomedcentral.com/1471-2148/7/106
[5,6]. FlaH harbors a domain similar to that found in bacterial RecA-like ATPases, and FlaH mutants are nonmotile
and nonflagellated [17]. FlaJ contains many transmembrane domains, while FlaI probably encodes an ATPase
that may be important for flagellins export, similarly to
the role of its bacterial homologue, the pilin export
ATPase PilT, and/or for providing the force for rotation.
No experimental data are presently available for FlaG and
FlaF, although FlaG may be a component of the anchoring
system between the hook and the filament [6]. It may be
possible that some of the multiple flagellin proteins have
different roles in flagellum substructures other than the
filament [6]. Finally, additional components of the
archaeal flagellum may be encoded by genes that have not
yet been identified.
The uniqueness of the archaeal flagellum in terms of components, structure, and assembly indicates that archaeal
and bacterial flagella have distinct origins (i.e. they are
analogous systems). Interestingly, homologues of most
bacterial chemotaxis genes are found in archaeal genomes
[5,18], suggesting that archaeal and bacterial chemotaxis
systems are evolutionary related. However, their interaction with the flagellum system in Archaea remains largely
unknown (for a recent review see [19]). In this work, we
sought to contribute to the research on archaeal flagella
and archaeal motility in general performing an accurate
phylogenomic study (sensu Eisen [20]) of the archaeal
motility apparatus in terms of taxonomic distribution of
the genes coding for its components, their genomic context, and their phylogeny. This allowed us to sketch a fairly
detailed image on the origin and evolution of this macromolecular structure.
Results
Taxonomic distribution and genomic context
The taxonomic distribution of the genes coding for
archaeal flagellum components is congruent with that
presented in a recent review[5] and is generally consistent
with species descriptions [21]. Gene homologues for all
components of the archaeal flagellum are found in the
complete genomes of Archaea that are described as motile
[5] (indicated by M and ● signs in Figure 1). Conversely,
no homologues of genes coding archaeal flagellum proteins are found in the complete genomes of Archaea that
are described as non motile [5] (indicated by NM and ❍
signs in Figure 1). Although the representative of the
Methanosarcina genus (i.e. Methanosarcina mazei, Methanosarcina acetivorans and Methanosarcina barkeri) are
described as non-motile [21], at least a full complement
of homologues of the genes coding for archaeal flagellum
components is present in the complete genomes of these
species [5,22] (Red rectangles in Figure 1). Conversely, no
homologues of the genes coding for archaeal flagellum
components are found in the complete genome of Pyrob-
Page 2 of 13
(page number not for citation purposes)
BMC Evolutionary Biology 2007, 7:106
aculum aerophilum [5] (Yellow rectangle in Figure 1). This
is surprising since the genus Pyrobaculum is described as
"motile due to flagellation" in the Bergey's manual and a
picture is included of a "platinum shadowed cell showing
flagella of P. aerophilum" [21]. Similarly, no homologues
of the genes coding for archaeal flagellum components are
present in the complete genome of Methanopyrus kandleri
[5] (Yellow rectangle in Figure 1), which is described as
motile in the Bergey's manual [21].
In all archaeal genomes harboring flagellum components,
the corresponding genes are always organized into one or
two very well conserved clusters [5] (fla clusters, Figure 2).
The only exception is the gene coding for the preflagellin
peptidase FlaK, which is located close to the fla cluster in
Methanococcus jannaschii only [5]. flaK homologues are
nevertheless always present, at least in single copy at different locations in the other archaeal genomes, to the
notable exception of Aeropyrum pernix, Thermoplasma acidophilum and Thermoplasma volcanium. We verified that
these species do not harbor any homologue of PilD -the
bacterial prepilin peptidase- that they may have recruited
by horizontal gene transfer, and how they cope with the
absence of this enzymatic activity (or which non-homologous enzyme performs the function) remains puzzling.
A careful observation of gene order within each cluster
revealed two types of organizations, that we will hereafter
call fla1 and fla2 (Figure 2). fla1 clusters are characterized
by the presence of flaC, plus one or few copies of flaD
(also annotated as flaE), or by a fusion of flaC and flaD
(Figure 2A), whereas fla2 clusters lack these genes (Figure
2B). Nevertheless, psi-Blast searches revealed that the
genes (hyp1 and hyp2, Figure 2B) lying between flaB and
flaG in fla2 clusters from Methanomicrobia (Methanosarcinales and the Methanomicrobiale Methanospirillum
hungatei) and Archaeoglobales may be a very distant
homologue of flaD, as already suggested [5]. A second
characteristic differentiates fla1 and fla2 clusters: they
show an inverted order of flaG and flaF. While the gene
order in fla1 clusters is flaB-flaC-flaD-flaF-flaG-flaH-flaIflaJ, fla2 clusters (lacking flaC/D) display the order flaBflaG-flaF-flaH-flaI-flaJ (Figure 2). Interestingly, all the
archaeal genomes contain only a single type of fla clusters
(i.e. fla1 or fla2), except M. burtonii that is the only species
that harbors both type I and II fla clusters. fla1 clusters are
present only in Euryarchaeota: Thermococcales (Pyrococcus abyssi, Pyrococcus furiosus, Pyrococcus horikoshii and
Thermococcus kodakarensis), Methanococcales (Methanocaldococcus jannashii and Methanococcus maripaludis), Thermoplasmatales
(Thermoplasma
acidophilum
and
Thermoplasma volcanium), Halobacteriales (Haloarcula
marismortuii, Halobacterium sp. and Natromonas pharaonis),
and Methanomicrobia(Methanococcoides burtonii) (Figure
2A); whereas fla2 clusters are present in Crenarchaeota:
http://www.biomedcentral.com/1471-2148/7/106
Desulfurococcales (Aeropyrum pernix) and Sulfolobales
(Sulfolobus solfataricus Sulfolobus tokodaii Sulfolobus acidocaldarius) and in some Euryarchaeota: Methanomicrobia
(Methanospirillum hungatei, Methanococcoides burtonii,
Methanosarcina acetivorans, Methanosarcina mazei and
Methanosarcina barkeri) and the Archaeoglobale Archaeoglobus fulgidus (Figure 2B). In Halobacteriales and Sulfolobales the clusters also include non-flagellum genes
(Figure 2A). In particular, a gene coding for a homologue
of a bacterial chemotaxis component (MCP domain signal transducer) is present in Halobacterium sp., and three
genes coding for components of the chemotaxis system
are present in H. marismortui (CheY, CheA, and CheD)
and in N. pharaonis (CheY, CheC, and CheD). In this
archaeon, the operon is disrupted, the second half lying at
~50 ORFs downstream (Figure 2A). Interestingly,M. mazei
and M. acetivorans each harbor two copies of the fla2 cluster (hereafter called fla2A and fla2B), that differ in the
number of flagellin gene copies (two in fla2A and one in
fla2B), and in the presence of two different hypothetical
genes (possibly very distant homologues of flaD, see
above) lying in between the genes coding for FlaB and
FlaG (Figure 2B). According to these characteristics, the A.
fulgidus, the M. hungatei and one of the M. burtonii gene
clusters resemble more to the fla2A cluster, while that
from M. barkeri resembles more to the fla2B cluster (Figure 2B). Multiple copies of flagellin genes (flaB/flaA) are
found in most archaeal genomes, especially in Thermococcales, whereas Thermoplasmatales and Sulfolobales
harbor single gene copies (Figure 2), confirming earlier
studies on the composition of the flagella from T. volcanium and S. shibatae [23]. In M. hungatei, the flagellin
genes lie in another region of the chromosome (two are
clustered together and an additional small one is isolated
(Figure 2B)), suggesting a disruption of the original cluster. This is also the case of the fla1 cluster of M. burtonii,
which presents no nearby flagellin genes (a single isolated
flaB gene was possibly part of this cluster before disruption, Figure 2A and see below). In S. solfataricus a transposase disrupts the gene coding for FlaG (Figure 2B).
However, both the N- and C-ter sequences of FlaG are still
very similar to FlaG homologues found in S. solfataricus
close relatives (e.g. S. acidocaldarius and S. tokodaii), suggesting that the disruption of flaG is recent. Indeed, the
sequenced genome of S. solfataricus presents indeed a high
number of insertion elements that may have recently
invaded this strain [24]. Interestingly, cells of this strain
appear non-flagellated under the electron microscope (P.
Redder, personal communication) even if the disruption
of the flaG gene does not affect the transcription of the
downstream operon genes [25]. This suggests that FlaG
(possibly involved in the flagellum anchoring system [6])
is an essential flagellum component.
Page 3 of 13
(page number not for citation purposes)
BMC Evolutionary Biology 2007, 7:106
http://www.biomedcentral.com/1471-2148/7/106
Pyrobaculum aerophilum
Picrophilus torridus
Ferroplasma acidarmanus
Thermoplasma acidophilum
Thermoplasma volcanium
Archaeoglobus fulgidus
Haloarcula marismortui
Halobacterium sp
Methanospirillum hungatei
Methanococcoides burtonii
Methanosarcina barkeri
Methanosarcina mazei
Methanosarcina acetivorans
Mesophilic Crenarchaea
Desulfurococcales
Crenarchaea
mésophiles
Crenarchaea
hyperthermophiles
Sulfolobales
Nanoarchaeota
Thermococcales
Methanopyrales
Methanobacteriales
Methanococcales
Euryarchaea
Thermoplasmatales
Archaeoglobales
Halobacteriales
Methanomicrobiales
Methanosarcinales
Methanomicrobia
Methanogen Class II
Natronomonas pharaonis
°
°
•
•
•
•
°
•
•
•
•
•
°
°
•
•
NM °
NM °
M•
M•
M•
M•
M•
M•
M•
M•
NM •
NM •
NM •
Methanogen Class I
NA
M
Aeropyrum pernix
M
Sulfolobus solfataricus
M
Sulfolobus tokodaii
M
Sulfolobus acidocaldarius
M
Nanoarchaeum equitans
NM
Thermococcus kodakarensis
M
Thermococcus gammatolerans
M
Pyrococcus furiosus
M
Pyrococcus horikoshii
M
M
Pyrococcus abyssi
Methanopyrus kandleri
M
Methanothermobacter thermautotrophicus NM
M
Methanocaldococcus jannaschii
Methanococcus maripaludis
M
Cenarchaeum symbiosum
Schematic
sequenced
Figure
1 phylogeny
genomes are
adapted
available
fromin[29]
databases
showing the relationship between the main archaeal phyla for which completely
Schematic phylogeny adapted from [29] showing the relationship between the main archaeal phyla for which completely
sequenced genomes are available in databases. Motility or non-motility by the mean of a flagellum of each organism according
to [21] is indicated by M and NM, respectively (NA is used when no information is available). Black circles and open circles
indicate the presence or the absence of flagellum components coding gene in the genomes of the considered organisms,
respectively. Red rectangles indicate the presence of flagellum component coding genes in organisms described as non-motile
whereas yellow rectangles indicate the absence of flagellum component coding genes in organisms described as motile.
Finally, additional copies of flagellum components lie in
a few instances outside of the clusters (examples are additional flaB genes in M. burtonii, the two Thermoplasmatales, H. marismortui and N. pharaonis; an additional flaG in
Halobacterium sp.; an additional flaD in H. marismortui, N.
pharaonis, and M. burtonii; an additional flaF in M. maripaludis, and an additional flaK in Methanococcales) (Figure 2).
Phylogenetic analysis
Phylogenetic analyses were performed on six amino acid
sequence datasets corresponding to FlaA/B, FlaD/E, FlaG,
FlaH, FlaI, and FlaJ. Phylogenetic analysis of FlaC, FlaF
and FlaK could not be performed due to a too restricted
phylogenetic distribution of FlaC, and the poor sequence
conservation of FlaF and FlaK.
FlaG, FlaH, FlaI, FlaJ
Among all archaeal flagellum components, FlaH, FlaI and
FlaJ are the most conserved at the sequence level and
always lie close to each other in all the analyzed genomes
(Figure 2) strengthening their likely fundamental role in
flagellum assembly and function (see above). FlaI has a
number of bacterial homologues belonging to type IV and
type II secretion systems, including the typeIV pili component PilT [26]. Moreover, FlaI has additional archaeal
homologues that are also probably part of yet to describe
secretion machineries [27,28]. In a phylogeny including
all these homologues FlaI sequences form a monophyletic
group and are most closely related to their archaeal counterparts (not shown). FlaH shares a RecA-like ATPase
domain with distant archaeal and bacterial homologues
that are not involved in motility structures. Psi-blast
Page 4 of 13
(page number not for citation purposes)
BMC Evolutionary Biology 2007, 7:106
http://www.biomedcentral.com/1471-2148/7/106
B
A
Desulfurococcales & Sulfolobales
Thermococcales
Thermococcus kodakarensis (TK0038-39-40-41-42-43-44-45-46-47-48-49) // +K TK0053
B
B
B
B
Pyrococcus furiosus
B
B
B
F
G
H
I
J
B
B
C
D
F
G
H
I
J
B
B
B
C
D
B
C
F
D
G
H
I
J
B
B
C
F
G
H
I
B
J
B
B
I
J
C
E
F
G
H
I
J
E
F
G
H
I
I
J
transposase
G
F
H
I
J
Methanomicrobia & Archaeoglobales
B
K
(MMP1666-1667-1668-1669-1670-1671-1672-1673-1674-1675-1676)
// +2K MMP0232-MMP0555 +F MMP0342
D
H
Methanosarcina barkeri (Mbar_A1970-1969-1968-1967-1966-1965-1964) +K Mbar_A1491
(MJ0891-892-893-894-895-896-897-898-899-900-901-902)
// +2K MJ0835.2 MJ1282.1
D
Methanococcus maripaludis
B
H
(SSO2323-2322-2320-2319-2319-2318-2316-2315-2314) +K SSO0131
Sulfolobus tokodaii (ST2518-2519-2520-2521-2522-2523-2524) +K ST2258
Sulfolobus acidocaldarius (Saci_1178-1177-1176-1175-1174-1173-1172) + K Saci_0139
Methanococcales & Thermoplasmatales
Methanocaldococcus jannashii
B
G
G
G F
transposase
(PH0546-548-549-550-551-552-553-553.1n-555-556-557-559) // +K PH0446
Pyrococcus abyssi (PAB1378-1379-1380-1381-1382-1383-1384-1385-1386-1387) // +K PAB1309
B
B
Sulfolobus solfataricus
B
B
D
(PF0338-337-336-335-334-333-332-331-330) // +K PF0471
Pyrococcus horikoshii
B
C
Aeropyrum pernix (APE1907-1905-1904-1903-1901-1899-1898-1896-1895)
G
F
H
I
J
fla2B
hyp2
Methanosarcina mazei
B
J
(MM0418-417-416-415-414-413-412) // K MM1678
(MM0323-322-321-320-319-318-317-316)
G
F
H
I
J
fla2B
G
F
H
I
J
fla2A
hyp2
Thermoplasma acidophilum
B
(Ta0553-554-555-556-557a-558-559-560) // +B Ta1407m
C
D
F
G
H
Thermoplasma volcanium (TVN0607-608-609-610-611-612-613-614)
B
C
D
F
G
H
I
B
J
// +B TVN1426
I
B
hyp1
Methanosarcina acetivorans (MA3077-3078-3079-3080-3081-3082-3083) // 2K MA0519-MA4505
(MA3062-3061-3060-3059-3058-3057-3056-3055)
B
J
G
F
H
I
J
fla2B
hyp2
Halobacteriales & Methanomicrobia
(rrnAC2198-2197-2195-2194-2193-2192-2191-2190-2187-2186-2184-2183)
// +2B pNG1026 -rrnB0018 +D rrnAC1482 + K rrnAC2525
Haloarcula marismortui
B
C/D
Halobacterium sp.
B
B
F
G
H
I
J
(VNG0962G-961G-960G-959H-958G-955G-954C-953C-950a-950Gm-949G-947G)
// +3B VNG1181G-VNG1009-VNG1008+G VNG2567C + K VNG2285C
B
D
Natromononas pharaonis
B
C/D
F
G
H
I
J
(NP2086A-2088A-2090A-2092A-2094A-2096A-2098A-2100A-2102A-2104A-2106A)
// (NP2152A-2154A-2156A-2158A-2160A)
// +K NP1276A +D NP2686A
B
B
H
B
I
J
B
G
F
H
I
J
fla2A
hyp1
Methanospirillum hungatei (Mhun_0100-0101-0102-0103-0104-0105-0106)
// +3B Mhun_1238 Mhun_3139 Mhun_3140 + K Mhun_2921
fla2A
G G
F
H
I
J
hyp1
Archaeoglobus fulgidus (AF1055-1054-1053-1052-1051-1050-1049-1048) +K AF0936
B
2
C/D
B
B
G
F
H
I
J
hyp1
Methanococcoides burtonii (Mbur_0346-0347-0348-0349-0350-0351-0352-0353)
fla2A
B
G
F
H
I
J
fla2A
hyp1
Methanococcoides burtonii (Mbur_1570-1571-1572-1573-1574-1575) // +B Mbur_0104 + D Mbur_1246 + K Mbur_0062
C
G
H
I
J
Figure
Genomic2 organization of the genes coding for flagellum components in complete archaeal genomes (fla clusters)
Genomic organization of the genes coding for flagellum components in complete archaeal genomes (fla clusters). Numbers
within brackets correspond to the locus tags of each gene. A // sign indicates that the following components are elsewhere in
the genome. The genes annotated as hyp1 and hyp2 are probable distant homologues of flaD/E. Genes colored in grey are
homologues of chemotaxis components, as discussed in the text. Remaining genes where no name is indicated are annotated as
hypothetical. A Genomic organization of type I clusters (fla1). These are found only in Euryarchaea and are characterized by
the presence of flaC and flaD/E and by a conserved gene order flaA/B, flaC, flaD, flaF, flaG, flaH, flaI, flaJ. B Genomic organization of the type II clusters (fla2). These are found in Crenarchaea and in some Euryarchaea and are characterized by the
absence of flaC and flaD/E and by a conserved gene order flaA/B, flaC, flaD, flaG, flaF, flaH, flaI, flaJ.
searches revealed (i) that FlaJ harbours a few distant
archaeal homologues annotated as involved in type II
secretion and (ii) weak similarities with the bacterial pilus
assembly protein TadC and TadB.
FlaJ shares the domain GSPII F with TadC and TadB. However, this may not be significant, given that the domain
was defined on the basis of an alignment that included
both archaeal and bacterial sequences. The similarity
between the FlaJ sequences and TadB and TadC sequences
is very weak (16% and 35% of identity and similarity with
TadB sequences, respectively and 14% and 32% of identity and similarity with TadC sequences, respectively) and
is mainly the result of the sharing of small hydrophobic
amino acids. To our point of view this sequence similarity
is too weak to definitively conclude that these sequences
are homologues although this has been claimed [27].
After removal of ambiguously aligned positions, 104,
193, 392 and 353 amino acids could be kept for phylogenetic analysis of FlaG, FlaH, FlaI, and FlaJ, respectively.
The resulting trees are strikingly congruent (Figure 3).
Notably, all major archaeal groups except Methanomicrobia (Methanomicrobiales plus Methanosarcinales) are
well defined and strongly supported statistically (Bootstrap Values -BV- > 990‰ and/or Posterior Probabilities PP- = 1), suggesting that no recent horizontal transfer of
flaH, flaG, flaI and flaJ genes occurred across these groups.
Page 5 of 13
(page number not for citation purposes)
BMC Evolutionary Biology 2007, 7:106
The sequences from Methanomicrobia form a well supported cluster (BV > 980‰ and/or Posterior Probabilities
-PP- = 1) that also includes sequences from A. fulgidus. It
is not possible for the time being to decide whether this is
due to a HGT or a hidden paralogy. Importantly, the corresponding trees are also strongly consistent with gene
cluster organization. In fact, homologues from fla1 and
fla2 clusters (characterized by a flaF-flaG and by a flaG-flaF
gene order, respectively) form two distinct strongly groups
(BV > 975‰ and PP = 1, Figure 3). In particular, in all
four trees, the homologues from the fla2 clusters of M.
hungatei, the four Methanosarcinales and A. fulgidus
appear close to Crenarchaeota (Figure 3). This is in contrast to their expected position as sister-group of Halobacteriales within Euryarchaeota (Figure 1, [29,30]).
Interestingly, such expected position is shown by the M.
burtonii sequences belonging to its fla1 cluster (Figure 2A).
This suggests that the fla1 and fla2 clusters from M. burtonii have different origins (see below). Independent species-specific duplications of flaG appear to have occurred
in Halobacterium sp., N. pharaonis, and M. hungatei (tandem gene duplications in these last two species). Moreover, in the FlaH, FlaI and FlaJ phylogenies, the sequences
from M. burtonii, M. hungatei, M. barkeri and A. fulgidus
fla2 clusters group with the sequences belonging to the
fla2B cluster from M. mazei and M. acetivorans (Figures 3B,
3C and 3D, respectively), supporting a close relationship
of these clusters, as suggested by their gene organization
(Figure 2B).
FlaD/E
As discussed above, homologues of FlaD/E genes are missing in all fla2 clusters from Crenarchaea. In fla2 clusters
from Methanomicrobia and A. fulgidus the two hypothetical proteins hyp1 and hyp2 could be distantly related to
FlaD/E (Figure 2B). However they are too distant to be
included in any phylogenetic analysis. The small number
of unambiguously aligned positions (77 amino acids
positions) that could be kept for phylogenetic analysis
gives a poor resolution of the relationships between major
euryarchaeal groups (Figure 4A). However, sequences
from phyla form monophyletic groups generally well supported (BV = 996‰, PP = 0.93, BV = 1000‰ and BV =
958‰ for Halobacteriales, Methanosarcinales, Thermoplasmatales and Thermococcales, respectively, Figure 4A).
This indicates that, similarly to FlaG, FlaI, FlaJ, and FlaH,
no recent HGT involving the flaD/E gene occurred among
these groups. Interestingly, a tandem duplication event
appears to have occurred in an ancestor of Methanococcales, with the two copies having been conserved within the
cluster. Halobacteriales also harbor two copies of flaD.
One of the two copies is fused with flaC and always resides
within the fla cluster, whereas the second copy resides
within the fla cluster only in Halobacterium sp. This suggests that the fusion of flaC and flaD genes occurred before
http://www.biomedcentral.com/1471-2148/7/106
the divergence of the three Halobacteriales, but after the
duplication event and that one of the two copies was displaced after the duplication event in the ancestor of H.
marismortui and N. pharaonis. A similar duplication of
FlaD followed by a fusion of one of the two copies of FlaD
and FlaC also appears to have occurred in M. burtonii. As
in most Halobacteriales one of the two copies resides outside the fla cluster (Figure 2A) suggesting its displacement
after the duplication event. The poor resolution of the
relationships between groups (due to the small number of
positions kept for the phylogenetic analysis) does not permit to determine if a single fusion event of FlaC and FlaD
occurred in ancestor of Halobacteriales and Methanomicrobia or if this event occurred two times independently
in both lineages.
FlaB
Given the use of only 72 unambiguously aligned positions for analysis, the FlaB tree presents a number of
poorly resolved nodes (Figure 4B). Nevertheless, the
monophyly of a number of groups is recovered and supported by BV > 850‰, except for Thermococcales, Methanococcales and Methanomicrobia. As for FlaG, FlaI, FlaJ,
and FlaH, the FlaB tree is again strongly consistent with
gene cluster organization (Figure 4B). In particular, the
FlaB from the fla2 clusters of M. hungatei, the four Methanosarcinales and A. fulgidus appear close to Crenarchaeota, while the isolated FlaB from M. burtonii appear close
to Halobacteria, and thus likely functions with the flagellum components of fla1 cluster (Figure 4B). Interestingly,
the multiple copies of flagellins group in a group-specific
manner (Figure 4B), suggesting that flagellin gene family
expansion occurred mainly by gene duplication and multiple times independently in different species, and not by
HGT.
Discussion and conclusion
The archaeal flagellum is a complex cellular structure
composed of multiple subunits and anchored to the
membrane. The striking conservation of the genomic context of the genes coding for these subunits indicates a
likely highly coordinated expression and assembly mechanisms. Coupled to the high sequence conservation of the
different subunits across archaeal species inhabiting very
different habitats, this underlines the importance for
structural maintenance of the archaeal flagellum. However, we highlighted an important dimorphism of the
genetic context organization, with two types of gene clusters harboring differences in both gene content and gene
order (Figure 2). In fact, most Euryarchaea exhibit the fla1
cluster, whereas all the Crenarchaea and some Euryarchaea have the fla2 cluster (e.g. Methanomicrobia and A.
fulgidus). The difference between the two clusters is
strongly supported by phylogenetic analysis, and indicates that Methanomicrobia and A. fulgidus flagellum
Page 6 of 13
(page number not for citation purposes)
BMC Evolutionary Biology 2007, 7:106
A
788
.93
881
.88
Type II clusters (fla2)
G
F
F
G
433
.6
641
.86
722
1
213
.48
977
1
458
999
1
901
1
532
.57
994/1
0.1
C
501
.6
Type II clusters (fla2)
G
F
F
G
697/.57
Type I
clusters
(fla1)
1000
1
539
.99
405
.98
0.1
B
983
1
1000
1 624
.94
892
1
Type I
clusters
(fla1)
Aeropyrum pernix BAA80906
Sulfolobus acidocaldarius AAY80523
Sulfolobus tokodaii BAB67633
Sulfolobus solfataricus
Methanospirillum hungatei ABD39880
Methanospirillum hungatei ABD39879
Methanosarcina mazei NP_632344
fla2A
943
Methanosarcina acetivorans AAM06432
497 .98
Methanococcoides burtonii ABE51344
.89
470
Archaeoglobus fulgidus NP_069885
442
.96
.91
Methanosarcina barkeri AAZ70904
1000
Methanosarcina acetivorans AAM06452
1
fla2B
542/.55
Methanosarcina mazei NP_632440
Methanococcoides burtonii EAN01051
Haloarcula marismortui AAV47027
Haloferax volcanii
Natronomonas pharaonis YP_326701
784
Natronomonas pharaonis YP_326700
.97
Halobacterium sp NP_444204
478
Halobacterium sp NP_281136
.98
Pyrococcus furiosus NP_578062
Thermococcus kodakarensis YP_182459
Pyrococcus abyssi NP_127164
985
Pyrococcus horikoshii NP_142523
1
Thermoplasma volcanium NP_111130
1000
1
Thermoplasma acidophilum NP_394031
Methanocaldococcus jannaschii NP_247893
Methanococcus maripaludis NP_988793
Methanothermococcus thermolithotrophicus AAG50076
816/1
577/.63
Methanococcus voltae O06640
http://www.biomedcentral.com/1471-2148/7/106
Aeropyrum pernix NP_148247
Sulfolobus acidocaldarius YP_255813
1000
Sulfolobus tokodaii NP_378527
1
731
Sulfolobus solfataricus NP_343681
.69
Methanosarcina barkeri YP_305481
Methanosarcina acetivorans NP_617975 fla2B
1000/1
863/.97 Methanosarcina mazei AAM30109
1000
Methanospirillum hungatei EAP16188
1
Archaeoglobus fulgidus AAB90188
641
Methanococcoides burtonii EAN00027 fla2A
.66
903
1
Methanosarcina mazei AAM30013
996
1
1000/1 Methanosarcina acetivorans NP_617949
Methanococcoides burtonii EAN01053
Haloferax volcanii
Haloarcula marismortui AAV47025
1000/1
Natronomonas pharaonis YP_326731
540/.85
407/.9
Halobacterium sp NP_279896
Methanocaldococcus jannaschii Q58310
Methanothermococcus thermolithotrophicus AAG50078
999
Methanococcus maripaludis NP_988795
1 998/1
961/1
Methanococcus voltae AAB57833
Thermoplasma acidophilum NP_394033
1000/1
Thermoplasma volcanium NP_111132
Thermococcus kodakarensis YP_182461
Pyrococcus furiosus NP_578060
999/1
Pyrococcus horikoshii NP_142525
923/1
965/1 Pyrococcus abyssi NP_127162
985
1
885
1
Type II clusters (fla2)
G
F
F
G
639
.96
Type I
clusters
(fla1)
1000
1
904
1
353
.73
0.1
.61
D
654
.73
Type II clusters (fla2)
G
F
F
G
Type I
clusters
(fla1)
1000
1
872
.99
823
1
442
.58
0.1
Aeropyrum pernix NP_148248
Sulfolobus acidocaldarius YP_255814
Sulfolobus tokodaii NP_378526
998
Sulfolobus solfataricus NP_343682
1
Methanosarcina barkeri YP_305482
1000
1
Methanosarcina mazei AAM30110
fla2B
734
Methanosarcina acetivorans NP_617974
.83
Methanospirillum hungatei EAP16187
Archaeoglobus fulgidus AAB90187
992
Methanococcoides burtonii EAN00028
1
fla2A
Methanosarcina mazei AAM30014
1000/1
999/1
Methanosarcina acetivorans NP_617950
Methanococcoides burtonii EAN01052
Haloferax volcanii
1000
Halobacterium sp NP_444203
1
Haloarcula marismortui AAV47026
731/.85
887/1
Natronomonas pharaonis YP_326730
Methanocaldococcus jannaschii Q58309
999
Methanothermococcus thermolithotrophicus AAG50077
1
Methanococcus maripaludis NP_988794
939
1 979
Methanococcus voltae O06641
1
Thermoplasma volcanium NP_111131
1000
1
Thermoplasma acidophilum NP_394032
Pyrococcus
horikoshii
NP_142524
628/.8
Pyrococcus abyssi NP_127163
Pyrococcus furiosus NP_578061
1000/1
607
Thermococcus kodakarensis YP_182460
1000
1
Aeropyrum pernix NP_148246
Sulfolobus solfataricus AAK42470
Sulfolobus tokodaii BAB67637
1000/1
320/.75
Sulfolobus acidocaldarius AAY80519
Methanosarcina barkeri YP_305480
Methanosarcina mazei AAM30108
1000/1
fla2B
884/1
Methanosarcina acetivorans NP_617976
1000
Methanospirillum hungatei EAP16189
1
Archaeoglobus fulgidus AAB90189
421
.69
Methanococcoides burtonii EAN00026
fla2A
1000
1 1000/1
Methanosarcina mazei AAM30012
1000/1 Methanosarcina acetivorans NP_617948
Halobacterium sp NP_279895
Haloferax volcanii
1000/1
Natronomonas pharaonis YP_326732
803/.99
694/.99
Haloarcula marismortui AAV47024
Methanococcoides burtonii EAN01054
Methanocaldococcus jannaschii NP_247896
Methanothermococcus thermolithotrophicus AAG50079
1000
Methanococcus maripaludis NP_988796
1 1000
1 998
Methanococcus voltae AAC19121
1
Thermoplasma volcanium NP_111133
Thermoplasma acidophilum NP_394034
1000/1
Thermococcus kodakarensis YP_182462
Pyrococcus furiosus NP_578059
1000/1
Pyrococcus horikoshii NP_142526
890/1
978/1 Pyrococcus abyssi NP_127161
Figure
Unrooted
3 maximum likelihood phylogenetic trees of FlaG (A), FlaH (B), FlaI (C) and FlaJ (D)
Unrooted maximum likelihood phylogenetic trees of FlaG (A), FlaH (B), FlaI (C) and FlaJ (D). Numbers at nodes indicate bootstrap values for 1000 replicates of the original dataset and posterior probabilities, computed by PHYML and MrBayes, respectively. The scale bar represents the average number of substitutions per site. The phylogenies show a clear separation between
the type I clusters (characterized by a FlaF FlaG order of genes) and type II clusters (characterized by a FlaG FlaF order of
genes).
components encoded by fla2 gene clusters are more
closely related to their crenarchaeal homologues than to
the homologues encoded by the fla1 gene cluster of their
close relatives (i.e. M. burtonii, Thermoplasmatales and
Halobacteriales, Figures 1, 3 and 4). Two different evolutionary scenarios can be proposed to explain our results.
In the first scenario (Figure 5), the last ancestor of Archaea
was flagellated and had the two types of clusters (i.e. both
fla1 and fla2, blue and red clusters, respectively, Figure 5),
resulting from the duplication of an ancestral cluster (purple cluster, Figure 5), and these were secondarily and differently lost during archaeal lineages evolution (seven
losses of the fla1 cluster and nine losses of the fla2 cluster).
Importantly, some of these losses would have also
occurred recently in the Euryarchaea (for example the loss
of the fla1 cluster in the Methanosarcinale lineage would
have occurred after the divergence of M. burtonii, that
would have kept the two types of clusters). Finally, a
duplication event of the whole fla2 cluster occurred in an
ancestor of Methanosarcina (red circle, Figure 5) leading to
the fla2B cluster (orange cluster, Figure 5). This first scenario involves 16 losses, and implies that the ancestor of
Archaea and the ancestor of each euryarchaeal group had
two types of flagella. Moreover, it does not explain the
position of A. fulgidus sequences within the Methanomicrobia group and not as sister of this group, as generally
indicated by molecular phylogeny of multiple markers
(Figure 1 and [30]). A second scenario involves less losses
(three losses of the fla2 cluster and seven losses of the fla1
cluster) (Figure 6 and Figure 7 for a more detailed scenario). Here, the ancestor of Archaea was also flagellated,
but had only a single type of fla cluster (either fla1, fla2, or
else, purple cluster in Figure 6 and Figure 7). After the
divergence of Crenarchaea and Euryarchaea (black circle,
Page 7 of 13
(page number not for citation purposes)
BMC Evolutionary Biology 2007, 7:106
http://www.biomedcentral.com/1471-2148/7/106
A
Sulfolobus acidocaldarius YP_255818
B
cluster
Sulfolobus tokodaii NP_378522
B
cluster
Sulfolobus solfataricus NP_343686
B
cluster
Aeropyrum pernix NP_148254
B
B
cluster
977/1
Aeropyrum pernix NP_148253
B
B
cluster
Archaeoglobus fulgidus O29207
B
cluster fla2A
547
M. hungatei EAP15115
B
800/1
868/1
Methanospirillum hungatei EAP15114
B
Methanosarcina acetivorans NP_617955
cluster
B
242/.58
Methanosarcina mazei AAM30019
cluster
B
Methanococcoides burtonii EAN00033
cluster
B
205/.68
620/.93
Methanococcoides
burtonii
EAN00032
cluster
B
B
fla2A
768
.84
100/.54
Methanosarcina acetivorans NP_617954
cluster
B
cluster
B
B
924/1Methanosarcina mazei AAM30018
334/.96
cluster
Archaeoglobus fulgidus O29208
B
B
cluster
Methanosarcina acetivorans NP_617970
B
512/.97
Type II clusters (fla2)
cluster
Methanosarcina barkeri YP_305486
B
fla2B
1000/1
G F
cluster
969/.97 Methanosarcina mazei AAM30114
B
Methanococcoides burtonii EAM99976
B
F G
Natronomonas pharaonis YP_326696
cluster
B
B
B
Type I
Natronomonas pharaonis YP_326697
B
B
B
cluster
1000/1
601
clusters
945
Natronomonas pharaonis YP_326695
B
B
B cluster
1
.81
(fla1)
Natrialba magadii O93719
974
Natrialba magadii Q9P9I3
1
744/.9
Natrialba magadii O93718
826
990/1 Natrialba magadii Q9P9I1
1
Haloarcula marismortuii AAV48232
B
782
.98
Haloarcula marismortuii AAV44283
530/.8
993/1
Haloarcula marismortuii AAV47035
cluster
400
Halobacterium
salinarum
AAA72644
876
B cluster
1000
.77
.97
.98
Halobacterium salinarum AAA72641
B cluster
1000/1 Halobacterium sp
s NP_279945
334/.31
Halobacterium sp
s NP_279905
716/.53
Halobacterium salinarum AAA72643
B
B
B cluster
Methanococcus voltae AAA73073
Pyrococcus abyssi NP_127170
cluster
B
B
B
317/Thermoplasma volcanium NP_111126
cluster
B
940
998 Thermoplasma acidophilum NP_394027
cluster
B
1
1
345
Thermoplasma volcanium NP_111945
B
984/1
Thermoplasma acidophilum NP_394861
B
Methanocaldococcus jannaschii NP_247886
cluster
B
B
B
285
Methanocaldococcus jannaschii NP_247887
cluster
B
B
B
786
Methanothermococcus thermolithotrophicus
504
Methanococcus voltae AAA73074
Methanococcus maripaludis AAG50058
968
167
Methanococcus maripaludis NP_988787
318
cluster
B
B
B
149
Methanococcus vannielii AAC27726
497
Methanothermococcus thermolithotrophicus
626
Methanococcus
voltae
AAA73075
14
Methanococcus vannielii AAC27725
238
Methanococcus maripaludis AAG50057
Methanococcus maripaludis NP_988786
998
B
B
B
cluster
Methanococcus voltae AAA73076
Methanococcus vannielii AAC27727
986
5
Methanothermococcus thermolithotrophicus
404
Methanocaldococcus jannaschii Q58303
cluster
B
B
B
564
382
Methanococcus maripaludis AAG50059
979 Methanococcus maripaludis NP_988788
473
B
B
B
cluster
Thermococcus kodakarensis BAA84106
cluster
B
B
B
B
Pyrococcus horikoshii NP_142515
cluster
B
B
B
B
B
Pyrococcus horikoshii NP_142517
558
cluster
B
B
B
B
B
321
Thermococcus kodakarensis BAA84105
154
B
B
B
B cluster
Pyrococcus horikoshii NP_142516
253
cluster
B
B
B
B
B
304
Thermococcus kodakarensis BAA84107
B
B
B
B cluster
179
Pyrococcus abyssi Q9UYL3
B
B
B cluster
922
Thermococcus
kodakarensis
BAA84108
cluster
B
B
B
B
885
1000 Pyrococcus furiosus NP_578067
cluster
B
B
Pyrococcus horikoshii NP_142518
520
cluster
B
B
B
B
B
Pyrococcus furiosus NP_578066
cluster
B
B
0.1
Pyrococcus horikoshii NP_142519
cluster
345
B
B
B
B
B
Pyrococcus abyssi Q9UYL4
297
cluster
B
B
B
270 Thermococcus kodakarensis BAA84109
cluster
B
B
B
B
B
B B C
D
F G H
I
J
B B C
D
F G H
I
J
B B B C
D
F G H
I
J
B B B C
D
F G H
I
J
Pyrococcus furiosus NP_578064
B B B
958 Pyrococcus horikoshii NP_142521
1
874 Pyrococcus abyssi NP_127166
.76
576
.37
Thermococcus kodakarensis YP_182457
B
B
Methanothermococcus thermolithotrophicus AAG50074
Methanocaldococcus jannaschii Q58306
962
1
706
.71
Methanococcus maripaludis NP_988791
900
.99
546
.73
D
E F G H
I
J
B B B C
D
E F G H
I
J
B B B C
D
D F G H
I
J
B B B C
D
D F G H
I
J
Methanococcus voltae O06638
Methanocaldococcus jannaschii Q58305
Methanococcus maripaludis NP_988790
870
1
B B B C
934
1
452
.6
1000
1
Methanococcus voltae O06637
Methanothermococcus thermolithotrophicus AAG50073
Thermoplasma volcanium NP_111128
B C D
F G H
I
J
Thermoplasma acidophilum NP_394029
B C D
F G H
I
J
C/D
F G H
I
J
D
F G H
I
J
C
C/D
F G H
I
Methanococcoides burtonii ABE52168
406
.56
782
.93
D
Methanococcoides burtonii EAN01049
Halobacterium sp NP_279899
498
.7
B B B
Natronomonas pharaonis
D
YP_326993
996
.7
Haloarcula marismortui
454
1
AAV47029
610
.99
328
.67
B
Natronomonas pharaonis
YP_326729
C/D
H
I
J
J
Haloarcula marismortui AAV
A 46407
0.1
508
.88
Halobacterium sp
s B B B
NP_279900
428
.63
Haloferax volcanii
C/D
F G H
I
J
997/1
987/.99
Figure
A.
Unrooted
4
maximum likelihood phylogenetic trees of FlaD/E
A. Unrooted maximum likelihood phylogenetic trees of FlaD/E. Numbers at nodes indicate bootstrap values for 1000 replicates of the original dataset and posterior probabilities, computed by PHYML and MrBayes, respectively. The scale bar represents the average number of substitutions per site. The light blue circles indicate duplication events. B. Unrooted maximum
likelihood phylogenetic trees of FlaA/B. Numbers at nodes indicate bootstrap values for 1000 replicates of the original dataset
and posterior probabilities, computed by PHYML and MrBayes, respectively. The scale bar represents the average number of
substitutions per site. Detailed cluster organizations are not shown. White arrows are used to schematize the clusters. The
phylogenies show a clear separation between the type I clusters (characterized by a FlaF FlaG order of genes) and type II clusters (characterized by a FlaG FlaF order of genes).
Figure 6 and Figure 7), evolution led to the fla1 and fla2
clusters. A single HGT of the fla2 cluster would have then
occurred from some ancestors of Sulfolobales and Desulfurococcales to an ancestor of Methanomicrobia (Figure 6
Figure 7, HGT 1), and was followed by a HGT from some
ancestors of Methanosarcinales to A. fulgidus (Figure 6 Figure 7, HGT 2). Methanosarcina, M. hungatei and A. fulgidus
would thus have lost their original fla1 gene cluster before
or after their replacement by a fla2 gene cluster of crenarchaeal origin. We favor a HGT from Crenarchaea to Methanomicrobia rather than the opposite, since in this case
we would expect to find the M. burtonii fla1 genes more
closely related to their paralogues belonging to the fla2
cluster than to their fla1 orthologues from Halobacteriales.
Both scenarios imply that the archaeal flagellum would
have appeared prior to the last archaeal ancestor
Apart these two likely cases of HGT of the entire fla2 gene
cluster, we found no clear evidence for recent transfers of
the genes coding for flagellum components among
archaeal species. Indeed, the poor resolution of interphyla relationships in some trees is more likely due to lack
of phylogenetic signal rather than horizontal gene transfer. One explanation for the rarity of HGT is that it is pos-
Page 8 of 13
(page number not for citation purposes)
BMC Evolutionary Biology 2007, 7:106
http://www.biomedcentral.com/1471-2148/7/106
Mesophilic
Crenarchaea
Cenarchaeum symbiosum
Pyrobaculum aerophilum
Desulfurococcales
Aero
r pyr
yrum pern
r iix
Sulfo
f lobus solfa
f tari
r cus
Type I cluster
T pe II cluster
Sulfolobales
Sulfo
f lobus tokodaiii
Sulfo
f lobus acidocaldarius
Nanoarchaeota
Nanoarc
r haeum equitans
Therm
T
r ococcus kodaka
k rrensis
i
Therm
T
r ococcus gammatolera
r ns
Thermococcales
Pyrococcus furiosus
Pyrococcus horikoshii
Pyyrococcus abys
P
y si
Meth
t anoth
t erm
r obacter therm
r autotr
trophicus
Meth
t anocaldococcus jjannaschiii
Methanococcus maripaludis
Picrophilus torridus
Methanogen class I
Meth
t anopyr
yrus kkandl
d eri
r
Ferroplasma acidarmanus
Methanopyrales
Methanobacteriales
Methanococcales
Thermoplasmatales
Thermoplasma acidophilum
Thermoplasma volcanium
Archaeoglobales
Archaeoglobus fulgidus
Haloarcula marismortui
Halobacteriales
Halobacterium sp
Natronomonas pharaonis
Methanococcoides burtonii
Methanosarcina barkeri
Methanosarcina mazei
Methanosarcina acetivorans
Methanomicrobia
Methanogen class II
Methanospirillum hungatei
Methanomicrobiales
Methanosarcinales
Figure
An
evolutionary
5
scenario for the origin and evolution of the archaeal flagellum
An evolutionary scenario for the origin and evolution of the archaeal flagellum. Blue, red, orange and purple clusters represent
fla1 cluster, fla2A cluster, fla2B cluster and their ancestor, respectively. The black circle represents the last common ancestor
of Archaea. The green circle represents the duplication event that led to the fla1 and fla2 clusters. This duplication event
occurred before the last common ancestor of Archaea, which thus harbored the two types of clusters. The red circle represents the recent duplication event of the fla2 cluster in ancestor of Methanosarcina. The blue and red crosses represent the loss
of the fla1 and fla2A clusters, respectively.
sible the result of the high level of integration of flagellum
components within the macromolecular structure. Importantly, this contradicts the generally assumed notion that
HGT of "informational" genes (i.e. those coding for proteins involved in the expression and the transmission of
genetic information) are less frequent than those involving the remaining ones ("operational") genes. Nevertheless, the acquisition of a whole flagellum component
coding gene cluster from distant donors seems possible.
Interestingly, even if two clusters coexist within a genome
(i.e. in M. burtonii) neither gene recombination is
observed between homologous genes belonging to the
two clusters, nor losses, suggesting that the components of
a cluster may interact preferentially due to coevolution,
although they form similar cellular structures. The presence in M. burtonii of two types of fla clusters (possibly
one native and one acquired by HGT from crenarchaeota)
represents an interesting experimental model to study. It
would be in fact particular interesting to know the difference between the components encoded by the two fla
clusters in terms of expression and assembly, and how two
different flagellum systems coexist in this archaeon.
Finally, it has been suggested that archaea are modified
Actinobacteria and that archaeal flagella are derived from
bacterial flagella following an adaptation to acidic environments by the recruitment of an already acid-stable
glycoprotein from the pilus machinery that would have
replaced the original flagellin while leaving intact the
basal rotary motor [31]. We find this hypothesis unlikely
for the fact that archaeal flagella do not resemble to bacterial flagella in major structure, assembly, and components, and not only concerning flagellin. Indeed, no even
distant homologues to any component, including the
basal rotary motor, of bacterial flagella can be recovered in
archaeal genomes, including the related type III secretion
system components [28]. Moreover, flagella of acidophilic bacteria (such as Thiobacillus) show a typical bacterial structure (e.g. a diameter of approximately 20 nm), so
they adapted to acidic conditions without radically modi-
Page 9 of 13
(page number not for citation purposes)
BMC Evolutionary Biology 2007, 7:106
http://www.biomedcentral.com/1471-2148/7/106
Type II cluster
Mesophilic Crenarchaea
Cenarchaeum symbiosum
Pyrobaculum aerophilum
Desulfurococcales
Aeropyrum pernix
Sulfolobus solfataricus
Sulfolobales
Sulfolobus tokodaii
Sulfolobus acidocaldarius
Nanoarchaeota
Nanoarchaeum equitans
Thermococcus kodakarensis
Therm
T
r ococcus gammatolerans
Ancestral cluster
Thermococcales
Pyyrococcus ffurriosus
P
Pyyrococcus hori
P
rikoshiii
Pyyrococcus abyssi
P
Meth
t anopyrus kandleri
Meth
t anoth
t erm
r obacter therm
r autotr
trophicus
Methanocaldococcus jannaschii
Methanococcus maripaludis
Picrophilus torridus
Methanogen class I
Type I cluster
Ferroplasma acidarmanus
Methanopyrales
Methanobacteriales
Methanococcales
Thermoplasmatales
Thermoplasma acidophilum
Thermoplasma volcanium
Archaeoglobales
Archaeoglobus fulgidus
Haloarcula marismortui
Halobacteriales
Halobacterium sp
HGT 1
Natronomonas pharaonis
Methanospirillum hungatei
Methanococcoides burtonii
Methanosarcina barkeri
Methanosarcina mazei
Methanosarcina acetivorans
Methanomicrobia
Methanogen class II
HGT 2
Methanomicrobiales
Methanosarcinales
Figure
An
evolutionary
6
scenario for the origin and the evolution of the archaeal flagellum
An evolutionary scenario for the origin and the evolution of the archaeal flagellum. Blue, red, orange and purple clusters represent fla1 cluster, fla2A cluster, fla2B cluster and their ancestor, respectively. The black circle represents the last common
ancestor of Archaea. The red circle represents the recent duplication event of the fla2 cluster in ancestor of Methanosarcina.
The blue and red crosses represent the loss of the fla1 and fla2A clusters, respectively. The green arrows represent horizontal
gene transfers.
fying their motility structures and components [28].
Indeed, the uncomplete genome of the extreme acidophilic bacterium (optimal pH<3) Acidobacterium capsulatum contains a clear homologue of bacterial flagellin,
indicating that adaptation to acidic environments was
possible without its replacement.
The lack of congruence between the description of
archaeal species as motile or non-motile and the taxonomic distribution of homologues of flagellum component coding genes[5] is particularly striking and
underlines the need to explore more in depth the motility
systems in Archaea. The presence of two gene clusters in
non motile Methanosarcinales is particularly puzzling.
Either these species can be flagellated under particular
conditions that have not yet been tested, or the flagellum
component homologues are involved in other functions
than motility (for example, they could be required for cellcell adhesion to form cell aggregates, a peculiarity of this
archaeal family). It will be extremely interesting to study
the expression and localization of the flagellum components in Methanosarcinales, in the light of the fact that the
flagellum genes of Methanosarcinales may have been
recruited from the distantly related Crenarchaeota, since
this event may have been an important step in the evolution of this archaeal lineage. Moreover, virtually nothing
is known about other types of motility than swimming in
archaea [28], while in bacteria these are starting being
investigated [4]. The fact that no homologues of flagellum
components are encoded in the genomes of at least two
archaeal species that are described as motile (M. kandleri
and P. aerophilum) is also puzzling. Although it is possible
that the strains used for genome sequencing have lost the
flagellum operon following lab cultivation (see for example [32], it would surely be interesting to test motility in
these archaeal species. Alternatively, this observation
could also suggest that other types of motility may occur
in archaea and are made possible by still unknown molecular structures. It would also be very useful to investigate
the relationship between the flagellum and the secretion
Page 10 of 13
(page number not for citation purposes)
BMC Evolutionary Biology 2007, 7:106
http://www.biomedcentral.com/1471-2148/7/106
C. symbiosum
Type II cluster
B
D/E
G
F
H
P. aerophilum
I
J
A. pern
r iix
B
G
F
H
I
G
J
H
I
G
B
B
G
F
H
I
B
I
J
I
J
H
I
J
B
S. acido
d calda
d rrius
J
H
F
I
J
S. solfa
f ttarricus
B
J
G
S. toko
daiii
k d
B
F
G
F
H
I
J
N. equita
N
t ns
J
B
Ancestral cluster
B
B
B
C
D
F
G
H
I
J
T. k
kod
dak
karrensis
T
i
B
B
B
B
B
C
D
F
G
H
I
J
B
B
C
D
F
G
H
I
J
B
B
B
C
D
F
G
H
I
J
B
B
B
C
D
F
G
H
I
J
P. furiosus
B
B
B
C
D
F
G
H
I
J
P. horikoshii
B
B
B
C
D
F
G
H
I
J
P. abyssi
B
B
M. kandleri
B
D/E
F
G
H
I
J
M. thermautotrophicus
C
B
Type I cluster
B
B
C
D
D
F
G
H
I
B
B
C
D
D
F
G
H
I
J
J
M. jjannaschii
i
M. mari
ripaludi
dis
B
B
B
C
D
D
F
F
G
H
I
J
torr
rridus
B
B
B
C
D
E
F
G
H
I
J
B
F
B
B
C
D
E
F
G
H
I
J
acida
d rrmanus
HGT 1
B
B
C
D
F
H
G
I
J
T. acido
T
d phil
i um
B
B
C
D
F
G
H
I
J
T. volcanium
B
B
C
D
F
G
H
I
J
B
B
G
F
H
I
J
F
G
H
I
J
A. fulgidus
H. volcanii
H
i
B
B
B
D
HGT 2
B
F
B
C
G
H
I
B
D
CD
G
F
F
B
F
B
D
F
I
D/E C D
B
G
G
H
I
G
H
H
I
I
B
B
G
G
F
H
I
J
H. mari
H
rismort
r ui
J
F
J
B
B
CD
G
J
G
H
I
J
J
B
D
B
B
CD
H. sp
D
CD
B
D
G
M. hungate
t iB
B
B
I
B
CD
B
B
B
D
B
M. barkeri
F
G
H
I
J
G
F
H
I
J
B
B
N. phara
N
r onis
i
M. burt
r onii
i
J
B
M. mazei
B
G
F
H
B
G
F
H
I
J
I
J
M. acetivorans
B
B
B
B
B
B
J
I
J
G
G
F
H
I
J
CD
F
G
G
F
H
H
I
I
J
J
G
F
H
I
J
G
F
H
I
J
G
G
F
F
H
I
J
H
I
J
G
F
H
I
J
A
Figure
detailed
7 evolutionary scenario for the origin and the evolution of the archaeal flagellum based on figure 6
A detailed evolutionary scenario for the origin and the evolution of the archaeal flagellum based on figure 6. The purple cluster
represents the ancestor of fla1 and fla2 clusters. The black circle represents the last common ancestor of Archaea. Blue
arrows represent the recent duplication events. Black arrows indicate gene movements to different locations from their original positions in the cluster. Dark green { symbols indicate gene insertions within the clusters. The blue, red and black crosses
represent the loss of the fla1 cluster, fla2A cluster, or of single genes, respectively. A red arrow indicates gene inversion. F indicates the fusion of FlaC and FlaD. The green arrows represent horizontal gene transfers.
systems in Archaea. Indeed, archaeal genomes harbor
only a few homologues of bacterial TypeII/IV secretion
systems [28], and it is not known whether they form pili,
despite rare observations [33-35] and evidence for conjugation [36-38].
To sum up, two radically different options for motility
were adopted at the divergence of the two prokaryotic
domains, and much still remain to be uncovered on
archaeal motility systems.
Methods
Data set construction
The list of archaeal flagellum components was retrieved
from the Kyoto Encyclopedia of Genes and Genomes [39].
Homologous sequences of each archaeal flagellum com-
ponent were retrieved from the nr database at the
National Center for Biotechnology Information [40]
using the BlastP program with different seeds [41] and the
ALIBABA program (P. Lopez personal communication).
For each dataset, additional searches using psi-Blast were
performed to search for divergent homologues (especially
from bacteria) [41]. tBlastN searches on the unfinished
genomes of the Halobacteriale Haloferax volcanii were performed at TIGR [42]. Multiple alignments were done with
ClustalW [43] and MUSCLE [44]. For each dataset, the
quality of the alignments obtained with CLUSTALW and
MUSCLE, was evaluated with T-COFFEE (CLUSTALW and
MUSCLE provided alignments of comparable quality, not
shown) [45]. All the alignments were edited and manually improved using the ED program from the MUST package [46]. Regions where the homology between sites was
Page 11 of 13
(page number not for citation purposes)
BMC Evolutionary Biology 2007, 7:106
doubtful were manually removed from the datasets for
phylogenetic analyses.
Phylogenetic analysis
Maximum Likelihood (ML) phylogenetic trees were computed with Phyml [47,48] and the JJT model of amino
acid substitution (Jones, Taylor and Thornton [49]. A
gamma correction with 4 discrete classes of sites was used
to take into account the heterogeneity of evolutionary
rates across sites. The alpha parameter and the proportion
of invariable sites were estimated for each dataset. The
robustness of each branch was estimated by non-parametric bootstrap analysis (1000 replicates) using PHYML. In
addition, bayesian analyses were performed using
MrBayes [50] with a mixed model of amino acid substitution. As for ML tree reconstruction, a gamma correction
with 4 discrete classes of sites was used and the alpha
parameter and the proportion of invariable sites were estimated. MrBayes was run with four chains for 1 million
generations and trees were sampled every 100 generations. To construct the consensus tree, the first 1500 trees
were discarded as "burnin".
Genomic context analysis
The genomic localization of each archaeal flagellum component was manually investigated in all archaeal complete genomes available at the NCBI.
http://www.biomedcentral.com/1471-2148/7/106
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
Authors' contributions
SG conceived the study and supervised the analyses. ED
carried out the analyses. CB and SG refined and completed the analyses and wrote the manuscript. All authors
read and approved the final manuscript.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Alam M, Oesterhelt D: Morphology, function and isolation of
halobacterial flagella. J Mol Biol 1984, 176(4):459-475.
Marwan W, Alam M, Oesterhelt D: Rotation and switching of the
flagellar motor assembly in Halobacterium halobium. J Bacteriol 1991, 173(6):1971-1977.
Macnab RM: The bacterial flagellum: reversible rotary propellor and type III export apparatus.
J Bacteriol 1999,
181(23):7149-7153.
Bardy SL, Ng SY, Jarrell KF: Prokaryotic motility structures.
Microbiology 2003, 149(Pt 2):295-304.
Ng SY, Chaban B, Jarrell KF: Archaeal flagella, bacterial flagella
and type IV pili: a comparison of genes and posttranslational
modifications. J Mol Microbiol Biotechnol 2006, 11(3-5):167-191.
Thomas NA, Bardy SL, Jarrell KF: The archaeal flagellum: a different kind of prokaryotic motility structure. FEMS Microbiol
Rev 2001, 25(2):147-174.
Faguy DM, Koval SF, Jarrell KF: Physical characterization of the
flagella and flagellins from Methanospirillum hungatei. J Bacteriol 1994, 176(24):7491-7498.
Metlina AL: Bacterial and archaeal flagella as prokaryotic
motility organelles. Biochemistry (Mosc) 2004, 69(11):1203-1212.
Cruden D, Sparling R, Markovetz AJ: Isolation and Ultrastructure
of the Flagella of Methanococcus thermolithotrophicus and
Methanospirillum hungatei.
Appl Environ Microbiol 1989,
55(6):1414-1419.
Cohen-Krausz S, Trachtenberg S: The structure of the archeabacterial flagellar filament of the extreme halophile Halobacterium salinarum R1M1 and its relation to eubacterial
23.
24.
25.
26.
27.
28.
29.
30.
31.
flagellar filaments and type IV pili.
J Mol Biol 2002,
321(3):383-395.
Trachtenberg S, Cohen-Krausz S: The archaeabacterial flagellar
filament: a bacterial propeller with a pilus-like structure. J
Mol Microbiol Biotechnol 2006, 11(3-5):208-220.
Bardy SL, Jarrell KF: FlaK of the archaeon Methanococcus maripaludis possesses preflagellin peptidase activity. FEMS Microbiol Lett 2002, 208(1):53-59.
Bardy SL, Jarrell KF: Cleavage of preflagellins by an aspartic acid
signal peptidase is essential for flagellation in the archaeon
Methanococcus voltae. Mol Microbiol 2003, 50(4):1339-1347.
Albers SV, Szabo Z, Driessen AJ: Archaeal homolog of bacterial
type IV prepilin signal peptidases with broad substrate specificity. J Bacteriol 2003, 185(13):3918-3925.
Wieland F, Paul G, Sumper M: Halobacterial flagellins are sulfated glycoproteins. J Biol Chem 1985, 260(28):15180-15185.
Mattick JS: Type IV pili and twitching motility. Annu Rev Microbiol
2002, 56:289-314.
Thomas NA, Pawson CT, Jarrell KF: Insertional inactivation of
the flaH gene in the archaeon Methanococcus voltae results
in non-flagellated cells.
Mol Genet Genomics 2001,
265(4):596-603.
Rudolph J, Oesterhelt D: Deletion analysis of the che operon in
the archaeon Halobacterium salinarium. J Mol Biol 1996,
258(4):548-554.
Szurmant H, Ordal GW: Diversity in chemotaxis mechanisms
among the bacteria and archaea. Microbiol Mol Biol Rev 2004,
68(2):301-319.
Eisen JA: A phylogenomic study of the MutS family of proteins. Nucleic Acids Res 1998, 26(18):4291-4300.
Garrity GM: Bergey's Manual of Systematic Bacteriology. 2nd
edition. Edited by: Garrity GM. New York , Springer-Verlag; 2001.
Galagan JE, Nusbaum C, Roy A, Endrizzi MG, Macdonald P, FitzHugh
W, Calvo S, Engels R, Smirnov S, Atnoor D, Brown A, Allen N, Naylor
J, Stange-Thomann N, DeArellano K, Johnson R, Linton L, McEwan P,
McKernan K, Talamas J, Tirrell A, Ye W, Zimmer A, Barber RD, Cann
I, Graham DE, Grahame DA, Guss AM, Hedderich R, Ingram-Smith C,
Kuettner HC, Krzycki JA, Leigh JA, Li W, Liu J, Mukhopadhyay B,
Reeve JN, Smith K, Springer TA, Umayam LA, White O, White RH,
Conway de Macario E, Ferry JG, Jarrell KF, Jing H, Macario AJ, Paulsen
I, Pritchett M, Sowers KR, Swanson RV, Zinder SH, Lander E, Metcalf
WW, Birren B: The genome of M. acetivorans reveals extensive metabolic and physiological diversity. Genome Res 2002,
12(4):532-542.
Faguy DM, Bayley DP, Kostyukova AS, Thomas NA, Jarrell KF: Isolation and characterization of flagella and flagellin proteins
from the Thermoacidophilic archaea Thermoplasma volcanium and Sulfolobus shibatae. J Bacteriol 1996, 178(3):902-905.
Redder P, Garrett RA: Mutations and rearrangements in the
genome of Sulfolobus solfataricus P2. J Bacteriol 2006,
188(12):4198-4206.
Albers SV, Driessen AJ: Analysis of ATPases of putative secretion operons in the thermoacidophilic archaeon Sulfolobus
solfataricus. Microbiology 2005, 151(Pt 3):763-773.
Planet PJ, Kachlany SC, DeSalle R, Figurski DH: Phylogeny of genes
for secretion NTPases: identification of the widespread tadA
subfamily and development of a diagnostic key for gene classification. Proc Natl Acad Sci U S A 2001, 98(5):2503-2508.
Peabody CR, Chung YJ, Yen MR, Vidal-Ingigliardi D, Pugsley AP, Saier
MH Jr.: Type II protein secretion and its relationship to bacterial type IV pili and archaeal flagella. Microbiology 2003,
149(Pt 11):3051-3072.
Albers SV, Szabo Z, Driessen AJ: Protein secretion in the
Archaea: multiple paths towards a unique cell surface. Nat
Rev Microbiol 2006, 4(7):537-547.
Gribaldo S, Brochier-Armanet C: The origin and evolution of
Archaea: a state of the art. Philos Trans R Soc Lond B Biol Sci 2006,
361(1470):1007-1022.
Brochier C, Forterre P, Gribaldo S: An emerging phylogenetic
core of Archaea: phylogenies of transcription and translation
machineries converge following addition of new genome
sequences. BMC Evol Biol 2005, 5(1):36.
Cavalier-Smith T: The neomuran origin of archaebacteria, the
negibacterial root of the universal tree and bacterial megaclassification. Int J Syst Evol Microbiol 2002, 52(Pt 1):7-76.
Page 12 of 13
(page number not for citation purposes)
BMC Evolutionary Biology 2007, 7:106
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
Labes A, Schonheit P: Sugar utilization in the hyperthermophilic, sulfate-reducing archaeon Archaeoglobus fulgidus
strain 7324: starch degradation to acetate and CO2 via a
modified Embden-Meyerhof pathway and acetyl-CoA synthetase (ADP-forming). Arch Microbiol 2001, 176(5):329-338.
Leadbetter JR, Breznak JA: Physiological ecology of Methanobrevibacter cuticularis sp. nov. and Methanobrevibacter curvatus sp. nov., isolated from the hindgut of the termite
Reticulitermes flavipes.
Appl Environ Microbiol 1996,
62(10):3620-3631.
Miroshnichenko ML, Gongadze GM, Rainey FA, Kostyukova AS,
Lysenko AM, Chernyh NA, Bonch-Osmolovskaya EA: Thermococcus gorgonarius sp. nov. and Thermococcus pacificus sp.
nov.: heterotrophic extremely thermophilic archaea from
New Zealand submarine hot vents. Int J Syst Bacteriol 1998, 48
Pt 1:23-29.
Weiss RL: Attachment of bacteria to sulfur in extreme environments. J Gen Microbiol 1973, 77:501-507.
Prangishvili D, Albers SV, Holz I, Arnold HP, Stedman K, Klein T,
Singh H, Hiort J, Schweier A, Kristjansson JK, Zillig W: Conjugation
in archaea: frequent occurrence of conjugative plasmids in
Sulfolobus. Plasmid 1998, 40(3):190-202.
Rosenshine I, Tchelet R, Mevarech M: The mechanism of DNA
transfer in the mating system of an archaebacterium. Science
1989, 245(4924):1387-1389.
Stedman KM, She Q, Phan H, Holz I, Singh H, Prangishvili D, Garrett
R, Zillig W: pING family of conjugative plasmids from the
extremely thermophilic archaeon Sulfolobus islandicus:
insights into recombination and conjugation in Crenarchaeota. J Bacteriol 2000, 182(24):7014-7020.
KEGG: [http://www.genome.jp/kegg/].
NCBI: [http://www.ncbi.nlm.nih.gov/].
Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ: Gapped BLAST and PSI-BLAST: a new generation of
protein database search programs. Nucleic Acids Res 1997,
25(17):3389-3402.
TIGR: [http://tigrblast.tigr.org/].
Thompson JD, Higgins DG, Gibson TJ: CLUSTAL W: improving
the sensitivity of progressive multiple sequence alignment
through sequence weighting, position-specific gap penalties
and weight matrix choice.
Nucleic Acids Res 1994,
22(22):4673-4680.
Edgar RC: MUSCLE: multiple sequence alignment with high
accuracy and high throughput.
Nucleic Acids Res 2004,
32(5):1792-1797.
Poirot O, O'Toole E, Notredame C: Tcoffee@igs: A web server
for computing, evaluating and combining multiple sequence
alignments. Nucleic Acids Res 2003, 31(13):3503-3506.
Philippe H: MUST, a computer package of Management Utilities for Sequences and Trees.
Nucleic Acids Res 1993,
21(22):5264-5272.
Guindon S, Gascuel O: A simple, fast, and accurate algorithm
to estimate large phylogenies by maximum likelihood. Syst
Biol 2003, 52(5):696-704.
Guindon S, Lethiec F, Duroux P, Gascuel O: PHYML Online--a
web server for fast maximum likelihood-based phylogenetic
inference.
Nucleic Acids Res 2005, 33(Web Server
issue):W557-9.
Jones DT, Taylor WR, Thornton JM: The rapid generation of
mutation data matrices from protein sequences. Comput Appl
Biosci 1992, 8(3):275-282.
Huelsenbeck JP, Ronquist F: MRBAYES: Bayesian inference of
phylogenetic trees. Bioinformatics 2001, 17(8):754-755.
http://www.biomedcentral.com/1471-2148/7/106
Publish with Bio Med Central and every
scientist can read your work free of charge
"BioMed Central will be the most significant development for
disseminating the results of biomedical researc h in our lifetime."
Sir Paul Nurse, Cancer Research UK
Your research papers will be:
available free of charge to the entire biomedical community
peer reviewed and published immediately upon acceptance
cited in PubMed and archived on PubMed Central
yours — you keep the copyright
BioMedcentral
Submit your manuscript here:
http://www.biomedcentral.com/info/publishing_adv.asp
Page 13 of 13
(page number not for citation purposes)