Genome Biology 2008, 9:R158
Open Access
2008Podaret al.Volume 9, Issue 11, Article R158
Research
A genomic analysis of the archaeal system Ignicoccus
hospitalis-Nanoarchaeum equitans
Mircea Podar
*
, Iain Anderson
†
, Kira S Makarova
‡
, James G Elkins
*
,
Natalia Ivanova
†
, Mark A Wall
§
, Athanasios Lykidis
†
,
Kostantinos Mavromatis
†
, Hui Sun
†
, Matthew E Hudson
§**
,
Wenqiong Chen
§††

, Cosmin Deciu
§
, Don Hutchison
§
, Jonathan R Eads
§
,
Abraham Anderson
§‡‡
, Fillipe Fernandes
§
, Ernest Szeto
†
, Alla Lapidus
†
,
Nikos C Kyrpides
†
, Milton H Saier Jr
¶
, Paul M Richardson
†
,
Reinhard Rachel
¥
, Harald Huber
¥
, Jonathan A Eisen
#
, Eugene V Koonin

‡
,
Martin Keller
*
and Karl O Stetter
¥
Addresses:
*
Biosciences Division, Oak Ridge National Laboratory, 1 Bethel Valley Rd, Oak Ridge, TN 37831, USA.
†
DOE Joint Genome Institute,
2800 Mitchell Drive, Walnut Creek, CA 94598, USA.
‡
National Center for Biotechnology Information, National Library of Medicine, National
Institutes of Health, 8600 Rockville Pike, Bethesda, MD 20894, USA.
§
Verenium Corporation, 4955 Directors Place, San Diego CA 92121, USA.
¶
Division of Biological Sciences, University of California San Diego, 9500 Gilman Drive, La Jolla, CA 92037, USA.
¥
Lehrstuhl für Mikrobiologie
und Archaeenzentrum, Universität Regensburg, Universitätstraße 31, Regensburg, D-93053, Germany.
#
Genome Center, University of
California Davis, One Shields Avenue, Davis, CA 95616, USA.
**
Current address: College of Agricultural, Consumer, and Environmental
Sciences University of Illinois at Urbana-Champaign, 1101 W Peabody Dr., Urbana, IL 61801, USA.
††
Current address: Biology Department, San

Diego State University, 5500 Campanile Drive San Diego, CA 92182, USA.
‡‡
Current address: Amgen Inc., One Amgen Center Drive, Thousand
Oaks, CA 91320, USA.
Correspondence: Mircea Podar. Email:
© 2008 Podar et al.; licensee BioMed Central Ltd.
This is an open access article distributed under the terms of the Creative Commons Attribution License ( which
permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
<p>Sequencing of the complete genome of Ignicoccus hospitalis gives insight into its association with another species of Archaea, Nanoar-chaeum equitans.</p>
Abstract
Background: The relationship between the hyperthermophiles Ignicoccus hospitalis and Nanoarchaeum equitans is the
only known example of a specific association between two species of Archaea. Little is known about the mechanisms
that enable this relationship.
Results: We sequenced the complete genome of I. hospitalis and found it to be the smallest among independent, free-
living organisms. A comparative genomic reconstruction suggests that the I. hospitalis lineage has lost most of the genes
associated with a heterotrophic metabolism that is characteristic of most of the Crenarchaeota. A streamlined genome
is also suggested by a low frequency of paralogs and fragmentation of many operons. However, this process appears to
be partially balanced by lateral gene transfer from archaeal and bacterial sources.
Conclusions: A combination of genomic and cellular features suggests highly efficient adaptation to the low energy yield
of sulfur-hydrogen respiration and efficient inorganic carbon and nitrogen assimilation. Evidence of lateral gene exchange
between N. equitans and I. hospitalis indicates that the relationship has impacted both genomes. This association is the
simplest symbiotic system known to date and a unique model for studying mechanisms of interspecific relationships at
the genomic and metabolic levels.
Published: 10 November 2008
Genome Biology 2008, 9:R158 (doi:10.1186/gb-2008-9-11-r158)
Received: 5 September 2008
Revised: 21 October 2008
Accepted: 10 November 2008
The electronic version of this article is the complete one and can be
found online at /> Genome Biology 2008, Volume 9, Issue 11, Article R158 Podar et al. R158.2

Genome Biology 2008, 9:R158
Background
The crenarchaeaote Ignicoccus hospitalis is a specific host for
Nanoarchaeum equitans in a relationship that is thus far
unique, involving two archaeal species [1-3]. Ignicoccus spe-
cies have a chemoautotrophic metabolism that couples CO
2
fixation with sulfur respiration using molecular hydrogen in
high temperature hydrothermal vent systems and thus might
resemble organisms that thrived on the primitive, hot and
anoxic Earth [4-8]. Uniquely among Archaea, Ignicoccus
cells are surrounded by two membranes separated by a wide
periplasmic space within which vesicles and tubular struc-
tures emerge from the cytoplasm [9]. Some of these struc-
tures reach and fuse with the outer membrane [10], which has
a distinct lipid composition and contains pores of a unique
type [11]. The physiological significance of these features and
their potential involvement in the relationship with N. equi-
tans are unknown.
With a highly reduced genome, N. equitans has virtually no
obvious metabolic or energetic capabilities and, using
unknown mechanisms, must obtain metabolites and energy
from I. hospitalis by attaching to its surface [3,12,13]. The
similarity of the lipid compositions between the cytoplasmic
membranes of I. hospitalis and N. equitans suggests specific
lipid partitioning and transport mechanisms [13]. In addi-
tion, carbon labeling and cell fractionation have demon-
strated the transfer of amino acids from I. hospitalis to N.
equitans [3]. In co-cultures with I. hospitalis, N. equitans
cells can be regularly observed detached and, for some time,

they appear to maintain their membrane integrity, at least
based on live-dead staining [3]. The mechanism of separation
from the host cell and the potential existence of a reattach-
ment process are still unknown. Attempts to propagate N.
equitans in co-cultures with other archaea, including other
species of Ignicoccus, have not been successful, suggesting
that the relationship with I. hospitalis is highly specific and
involves a recognition mechanism [3]. While under labora-
tory conditions the effects exerted by N. equitans on its host
range from mildly to moderately inhibitory [1,3], Nanoar-
chaeum might confer on Ignicoccus an advantage in coloniz-
ing hydrothermal vents [14]. As its exact nature remains
elusive, provisionally describing this relationship as a symbi-
osis is compatible with representing either a novel type of
interspecific association or fitting within recognized catego-
ries of microbial interactions [15].
It has been proposed that genomic characteristics of N. equi-
tans such as the numerous split genes and extremely compact
genome might be signatures of an ancient lineage [12,16],
although a viable alternative seems to be that at least some of
these features are secondarily derived [17]. The age of the
Ignicoccus-Nanoarchaeum relationship is unknown,
although both organisms represent hyperthermophilic line-
ages and inhabit types of ecosystems that are often consid-
ered to be ancient [7,18]. This system provides insights into
physiological mechanisms of interaction between unicellular
organisms and can offer clues to evolutionary events that
shape the genomes of symbionts leading to physiological
interdependence. The Ignicoccus-Nanoarchaeum relation-
ship might even serve as an analogous model to proposed

symbiotic events that could have led to the formation of
eukaryotic cells [19]. To advance the study of this relationship
at the genomic level, we sequenced the complete genome of I.
hospitalis, complementing that of N. equitans [12]. In this
study, in conjunction with the available physiological and
morphological data, we performed the genomic analysis and
metabolic reconstruction of I. hospitalis, as a step to deci-
phering the evolutionary history and the molecular mecha-
nisms that enable the symbiotic relationship between the two
archaea.
Results and discussion
A minimal genome
The genome of I. hospitalis consists of a single circular chro-
mosome (Table 1). At 1,297,538 bp, the genome of I. hospi-
talis is the smallest among free-living organisms, which do
not require a continuous association with another species and
can replicate independently (Figure 1). Even the combined
gene complement of I. hospitalis and N. equitans (1,434 and
556 protein-coding genes, respectively) is significantly
smaller than that of average free-living bacteria (approxi-
mately 3,600 genes) or archaea (approximately 2,300 genes),
based on the available completed genomes. The size distribu-
tion of 623 complete microbial genomes indicates that the 1-
2 Mbp range includes both obligate symbionts/parasites as
well as free living bacteria and archaea (Figure 1). The mini-
mal genome for free-living organisms may therefore be on the
Table 1
General features of the I. hospitalis genome
Parameter Value %
Chromosome size (bp) 1,297,538

Chromosome G+C content 56.5
Total number of genes 1,494 100
Protein coding genes .1,444 96.6
RNA genes .50 3.3
Genes with function prediction .885 59.2
Genes without function prediction 559 37.4
Genes in ortholog clusters .1,149 76.9
Genes in paralog clusters .406 27.2
Fusion genes .27 1.8
Genes assigned to COGs .972 65.1
Genes assigned to arCOGs 1,155 80.5
Genes assigned to Pfam domains .875 58.6
Genes with signal peptides .213 14.3
Genes with transmembrane helices .216 14.5
Putative pseudogenes (RNA + proteins) 12 0.8
Genome Biology 2008, Volume 9, Issue 11, Article R158 Podar et al. R158.3
Genome Biology 2008, 9:R158
order of 1 Mbp, several taxonomically and metabolically dis-
tant archaeal and bacterial lineages having independently
reached near-minimal functional gene sets for their respec-
tive ecological niches.
The sizes of microbial genomes are the result of dynamic
equilibria between contraction by deletions and expansion
due to duplications, lateral gene transfer and insertion of
mobile DNA. For free-living organisms with very large effec-
tive population sizes, genome streamlining is likely to be a
selective consequence of reducing the metabolic burden to
maintain DNA of little adaptive value, as illustrated by the
genomes of such highly successful and widespread lineages as
Prochlorococcus and Pelagibacter [20,21]. An alternative

(but not necessarily exclusive) hypothesis links genome
reduction to elevated mutation rates in large populations.
Accumulation of mutations could lead to inactivation and loss
of genes that make weak contribution to the fitness of the
respective organisms [22]. Ignicoccus, however, inhabits het-
erogeneous, geographically dispersed and relatively ephem-
eral hydrothermal marine environments. Such organisms
generally have small effective populations and experience
periodic bottlenecks and limited gene flow [23]. Conceivably,
in a case like this, genome contraction might have to do with
the very active recombination and DNA repair that organisms
inhabiting extreme environments employ for maintaining
genomic integrity. Frequent recombination might not only
efficiently remove deleterious mutations induced by the envi-
ronmental conditions but also generate diversity and increase
the fixation rate of adaptive alleles [24,25]. A high frequency
of illegitimate, intra-chromosome recombination could also
be effective in preventing genome expansion by increasing
the frequency of deletions and counteracting gene duplica-
tion. This might explain the reduced genome size in many
members of the Archaea and contribute to their proposed
higher adaptability to chronic energy stress [26]. While we
expect these general principles to be valid in the Nanoar-
chaeum-Ignicoccus system as well, the co-evolution of these
two organisms also left unique imprints on their physiology
[2,3]. The most striking effect of this co-evolution, however,
is the massive gene loss in N. equitans, resembling that of
Relationship between the genome size and the number of protein-coding genes in 623 complete archaeal and bacterial genomes, based on data in IMG version 2.5 (March 2008)Figure 1
Relationship between the genome size and the number of protein-coding genes in 623 complete archaeal and bacterial genomes, based on data in IMG
version 2.5 (March 2008). The line points to I. hospitalis having the smallest genome among independently replicating organisms. The genomes of obligate

parasites/symbionts are represented by grey circles. The shaded region of genome sizes spans the transition between obligate symbionts/parasites and
free-living organisms.
100
1000
10000
0.1 1 10
Nanoarchaeum equitans
Buchnera aphidicola BCc
Cand. Carsonella ruddii
Mycoplasma genitalium
Pelagibacter ubique
HTCC1062
Rickettsia conorii
Ignicoccus hospitalis
Sulfolobus solfataricus
Thermoplasma acidophilum
Staphylothermus marinus
Sodalis glossinidius morsitans
Trichodesmium
erythraeum
Burkholderia xenovorans
Solibacter
usitatus
0.5 5
500
5000
Methanosarcina
acetivorans
Haloarcula marismortui
Methanosarcina barkeri

Pyrobaculum aerophilum
Hyperthermus butylicus
Aeropyrum pernix
Genome size (Mbp)
Number of protein encoding genes
Bacteria, obligate symbionts/parasites
Bacteria, free living and facultative symbionts
Euryarchaeota
Crenarchaeota
Cand. Sulcia muelleri
Rhodococcus sp. RHA1
Sorangium cellulosum
Bartonella henselae
1.297
Genome Biology 2008, Volume 9, Issue 11, Article R158 Podar et al. R158.4
Genome Biology 2008, 9:R158
obligate intracellular bacterial symbionts and, as an extreme
case, that of eukaryotic organelles [12].
The recently published database of archaeal clusters of
orthologous genes (arCOGs) provides a framework for com-
paring the I. hospitalis genomic data to genes from 41 previ-
ously sequenced archaeal genomes organized into sets of
probable orthologs [27]. Of the 1,434 annotated I. hospitalis
protein-coding genes, 1,155 (80.5%) were assigned to
arCOGs, a coverage that is the lowest among the Desulfuro-
coccales (85% on average) and overall among thermophilic
Crenarchaeota.
I. hospitalis lacks orthologs of 19 genes from the Crenarchae-
ota core (that is, genes that are represented in all 12 available
genomes of thermophilic species of Crenarchaeota included

in the arCOGs) [27] (Table S1 in Additional data file 1). None
of these genes include components of information processing
systems, indicating that these systems are largely intact in I.
hospitalis despite the small genome. The missing genes
encode, primarily, diverse metabolic enzymes, some of which
- for example, thymidylate kinase - catalyze essential reac-
tions. Conceivably, these enzymes are substituted for by dis-
tant homologs that so far remain undetected or by analogs.
Using the assignment of I. hospitalis genes to arCOGs, we
applied weighted parsimony to perform a reconstruction of
gene gain and loss events in archaea [27,28], with an empha-
sis on the I. hospitalis lineage. The small genome size appears
to be a result of gene loss that has vastly predominated the
evolution of this lineage: it was inferred that approximately
484 arCOGs were lost, compared to the inferred gain of only
56. Approximately 946 arCOGs (1,094 genes, representing
76% of the I. hospitalis gene set) appear to have been inher-
ited from the last common ancestor of the Desulforococcales,
the order to which Ignicoccus belongs, together with Aero-
pyrum pernix, Hyperthermus butylicus and Staphylother-
mus marinus. The functional distribution of the lost genes is
consistent with the fact that I. hospitalis is an obligate anaer-
obic autotroph. In contrast to A. pernix, numerous genes
related to aerobic metabolism as well as catabolism and
transport of amino acids, sugar and nucleotides were lost,
along with many transcriptional regulators (Figure 2; Table
S2 in Additional data file 1). An analysis of arCOGs that are
present in N. equitans but absent in I. hospitalis does not sug-
gest that the inferred gene loss in I. hospitalis was accompa-
nied by transfer of potentially essential functions to the

symbiont (Table S3 in Additional data file 1). Ignicoccus is far
removed from the root of the tree of thermophilic Crenar-
chaeota (whether the tree is constructed for rRNA or various
informational proteins), and the tree, including basal
branches, is dominated by heterotrophs and mixotrophs (Fig-
ure S1 in Additional data file 2). Thus, the alternative sce-
nario, namely, that Ignicoccus reflects the ancestral state for
this entire group, is not supported by the phylogenetic analy-
ses. However, this might reflect our incomplete sampling of
the archaeal diversity and the bias towards isolation and char-
acterization of heterotrophs. A better understanding of the
direction of evolution in archaeal genome size and architec-
ture will require a significant increase in the number and
diversity of cultivated species and sequenced genomes,
including close relatives of I. hospitalis and additional
chemolithoautotrophs.
The reduced frequency of duplicated genes (paralogs) in I.
hospitalis compared to all other archaea except N. equitans
(Figure 3) and the absence of transposable elements support
the hypothesis of genome streamlining. Furthermore,
approximately 180 chromosomal gene clusters that are typi-
cally conserved in archaea are disrupted in the genome,
including some of the ribosomal operons as well as those
encoding the proteasome components, ATP synthase and
DNA topoisomerase VI. As it is unlikely that so many gene
clusters and operons have independently assembled in
archaeal lineages not directly related, the architecture of the
I. hospitalis genome suggests that recombination events have
resulted in gene cluster fragmentation, deletions, and may
have restricted gene family expansion. On the other hand, it

is notable that several families of paralogous genes are
uniquely expanded in I. hospitalis (Table S4 in Additional
data file 1). The most intriguing is the presence of 10 genes
that encode WD40-repeat-containing proteins. Proteins con-
Numbers of arCOGs in different functional categories (COG classification) lost or gained in the I. hospitalis lineageFigure 2
Numbers of arCOGs in different functional categories (COG
classification) lost or gained in the I. hospitalis lineage. The sets of lost and
gained genes were derived on the basis of a comparison of the I. hospitalis
gene compliment with the reconstructed gene set of the last common
ancestor of Desulfurococcales [27] (see Additional data files). The
numbers of arCOGs in each category that are present in N. equitans but
are absent in I. hospitalis are also indicated. The one letter code for COG
categories is the following: amino acid transport and metabolism (E);
carbohydrate transport and metabolism (G); cell cycle control, cell
division, chromosome partitioning (D); cell motility (N); cell wall/
membrane/envelope biogenesis (M); coenzyme transport and metabolism
(H); defense mechanisms (V); energy production and conversion (C);
inorganic ion transport and metabolism (P); intracellular trafficking,
secretion, and vesicular transport (U); lipid transport and metabolism (I);
nucleotide transport and metabolism (F); posttranslational modification,
protein turnover, chaperones (O); replication, recombination and repair
(L); secondary metabolites biosynthesis, transport and catabolism (Q);
signal transduction mechanisms (T); transcription (K); and translation,
ribosomal structure and biogenesis (J).
0
5
10
15
20
25

30
35
40
45
50
JKLD_UO C E F G H I P M N Q T V
Loss
Gain
N.equitans minus I. hospitalis
COG categories
Genome Biology 2008, Volume 9, Issue 11, Article R158 Podar et al. R158.5
Genome Biology 2008, 9:R158
taining WD40 repeats are among the most abundant and
highly conserved in eukaryotes, where they are key structural
components of a variety of macromolecular complexes [29].
Proteins containing these repeats are also widely scattered
among archaea and bacteria but are mostly encoded in (rela-
tively) large genomes [30]. In particular, among archaea, we
have detected comparable expansions of WD40-containing
proteins only in Methanosarcinales, a group of Euryarchaeota
that displays significant gene gain [27]. Conceivably, the
WD40-proteins of I. hospitalis are involved in the organiza-
tion of specific protein complexes and/or cellular compart-
ments, and potentially might contribute to the interaction
with N. equitans. Similarly, I. hospitalis encodes 9 proteins
containing the V4R domain and 12 proteins containing the
CBS domain, both small-molecule-binding domains that are
likely to be involved in metabolic regulation and signaling
[31,32] (Table S4 in Additional data file 1). Considering the
homology identified between the V4R domain and a compo-

nent of the eukaryotic Golgi vesicle transport machinery [33],
some of the expanded V4R gene family members also might
be implicated in the unique vesicle formation process that has
been observed in Ignicoccus [9].
In addition to streamlining, selection for reducing metabolic
cost in I. hospitalis may have impacted its proteome compo-
sition. In hyperthermophiles, certain biases in amino acid
usage have been associated with side chain physical and
chemical properties that contribute to increased protein sta-
bility [34,35]. For example, a preference for lysine over
arginine has been attributed to a greater flexibility of the
lysine side chain, which entropically stabilizes the folded state
of proteins [36]. While the overall amino acid usage in the N.
equitans-I. hospitalis proteomes follows the distribution
observed for other hyperthemophiles, there is a significant
increase in lysine over arginine usage in I. hospitalis relative
to the values that could be predicted from the GC content
(Figure 4; note that the two positively charged amino acids,
lysine and arginine, are often interchangeable in proteins but
are encoded by contrasting codons, namely AAA/G for lysine,
and CGX and AGA/G for arginine, hence the strong correla-
tion of the abundance of these amino acids with the GC con-
tent). This discrepancy could be explained by selection at the
genomic level against using the metabolically more expensive
arginine. Arginine biosynthesis in Ignicoccus is predicted to
proceed via carbamoyl-phosphate and would require five ATP
equivalents, whereas lysine, synthesized from 2-oxoglutarate
via the aminoadipate pathway, would use two ATP equiva-
lents (Figure 4). Metabolic cost and nutrient availability have
been proposed to play a selective role in the evolution of

genome size, GC content and amino acid use in organisms
that inhabit oligotrophic or energetically poor environments
[20,26,37]. Since sulfur-hydrogen respiration is energetically
weak [38], such genomic and proteomic adaptations allow I.
hospitalis not only to be a competitive vent colonizer but also
to support N. equitans. At present, in the absence of sequence
data from other species of Ignicoccus, we cannot distinguish
Paralog distribution in completely sequenced archaeal genomesFigure 3
Paralog distribution in completely sequenced archaeal genomes. (a) The average number of paralogs in arCOGs for completely sequenced archaeal
genomes. The arrows point to the vales for N. equitans and I. hospitalis genomes, respectively. (b) Paralog density in completed genomes of species from
the order Desulfurococcales and in N. equitans, determined by blastclust using a variable identity threshold over at least 50% of the aligned pairs of
sequences.
(a)
1.00
1.10
1.20
1.30
1.40
1.50
1.60
1.70
1.80
1.90
2.00
N. equitans
I. hospitalis
Metmp
Theac
MetmC
Metka

Pyrca
Hypbu
Thevo
Aerpe
Stama
Metja
Pyris
Pyrho
Picto
Thete
Metth
Metsa
Pyrab
Pyrae
Metst
Metla
Sulac
Pyrfu
Calma
Thepe
Theko
Metbu
Halwa
Metcu
Natph
Halsp
Arcfu
Sulto
Censy
Metma

Metba
Uncme
Sulso
Halma
Metac
Methu
Euryarchaeota
Crenarchaeota
Paralog density
0
50
100
150
200
250
25 30 35 40 45 50 55 60
% pairwise identity (>50% gene)
Paralogs / Mbp
Ignicoccus
Hyperthermus
Staphylothermus
Aeropyrum
Nanoarchaeum
(b)
Genome Biology 2008, Volume 9, Issue 11, Article R158 Podar et al. R158.6
Genome Biology 2008, 9:R158
if the relationship with N. equitans has directly influenced
these genomic features of I. hospitalis.
Lateral gene transfer
The cell-cell contact between I. hospitalis and N. equitans

seems to present an opportunity for extensive lateral gene
transfer (LGT). LGT is considered to play a major role in
microbial genome evolution and is well-documented in sym-
biotic systems and in environmental microbial communities
[39-42]. Recent LGT events are readily detected with various
methods based on nucleotide composition or codon usage,
but methods that rely on protein sequence similarity and phy-
logenetic trees are more informative for ancient LGT events
[43]. To analyze the I. hospitalis genome for potential LGT
events, we therefore combined automatic genome-wide phyl-
ogenetic reconstruction using PyPhy [44] with similarity
searches and COG distribution analysis. The LGT candidates
were further analyzed using hand-curated alignments and
maximum likelihood phylogenetic analyses. Identifying the
LGT direction requires analysis of conflicts between the
topologies of the corresponding gene trees and the adopted
species tree. The position of N. equitans within the Archaea is
controversial and ranges from representing a distinct and
basal phylum [1,12,16] to being a derived member of order
Thermococcales from the Euryarchaeota [17]. Many gene
trees identify the Thermococcales as an early diverging line-
age, which further complicates this distinction. Ignicoccus on
the other hand has been confidently assigned to order Desul-
furococcales from the Crenarchaeota based on phylogenetic
and arCOG analysis. Therefore, when attempting to infer
direction of LGT, we relied on the phylogenetic placing of N.
equitans and I. hospitalis genes relative to other crenarchaeal
homologues, especially those from the Desulfurococcales
(Aeropyrum, Hyperthermus and Staphylothermus).
A small fraction of I. hospitalis genes (approximately 6%)

appear to have been transferred from lineages within Euryar-
chaeota, while approximately 4% seem to be of bacterial ori-
gin (Figure 5). Many of those genes encode subunits of
protein complexes involved in energy metabolism or trans-
porters and might have been acquired by I. hospitalis as small
clusters or operons. Examples of putative 'bacterial' gene
clusters include those encoding bacterial type polysulfide
reductase (Igni528-530), the multisubunit putative Ech
hydrogenase (Igni542-546, Igni1144-148) and a nitrate
reductase-like complex (Igni1377-1379). Among the clusters
of apparent origin from Euryarchaeota are genes encoding
ABC-type transporters for antibiotics and molybdate
(Igni146-147, Igni1340-1343) as well as a 2-oxoacid:ferre-
doxin oxidoreductase complex (Igni1075-1078). Other genes
Lysine and arginine use in archaeal proteomes, relative to genome G+C contentFigure 4
Lysine and arginine use in archaeal proteomes, relative to genome G+C content. The dotted lines represent the linear fit to the hyperthermophile data and
the goodness of fit values. The archaeal classification as hyperthermophiles, thermophiles and mesophiles follows that of the NCBI Genome Project
database [100]. The proposed pathways for the biosynthesis of the two amino acids, the genes predicted to be involved and the metabolic costs of the two
reactions are shown below the graphs.
r
2
= 0.87
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0

9.0
25 30 35 40 45 50 55 60 65 70
GC content (%)
Arg %
Hyperthermophiles
Mesophiles
Thermophiles
r
2
= 0.88
2.0
4.0
6.0
8.0
10.0
12.00
Lys %
Arg
Lys
Ignicoccus
Ignicoccus
Nanoarchaeum
Nanoarchaeum
/Gln
Asp
Carbamoyl-P
CO
1399,
1400
635,

1430
2
NH
4
+
Arg
2 ATP
57,621,
728,944,315
AcCoA
ATP,
NADPH
Ornithine
Citrulline
1387
H O
2
ATP
Glutamate
Homo
citrate
1249
Lys
944,315
NADPH
Glu
2-amino
adipate
621
2-oxo

glutarate
AcCoA
2-oxoglutarate
859
Glu
2-oxoglutarate
AcCoA
57
ATP
728
acetate
25 30 35 40 45 50 55 60 65 70
GC content (%)
Genome Biology 2008, Volume 9, Issue 11, Article R158 Podar et al. R158.7
Genome Biology 2008, 9:R158
encoding characteristic proteins of Euryarchaeota are scat-
tered in the genome (for example, the CrcB-like integral
membrane protein Igni921, a 6Fe-6S prismane cluster-con-
taining protein Igni960, micrococcal thermonuclease
Igni1343, thermophilic glucose-6-phosphate isomerase
Igni415). If N. equitans is a derived member of
Thermococcales, as some gene trees and genomic analyses
suggest [17,27], then some of the putative euryarchaeal LGTs
in the I. hospitalis genome might actually represent transfers
from N. equitans. Such transfers could have occurred during
extensive genome degradation suffered by N. equitans associ-
ated with elimination of metabolic functions, similar to cases
of nuclear transfer of symbiont genes during eukaryotic
organelle formation. Additional LGTs from bacteria and/or
archaea, including N. equitans, might be hidden in the large

number of genes (>600 or approximately 40% of the open
reading frames) that either lack detectable homologs or are
placed unresolved within the Archaea due to insufficient phy-
logenetic signal.
One of the possible outcomes of LGT in symbiotic associa-
tions involves orthologous gene displacement in the recipient
genome and maintenance of the gene in the donor genome as
well. In the N. equitans-I. hospitalis system, we identified 13
such cases, in which the orthologs in both genomes are each
other's closest homologues (Figure 5). Several of the genes
appear to have been transferred from N. equitans to I. hospi-
talis, including ones encoding valyl-tRNA synthetase
(Igni220-Neq252), tyrosyl-tRNA synthetase (Igni347-
Neq389) and a type IV endonuclease (Igni1092-Neq77a)
Taxonomic classification of I. hospitalis protein-coding genes based on phylogenetic and COG distribution analysesFigure 5
Taxonomic classification of I. hospitalis protein-coding genes based on phylogenetic and COG distribution analyses. Genes labeled in green or blue-green
are of Crenarchaeota-type or are of unresolved archaeal nature, respectively. Genes that could represent horizontal gene transfers from Euryarchaeota or
Bacteria are labeled in purple and yellow, respectively. Genes that have their closest ortholog in N. equitans are labeled red and are described in the table.
Genes labeled in gray lack recognizable homologues in other microbial genomes or have unresolved phylogenies preventing confident affiliation to either
Archaea or Bacteria.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100
101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150
151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200
201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250
251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300
301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350
351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400
401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450
451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500

501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550
551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600
601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650
651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700
701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750
751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800
801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850
851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900
901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950
951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000
10011002 1003 1004 10051006100710081009 1010 1011 1012101310141015 1016 1017 1018 1019102010211022 1023 1024 1025102610271028 1029 1030 1031 1032103310341035 1036 1037 1038 1039104010411042 1043 1044 1045104610471048 1049 1050
10511052 1053 1054 10551056105710581059 1060 1061 1062106310641065 1066 1067 1068 1069107010711072 1073 1074 1075107610771078 1079 1080 1081 1082108310841085 1086 1087 1088 1089109010911092 1093 1094 1095109610971098 1099 1100
11011102 1103 1104 11051106110711081109 1110 1111 1112111311141115 1116 1117 1118 1119112011211122 1123 1124 1125112611271128 1129 1130 1131 1132113311341135 1136 1137 1138 1139114011411142 1143 1144 1145114611471148 1149 1150
11511152 1153 1154 11551156115711581159 1160 1161 1162116311641165 1166 1167 1168 1169117011711172 1173 1174 1175117611771178 1179 1180 1181 1182118311841185 1186 1187 1188 1189119011911192 1193 1194 1195119611971198 1199 1200
12011202 1203 1204 12051206120712081209 1210 1211 1212121312141215 1216 1217 1218 1219122012211222 1223 1224 1225122612271228 1229 1230 1231 1232123312341235 1236 1237 1238 1239124012411242 1243 1244 1245124612471248 1249 1250
12511252 1253 1254 12551256125712581259 1260 1261 1262126312641265 1266 1267 1268 1269127012711272 1273 1274 1275127612771278 1279 1280 1281 1282128312841285 1286 1287 1288 1289129012911292 1293 1294 1295129612971298 1299 1300
13011302 1303 1304 13051306130713081309 1310 1311 1312131313141315 1316 1317 1318 1319132013211322 1323 1324 1325132613271328 1329 1330 1331 1332133313341335 1336 1337 1338 1339134013411342 1343 1344 1345134613471348 1349 1350
13511352 1353 1354 13551356135713581359 1360 1361 1362136313641365 1366 1367 1368 1369137013711372 1373 1374 1375137613771378 1379 1380 1381 1382138313841385 1386 1387 1388 1389139013911392 1393 1394 1395139613971398 1399 1400
14011402 1403 1404 14051406140714081409 1410 1411 1412141314141415 1416 1417 1418 1419142014211422 1423 1424 1425142614271428 1429 1430 1431 1432143314341435 1436 1437 1438 1439144014411442
Nanoarchaeum 13 0.902
Bacteria 60 4.1
Euryarchaeota 87 5.9
Archaea (unresolved) 247 17.2
Crenarchaeota 659 45.7
Unknown 376 26.2
Genes % total
Igni_Nano Reciprocal top orthologues:
Igni0112_Neq368 : COG0648, endo IV
Igni0145_Neq369 : COG1061, helicase
Igni0220_Neq252: COG0525, Val-tRNA synthetase

Igni0347_Neq389: COG0162, Tyr-tRNA synthetase
Igni0701_Neq453: COG1719, V4R
Igni0719_Neq028: Aminopeptidase, Iap family
Igni0738_Neq412 : COG0260, leucyl aminopeptidase
Igni0899_Neq024 : COG1252, dehydrogenase
Igni1092_Neq077a: COG0648, endo IV
Igni1332_Neq453: COG1719, V4R
Igni1353_Neq526/049: Fusion of radical SAM family
enzyme and queuine/archaeosine tRNA-ribosyltransferase
Igni1357_Neq233: unknown
Igni1397_Neq009: unknown
Genome Biology 2008, Volume 9, Issue 11, Article R158 Podar et al. R158.8
Genome Biology 2008, 9:R158
(Figure 6a; Figure S2 in Additional data file 2). Two genes
involved in recombination and repair that form a predicted
operon in N. equitans (an AP endonuclease 2 family and a
DEAD/DEAH box helicase, NEQ368-369) have also been
transferred to I. hospitalis, either as independent events or
becoming separated later by genomic rearrangement
(Igni0112 and 0145). Genes encoding aminoacyl tRNA syn-
thetases and recombination and repair proteins are fre-
quently exchanged in microbial communities and might
increase the fitness of recipient organisms, for example, by
conferring antibiotic resistance in the case of aminoacyl-
tRNA synthetases [45,46].
A similar case of lateral transfer likely involved the gene
encoding leucyl aminopeptidase (LAP), Igni738-Neq412 (Fig-
ure 6b). LAPs are ubiquitous in bacteria and eukaryotes but
their presence in archaea is so far strictly limited to the Des-
ulfurococcales and the Cenarchaeales. While no specific func-

tion has been described so far for archaeal LAPs, in bacteria
they are multifunctional proteins, with roles in protein turn-
over as well as in transcription control and recombination
[47]. The absence of LAP in Euryarchaeota, in Korarchaeum
cryptofilum (a potentially basal archaeal lineage with affini-
ties with the Crenarchaeota [48]) as well as in two of the four
Crenarchaeota orders for which genomic data are available
may suggest that the Desulfurococcales and Cenarchaeales
acquired the gene via LGT from bacteria. The phylogenetic
analysis places the I. hospitalis gene close to that of N. equi-
tans but not part of the Desulfurococcales clade. The high
level of sequence similarity between the N. equitans and I.
hospitalis LAP genes (40%) surpasses that between any of the
other Desulfurococcales (approximately 30%). However, the
direction of the transfer is uncertain. The exclusion of the I.
hospitalis LAP from the clade formed by the other Crenar-
chaeota homologs suggests that the Ignicoccus gene may
have been acquired from N. equitans followed by orthologous
gene displacement. Based on this scenario, the original pres-
ence of LAP in N. equitans would be at odds with its pur-
ported affiliation with the Euryarchaeaota and specifically the
Thermococcales, which are lacking leucyl aminopeptidases.
The alternative hypothesis, transfer of the LAP gene from I.
hospitalis to N. equitans, is challenged by the separation of
the Ignicoccus-Nanoarchaeum clade from the other Desul-
furococcales. Complete genome sequences of other Ignicoc-
cus or Nanoarchaeota species may help distinguish between
these competing hypotheses.
Genetic information processing in I. hospitalis, as inferred
from the genome sequence, is typical of the Crenarchaeota.

Orthologs of two family B DNA polymerases are present in
the genome (Igni62, 690); one corresponds to the aphidico-
Maximum likelihood phylogenetic trees (a) of archaeal valyl-tRNA synthetases and (b) of leucyl aminopeptidases representing the three domains of life and including all the known archaeal sequencesFigure 6
Maximum likelihood phylogenetic trees (a) of archaeal valyl-tRNA synthetases and (b) of leucyl aminopeptidases representing the three domains of life
and including all the known archaeal sequences. Numbers indicate bootstrap support based on 100 replicates. Where the value was <50, the branch was
collapsed. The scale bar indicates the inferred number of substitutions per site. The sequence alignments used to generate the trees are provided in the
Additional data file 4.
Archaeoglobus fulgidus
Methanospirillum hungatei
Methanosarcina mazei
Methanosaeta thermophila
Natronomonas pharaonis
Halobacterium sp. NRC-1
Methanosphaera stadtmanae
Methanococcus maripaludis
Methanocaldococcus jannaschii
uncultured Alv FOS1
Nitrosopumilus maritimus
Cen
archaeum symbiosum
C. Korarchaeum cryptofilum
Nanoarchaeum equitans
Ignicoccus hospitalis
Pyrobaculum calidifontis
Thermoproteus neutrofilus
Pyrobaculum islandicus
Pyrobaculum islandicum
Pyrobaculum aerophilum
Caldivirga maquilingensis
Stap

hylothermus marinus
Aeropyrum pernix
Hyperthermus butylicus
Pyrococcus furiosus
Thermococcus kodakaraensis
Pyrococcus abyssi
Metallosphaera sedula
Sulfolobus solfataricus
Sulfolobus tokodaii
Sulfolobus acidocaldarius
0.2
99
100
100
100
100
92
100
93
100
100
100
85
Thermoproteales
Desulfurococcales
Cenarchaeales
Sulfolobales
EURYARCHAEOTA
CRENARCHAEOTA
Thermococcales

Schizosaccharomyces pombe
Homo sapiens
Nematostella vectensis
Thiomicrospira crunogena
Bacillus subtilis
Solibacter usitatus
Aqu
if
ex aeolicus
Solanum tuberosum
Arabidopsis thaliana
Ostreococcus tauri
Anabaena variabilis
Proc
h
l
o
rococcus marinus
Nanoarchaeum equitans
Ignicoccus hospitalis
Staphylothermus marinus
Hyperthermus butylicus
Pyrolobus fumari
Aeropyrum pernix
Nitrosopumilus maritimus
Ce
n
archaeum symbiosum
0.2
Cyanobacteria

Viriplantae
Metazoa
Fungi
Desulfurococcales
Cenarchaeales
ARCHAEA
EUKARYA
BACTERIA
100
80
100
61
98
71
100
54
69
(a) (b)
Genome Biology 2008, Volume 9, Issue 11, Article R158 Podar et al. R158.9
Genome Biology 2008, 9:R158
lin-resistant DNA polymerase I (polA), and the other to the
aphidicolin-sensitive DNA polymerase II (polB) of Aero-
pyrum pernix [49]. No orthologs of the third family B DNA
polymerase or Euryarchaeota-type heterodimeric DNA
polymerase were found. Unlike other archaeal genomes, the
genes coding for replication initiation/origin recognition fac-
tor (Orc1/Cdc6) are not co-localized with the predicted origin
of replication [50,51], a characteristic potentially related to
general operon fragmentation in I. hospitalis. Unlike other
archaea, including I. hospitalis, that possess DNA primases

consisting of a small (catalytic) and large (structural) subu-
nits, N. equitans seems to encode a single-subunit primase
(NEQ395) in which the small subunit is fused to the carboxy-
terminal domain of the large subunit [52] (EVK, unpublished
observations). This may be the result of extreme genome con-
traction in this organism, possibly linked to its symbiotic life-
style. Similarly, an important molecular machine absent in N.
equitans but present in I. hospitalis is the RNase P complex
(RNA and four separate proteins subunits, rpp14, 21, 29 and
30). It has been recently shown that tRNA processing in N.
equitans is RNase P-independent, most likely because
genome shrinkage led to the evolution of leaderless tRNAs
that was followed by the loss of all five RNAse P complex
genes [53].
Transport processes
The membrane composition of hyperthermophiles is specifi-
cally adapted to reduce proton and ion permeability, which
increase with temperature [54]. Cyclic tetraether-type lipids
(caldarchaeol) that are present in the cytoplasmic membrane
of I. hospitalis and in the cell membrane of N. equitans are
especially associated with low permeability [13]. In contrast,
the absence of caldarchaeol in the outer membrane of Ignic-
occus and the presence of protein pores [11] indicate poten-
tially less restrictive exchanges with the environment through
the outer membrane. With only eight types of transporters,
almost all predicted to be specific for inorganic ions or export
of intracellular solutes (Figure 7), N. equitans is unlikely to
import by itself all of the required metabolic precursors from
its host. Consistent with its streamlined genome and
autotrophic lifestyle, I. hospitalis also encodes very few trans-

porters (<3% of its proteome), the lowest number among the
sequenced species of Crenarchaeota. The types of transport-
ers and their inferred specificities are described in Figure 7. A
number of inferred subunits of ABC transporters were found
in membrane preparations of I. hospitalis cells, showing that
these proteins are expressed in significant amounts [55]. An
unexpected finding for an obligate autotroph was the pres-
ence of genes encoding two ABC transporters for oligopep-
tides and branched amino acids. Under laboratory
conditions, it was indeed found that addition of peptides
improved growth of I. hospitalis [2], suggesting that, in its
natural environment, this organism might be opportunistic in
utilizing such resources. The different lipid and protein com-
positions between the cytoplasmic membrane and the outer
membrane of I. hospitalis [10,13] suggest the existence of
specific partitioning mechanisms. The genome encodes a pre-
dicted gene (Igni479) from the LolE permease family, an
ATP-dependent transport system involved in lipoprotein
release that has been shown in Buchnera to transport lipids
targeted to the outer membrane across the inner membrane
[56] and that might play a role in I. hospitalis membrane
synthesis.
While some proteins may spontaneously insert in the mem-
brane, most transport into and across the membrane requires
the function of specialized cellular systems [57]. All the com-
ponents of the Sec pathway were identified in the I. hospitalis
genome, including the 7S RNA gene component of the signal
recognition particle (Figure 7). Even though potentially func-
tional protein secretion complexes, including the euryar-
chaeal-specific SecDF are encoded in its genome, N. equitans

lacks an identifiable 7S RNA gene. Since that component is
critical for the assembly of a functional signal recognition
particle, it might be synthesized as two separate transcripts,
such as some of the tRNAs [58], or might be imported from
the host. The Tat system, which transports folded protein
across the membrane, is present in I. hospitalis but absent in
N. equitans. For all proteins that are targeted for transloca-
tion, the signal peptide has to be removed either during or
after translocation. The protease that removes some of the
signal peptides in archaea, signal peptidase I, was identified
in both genomes (Igni153 and Neq432). I. hospitalis
also
encodes a type IV prepillin peptidase (PibD, MEROPS family
A24A, Igni1405), which processes membrane and secreted
proteins with a class III signal peptide, including proteins
involved in motility (flagellin) and pili formation [59]. Nei-
ther I. hospitalis nor N. equitans appear to have flagellins,
although several genes potentially associated with archaeal
flagellar or pili assemblies were identified in both genomes
(flaI, flaJ). While the cells do not appear to be motile, certain
appendages and pili-like structures have been observed in
electron micrographs [60-62] and might play a role in the
interaction between the two organisms.
Central metabolism
I. hospitalis is the first archaeon with sulfur-based autotro-
phy for which a complete genome sequence is hereby
reported. Metabolic reconstruction (Figure 7) points to sim-
ple and efficient strategies that fit a streamlined genome.
Nitrogen assimilation is predicted to rely on readily available
ammonia, the most economical strategy in reduced environ-

ments [63]. Ammonia could be acquired through an AmtB
transporter (Igni1293), which is apparently co-transcribed
with the gene for the nitrogen regulatory protein PII (glnK,
Igni1294). These genes are widely present in bacteria and
most members of the Euryarchaeota but are nearly absent
from Crenarchaeota, and probably have been laterally trans-
ferred to I. hospitalis from a euryarchaeon (Figure 5). GlnK
controls the transport of ammonium ions by interacting with
AmtB and also activates a type of glutamine synthase (GS)
that fixes the ammonia into glutamine. GS is present in all
Genome Biology 2008, Volume 9, Issue 11, Article R158 Podar et al. R158.10
Genome Biology 2008, 9:R158
Predicted functional systems and metabolic pathways of the I. hospitalis-N. equitans systemFigure 7
Predicted functional systems and metabolic pathways of the I. hospitalis-N. equitans system. The numbers refer to the corresponding genes in the I. hospitalis
and N. equitans genome (green and red, respectively). Some of the biochemical pathways (carbon fixation, amino acid biosynthesis and sugar metabolism)
have been experimentally validated [66,69]. Specific subcellular compartments and structures (periplasmic space, vesicles, tubules, pores, fibers) [9,11,62]
are indicated and speculative functions are indicated with question marks. Scissors indicate proteases. Stars indicate specific regulatory proteins. Different
transporter categories and their individual subunits are indicated by shape symbols and the direction of transport of specific substrates across the
membrane is shown by arrows.
amtB
FEV APS
feoB
trkAH
Fe complex
3+
Fe
2+
K
+
Na

+
H
+
Mg
MgtE
Heme
As
MoO
4
2-
Pi
antibiotics
macrolide export
MacB
AA+
MscS
oligo/
dipeptides,
Peptide/Ni
(Opp)
pstABCS
modABC
arsAB
ccmC
ABC-2
multidrug,
ATP
ADP, P
i
H

+
AA-ATPase
ATP synthase
isopentenyl-PP
1174
377,
1401
688
HMG-CoA
476
mevalonate
758
mevalonate-5P
804
dimethylallyl-PP
1168
geranyl-PP (GPP)
338
GGPP
425
GGGP
DGGGP
(archaeol)
GDGT
(caldarchaeol)
glycosylated
derivatives
(mannose,
glucose)
52,

626?
975
106 -108
S
B
C
A
phoU
588
F
K
G
1270
1269
B
C
A
1339-
1341
Zn
2+
Mn
2+
znuABC
A
B
C
913
914
118

A
S
P
550
1203
1204
146, 161
147, 163
47
539, 829,1326
Cl
-
(anions)
eriC
1222
192
236,
886
1002
tungstate
tupAPS
P
A
S
470-472
1293
CPA2
antiporter
394
454

Co
Cbi
APC
porter
solute
+
+
solute
573
893
895
2+
LysE
1398
Fe-S cluster
assembly
suf
1219 1220
?
MarC
128
614
56
235
MFS
302
391
555
1008
800

LPT
479
480
SSS
feoB
Fe
2+
514
Na
+
Ca
2+
CaCA
486
corA
Mg
2+
501
TDT
Dicarboxylate (C4)
14
H
+
Tellurite
MOP
90
oligosaccharides,
lipids
436, 437
MscS

198
531
suf
129
74
LPT
175
S-layer protein
ABC-2
multidrug,
other
159
2+
glnK
1294
acetate
256, 257
pyruvate
1256-1259
1113
2-oxoglutarate
isocitrate
citrate
341
1075-1078
1064
glutamate
Alanine
2-oxoglutarate
Alanine

metabolism
glutamate
metabolism
aspartate
metabolism
1228
glutamine
406
407
408
asparagine
570
Thr, Asn, Met
Pyrimidines
glycerate-1,3 diP
sn-glycerol-1P
274
Asp
carbamoyl-P
CO
2
1399,1400
635
1430
Glutamate
2-OxGlutarate
NH
3
+
77

Val, Leu
H
+
679
B
607
A
682
c
1080
E
D
680
1081
F
a
609
C
1214
ATP
?
ADP , P
i
H
+
AA-ATPase
ATP synthase
H
+
263

B
103
A
217
c
166
E
a
410
PP
i
461
NADH
H
+
+
NAD
+
ADP
+ NTP
ATP + NDP
307
dCTP
316
dUTP
453
V4R
ClpP
533
DMT

?
381
RND
?
962
Na
+
1084
Na
+
citramalate
983
(1249,645,261)
Ile
421?300
268
169,425
SecYE
168,376a
SecDF
1199,1178
ClpP
315
Sec
YE
FtsY
1177
SRP
768
1044

SpI
153
348 374
7S RNA
Ribosome
SpI
432
1009
1010
267
GSPII_F
1011
Snf7
101,1156
701
1332
?
?
mRNA
1266
pores
HCO
3
CO
2
CO
2
NH
4
+

480
1072
1324
CstA
+
?
Tat
A,C
flaJ
flaI
TadC
flaJ
flaI
2Fd
red
2Fd
ox
2Fd
red
2Fd
ox
Fd
red
H
+
1377
1379
1378
+ NO
3

Nitrate reductase?
(Formate DH?)
H
+
NO
2
-

AG
H
1380
D
Branched AA
(Liv)
K
H
M
G
F
139
140
729
730
731
732
I,L,V
A
B
C
D

F
1021
1201
1019,
1336
1337
1338
1292
+
S
0
HS
-
S
n
2-
Polysulfide
reductase (Sre)
530
803
802
801
HS
-
H
+
A
C
528
529

cyt?
cyt?
L
S
FeS
NiFe
hydrogenase
H
+
1367
1366
1368
1369
B
43
44
HypF
CN
HypE
HypD
HypC
S
S
Fe
II
HypD
HypC
S
Fe
III

S
CN
HynE
pre-
HynE
HypC
Fe
III
S
CN
S
Ni
2+
Ni
S
S
?
HynG
HypB
HypA
355
413
572
572
490
674
1367
HynG
1366
504

HycI
825
H
2
S
n
2-
+
e
-
H
+
H
+
Q
Q
276
e
-
H
+
H
+
Q
694,695
Fe-S cluster
assembly
suf
74 129
acetyl-CoA

PEP
696
fructose-1,6 diP
glyceraldehyde-3P
415
glycerate-3P
glucose-6P
fructose-6P
glycerone-P
363
glycerate-2P
1079?
41
1001
1374
1007
Phe,
T
yr
,
Trp
Ser, Gly, Cys
2-deoxy ribose
1079, 41
Pentose-P
PRPP
Purines and
pyrimidines
Histidine
metabolism

1032
Fd
ox
PEP
?
Lon
110
281
349
414
vesicle
fusion
cell recognition
and interaction
CO
2
1266
1266
1266
?
?
V4R
WH
acetate
1085
Na
+
1084
1085
Arg

Lys
Outer membrane
Wide periplasmic space
transport and
fusion
respiratory
complexes
processing?
57,621,728,944,315
ornithine
ornithine
citrulline
1387
transcription
vesicle formation
H
+
respiratory
tubes/vesicles?
H
+
/e
-
/e
-
H
+
/e
-
H

+
/e
-
H
+
/e
-
H
+
/e
-
signalling?
H
+
/e
-
??
?
?
566
4-hydroxybutyrate
1263
678
85,86
276,
445
475
malate
fumarate
succinate

succinyl-CoA
succinic
semialdehyde
oxaloacetate
4-hydroxybutyryl-CoA
crotonyl-CoA
3-hydroxybutyryl-CoA
acetoacetyl-CoA
ATP+CoA
ATP+CoA
595
1058
1058
?
?
CO
2
261
ATP
NADH+H
+
Carbon
fixation
CoA
752
?
NAD(P)H+H
+
CoA
-

?
acetyl-CoA
exogenous
aminoacids
C
4Fe-4S
F
4Fe-4S
1145
542
543
B
1144
A
544-546,1146
H
+
H
2
E
Ni-Fe
Fd
red
Fd
ox
Ech hydrogenase
cell fusion
narrow
periplasmic
space

tranfered
complexes?
Cytoplasmic
membrane
fibrilar structures
I. hospitalis
N. equitans
Genome Biology 2008, Volume 9, Issue 11, Article R158 Podar et al. R158.11
Genome Biology 2008, 9:R158
sequenced archaea with the exception of N. equitans and, by
assimilating ammonia into the amide group of L-glutamine,
makes it available to downstream glutamine-dependent ami-
dotransferases. One such enzyme is glutamate synthase
(GltS, Igni408), which is predicted to catalyze the reductive
transfer of the amide group from glutamine to 2-oxoglutar-
ate, resulting in glutamate and an amino group donor for
transamination reactions. The domain architecture of GltS in
I. hospitalis is unique and contains a GXGXG structural
domain [64] followed by a ferredoxin (4Fe-4S) and a
glutamine synthetase FMN-binding domain. The glutamate
synthase domain I (GlxB domain) is expressed as a separate
polypeptide (Igni407) and it is unclear if these two proteins
actually assemble to form glutamine synthase or if GlxB func-
tions independently as a type II glutamine aminotransferase.
The two genes are likely cotranscribed with an aspartate/aro-
matic aminotransferase (Igni406), suggesting a tight cou-
pling of the transamination processes. The GS/GltS operates
very efficiently at low concentrations of ammonia and substi-
tutes for the alternative ammonia incorporation mechanisms
that use glutamate dehydrogenase (GDH) [63]. GDH cata-

lyzes the reversible conversion of glutamate to 2-oxoglutarate
and represents an alternative route for both deamination and
ammonium incorporation. It has been suggested that GDH
provides the major nitrogen assimilation mechanism in most
hyperthermophiles [65], but this is clearly not the case with
Ignicoccus. This is likely due to the absence of a steady source
of exogenous amino acids; thus, the cells must rely on free
ammonia, present at concentrations too low for GDH to oper-
ate at. However, in N. equitans, which lacks the ammonium
transporter or a GS/GtlS enzyme pair, limited nitrogen
metabolism could rely on a GDH (NEQ077), likely using
glutamate imported from the host. The transfer of glutamate
from I. hospitalis to N. equitans has been detected experi-
mentally [3]. It is not clear why N. equitans has retained a
GDH gene among its very few encoding metabolic enzymes.
One possibility could be that GDH would contribute to the cell
redox potential by oxidative deamination of glutamate.
Pathways for the synthesis of almost all amino acids can be
recognized in the I. hospitalis genome, with the exception of
proline and homocysteine. Some of the enzymatic activities
involved in I. hospitalis amino acid biosyntheses have been
detected experimentally and labeling experiments have been
used to reconstruct most the pathways [66]. The genome also
encodes the predicted enzymes of purine, pyrimidine, NAD,
riboflavin/FAD, pyridoxal and CoA biosynthesis. The meval-
onate pathway for the synthesis of the characteristic archaeal
membrane archaeol- and caldarchaeol-type lipids appears to
be complete (Figure 7), although enzymes involved in some of
the steps have not yet been characterized in archaea [67,68].
I. hospitalis utilizes a novel and so far unique autotrophic

CO
2
fixation pathway, termed the dicarboxylate/4-hydroxy-
butyrate cycle [69]. The individual steps of the pathway have
been investigated experimentally in detail and most have
been confirmed biochemically [66,69,70] (Figure 7). Acetate
in the form of acetyl-CoA is carboxylated by a pyruvate ferre-
doxin oxidoreductase enzyme complex (Igni1256-1259) to
form pyruvate, which is then converted to phosphoenolpyru-
vate by pyruvate:water dikinase (Igni1113). The source of
acetyl-CoA may be linked to two adjacent genes potentially
encoding an acetyl-CoA synthase (Igni256, 257). Normally,
that enzyme is encoded as a single polypeptide. The two genes
in I. hospitalis may encode the enzyme as two subunits
requiring post translational assembly or, alternatively, the
two open reading frames could indicate a pseudogene. The
level of acetate in the environment where I. hospitalis has
been isolated from is not known, but the genome encodes a
putative sodium-acetate symporter (Igni454) and uptake of
acetate has been confirmed experimentally [66]. Since Ignic-
occus can grow in the laboratory in the absence of acetate, the
genome might also encode an acetogenesis mechanism using
CO
2
. One potential route is direct reduction of CO
2
to formate
using hydrogen, catalyzed by a putative membrane formate
dehydrogenase complex. Genes encoding a membrane com-
plex with similarity to both nitrate reductase and formate

dehydrogenase were identified as a likely operon acquired
from a bacterium (Igni 1377-1380). However, since neither
nitrate respiration in I. hospitalis cultures [2] nor the bio-
chemical activity of formate dehydrogenase in cell extracts
[66] were detected, the cellular function of that complex
remains unclear.
The archaeal-type PEP carboxylase [71] catalyzes the second
CO
2
incorporation reaction, which results in the formation of
oxaloacetate, an important precursor for amino acid biosyn-
thetic pathways (Figure 7). Reactions catalyzed by malate
dehydrogenase, fumarase, succinate dehydrogenase and suc-
cinyl CoA-ligase lead to the synthesis of succinyl-CoA. Until
recently, the fate of succinyl-CoA was unclear and, as the
reactions that would close the cycle were not apparent based
on experimental data or genomic information, the mecha-
nism of acetyl-CoA regeneration remained unknown. Huber
et al. [69] recently discovered that I. hospitalis uses a novel
strategy to connect, through succinyl-CoA, the partial reduc-
tive citric acid cycle with the 4-hydroxybutyrate route of
acetyl-CoA regeneration (Figure 7). Based on this finding, the
proposed dicarboxylate/4-hydroxybutyrate cycle appears to
be energetically less costly than other carbon fixation cycles
operating in archaea [69], further supporting the notion that
I. hospitalis combines a streamlined genome with efficient
metabolic strategies.
Phylogenetic analysis of the two I. hospitalis gene clusters
encoding oxoacid:ferredoxin oxidoreductase complexes indi-
cates that one of them (Igni1256-1259) belongs to the pyru-

vate:ferredoxin oxidoreductase family and, therefore, is the
likely catalyst for acetyl-CoA carboxylation. The other com-
plex (Igni1075-1078) has a close affinity to a family with
oxoglutarate specificity with no close homologs in Crenar-
chaeota (Figure S3 in Additional data file 2), suggesting
Genome Biology 2008, Volume 9, Issue 11, Article R158 Podar et al. R158.12
Genome Biology 2008, 9:R158
acquisition by lateral transfer. The functional inference is
based on phylogenetic partitioning of archaeal oxoacid:ferre-
doxin oxidoreductase genes into distinct clades that corre-
spond to enzymes specific for pyruvate, valerate/isovalerate,
or 2-oxoglutarate or that have mixed specificity [72-74]. In
addition, alignments of the I. hospitalis alpha and beta subu-
nit sequences revealed the presence of motifs conserved in
archaeal and bacterial enzymes specific for pyruvate (Igni
1258-1259) or 2-oxoglutarate (Igni 1077-1078) [75-77] (Fig-
ure S4 in Additional data file 2).
The function of the predicted 2-oxoglutarate:ferredoxin oxi-
doreductase (OGOR) complex in I. hospitalis remains
unclear. 2-Oxoglutarate serves as an entry point in glutamate
and lysine biosynthesis and is also linked to the biosynthesis
of several other amino acids as shown by carbon tracing and
inferred from genomic data [66] (Figure 7). In heterotrophic
archaea and bacteria, oxoacid:ferredoxin oxidoreductases are
involved in amino acid and sugar fermentation reactions that
generate reduced ferredoxin and ATP, although OGOR has
been ascribed a biosynthetic function, namely, generation of
succinyl-CoA from 2-oxoglutarate [78,79]. By contrast, in
anaerobic autotrophs, OGOR is a key enzyme in the reverse
citrate cycle, where it catalyzes the fixation of CO

2
on succi-
nyl-CoA with the formation of 2-oxoglutarate [70]. However,
this reaction has not been detected in I. hospitalis cell extracts
and carbon isotope tracing does not support its occurrence in
laboratory cultures [66]. In fact, succinyl-CoA produced by
the first half of the carbon fixation cycle is reduced by succi-
nyl-CoA reductase to succinic semialdehyde and channeled
into the hydroxybutyrate pathway [69]. The same reaction
has been shown to connect the 3-hydroxypropionate with the
4-hydroxybutyrate pathways in another recently discovered
novel carbon fixation cycle in the crenarchaeaote Metal-
losphaera sedula [80]. However, unlike the succinyl-CoA
reductase from M. sedula, which uses NADPH as electron
donor, the enzyme in I. hospitalis requires reduced ferre-
doxin [69]. A speculative role for OGOR could be to provide a
ferredoxin-based electron shuttle between oxoglutarate and
succinyl-CoA, at the expense of a fixed carbon (Figure 7).
Such coupling has been shown to be important in the anaero-
bic metabolism based on aromatic compounds in Thauera
aromatica, OGOR providing benzoyl-CoA reductase with
reduced ferredoxin [75]. In Ignicoccus, under active growing
conditions, such a reaction based on de novo synthesized 2-
oxoglutarate does not seem advantageous as it would increase
the succinyl-CoA pool at the loss of an acetyl-CoA while suffi-
cient reduced ferredoxin may be supplied by a hydrogenase.
However, under limited CO
2
and H
2

conditions, 2-oxoglutar-
ate from the internal pool or derived from exogenous amino
acids and peptides could keep the 4-hydroxybutyrate part of
the cycle active and generate acetyl-CoA for maintenance
functions. Experimental studies will be needed to test this
hypothesis and identify the specificity of the predicted OGOR
complex.
Respiration and energetic metabolism
Under laboratory conditions, the only energy yielding reac-
tion that sustains the metabolism of I. hospitalis is the oxida-
tion of molecular hydrogen coupled to the reduction of
elemental sulfur. While energetically weak (-6.7 kcal/mol)
[38], there are indications that this type of respiration might
have been used by ancient microbes of the early Archaean [5].
Details of bioenergetic reactions and the mechanisms for gen-
erating the membrane chemiosmotic potential in anaerobic
hyperthermophilic archaea are still not well understood. Min-
imal enzymatic components that are required include a mem-
brane hydrogenase complex, a sulfur reductase and an
electron transport chain between them. In I. hospitalis, there
appear to be two clusters of genes encoding subunits of the
sulfur/polysulfide reductase complex. The first such cluster
(Igni801-803) contains the catalytic reductase (SreA), a 4Fe-
4S ferredoxin (SreB) and the membrane anchoring compo-
nent NrfD (SreC) with eight transmembrane domains. NrfD
is thought to participate in the transfer of electrons from the
quinone pool into the terminal components of the Nrf path-
way. Elsewhere in the genome, a gene cluster (Igni528-530)
that appears to be of bacterial origin contains a different
NrfD, a periplasmic FeS ferredoxin, as well as a membrane

protein with four putative heme binding sites that may serve
in the electron transfer chain through the membrane, possi-
bly binding menaquinone. This gene cluster is also present in
the related archaeon Hyperthermus butylicus [81], suggest-
ing the possibility that it was transferred between the two
archaeal lineages after one of them likely acquired it from a
delta proteobacterium. Two types of reductase complexes
might therefore assemble in I. hospitalis, archaeal and bacte-
rial. In other sulfur reducers a periplasmic polysulfide-sulfur
transferase (a member of the rhodanese family) facilitates the
transfer of low concentrations of polysulfide to the reductase.
I. hospitalis is the only crenarchaeote that is missing a rhod-
anese family gene. This could be a result of growing under rel-
atively neutral pH, where polysulfide concentrations may be
high enough. Therefore, access of polysulfide to the cytoplas-
mic membrane, where the reductase complex is likely located,
could occur by diffusion across the large periplasmic space
after passage though the outer membrane pores.
Ignicoccus depends on molecular hydrogen as the sole elec-
tron donor. A single predicted operon contains the genes
encoding the large and small subunits of a hydrogen uptake
NiFe hydrogenase, including the large and small subunits
(Igni1366-1369). The heterodimer is exported to the peri-
plasm through the twin-arginine translocation (TAT) system
and is assembled with a 4Fe-4S ferredoxin and a membrane
protein anchor containing histidine residues that might bind
a b-type heme [82]. The formation of the metal-containing
active site and the assembly of the hydrogenase is a complex
process requiring multiple accessory proteins [83], all of
which appear to be encoded in the I. hospitalis genome (Fig-

ure 7). Hydrogen oxidation is coupled with electron transfer
through FeS centers and a putative membrane cytochrome to
Genome Biology 2008, Volume 9, Issue 11, Article R158 Podar et al. R158.13
Genome Biology 2008, 9:R158
the quinone pool of the respiratory chain, which contributes
to the generation of a membrane potential that drives ATP
synthesis. The quinone appears to be associated with the
membrane component of the hydrogenase and that of
polysulfide reductase, with the exchange of the electrons
likely involving formation of respiratory 'supracomplexes'
[84]. A separate 'energy-converting' Ni-Fe hydrogenase fam-
ily complex (Ech), which is evolutionarily related to the
energy-conserving NADH:quinone oxidoreductase (complex
1), appears to be encoded by genes in two clusters (Igni542-
546 and Igni1144-1146). This hydrogenase is the likely cata-
lyst in maintaining the pool of reduced ferredoxin.
The I. hospitalis genome also contains a four gene putative
operon with close homologues among the bacterial respira-
tory periplasmic nitrate reductases (Igni1377-1380). Similar-
ity to formate dehydrogenases was also detected, so the
function of the complex is not clear, as nitrate cannot serve as
a sole electron acceptor in Ignicoccus [2,60]. In bacteria,
depending on the composition of the complex, periplasmic
nitrate reduction can either contribute to the generation of
the proton gradient or serve as an electron sink, eliminating
excess reducing equivalents from the cytoplasm [85].
A complete membrane A-type ATPase is predicted to be
encoded in the genome of I. hospitalis, in contrast with only a
subset of subunits in N. equitans [12]. While N. equitans
might be unable to synthesize ATP, the presence of a pre-

dicted nucleoside diphosphate kinase (Neq307) suggests that
regeneration of the NDP pool is feasible, which might reduce
its host dependency by recycling (Figure 7). Since it has few
ion transporters and no genes encoding membrane hydroge-
nases or oxidoreductases, it is unknown if N. equitans can
independently maintain a membrane potential or whether it
needs to acquire such capabilities from its host.
As an obligate anaerobe, I. hospitalis requires a mechanism
to deal with the toxicity of reactive oxygen species. A superox-
ide reductase is present (Igni1348) and could detoxify super-
oxide resulting from oxygen reduction by transition metals.
According to a recently proposed mechanism [86], a ferrocy-
anide complex bound within the superoxide reductase active
site may scavenge the superoxide by one-electron redox
chemistry while the superoxide reductase iron site remains
reduced. The resulting peroxide could be transferred to solu-
ble organic compounds, resulting in the formation of alkyl
peroxides that can be reduced by peroxiredoxin. A gene
encoding a member of this family is encoded in the genome
(Igni459) and a recent proteomic analysis of I. hospitalis in
laboratory cultures has shown that its product is an abundant
cytosolic protein [55].
Potential molecular and structural determinants of the
I. hospitalis-N. equitans interaction
Although the recognition and exchange mechanisms between
I. hospitalis and N. equitans remain elusive, the available
genomic and ultra-structural data suggest some possible
ways of interaction between the two organisms. Since the
transporters in both species are few and provide limited spe-
cificities, they are unlikely to comprise the main route of

metabolite acquisition by N. equitans. Similarly, transfer of
protein complexes to N. equitans from the host by secretion,
especially for membrane components, would violate topolog-
ical and signal sequence constraints of the translocation
machinery. Potential vehicles for the transport of metabolites
and proteins from I. hospitalis to N. equitans appear to be the
large and variably shaped vesicles and tubes that emerge from
the host's cytoplasm [9,10]. Such structures could provide
transient or even constant contact between the two cyto-
plasms once the physical contact between the cells has been
established, possibly fulfilling the metabolic and energetic
requirements of N. equitans. This would also allow it to carry
out limited respiration, transport and ATP synthesis and may
explain how detached N. equitans cells or cells not in direct
contact with the host can survive for some time. Electron
microscopy studies have indicated that some of the I. hospi-
talis periplasmic vesicles fuse with the outer membrane,
which likely results in their contents being released into the
environment [9,10]. This release of small molecules and, per-
haps, peptides might provide chemical cues to N. equitans for
host recognition and attachment. Since neither of the two
organisms appears to be motile, the actual mechanism by
which they find each other and become attached in the turbu-
lent hydrothermal vent environment remains enigmatic.
Recent ultra-structural and physiological studies have shown
that a physical connection can form between the two organ-
isms [3,62]. Three-dimensional reconstructions point to a
dynamic type of interaction, some N. equitans cells contact-
ing the outer membrane of I. hospitalis in places where the
host periplasmic space is wide and contains cytoplasmic ves-

icles while others are attached to regions with a very narrow
periplasm and displaying fibrilar structures [62]. The steps
and molecular determinants of the cell-cell recognition and
interaction and the membrane and periplasm dynamics
remain uncharacterized. The cytoplasmic membrane of Igni-
coccus itself is highly 'corrugated', as shown in sections and
three-dimensional reconstructions, thereby increasing its
surface significantly; in addition, it spontaneously evaginates
in the absence of N. equitans [2,9,10,62]. Therefore, the phys-
iological role of the conglomerate of tubes and vesicles and
the significance of the wide periplasmic space probably
extends beyond their possible connection to N. equitans. As
energy generation resides at the level of the cytoplasmic
membrane, these structures could provide a substantially
increased respiratory surface confined in the space sur-
rounded by the outer membrane, analogous to the eukaryotic
mitochondrial cristae. Vesicles might concurrently transport
specific lipids and proteins to the outer porous membrane,
which in this case would serve not only as a protective barrier
but also for controlling gas and solute exchange. This could
represent a mechanism enabling Ignicoccus species to rely
Genome Biology 2008, Volume 9, Issue 11, Article R158 Podar et al. R158.14
Genome Biology 2008, 9:R158
exclusively on the low energetic yield of the sulfur-hydrogen
respiration to sustain an elevated turnover of cellular compo-
nents at high temperature. Combined with the obligate CO
2
autotrophy and efficient metabolism, such adaptations might
allow Ignicoccus to outcompete heterotrophs in colonizing
emerging hydrothermal vent niches that are still poor in dis-

solved organic compounds.
Conclusion
The combinations of ecophysiological and morphological fea-
tures that collectively enable the I. hospitalis-N. equitans
relationship are encoded within a surprisingly simple
genomic blueprint. The genome of I. hospitalis is the smallest
among free-living bacteria and archaea, shows evidence of
gene exchange with N. equitans and encodes streamlined bio-
chemical functions necessary for a chemoautotrophic metab-
olism relying on carbon dioxide, hydrogen and sulfur. Aside
from selection pressure against genome expansion in a
restrictive environmental niche, the two organisms have coe-
volved, leading to symbiotic specificity and gene exchange. In
addition, I. hospitalis appears to have acquired a significant
number of genes and predicted operons from Bacteria and
Euryarchaeota, some of them encoding membrane-associ-
ated complexes involved in transport and energetic metabo-
lism. This unicellular symbiotic system might resemble
relationships that gave rise to eukaryotic organelles. The
availability of complete genomic data for both organisms
opens the possibility to study interspecific gene regulatory
networks and identify proteins that might be exchanged
between interacting cells.
Materials and methods
Genome sequencing and functional annotation
I. hospitalis KIN4I cells (DSMZ strain 18386) were grown as
described in [2]. DNA was isolated from frozen cells using an
alkaline lysis followed by proteinase K digestion method [87].
Sequencing and assembly were performed at the DOE Joint
Genome Institute, Walnut Creek, CA, USA using the standard

microbial genome sequencing pipeline [88] based on a com-
bination of 3-, 6- and 40-kb (fosmid) DNA libraries. The
Phred/Phrap/Consed software package was used to assemble
and assess quality [89]. Possible miss-assemblies were cor-
rected and gaps between contigs were closed by editing in
Consed, custom primer walks or PCR amplification and
sequencing. The estimated error rate in the completed
genome sequence of I. hospitalis is less than 1 in 50 kbp.
Automated gene prediction was performed by using the out-
put of Critica complemented with the output of Glimmer as
part of the genome annotation pipeline at Oak Ridge National
Laboratory (ORNL), Oak Ridge, TN, USA. The predicted cod-
ing sequences were translated and used to search the
National Center for Biotechnology Information (NCBI) non-
redundant database, UniProt, TIGRFam, Pfam, PRIAM,
KEGG, COG, and InterPro databases. The tRNAScanSE tool
[90] was used to find tRNA genes, whereas ribosomal RNAs
were found by using BLASTn against the ribosomal RNA
databases. The RNA components of the protein secretion
complex and the RNaseP were identified by searching the
genome for the corresponding Rfam profiles using INFER-
NAL [91]. Transporter proteins were initially identified based
on similarity to transporter categories in GOG and Pfam and
were further analyzed using the Transporter Classification
Database [92]. Additional gene prediction analysis and man-
ual functional annotation was performed within the Inte-
grated Microbial Genomes (IMG) platform developed by the
Joint Genome Institute, Walnut Creek, CA, USA [93]. The
complete genome sequence has been deposited in GenBank
[GenBank:CP000816.1

].
Comparative genomic analysis
Analysis of the I. hospitalis and N. equitans genomes were
carried out using the IMG system [93]. The genes referred to
throughout the text and figures correspond to the assigned
open reading frame numbers in the two genomes. Putative
operons were identified using the method of Overbeek et al.
[94]. Structure fold prediction of membrane proteins with no
detectable similarity to other database sequences was per-
formed using Phyre [95]. To calculate the frequency of para-
logs in the different archaeal genomes, blastclust analyses
were performed using the translated coding sequences and
varying the similarity threshold for sequence inclusion into
clusters. To derive the amino acid usage statistics for archaeal
genomes, the percentages of amino acids encoded within each
protein were first calculated and used to determine the over-
all percentage for the whole proteome. The frequency for each
amino acid use was then analyzed graphically relative to the
GC content of the genome considering that the GC content
can influence codon usage. Archaeal COG analyses for I. hos-
pitalis and N. equitans were performed as described [27].
Analysis of genome sizes of bacteria and archaea was based
on genomic data available in IMG (March 2008 version). A
table containing the accession numbers for all the genomes as
well as all the numerical parameters and classification used in
the analysis is provided as Additional data file 3.
Phylogenetic analysis
To identify the potential presence of laterally transferred
genes in I. hospitalis, we first used the Pyphy system [44] to
automatically calculate individual phylogenetic trees for

every gene in the genome. Briefly, each protein sequence was
blasted against a local version of a non-redundant protein
database (SWISS and TREMBL) and sequences with signifi-
cant hits (<10e-6) were retrieved and aligned with the query
sequence using CLUSTALW. Phylogenetic trees were then
constructed using PAUP* with the neighbor joining and par-
simony methods with 100 bootstrap replicates. Because the
automatic 'phylogenetic connection' calculated by Pyphy and
displayed as the phylome map of the genome was at times
affected by poor bootstrap support values or unresolved trees,
Genome Biology 2008, Volume 9, Issue 11, Article R158 Podar et al. R158.15
Genome Biology 2008, 9:R158
we visually inspected each tree and, when sufficient confi-
dence was present, a broad phylogenetic connection to the
Crenarchaeaota, Euryarchaeota or Bacteria was assigned to
the I. hospitalis gene. For numerous genes, although the Igni-
coccus gene was clearly of archaeal type, either the phyloge-
netic signal was insufficient or the evolutionary history of that
gene across Archaea involved numerous potential LGTs. Such
genes have been generically designated as 'archaeal'. When
no close homologues for I. hospitalis genes were found or the
phylogenetic trees included archaeal and bacterial genes but
were not sufficiently resolved, such genes were designated as
'unknown' phylogenetic type. Finally, when the closest hit and
the resulting phylogenetic trees indicated a N. equitans gene
as the closest homologue, those genes were designated as
potential LGTs within the N. equitans-I. hospitalis system.
We also used the arCOG analysis to improve the phylotyping
information for some of the functional gene categories that
were not resolved by phylogenetic analysis.

Genes representing potential LGTs within the N. equitans-I.
hospitalis system were subjected to a more extensive phylo-
genetic analysis. Sequence alignments were obtained using a
combination of alternative methods as implemented on the
M-Coffee web server [96]. Following manual alignment cura-
tion and masking of regions with high variability that could
not be confidently aligned, the amino acid substitution model
best fit for each gene was chosen using the software Model-
generator v84 [97]. Maximum likelihood phylogenetic trees
were constructed using PhyML v2.4.4 [98] using the parame-
ters identified by Modelgenerator. Alternative tree topologies
were also explored using a combination of the software Tree
Puzzle and PROML/PHYLIP, as previously described [99].
The protein sequence alignments used to generate the trees
for several inferred laterally transferred genes are provided as
Additional data file 4.
Abbreviations
arCOGs: archaeal cluster of orthologous groups; GDH: gluta-
mate dehydrogenase; GS: glutamine synthase; IMG: Inte-
grated Microbial Genomes; LAP: leucyl aminopeptidase;
LGT: lateral gene transfer; OGOR: 2-oxoglutarate:ferredoxin
oxidoreductase.
Authors' contributions
MP and KOS conceived and coordinated the study. HS, DH,
JRE, ES, AL and PR coordinated and conducted genome
sequencing, assembly and sequence data management. MP,
IA, KSM, JGE, NI, MEH, MW, AL, KM, WC, AA, NK, MS and
EVK performed sequence annotation, comparative genomics
and functional inference analyses. MP, KSM, CD and FF per-
formed phylogenetic and phylogenomic analyses. All authors

analyzed the results and participated in writing sections of the
manuscript. MP assembled and wrote the final version of the
manuscript.
Additional data files
The following additional data are available with the online
version of this paper. Additional data file 1 contains a table
listing the inferred creanarchaeal core genes lost by I. hospi-
talis (Table S1), a table listing functional categories gained
and lost in the I. hospitalis genome (Table S2), a table of func-
tional gene categories (arCOGs) present in N. equitans but
absent in I. hospitalis and their distribution in archaeal
genomes (Table S3) and a table listing the gene family expan-
sions in the I. hospitalis genome (Table S4). Additional data
file 2 contains a phylogenetic tree of cultivated thermophilic
species of Crenarchaeota based on SSU rRNA sequences (Fig-
ure S1), phylogenetic trees of archaeal tyrosyl-tRNA syn-
thetases and of family IV endonucleases (Figure S2), a
phylogenetic tree of the alpha subunit of archaeal 2-oxoacid:
ferredoxin oxidoreductases (Figure S3) and an amino acid-
based sequence alignment of conserved regions of the alpha
and beta subunits of pyruvate:ferredoxin oxidoreductases
and OGORs (Figure S4). Additional data file 3 contains
numerical and classification data associated with all the bac-
terial and archaeal genomes used in genome size analysis.
Additional data file 4 contains the protein sequence align-
ments used to infer lateral gene transfer of valyl t-RNA syn-
thetase, leucyl aminopeptidase, tyrosyl t-RNA synthetase and
endonuclease IV, in phylip format.
Additional data file 1Tables S1-S4Table S1: the inferred creanarchaeal core genes lost by I. hospitalis. Table S2: functional categories gained and lost in the I. hospitalis genome. Table S3: functional gene categories (arCOGs) present in N. equitans but absent in I. hospitalis and their distribution in archaeal genomes. Table S4: the gene family expansions in the I. hospitalis genome.Click here for fileAdditional data file 2Figures S1-S4Figure S1: a phylogenetic tree of cultivated thermophilic species of Crenarchaeota based on SSU rRNA sequences. Figure S2: phyloge-netic trees of archaeal tyrosyl-tRNA synthetases and of family IV endonucleases. Figure S3: a phylogenetic tree of the alpha subunit of archaeal 2-oxoacid: ferredoxin oxidoreductases. Figure S4: amino acid-based sequence alignment of conserved regions of the alpha and beta subunits of pyruvate:ferredoxin oxidoreductases and OGORs.Click here for fileAdditional data file 3Numerical and classification data associated with all the bacterial and archaeal genomes used in genome size analysisNumerical and classification data associated with all the bacterial and archaeal genomes used in genome size analysis.Click here for fileAdditional data file 4Protein sequence alignments used to infer lateral gene transfer of valyl t-RNA synthetase, leucyl aminopeptidase, tyrosyl t-RNA syn-thetase and endonuclease IV, in phylip formatProtein sequence alignments used to infer lateral gene transfer of valyl t-RNA synthetase, leucyl aminopeptidase, tyrosyl t-RNA syn-thetase and endonuclease IV, in phylip format.Click here for file
Acknowledgements

We thank Diversa/Verenium Corporation (San Diego, CA), JGI production
sequencing group and the Computational Biology Group at Oak Ridge
National Laboratory (Oak Ridge, TN) for sequencing and annotation sup-
port. MP, JGE and MK were supported by the US Department of Energy,
Office of Science, Biological and Environmental Research programs at Oak
Ridge National Laboratory (ORNL). ORNL is managed by UT-Battelle,
LLC, for the US Department of Energy under contract DE-AC05-
00OR22725. Support for sequencing and data analysis was provided by the
Joint Genome Institute, the US Department of Energy (IA, NI, AL, KM, HS,
ES, AL, NK and PR). Diversa Corporation provided support for MP, MW,
WC, CD, DH, JRE, AA and FF. KSM and EVK are supported by the Intra-
mural Research Program of the National Institutes of Health, National
Library of Medicine. HH, RR and KOS were supported by grants from the
Deutsche Forschungsgemeinschaft.
References
1. Huber H, Hohn MJ, Rachel R, Fuchs T, Wimmer VC, Stetter KO: A
new phylum of Archaea represented by a nanosized hyper-
thermophilic symbiont. Nature 2002, 417:63-67.
2. Paper W, Jahn U, Hohn MJ, Kronner M, Nather DJ, Burghardt T,
Rachel R, Stetter KO, Huber H: Ignicoccus hospitalis sp. nov., the
host of ' Nanoarchaeum equitans '. Int J Syst Evol Microbiol 2007,
57:803-808.
3. Jahn U, Gallenberger M, Paper W, Junglas B, Eisenreich W, Stetter
KO, Rachel R, Huber H: Nanoarchaeum equitans and Ignicoccus
hospitalis: new insights into a unique, intimate association of
two archaea. J Bacteriol 2008, 190:1743-1750.
4. Martin W, Russell MJ: On the origin of biochemistry at an alka-
line hydrothermal vent. Phil Trans R Soc Lond B Biol Sci 2007,
362:1887-1925.
5. Philippot P, Van Zuilen M, Lepot K, Thomazo C, Farquhar J, Van

Kranendonk MJ: Early archaean microorganisms preferred ele-
mental sulfur, not sulfate. Science 2007, 317:1534-1537.
6. Schwartzman DW, Lineweaver CH: The hyperthermophilic ori-
gin of life revisited. Biochem Soc Trans 2004, 32:168-171.
7. Stetter KO: Hyperthermophiles in the history of life. Philos
Trans R Soc Lond B Biol Sci 2006, 361:1837-1842.
Genome Biology 2008, Volume 9, Issue 11, Article R158 Podar et al. R158.16
Genome Biology 2008, 9:R158
8. Wächtershäuser G: From volcanic origins of chemoau-
totrophic life to Bacteria, Archaea and Eukarya. Philos Trans R
Soc Lond B Biol Sci 2006, 361:1787-1806.
9. Rachel R, Wyschkony I, Riehl S, Huber H: The ultrastructure of
Ignicoccus: evidence for a novel outer membrane and for
intracellular vesicle budding in an archaeon. Archaea 2002,
1:9-18.
10. Näther DJ, Rachel R: The outer membrane of the hyperther-
mophilic archaeon Ignicoccus: dynamics, ultrastructure and
composition. Biochem Soc Trans 2004, 32:199-203.
11. Burghardt T, Nather DJ, Junglas B, Huber H, Rachel R: The domi-
nating outer membrane protein of the hyperthermophilic
Archaeum Ignicoccus hospitalis: a novel pore-forming
complex. Mol Microbiol 2007, 63:166-176.
12. Waters E, Hohn MJ, Ahel I, Graham DE, Adams MD, Barnstead M,
Beeson KY, Bibbs L, Bolanos R, Keller M, Kretz K, Lin X, Mathur E,
Ni J, Podar M, Richardson T, Sutton GG, Simon M, Soll D, Stetter KO,
Short JM, Noordewier M: The genome of Nanoarchaeum equi-
tans: insights into early archaeal evolution and derived
parasitism. Proc Natl Acad Sci USA 2003, 100:12984-12988.
13. Jahn U, Summons R, Sturt H, Grosjean E, Huber H: Composition of
the lipids of Nanoarchaeum equitans and their origin from its

host Ignicoccus sp. strain KIN4/I. Arch Microbiol 2004,
182:404-413.
14. McCliment EA, Voglesonger KM, O'Day PA, Dunn EE, Holloway JR,
Cary SC: Colonization of nascent, deep-sea hydrothermal
vents by a novel Archaeal and Nanoarchaeal assemblage.
Environ Microbiol 2006, 8:114-125.
15. Moran NA: Symbiosis. Curr Biol 2006, 16:
R866-871.
16. Di Giulio M: Nanoarchaeum equitans is a living fossil. J Theor Biol
2006, 242:257-260.
17. Brochier C, Gribaldo S, Zivanovic Y, Confalonieri F, Forterre P:
Nanoarchaea: representatives of a novel archaeal phylum or
a fast-evolving euryarchaeal lineage related to
Thermococcales? Genome Biol 2005, 6:R42.
18. Reysenbach AL, Shock E: Merging genomes with geochemistry
in hydrothermal ecosystems. Science 2002, 296:1077-1082.
19. Embley TM, Martin W: Eukaryotic evolution, changes and
challenges. Nature 2006, 440:623-630.
20. Giovannoni SJ, Tripp HJ, Givan S, Podar M, Vergin KL, Baptista D,
Bibbs L, Eads J, Richardson TH, Noordewier M, Rappe MS, Short JM,
Carrington JC, Mathur EJ: Genome streamlining in a cosmopol-
itan oceanic bacterium. Science 2005, 309:1242-1245.
21. Dufresne A, Garczarek L, Partensky F: Accelerated evolution
associated with genome reduction in a free-living
prokaryote. Genome Biol 2005, 6:R14.
22. Marais GA, Calteau A, Tenaillon O: Mutation rate and genome
reduction in endosymbiotic and free-living bacteria. Genetica
2007, 134:205-210.
23. Whitaker RJ, Grogan DW, Taylor JW: Geographic barriers iso-
late endemic populations of hyperthermophilic archaea. Sci-

ence 2003, 301:976-978.
24. Allen EE, Tyson GW, Whitaker RJ, Detter JC, Richardson PM, Ban-
field JF: Genome dynamics in a natural archaeal population.
Proc Natl Acad Sci USA 2007, 104:1883-1888.
25. Whitaker RJ, Grogan DW, Taylor JW: Recombination shapes the
natural population structure of the hyperthermophilic
archaeon Sulfolobus islandicus. Mol Biol Evol 2005, 22:2354-2361.
26. Valentine DL: Adaptations to energy stress dictate the ecology
and evolution of the Archaea. Nat Rev Microbiol 2007, 5:316-323.
27. Makarova KS, Sorokin AV, Novichkov PS, Wolf YI, Koonin EV: Clus-
ters of orthologous genes for 41 archaeal genomes and
implications for evolutionary genomics of archaea. Biol Direct
2007, 2:33.
28. Mirkin BG, Fenner TI, Galperin MY, Koonin EV: Algorithms for
computing parsimonious evolutionary scenarios for genome
evolution, the last universal common ancestor and domi-
nance of horizontal gene transfer in the evolution of
prokaryotes. BMC Evol Biol 2003, 3:2.
29. Smith TF, Gaitatzes C, Saxena K, Neer EJ: The WD repeat: a com-
mon architecture for diverse functions. Trends Biochem Sci
1999, 24:181-185.
30. Ponting CP, Aravind L, Schultz J, Bork P, Koonin EV: Eukaryotic sig-
nalling domain homologues in archaea and bacteria. Ancient
ancestry and horizontal gene transfer. J Mol Biol 1999,
289:729-745.
31. Anantharaman V, Koonin EV, Aravind L: Regulatory potential,
phyletic distribution and evolution of ancient, intracellular
small-molecule-binding domains. J Mol Biol 2001,
307:1271-1292.
32. Bateman A: The structure of a domain common to archaebac-

teria and the homocystinuria disease protein. Trends Biochem
Sci 1997, 22:12-13.
33. Podar M, Wall MA, Makarova KS, Koonin EV: The prokaryotic V4R
domain is the likely ancestor of a key component of the
eukaryotic vesicle transport system. Biol Direct 2008, 3:2.
34. Saelensminde G, Halskau O Jr, Helland R, Willassen NP, Jonassen I:
Structure-dependent relationships between growth temper-
ature of prokaryotes and the amino acid frequency in their
proteins. Extremophiles 2007, 11:585-596.
35. Zeldovich KB, Berezovsky IN, Shakhnovich EI: Protein and DNA
sequence determinants of thermophilic adaptation. PLoS
Comput Biol 2007, 3:e5.
36. Berezovsky IN, Zeldovich KB, Shakhnovich EI: Positive and nega-
tive design in stability and thermal adaptation of natural
proteins. PLoS Comput Biol
2007, 3:e52.
37. Swire J: Selection on synthesis cost affects interprotein amino
acid usage in all three domains of life. J Mol Evol 2007,
64:558-571.
38. Thauer RK, Jungermann K, Decker K: Energy conservation in
chemotrophic anaerobic bacteria. Bacteriol Rev 1977,
41:100-180.
39. Koonin EV, Makarova KS, Aravind L: Horizontal gene transfer in
prokaryotes: quantification and classification. Annu Rev
Microbiol 2001, 55:709-742.
40. Boucher Y, Douady CJ, Papke RT, Walsh DA, Boudreau ME, Nesbo
CL, Case RJ, Doolittle WF: Lateral gene transfer and the origins
of prokaryotic groups. Annu Rev Genet 2003, 37:283-328.
41. Hoffmeister M, Martin W: Interspecific evolution: microbial
symbiosis, endosymbiosis and gene transfer. Environ Microbiol

2003, 5:641-649.
42. Hotopp JC, Clark ME, Oliveira DC, Foster JM, Fischer P, Torres MC,
Giebel JD, Kumar N, Ishmael N, Wang S, Ingram J, Nene RV, Shepard
J, Tomkins J, Richards S, Spiro DJ, Ghedin E, Slatko BE, Tettelin H,
Werren JH: Widespread lateral gene transfer from intracellu-
lar bacteria to multicellular eukaryotes. Science 2007,
317:1753-1756.
43. Poptsova MS, Gogarten JP: The power of phylogenetic
approaches to detect horizontally transferred genes. BMC
Evol Biol 2007, 7:45.
44. Sicheritz-Ponten T, Andersson SG: A phylogenomic approach to
microbial evolution. Nucleic Acids Res 2001, 29:545-552.
45. Wolf YI, Aravind L, Grishin NV, Koonin EV: Evolution of aminoa-
cyl-tRNA synthetases analysis of unique domain
architectures and phylogenetic trees reveals a complex his-
tory of horizontal gene transfer events. Genome Res 1999,
9:689-710.
46. Woese CR, Olsen GJ, Ibba M, Soll D: Aminoacyl-tRNA syn-
thetases, the genetic code, and the evolutionary process.
Microbiol Mol Biol Rev 2000, 64:202-236.
47. Matsui M, Fowler JH, Walling LL: Leucine aminopeptidases:
diversity in structure and function.
Biol Chem 2006,
387:1535-1544.
48. Elkins JG, Podar M, Graham DE, Makarova KS, Wolf Y, Randau L,
Hedlund BP, Brochier-Armanet C, Kunin V, Anderson I, Lapidus A,
Goltsman E, Barry K, Koonin EV, Hugenholtz P, Kyrpides N, Wanner
G, Richardson P, Keller M, Stetter KO: A korarchaeal genome
reveals insights into the evolution of the Archaea. Proc Natl
Acad Sci USA 2008, 105:8102-8107.

49. Cann IK, Ishino S, Hayashi I, Komori K, Toh H, Morikawa K, Ishino Y:
Functional interactions of a homolog of proliferating cell
nuclear antigen with DNA polymerases in Archaea. J Bacteriol
1999, 181:6591-6599.
50. Lopez P, Philippe H, Myllykallio H, Forterre P: Identification of
putative chromosomal origins of replication in Archaea. Mol
Microbiol 1999, 32:883-886.
51. Robinson NP, Dionne I, Lundgren M, Marsh VL, Bernander R, Bell SD:
Identification of two origins of replication in the single chro-
mosome of the archaeon Sulfolobus solfataricus. Cell 2004,
116:25-38.
52. Iyer LM, Koonin EV, Leipe DD, Aravind L: Origin and evolution of
the archaeo-eukaryotic primase superfamily and related
palm-domain proteins: structural insights and new
members. Nucleic Acids Res 2005, 33:3875-3896.
53. Randau L, Schroder I, Soll D: Life without RNase P. Nature 2008,
453:120-123.
Genome Biology 2008, Volume 9, Issue 11, Article R158 Podar et al. R158.17
Genome Biology 2008, 9:R158
54. Mathai JC, Sprott GD, Zeidel ML: Molecular mechanisms of
water and solute transport across archaebacterial lipid
membranes. J Biol Chem 2001, 276:27266-27271.
55. Burghardt T, Saller M, Gurster S, Muller D, Meyer C, Jahn U, Hoch-
muth E, Deutzmann R, Siedler F, Babinger P, Wirth R, Huber H,
Rachel R: Insight into the proteome of the hyperthermophilic
Crenarchaeon Ignicoccus hospitalis: the major cytosolic and
membrane proteins. Arch Microbiol 2008, 190:379-394.
56. Shigenobu S, Watanabe H, Hattori M, Sakaki Y, Ishikawa H: Genome
sequence of the endocellular bacterial symbiont of aphids
Buchnera sp. APS. Nature 2000, 407:81-86.

57. Albers SV, Szabo Z, Driessen AJ: Protein secretion in the
Archaea: multiple paths towards a unique cell surface. Nat
Rev Microbiol 2006, 4:537-547.
58. Randau L, Munch R, Hohn MJ, Jahn D, Soll D: Nanoarchaeum equi-
tans creates functional tRNAs from separate genes for their
5'- and 3'-halves. Nature 2005, 433:537-541.
59. Szabo Z, Stahl AO, Albers SV, Kissinger JC, Driessen AJ, Pohlschro-
der M: Identification of diverse archaeal proteins with class III
signal peptides cleaved by distinct archaeal prepilin
peptidases. J Bacteriol 2007, 189:772-778.
60. Huber H, Burggraf S, Mayer T, Wyschkony I, Rachel R, Stetter KO:
Ignicoccus gen. nov., a novel genus of hyperthermophilic,
chemolithoautotrophic Archaea, represented by two new
species, Ignicoccus islandicus sp nov and Ignicoccus pacificus sp
nov. and Ignicoccus pacificus sp. nov. Int J Syst Evol Microbiol 2000,
50(Pt 6):2093-2100.
61. Huber H, Hohn MJ, Stetter KO, Rachel R: The phylum
Nanoarchaeota: present knowledge and future perspectives
of a unique form of life. Res Microbiol 2003, 154:165-171.
62. Junglas B, Briegel A, Burghardt T, Walther P, Wirth R, Huber H,
Rachel R: Ignicoccus hospitalis and Nanoarchaeum equitans:
ultrastructure, cell-cell interaction, and 3D reconstruction
from serial sections of freeze-substituted cells and by elec-
tron cryotomography. Arch Microbiol 2008.
63. Cabello P, Roldan MD, Moreno-Vivian C: Nitrate reduction and
the nitrogen cycle in archaea. Microbiology 2004, 150:3527-3546.
64. Binda C, Bossi RT, Wakatsuki S, Arzt S, Coda A, Curti B, Vanoni MA,
Mattevi A: Cross-talk and ammonia channeling between
active centers in the unexpected domain arrangement of
glutamate synthase. Structure 2000, 8:1299-1308.

65. Diruggiero J, Robb FT: Enzymes of central nitrogen metabolism
from hyperthermophiles: characterization, thermostability,
and genetics. Adv Protein Chem 1996, 48:311-339.
66. Jahn U, Huber H, Eisenreich W, Hugler M, Fuchs G: Insights into
the autotrophic CO2 fixation pathway of the archaeon Ignic-
occus hospitalis: comprehensive analysis of the central car-
bon metabolism. J Bacteriol 2007, 189:4108-4119.
67. Boucher Y, Kamekura M, Doolittle W: Origins and Evolution of
isoprenoid lipid biosynthesis in archaea. Mol Microbiol 2004,
52:515-527.
68. Koga Y, Morii H: Biosynthesis of ether-type polar lipids in
Archaea and evolutionary considerations. Microbiol Mol Biol Rev
2007, 71:97-120.
69. Huber H, Gallenberger M, Jahn U, Eylert E, Berg IA, Kockelkorn D,
Eisenreich W, Fuchs G: A dicarboxylate/4-hydroxybutyrate
autotrophic carbon assimilation cycle in the hyperther-
mophilic Archaeum Ignicoccus hospitalis. Proc Natl Acad Sci USA
2008, 105:7851-7856.
70. Hugler M, Huber H, Stetter KO, Fuchs G: Autotrophic CO2 fixa-
tion pathways in archaea (Crenarchaeota). Arch Microbiol 2003,
179:
160-173.
71. Ettema TJ, Makarova KS, Jellema GL, Gierman HJ, Koonin EV, Huynen
MA, de Vos WM, Oost J van der: Identification and functional
verification of archaeal-type phosphoenolpyruvate carboxy-
lase, a missing link in archaeal central carbohydrate
metabolism. J Bacteriol 2004, 186:7754-7762.
72. Fukuda E, Wakagi T: Substrate recognition by 2-oxoacid:ferre-
doxin oxidoreductase from Sulfolobus sp. strain 7. Biochim Bio-
phys Acta 2002, 1597:74-80.

73. Kletzin A, Adams MW: Molecular and phylogenetic characteri-
zation of pyruvate and 2-ketoisovalerate ferredoxin oxidore-
ductases from Pyrococcus furiosus and pyruvate ferredoxin
oxidoreductase from Thermotoga maritima. J Bacteriol 1996,
178:248-257.
74. Nishizawa Y, Yabuki T, Fukuda E, Wakagi T: Gene expression and
characterization of two 2-oxoacid:ferredoxin oxidoreduct-
ases from Aeropyrum pernix K1. FEBS Lett 2005, 579:2319-2322.
75. Dorner E, Boll M: Properties of 2-oxoglutarate:ferredoxin oxi-
doreductase from Thauera aromatica and its role in enzy-
matic reduction of the aromatic ring. J Bacteriol 2002,
184:3975-3983.
76. Yoon KS, Ishii M, Kodama T, Igarashi Y: Purification and charac-
terization of pyruvate:ferredoxin oxidoreductase from
Hydrogenobacter thermophilus TK-6. Arch Microbiol 1997,
167:275-279.
77. Yun NR, Arai H, Ishii M, Igarashi Y: The genes for anabolic 2-
oxoglutarate: ferredoxin oxidoreductase from Hydrogeno-
bacter thermophilus TK-6. Biochem Biophys Res Commun 2001,
282:589-594.
78. Kelly RM, Adams MW: Metabolism in hyperthermophilic
microorganisms. Antonie Van Leeuwenhoek 1994, 66:247-270.
79. Mai X, Adams MW: Characterization of a fourth type of 2-keto
acid-oxidizing enzyme from a hyperthermophilic archaeon:
2-ketoglutarate ferredoxin oxidoreductase from Thermococ-
cus litoralis. J Bacteriol 1996, 178:5890-5896.
80. Berg IA, Kockelkorn D, Buckel W, Fuchs G: A 3-hydroxypropion-
ate/4-hydroxybutyrate autotrophic carbon dioxide assimila-
tion pathway in Archaea. Science 2007, 318:1782-1786.
81. Brugger K, Chen L, Stark M, Zibat A, Redder P, Ruepp A, Awayez M,

She Q, Garrett RA, Klenk HP: The genome of Hyperthermus
butylicus: a sulfur-reducing, peptide fermenting, neutrophilic
Crenarchaeote growing up to 108 degrees C. Archaea 2007,
2:127-135.
82. Laska S, Lottspeich F, Kletzin A: Membrane-bound hydrogenase
and sulfur reductase of the hyperthermophilic and
acidophilic archaeon Acidianus ambivalens. Microbiology 2003,
149:2357-2371.
83. Bock A, King PW, Blokesch M, Posewitz MC: Maturation of
hydrogenases. Adv Microb Physiol 2006, 51:1-71.
84. Schagger H: Respiratory chain supercomplexes of mitochon-
dria and bacteria. Biochim Biophys Acta 2002, 1555:154-159.
85. Gonzalez PJ, Correia C, Moura I, Brondino CD, Moura JJ: Bacterial
nitrate reductases: Molecular and biological aspects of
nitrate reduction. J Inorg Biochem 2006, 100:1015-1023.
86. Molina-Heredia FP, Houee-Levin C, Berthomieu C, Touati D, Tremey
E, Favaudon V, Adam V, Niviere V: Detoxification of superoxide
without production of H2O2: antioxidant activity of super-
oxide reductase complexed with ferrocyanide. Proc Natl Acad
Sci USA 2006, 103:14750-14755.
87. Ramakrishnan V, Adams MW: Preparation of genomic DNA
from sulfur-dependent hyperthermophilic Archaea. In
Archaea, A Laboratory Manual: Thermophiles Edited by: Robb FT, Place
AR. Plainview, NY: Cold Spring Harbor Laboratory Press; 1995:95-96.
88. DOE Joint Genome Institute Sequencing [http://
www.jgi.doe.gov/sequencing/index.html]
89. Phrap and Phred []
90. Schattner P, Brooks AN, Lowe TM: The tRNAscan-SE, snoscan
and snoGPS web servers for the detection of tRNAs and
snoRNAs. Nucleic Acids Res 2005, 33:W686-689.

91. INFERNAL [ />92. Transporter Classification Database []
93. Markowitz VM, Szeto E, Palaniappan K, Grechkin Y, Chu K, Chen IM,
Dubchak I, Anderson I, Lykidis A, Mavromatis K, Ivanova NN, Kyrpi-
des NC: The integrated microbial genomes (IMG) system in
2007: data content and analysis tool extensions. Nucleic Acids
Res 2008, 36:D528-533.
94. Overbeek R, Fonstein M, D'Souza M, Pusch GD, Maltsev N: The use
of gene clusters to infer functional coupling. Proc Natl Acad Sci
USA 1999, 96:2896-2901.
95. Bennett-Lovsey RM, Herbert AD, Sternberg MJ, Kelley LA: Explor-
ing the extremes of sequence/structure space with ensemble
fold recognition in the program Phyre. Proteins 2008,
70:611-625.
96. Moretti S, Armougom F, Wallace IM, Higgins DG, Jongeneel CV,
Notredame C: The M-Coffee web server: a meta-method for
computing multiple sequence alignments by combining
alternative alignment methods. Nucleic Acids Res 2007,
35:W645-648.
97. Keane TM, Creevey CJ, Pentony MM, Naughton TJ, McLnerney JO:
Assessment of methods for amino acid matrix selection and
their use on empirical data shows that ad hoc assumptions
for choice of matrix are not justified. BMC Evol Biol 2006, 6:29.
98. Guindon S, Gascuel O: A simple, fast, and accurate algorithm
to estimate large phylogenies by maximum likelihood. Syst
Biol 2003, 52:696-704.
Genome Biology 2008, Volume 9, Issue 11, Article R158 Podar et al. R158.18
Genome Biology 2008, 9:R158
99. Randau L, Calvin K, Hall M, Yuan J, Podar M, Li H, Soll D: The heter-
omeric Nanoarchaeum equitans splicing endonuclease
cleaves noncanonical bulge-helix-bulge motifs of joined

tRNA halves. Proc Natl Acad Sci USA 2005, 102:17934-17939.
100. National Center for Biotechnology Information, Microbial
Genome Projects [ />lproks.cgi]