Full text of DOI:10.1089/153110702762027835

ASTROBIOLOGY
Volume 2, Number 3, 2002
© Mary Ann Liebert, Inc.
Archaeal Genome Organization and Stress Responses:
Implications for the Origin and Evolution
of Cellular Life
DAVID MUSGRAVE, XIAOYING ZHANG,p and MARCEL DINGER
ABSTRACT
For DNA to be used as an informational molecule it must exist in the cell on the edge of sta-
bility because all genomic processes require local controlled melting. This presents mecha-
nistic opportunities and problems for genomic DNA from hyperthermophilic organisms,
whose unpackaged DNA could melt at optimal temperatures for growth. Hyperthermophiles
are suggested to employ the novel positively supercoiling topoisomerase enzyme reverse gy-
rase (RG) to form positively supercoiled DNA that is intrinsically resistant to thermal denat-
uration. RG is presently the only archaeal gene that is uniquely found in hyperthermophiles
and therefore is central to hypotheses suggesting a hypothermophilic origin of life. However,
the suggestion that RG has evolved by the fusion of two pre-existing enzymes has led to hy-
potheses for a lower temperature for the origin of life. In addition to the action of topoiso-
merases, DNA packaging and the intracellular ionic environment can also manipulate DNA
topology significantly. In the Euryarchaeota, nucleosomes containing minimal histones can
adopt two alternate DNA topologies in a salt-dependent manner. From this we hypothesize
that since internal salt concentrations are increased following an increase in temperature, the
genomic effects of temperature fluctuations could also be accommodated by changes in nu-
cleosome organization. In addition, stress-induced changes in the nucleoid proteins could also
play a role in maintaining the genome in the optimal topological state in changing environ-
ments. The function of these systems could therefore be central to temperature adaptation
and thus be implicated in origin of life scenarios involving hyperthermophiles. Key Words:
Astrobiology
Archaeal histones—Genome packaging—Genome topology—Stationary phase.
2, 241–253.
1990). Archaea are morphologically and meta-
bolically similar to Bacteria, but their genes in-
volved in the information processing systems,
including transcription, translation, DNA repli-
cation, and recombination, most closely resemble
INTRODUCTION
NA SEQUENCE COMPARISONS of genes encoding
the small subunit rRNA component of ribo-
somes (16S rRNA in prokaryotes and 18S rRNA
in eukaryotes) have revealed that cellular life can
be placed into one of three domains: Bacteria, Ar-
D
eucaryal orthologues (Rivera
1998). How-
et al.,
ever, the eucaryal-like nature of the informational
genes in Archaea could also be the result of dif-
chaea, or Eucarya (Woese, 1987; Woese
et al.,
Department of Biological Sciences, University of Waikato, Hamilton, New Zealand.
pPresent address: Nucleics, University of New South Wales, Sydney, Australia.
241

242
MUSGRAVE ET AL.
ferential rates of evolution for metabolic and in-
formational processing genes (Forterre and
Phillipe, 1999). The phylogenetic trees based on
16S rRNA sequences also highlight two pheno-
typically diverse groups in the archaeal domain:
the Euryarchaeota and the Crenarchaeota (Woese
saicism found in many genomes may be due to
differential gene loss or nonorthologous replace-
ment.
The potential nature of the last universal com-
mon ancestor can also be predicted by under-
standing the physical conditions that could have
provided environments for the development and
evolution of cellular life. The discovery of deep-
1990). One of the major differences between
et al.,
them is that most euryarchaeal species, like all of
the Eucarya, have a histone-containing, nucleo-
some-based DNA condensation system (Ronimus
and Musgrave, 1996a; Sandman and Reeve, 2001).
This observation, together with the existence of a
minimal eucaryal information processing system,
predicts that the archaeal and eucaryal domains
shared an important evolutionary history subse-
quent to the universal ancestor stage (Olsen and
Woese, 1997).
sea hydrothermal vents (Corliss
1979) has
et al.,
provided such an environment, and the physiol-
ogy of the Archaea and Bacteria that inhabit hy-
drothermal vents continues to challenge our
preconceptions about the physical limits of con-
ventional life and Darwin’s suggestion, found in
a private letter to Hooker, of a “warm little pond”
for the site of the origin of life itself. Investiga-
tions of hyperthermophilic microorganisms have
formed the basis of the “hot” origin of life for a
number of reasons other than because phyloge-
netic studies place them close to the root of the
universal tree of life. Firstly, they exist in condi-
tions expected to have been common during for-
mation of the earth. Secondly, they could have
survived meteorite impacts at the earth’s surface.
And finally, many Archaea have novel resistance
mechanisms, such as extreme radiation resis-
tance, that could have been used to protect them
For many biologists, an interesting outcome of
phylogenetic trees is that they predict that the last
universal common ancestor was a hyperther-
mophilic archaeon because these organisms are
the closest extant organisms to the root of these
trees. However, a vast body of work has cast
doubt on the validity of these trees, particularly
the position of the root (Forterre and Phillipe,
1999). Doolittle (1999) goes further and argues for
a net-like rather than a tree-like structure to de-
scribe prokaryotic evolution, describing prokary-
otic taxa as being “imprecisely bounded and
ephemeral.” In 1998, Woese proposed a genetic
annealing model for early cellular life in which
life originated as a genetically plastic progenote.
He described this by saying “the universal an-
cestor is not a discrete entity. It is, rather a diverse
community of cells that survives and evolves as
a biological unit.” This harks back to the sugges-
tion of Reanney (1974), several decades ago, that
we should more realistically view prokaryotes as
one global superorganism within which genes are
exchanged at different frequencies. The extent
and importance of horizontal transfer in early
evolution are, in part, demonstrated by the mo-
saic nature of the genomes of extant prokaryotes.
For example, the fully sequenced genome of the
in these extreme environments (Peak
1995;
et al.,
Gerard
2001).
et al.,
Comparative genomic studies of prokaryotes
have shown that only one gene, encoding a type
I topoisomerase enzyme, reverse gyrase (RG), is
uniquely found in hyperthermophiles. RG is a
topoisomerase that has the unique potential to
use ATP to produce positively supercoiled DNA
(Kikuchi and Asai, 1984), and it has been corre-
lated with the presence of positively supercoiled
plasmid and viral DNA in hyperthermophilic
organisms. Positively supercoiled DNA is sug-
gested to have greater resistance to thermal
denaturation than the negatively supercoiled
DNA found in all other cells (Lopez-Garcia and
Forterre, 2000). However, both the formation of
archaeal nucleosomes and the extreme intracel-
lular salt concentrations also have a significant af-
fect on the structural stability of DNA (Anderson
and Bauer, 1978; Ronimus and Musgrave, 1996b).
Here we propose that the ability of archaeal nu-
cleosomes to form both positive and negative su-
percoils in a salt-dependent manner (Musgrave
deeply rooted thermophilic bacterium
Thermotoga
has shown that 24% of its genes have
maritima
greater sequence similarity with archaeal genes
than any other bacterial or eucaryal gene and that
these genes are in defined clusters with conserved
gene order and have an atypical mol percent
G C compared with the rest of the genes in the
1
2000) and a novel stress response in hyper-
et al.,
thermophilic Archaea (Dinger
genome (Nelson
1999). However, Forterre
et al.,
and Phillipe (1999) have argued that the mo-
2000) give us
et al.,
additional clues to how these organisms can

243
ARCHAEAL GENOMIC STRESS RESPONSES
adapt their genome topology to changing envi- irradiating the membrane for 45 s with an UV
ronments and act as an alternate system to RG, transilluminator. Membrane hybridization was
allowing hypotheses for hyperthermophilic ori- performed essentially as described by Dinger and
gins of life to be revisited.
Musgrave (2000).
Electroblotting from polyacrylamide gels onto
PVDF membranes for N-terminal sequencing
MATERIALS AND METHODS
Electroblotting of proteins from polyacry-
lamide gels onto a retentive membrane was used
for western blotting and N-terminal protein se-
Reference strains
Pyrococcus abyssi
GE5 was obtained from Patrick
quencing (Ausubel
1998). For N-terminal se-
et al.,
Forterre, University of Paris XI, Paris, France, and
was obtained from the Ther-
quencing Trans-Blot membranes (Bio-Rad) were
used to transfer proteins from sodium dodecyl
sulfate-containing polyacrylamide electrophore-
sis gels.
Thermococcus zilligii
mophile Research Unit, University of Waikato,
Hamilton, New Zealand.
The transfer procedure was as described by
T. zilligii
P. abyssi
and
Culturing of
Ausubel
(1998). After blotting, the gel was
et al.
was grown in 850-mL volumes, in 1.0-
T. zilligii
L Schott bottles as previously described (Klages
and Morgan, 1994), and cells were culti-
stained with silver or Coomassie Brilliant Blue
to check transfer efficiency and photographed.
The PVDF membrane was stained with 10%
(wt/vol) amido black (naphthol blue black 10B,
Sigma) in 10% (vol/vol) acetic acid, destained
with 10% acetic acid/30% ethanol, and then
drip-dried. The desired band on the membrane
was excised and sent for protein N-terminal se-
quencing. Sequencing was carried out using Ed-
man degradation chemistry on an Applied
Biosystem Procise 492 protein sequencer at the
Sequencing Facility, Auckland University,
Auckland, New Zealand.
P. abyssi
vated in “YPS” medium (Erauso
1993). The
et al.,
OD600 of the samples was measured and con-
verted to cell density by multiplying the OD600
by a conversion factor empirically determined by
cell counting as described (Koch, 1994).
Isolation of RNA
Total RNA was isolated using TRIZOL LS
Reagent (GibcoBRL, Life Technologies, Ltd.), ac-
cording to the manufacturer’s instructions. RNA
was recovered after phenol (pH 4.0)/chloroform
and chloroform extraction, then precipitated and
resuspended in RNase-free water, and stored at
Radioactive labeling of DNA probes and ladders
DNA probes were prepared from polymerase
chain reaction (PCR) products by random primer
labeling using the Rediprime™ II DNA labeling
system (Amersham Pharmacia Biotech), follow-
ing the manufacturer’s instructions. The labeled
70°C. The concentration and quality of RNA
were estimated by absorbance ( ,
2
,
/
A260 A260 A280
and
/
) measurement.
A260 A230
DNA was denatured by adding 7.32 L of 10
NaOH 10 min prior to use. Single-stranded
oligonucleotides were end-labeled using T4 poly-
m
M
Northern blot procedures
For northern blots, total RNA samples were de-
natured by incubation for 15 min at 65°C in an
equal volume of a solution containing 20% for-
malin, 50% formamide, and 0.1
morpholino)propanesulfonic acid (MOPS) and
then separated by electrophoresis through 2%
nucleotide kinase as described by Sambrook
et al.
(1989), stored at 4°C, and used within 3 months
of preparation.
sodium 3-( -
M
N
Preparation of archaeal histone proteins
(wt/vol) agarose gels containing 2.2 formalde-
M
hyde, 20 m MOPS, and 8 m sodium acetate
M
M
Native HTz was purified as previously de-
scribed (Ronimus and Musgrave, 1996b). Recom-
at 4.2 V cm21. After washing the gels for 30 min
in 10 saline–sodium citrate (SSC), the RNA was
3
Es-
transferred to Hybond1 nylon membranes by
capillary blotting for 12 h in 10 SSC. Nucleic
3
acids were UV-cross-linked to the membrane by
as described (Fahrner
2001).
et al.,

244
MUSGRAVE ET AL.
bility when incorporated into chromatin, and the
ability of nucleosomes to wrap DNA in two op-
posite configurations forming positive and nega-
Preparation of DNA substrates
pUC18 plasmid was isolated by alkaline lysis
(Sambrook
1989) and CsCl density gradient
et al.,
tive toroidal supercoils (Hamiche
1996;
et al.,
Hamiche and Richard-Foy, 1998; Santisteban
centrifugation followed by treatment with pro-
teinase K (Sigma), phenol/chloroform extraction,
and ethanol precipitation. Relaxed pUC18 plas-
mids were produced by treating negatively su-
percoiled plasmid DNA with Topo V (Fidelity
Systems) at 70°C or 85°C or with wheat germ
Topo I (BRL) at 37°C followed by phenol/chlo-
roform extraction and ethanol precipitation.
pUC18 topoisomers of different superhelical den-
sities were produced by relaxation in the pres-
ence of ethidium bromide as previously de-
et
2000). We have suggested that archaeal nu-
al.,
cleosomes are also structurally plastic because of
their ability to form both positive and negative
supercoils in a salt-dependent manner and, in
T.
possibly by direct modifications to the nu-
zilligii,
cleoid (Dinger and Musgrave, 2000; Musgrave
et
2000).
al.,
To examine the effect that archaeal chromatin
has on global gene expression control we have
tested stationary-phase changes in the archaeal
nucleoid and have examined the level of tran-
scripts of the histone genes and genes encoding
scribed (Pugh
1989) except that Topo V was
et al.,
used. Positively supercoiled pUC18 control topoi-
somers were produced by incubation of nega-
tively supercoiled pUC18 DNA with RG (Kikuchi
and Asai, 1984).
nucleoid proteins at various growth phases in
P.
another member of the Thermococcales. To
abyssi,
test whether both the
histone genes (
P. abyssi
) and the stress response genes (
hpa1
and
and
hpa2
srbPa2
srbPa1
DNA topology assays
) were being transcribed in a growth
phase-dependent manner, as was found for the
homologous genes in (Dinger, 1998), to-
Topology assays and two-dimensional gels to
determine the handedness of topoisomers were
T. zilligii
tal RNA samples were isolated from cells at dif-
ferent growth phases as depicted in Fig. 1. The
RNA samples shown in Fig. 2 were subject to
northern blotting and hybridization analysis. In-
dividual duplicate N1 membranes containing to-
tal RNA samples were hybridized with one of the
probes listed in Table 1. Figure 2 shows that RNA
degradation was minimal, as indicated by the
equivalent intensities and lack of smearing of the
23S and 16S rRNA bands for each RNA sample.
performed as described (Musgrave
2000).
et al.,
Gel visualization and photography
Nucleic acids, stained with ethidium bromide,
were visualized under UV light, with a TFx-35 UV
transilluminator (Life Technologies). The gels were
photographed using a COHU high-performance
CCD camera and printed on a Sony UP-D890 Dig-
ital Graphic Printer. A digital imaging program,
Scion Image (release beta 2 12/5/97, Scion Corp.)
was used for manipulation of the images.
The northern hybridization results for
hpa1,
and
are shown in Figs. 3–5, re-
hpa2,
spectively. A single transcript of 206 bp for
and
srbPa1/srbPa2
hpa1
and two transcripts of the predicted
hpa2
sizes of
(585 bp) and
(417 bp) were
srbPa1
srbPa2
RESULTS
detected. Southern blots were hybridized at the
same time in the same hybridization tube, with
the same probe used for northern hybridization
as a control for the integrity of the probes used
Recently it has become clear that in eukaryotes
chromatin and the packaging of DNA in nucleo-
somes are intimately involved in the control of
gene expression (Wu and Grunstein, 2000). Also,
the discovery of histone acetyltransferases, his-
tone deacetylases, and other chromatin remodel-
ing complexes, which can specifically modify
the structure of chromatin, has suggested that
eukaryotic chromatin is structurally dynamic
(Jenuwein and Allis, 2001). The structural plas-
ticity of chromatin is also evidenced by the pro-
duction of nucleosomes containing H2A.Z, a
novel histone that promotes nucleosome insta-
and to confirm the
genome sequence an-
P. abyssi
notation (Figs. 3B and 4B). The northern hy-
bridization autoradiographs were quantified, and
the relative level of each transcript was deter-
mined in relation to the level of 23S rRNA in each
sample from Fig. 2 (Figs. 3C, 4C, and 5B). The in-
tensity of
and that of
did not show any
hpa1
hpa2
significant change at any of the points in the
growth phase that were tested.
In contrast, the level of both
and
srbPa1
srbPa2

245
ARCHAEAL GENOMIC STRESS RESPONSES
for HMfA B and HTz). Furthermore, positive
1
supercoils are produced at protein/DNA mass
ratios above this (lanes f–j for HMfB and lanes g–j
and g–i for HMfA B and HTz, respectively). We
1
describe the ability of the archaeal nucleosome to
adopt two conformations as “nucleosome flip-
ping” in order to suggest that one conformation
can be converted to the other, as is seen in the
production of both conformations in the same
tube. This is most evident in lane g of Fig. 6, in
which topoisomers of pUC18 are produced by
FIG. 1. Growth curve of a representative batch culture
The
of P. abyssi cells grown in “YPS” medium at 80°C.
the binding of the HTz histone isolated from
T.
In contrast, only negative supercoils
cell density of the culture versus culture age is indicated.
zilligii.
were produced using the histone MkaH from
MkaH is an unusual ar-
transcripts increased significantly in stationary-
Methanopyrus kandleri.
chaeal histone in that it is two histone monomers
fused in a single polypeptide (Fahrner 2001).
phase cells, indicative of induction of
and
srbPa1
gene transcription in these cells. Both
srbPa2
srbPa1
et al.,
and
transcripts were able to be de-
srbPa2
We have proposed that this fusion prevents the
rearrangement of the monomers that are neces-
sary for the structural changes that allow two nu-
tected using the
gene probe because these
srbPa1
two genes are 40% identical, but because a sin-
,
gle probe was used the relative levels of
srbPa1 cleosome conformations to occur (Musgrave
et al.,
and
could not be determined.
srbPa2
Three other genes of
2000). Models aimed at explaining the signifi-
cance of nucleosomal flipping are presented in
Figs. 7 and 8. In 1987, Liu and Wang proposed
that the tracking of a macromolecule along a
DNA molecule would result in the production of
positive supercoils in front of the advancing com-
plex and that the equivalent number of negative
(
and
P. abyssi srcPa, srdPa,
described as encoding stationary-phase re-
srePa,
sponse proteins in
stitutively (results not shown). Constitutive ex-
pression of these genes was surprising because
) were expressed con-
T. zilligii
and
are both members of a
T. zilligii
P. abyssi
same archaeal family, the Thermococcales, and
each contain the same suite of highly conserved
stationary-phase stress genes (Dinger and Mus-
grave, 2000). The significantly different expres-
sion of the histone and stationary-phase stress
genes in these species suggests that the response
to stationary-phase stress in these organisms is
significantly different. The results also suggest
that the depletion of nucleosomes is not a sig-
nificant factor in the control of global gene ex-
pression in
as has been suggested for
P. abyssi,
T. zilligii.
The dynamic behavior of archaeal nucleosomes
has been demonstrated by the ability of the ar-
chaeal histones to wrap DNA in both positive and
negative supercoils in a salt-dependent manner.
This is also seen in Fig. 6, which shows the
toroidal wrapping of a number of different his-
tones at a salt concentration in which both posi-
tive and negative topoisomers are produced. The
results for the HMf histone from
and the HTz histone from
fervidus
that, at a salt concentration of 150 m
Methanothermus
FIG. 2. Total RNA samples extracted from P. abyssi
cells at various stages of growth, resolved by TBE
show
T. zilligii
negative
M,
Lane 1 contains a 1-kb DNA
agarose gel electrophoresis.
supercoils are produced at protein/DNA mass
ratios below 0.8 (lanes a–e for HMfB, lanes a–f from cells at the times indicated.
ladder. Lanes 2–5 each contain 2 g of total RNA, isolated
m
,

246
MUSGRAVE ET AL.
TABLE 1. PRIMERS USED FOR PCR AMPLIFICATION OF
GENE PROBES USED IN NORTHERN ANALYSES
P. ABYSSI
HPa1-F
HPa1-R
HPa2-F
HPa2-R
SrbPa1-F
SrbPa1-R
ATGGGAGAGTTG CCAATT
TCAGCTCTTAATA GCCAA
ATGGCTGAGTTGC CAATT
GTCAGCTCCTAAT TGCAA
GTGATG
A GGATCGCGGTTCC AACTAA
CATATG
AAAAAG
CCACCACCA GCCGTAGAGCCACC
GCTCTTC CGCA
The oligonucleotide primers listed were used to PCR-amplify the histone and stress
response protein B genes from genomic DNA. PCR products were purified,
P. abyssi
labeled with -33P, and used as a probes for the
and
P. abyssi
transcripts in
g
P. abyssi hpa
srb
cells of various ages.
northern hybridizations using total RNA extracted from
The start and stop codons of the respective open reading frames are indicated in boxes,
and restriction recognition sites are underlined.
supercoils would be produced behind. This could coiled DNA templates
Therefore, tran-
in vitro.
be envisaged to occur when DNA is replicated or scriptionally mediated supercoiling is thought to
transcribed by the tracking of DNA or RNA poly- be another way in which the repressive effects of
merase. The production of positive supercoils nucleosomes on transcription could be overcome
ahead of a DNA tracking complex is expected to by clearing nucleosomes from in front of the ad-
modulate gene expression by decreasing the sta- vancing RNA polymerase. This could also con-
bility of nucleosomes because nucleosomes have tribute to polar effects, when the transcription of
been shown to be unstable on positively super- one gene is seen to affect the expression of a gene
Northern hybridization of total
cells after various periods
FIG. 3. A:
RNA from
P. abyssi
of growth, with the
gene probe. Lane 1
hpa1
contains 1.5 g of -33P-5 end-labeled DNA
m
g
9
marker, with sizes indicated. Lanes 2–6 each
contain 10 g of total RNA from cells of the
m
ages indicated. Lane 7 contains 1 ng of
PCR product, and lane 8 contains 1 ng of
hpa1
hpa2
PCR product, as a positive control.
South-
Lane1
B:
hpa1.
ge-
P. abyssi
ern hybridization controls for
contains 3 g of
I-digested
m
Sac
nomic DNA, and lane 2 contains 0.5 ng of the
PCR product.
Levels of
mRNA
hpa1
hpa1
C:
relative to the corresponding 23S rRNA lev-
els versus culture age. Levels were deter-
mined by densitometric analysis of the north-
ern blots shown in (A) and were compared
with the level of 23S rRNA stained with meth-
ylene blue.

247
ARCHAEAL GENOMIC STRESS RESPONSES
Northern hybridization of total
FIG. 4. A:
RNA from
cells after various peri-
P. abyssi
ods of growth, with the
gene probe
.
hpa2
Lane 1 contains 1.5 g of -33P-5 end-la-
m
g
9
beled DNA marker, with sizes indicated.
Lanes 2–6 each contains 10 g of total RNA
m
from cells of the ages indicated. Lane 7 con-
tains 1 ng of
PCR product, and lane 8
hpa1
contains 1 ng of
PCR product as posi-
hpa2
tive controls.
Southern hybridization
B:
hpa2.
controls for
Lane 1 contains 3 g of
m
I-digested
Sac P. abyssi
lane 2 contains 0.5 ng of
genomic DNA, and
PCR product.
hpa2
mRNA to the cor-
hpa2
Relative levels of
C:
responding 23S rRNA levels versus their
culture ages. The mRNA levels were deter-
mined as described for Fig. 3.
downstream. In Fig. 7 we propose a topology growth temperatures in archaeal cells (Hensel
sensing model in which the ability for nucleo- and Konig, 1988; Kurr 1991). Since the wrap-
et al.,
somes to flip from one conformation to another ping of DNA in archaeal nucleosomes is sensitive
could negate the topological effects produced by to salt concentration (Musgrave 2000) and
et al.,
protein tracking. Flipping from a negative to a since negative supercoils are substituted for pos-
positive supercoil within the archaeal nucleo- itive supercoils as the salt concentration is in-
some could minimize the positive supercoils pro- creased, we envisage that nucleosome flipping
duced ahead of a tracking complex, and, con- could be utilized by the cell to bring about a rapid
versely, flipping in the opposite direction could change in the topology of genomic DNA when
maximize it. The former change could be used by cells are required to adapt to different growth
the cell to minimize the topological effects pro- temperatures. The ability of both archaeal and eu-
duced by transcription or replication, whereas the caryal nucleosomes to wrap DNA in both posi-
latter change could maximize the ability of the tive and negative supercoils is also another indi-
advancing complex to clear nucleosomes ahead cator of the shared evolutionary history of the
of it and hence modify the topology of down- Euryarchaeota and eukaryotes. Nucleosome flip-
stream genes. A temperature sensing model is ping of the eucaryal H3/H4 core tetramer also
proposed in Fig. 8 in which nucleosome flipping suggests that the evolution of the eucaryal nu-
could be used by the cell to accommodate cleosome, by the addition of flanking H2A/H2B
changes in the temperature of its surrounding en- dimers on each side, has stabilized the octamer
vironment. Many hyperthermophilic archaeal in a negative orientation. This may be required to
species contain concentrations of salts in their cy- limit its structural flexibility, given the seemingly
toplasm as high as 1
and increased internal unlimited posttranslational modifications that
M,
salt concentrations are correlated with increased are possible for eucaryal histone proteins and the

248
MUSGRAVE ET AL.
ment, and by the observation that the two do-
mains can function separately (Collin
1988;
et al.,
2000). RG has been at
Collin, 1990; Declais
et al.,
the center of many “hot origin” hypotheses for
the origin of cellular life because it is, to date, the
only protein that is exclusively found in hyper-
thermophilic species and is thus seen to be es-
sential to life at high temperature. However, since
RG is considered to be a modern enzyme, con-
structed by the fusion of a topoisomerase I and a
DNA helicase (Declais
2000), it has been pro-
et al.,
posed that life could not have evolved at the tem-
peratures inhabited by hyperthermophilic mi-
croorganisms. In “lower temperature” origin
hypotheses, cellular life is proposed to have
evolved both up temperature as well as down
temperature (Forterre, 1995). However, because
of the lack of a robust genetic system in hyper-
thermophilic organisms, it has yet to be proven
that RG is essential for life at all, let alone essen-
tial for hyperthermophily. Also, positively su-
percoiled DNA could be produced in a number
of other ways in the absence of RG. Firstly, the
negative wrapping of DNA by archaeal histones,
as seen in Fig. 6, produces positive supercoils in
nucleosome-free DNA. Secondly, positive super-
coils can be produced by the removal of negative
supercoils by a conventional type I topoiso-
merase. Thirdly, as suggested here, positive
wrapping could also be the result of nucleosome
flipping caused by an increase in intracellular salt
concentrations or the selective removal of nega-
tive supercoils by a conventional type I topoiso-
merase following the tracking of a complex along
the DNA. These methods could have been used
to stabilize DNA molecules in the early evolution
Northern hybridization of total RNA from
FIG. 5. A:
abyssi
P.
cells after various periods of growth, with the
srbPa1
gene probe. Lanes 1–5 each contain 10 g of total RNA
m
corresponding from cells of the ages indicated. Lane 6 con-
tains 1 ng of
PCR product as a positive control.
srbPa1
B:
mRNA to the corre-
srbPa2
Relative levels of
and
srbPa1
sponding 23S rRNA levels versus their culture ages. The
mRNA levels were determined as described in Fig. 3.
fact that topoisomerases capable of increasing of cellular life at temperatures currently inhab-
the superhelicity of DNA have not been found in ited by hyperthermophiles in the absence of RG.
the Eucarya.
RG is thus proposed to be a modern solution to
the DNA stability problem in hyperthermophiles.
In addition to modulating topology and the pro-
duction of inherently stable positive supercoils in
nucleosome-free DNA, the wrapping of DNA in
DISCUSSION
RG is a type I DNA topoisomerase that has the nucleosomes has a significant effect on the tem-
unique ability to use ATP to drive the production perature stability of double-stranded DNA in its
of positive supercoils. It is proposed that the pos- own right. The melting temperature of DNA is
itive supercoils produced by RG result from the increased by 20°C by the formation of HTz his-
tracking of the enzyme along the DNA in a heli- tone nucleosomes
(Ronimus and Mus-
in vitro
case-like fashion and the consequent removal of grave, 1996b). The melting temperature of DNA
negative supercoils by its topoismerase I-like ac- can also be increased significantly by charge neu-
tivity, resulting in the production of net positive tralization resulting from an increase in the ionic
supercoils in the DNA (Declais
2000). This strength of the cytoplasm or an increase in the
et al.,
view of RG is influenced by its low ATP require- concentration of basic proteins, peptides, or other

249
ARCHAEAL GENOMIC STRESS RESPONSES
FIG. 6. DNA topology assays of the archaeal histones: recombinant HMfB homodimer from M. fervidus, recom-
pUC18
binant HMfA/B heterodimers, HTz native heterodimer from T. zilligii, and native MkaH from M. kandleri.
(150 ng) relaxed at 70°C in 150 m potassium glutamate was incubated in these conditions with archaeal histone
M
proteins at protein/DNA mass ratios of (a) 0:1, (b) 0.1:1, (c) 0.15:1, (d) 0.3:1, (e) 0.6:1, (f) 0.8:1, (g) 1:1, (h) 1.5:1, and (i)
2:1. Topoisomerase V (20 U) was added, and the incubation was continued for 15 min. Reaction mixes were incu-
bated with 1% sodium dodecyl sulfate for 2 min at 90°C and treated with 25 g/ml proteinase K for 30 min at 40°C.
m
Topoisomers were separated by 1.5% agarose gel electrophoresis at 1 V/cm for 16 h with buffer circulation. Lanes
labeled “rel” are untreated substrate molecules, and those labeled SC are pUC18 positively supercoiled topoisomer
1
controls.
positively charged species. However, considera- positively supercoiled molecules produced at low
tions of the
stability of DNA must be con- protein/DNA mass ratios and negatively super-
in vivo
sidered with some caution, as its cellular con- coiled molecules produced at higher protein/
centration cannot be approached For DNA mass ratios in the topology assay shown in
in vitro.
example, a bacterial chromosome must be con- Fig. 6. At 150 m potassium glutamate a switch
M
densed 1,000-fold to fit inside the cell, resulting from positive to negative wrapping occurs with
in a cellular concentration of DNA of between 14 a very small increase in added protein. We sug-
and 34 g/mL (Bohrmann
1991).
gest here that this structural plasticity could be
m
et al.,
The ability of archaeal histones to wrap DNA used by the cell to modulate the topology of chro-
in two alternate topologies, positive and negative mosomal DNA in a rapid and energetically effi-
toroidal supercoils, allows archaeal nucleosomes cient way. We propose that this “nucleosome flip-
to be structurally dynamic. Initially we suggested ping” could allow Archaea to rapidly adjust the
that positive toroidal supercoils might be neces- topology of their genome in response to envi-
sary to counteract the affect that cytoplasms ronmental changes. Because Archaea have been
containing high salt might have on the cellular shown to increase their cytoplasmic salt concen-
DNA in hyperthermophiles (Musgrave
trations in response to temperature increases
et al.,
1991, 1992). We reasoned that positive wrapping (Hensel and Konig, 1988; Kurr 1991) we pro-
et al.,
would result in an increase in the negative su- pose here that nucleosome flipping could also re-
perhelicity of free DNA and that this wrapping sult from such an increase. The result of this
would have the opposite effect to that exerted by would be to switch the toroidal wrapping in the
intracellular salt. At low salt concentrations all ar- nucleosomes from positive to negative as the tem-
chaeal histones wrap DNA in positive toroidal perature increased. This would cause the free
supercoils. However, as the salt concentration is DNA to become more positively supercoiled, a
increased, the wrapping becomes both positive result that is similar to the action of RG, and con-
and negative in the same reaction and possibly sequently the free DNA would be less prone to
on the same molecule. This is evidenced by the denaturation. An example of a drastic topologi-

250
MUSGRAVE ET AL.
around the DNA helix during transcription.
Therefore, for transcription to proceed the DNA
must effectively rotate, resulting in the produc-
tion of supercoils in the DNA (Liu and Wang,
1987; Wu
1988). Because DNA strands are
et al.,
not broken, the positive supercoils produced
ahead of RNA polymerase result in an equal
number of negative supercoils behind the com-
plex. A change in the toroidal supercoiling in ar-
chaeal nucleosomes from negative to positive
would decrease the positive superhelicity ahead
of tracking complex, and
With eucaryal
vice versa.
nucleosomes it has been proposed that the posi-
tive supercoils produced ahead of RNA poly-
merase could help to remove nucleosomes form
the DNA, leading to more efficient elongation of
the transcript. This model is based on the insta-
bility of nucleosomes on positively supercoiled
DNA and thermodynamics suggesting that for-
mation of negatively wrapped nucleosomes on
negatively supercoiled DNA is favored because
of the effective reduction in net negative super-
coils. In Eucarya, the dynamic nature of the nu-
cleosome is modulated by remodeling. However,
it appears that archaeal nucleosomes are not sub-
ject to the same remodeling because archaeal hi-
stones do not have the required N-terminal tails,
FIG. 7. Topology sensing via protein tracking-induced
conformational change of the DNA in archaeal nucleo-
somes.
A model is presented to propose that archaeal nu-
cleosome conformational changes could modulate the
topological changes that occur in DNA as a result of the
tracking of a protein complex. Archaeal nucleosomes, de-
picted as half-shaded circles, can wrap the DNA in posi-
tive and negative toroidal supercoils. Negative and pos-
itive plectonemic supercoils are produced in the free DNA
by protein tracking along the DNA. Conformational
changes in the mode of wrapping of the DNA by the ar-
chaeal nucleosomes, from positive to negative on the left
and negative to positive on the right, would negate the
plectonemic supercoils in the free DNA.
and, in
histones with altered mass have
T. zilligii,
not been isolated (Dinger
2000).
et al.,
In addition to the role that the wrapping of ar-
chaeal chromosomes in nucleosomes might play
in allowing cellular life at high temperatures, we
have described a stationary-phase stress response
protein that has a phylogenetic profile that mim-
cal change brought about by a change in DNA ics that seen for RG (Dinger
2000). This gene,
et al.,
is found in the genomes of both ar-
packaging is the positive wrapping of the DNA named
srb,
Positive chaeal and bacterial thermophiles. We previously
found in the spores of
Bacillus subtilis.
wrapping, as opposed to the negative wrapping detected SRB protein as a stationary phase-
of bacterial chromosomal proteins found in veg- induced acid-soluble protein in the hyperther-
etative cells, is induced by the small acid-soluble mophile
(Dinger
2000). In accord
T. zilligii
spore proteins (SASPs). This wrapping is sug- with this, we show here that transcription of the
gested to account for the resistance of spore DNA paralogues is also stationary phase-induced
to irradiation by UV light because SASP/DNA in
et al.,
srbPa
P. abyssi,
prevent the production of of the Thermococcales. SRB protein has been crys-
another hyperthermophilic member
complexes
in vitro
DNA-damaging thymine dimers (Nicholson
tallized from
et
Methanobacterium thermoautotroph-
and the structure has been used to search
icum,
1991; Griffith
1994).
al.,
et al.,
Another potential function of nucleosome flip- for structural homologues. Although highly sig-
ping could be to compensate for changes in nificant matches were not made, the closest struc-
genome topology that accompany DNA replica- tural homologues all have roles in nucleic acid
tion and transcription as described in Fig. 7. Tran- metabolism or processing, a role that agrees with
scriptionally mediated supercoiling occurs be- a GR repeat found in the C-terminus of some SRB
cause the mass of the RNA transcript with its proteins (Cort
2000). In addition to testing
et al.,
attached ribosomes is too great to be rotated the transcription of the
genes in stationary-
srb

251
ARCHAEAL GENOMIC STRESS RESPONSES
A
FIG. 8. Temperature sensing via a salt-induced conformational change of the DNA in archaeal nucleosomes.
model is presented to propose that archaeal nucleosome conformational changes could modulate the topological
changes that occur in DNA as a result of a change in intracellular salt concentrations induced by a change in tem-
perature of the environment, or another physical parameter that would influence the topology of genomic DNA.
Archaeal nucleosomes, depicted as hatched circles labeled ( ) or ( ), can wrap the DNA in positive and negative
1
2
toroidal supercoils, respectively. An increase in temperature, salt, or both would cause archaeal nucleosomes to
“flip” from a positive toroidal supercoil to a negative toroidal supercoil, and An increase in temperature
vice versa.
would therefore result in the free DNA more positively supercoiled, a response similar to that effected by the topo-
isomerase RG.
phase cells we have shown that there are con- topoisomerases in discussing the role that the
trasting transcription patterns of the histone par- physical and genetic stability of the genome plays
alogues in
and
We previously in cellular life and the evolution of that life. We
T. zilligii
reported that the histone proteins were rapidly also conclude that the systems induced to allow
depleted from cells in the stationary phase in Archaea to survive stationary phase show signif-
gene transcript icant diversity given that and
P. abyssi
P. abyssi.
T.
and that the level of the
zilligii
htz1
T. zilligii
was decreased as cells entered the stationary are members of the same archaeal family. This
phase and was not able to be detected by north- should promote a wider examination of these
ern hybridization using stationary-phase cells mechanisms, particularly since the recently de-
(Dinger
2000). In this study we show that, scribed archaeal symbiont
et al.,
in contrast with the similar stationary-phase in-
duction of expression of the
NanoArchaeaum equi-
whose 16S rRNA sequence suggests it be-
tans,
genes, the histone longs to a new phylum, are found associated
srb
genes (
and
hpa2
) are not transcribed in such predominantly with stationary-phase cells (Hu-
hpa1
an obvious stationary phase-repressed manner in ber
2002).
et al.,
as was shown for
and
Zhang,
P. abyssi
htz1
htz2 (
2001 Consistent with this result, the levels of the
).
gene transcript, which may encode a histone
sre
replacement protein, is likewise not stationary
phase-induced in From these results we
ABBREVIATIONS
P. abyssi.
conclude that the nucleiod and hence the topol-
ogy of genomic DNA is dynamic in the Archaea.
Therefore the mechanisms for genome packaging
should be considered along with the function of
MOPS, sodium 3-( -morpholino)propanesul-
N
fonic acid; PCR, polymerase chain reaction; RG,
reverse gyrase; SASP, small acid-soluble spore
protein; SSC, saline–sodium citrate.

252
MUSGRAVE ET AL.
dioresistance of the hyperthermophilic Archaea
Pyro-
266,
P. furiosus. Mol. Genet. Genomics
REFERENCES
and
coccus abyssi
72–78.
Anderson, P. and Bauer, W. (1978) Supercoiling in closed
circular DNA: dependence upon ion type and concen-
Griffith, J., Makhov, A., Santiago-Lara, L., and Setlow, P.
(1994) Electron microscopic studies of the interaction
between a
soluble spore protein with DNA: protein binding is co-
operative, stiffens the DNA and induces negative su-
tration
17, 594–601.
. Biochemistry
Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D.,
Seidman, J.G., Smith, J.A., and Struhl, K. (1998)
alpha/beta-type small, acid
Bacillus subtilis
Current
John Wiley & Sons, New
Protocols in Molecular Biology,
percoiling.
91, 8224–8228.
Proc. Natl. Acad. Sci. USA
York.
Hamiche, A. and Richard-Foy, H. (1998) The switch in he-
lical handedness of the histone (H3-H4)2 tetramer
within a nucleoprotein particle requires a re-orientation
Bohrmann, B., Villiger, W., Johansen, R., and Kellen-
berger, E. (1991) Coralline shape of the bacterial nu-
cleoid after cryofixation
. J. Bacteriol.
173, 3149–3158.
of the H3-H3 interface.
273, 9261–9268.
J. Biol. Chem.
Collin, R.G. (1990) Reverse gyrase from Archaebacteria
[Ph.D. Thesis]. University of Waikato, Hamilton, New
Zealand.
Collin, R.G., Morgan, H.W., Musgrave, D.R., and Daniel,
R.M. (1988) Distribution of reverse gyrase in represen-
Hamiche, A., Carot, V., Alilat, M., De Lucia, F., O’Dono-
hue, M.-F., Révet, B., and Prunell, A. (1996) Interaction
of the histone (H3-H4)2 tetramer of the nucleosome
with positively supercoiled DNA minicircles: potential
flipping of the protein from a left- to a right-handed su-
tative species of eubacteria and Archaeabacteria.
FEMS
perhelical form.
Hensel, R. and König, H. (1988) Thermoadaptation of
methanogenic bacteria by intracellular ion concentra-
93, 7588–7593.
Proc. Natl. Acad. Sci. USA
55, 235–240.
Microbiol. Lett.
Corliss, J.B., Dymond, J., and Gordo, L.I. (1979) Subma-
rine thermal springs on the Galapagos Rift.
203,
Science
tion.
49, 75–79.
FEMS Microbiol. Lett.
1073–1078.
Huber, H., Hohn, M.J., Rachel, R., Fuchs, T., Wimmer,
V.C., and Stetter, K.O. (2002) A new phylum of Archaea
represented by a nanosized hyperthermophilic sym-
Cort, J.R., Yee, A., Edwards, A.M., Arrowsmith, C.H., and
Kennedy, M.A. (2000) NMR structure determination
and structure-based functional characterization of
biont.
417, 63–67.
Nature
Jenuwein, T. and Allis, C.D. (2001)Translating the histone
code. 293, 1074–1080.
conserved hypothetical protein MTH1175 from
anobacterium thermoautotrophicum. J. Struct. Funct. Genom.
1, 15–25.
Meth-
Science
Kikuchi, A. and Asai, K. (1984) Reverse gyrase: a topoi-
somerase which introduces positive superhelical turns
Declais, A.C., Marsault, J., Confalonieri, F., de La Tour,
C.B., and Duguet, M. (2000)Reverse gyrase, the two do-
mains intimately co-operate to promote positive su-
into DNA.
309, 677–681.
Nature
Klages, K.U. and Morgan, H.W. (1994) Characterization
of an extremely thermophilic sulfur-metabolizing ar-
percoiling
275, 19498–19504.
. J. Biol. Chem.
Dinger, M. (1998) Histones and stationary phase of the
Thermophilic archaeon [M.Sc. The-
chaebacterium belong to the thermococcales.
Arch. Mi-
Thermococcus zilligii
162, 261–266.
sis]. University of Waikato, Hamilton, New Zealand.
Dinger, M.E. and Musgrave, D.R. (2000) Identification of
archaeal genes encoding a novel stationary phase-re-
crobiol.
Koch, A.L. (1994) Growth measurement. In
Methods for
edited by P. Gerhard,
General and Molecular Bacteriology,
R.G.E. Murray, W.A. Woord, and N.R. Krieg, ASM
Press, Washington, DC, pp. 248–277.
Kurr, M., Huber, H., Jannasch, H.W., Fricke, H., and
sponse protein.
Dinger, M.E., Baillie, G.J., and Musgrave, D.R. (2000)
Growth phase-dependent expression and degradation
1490, 115–120.
Biochim. Biophys. Acta
Tricone, A. (1991)
gen sp Nov.
of histones in the thermophilic Archaeaon
Methanopyrus kandleri,
represents a novel group of hyperthermophilic me-
Thermococ-
36, 876–885.
cus zilligii. Mol. Microbiol.
Doolittle, W.F. (1999) Phylogenetic classification and the
universal tree. 284, 2124–2129.
thanogens, growing at 110°C.
239–242.
156,
Arch. Microbiol.
Science
Liu, L.F. and Wang, J.C. (1987) Supercoiling of the DNA
Erauso, G., Reysenbach, A.L., Godfroy, A., Meunier, J.R.,
template during transcription.
84, 7024–7027.
Lopez-Garcia, P. and Forterre, P. (2000) DNA topology
and the thermal stress response, a tale from mesophiles
Proc. Natl. Acad. Sci. USA
Crump, B., Partensky, F., Baross, J.A., Marteinsson, V.,
Barbier, G., Pace, N.R., and Prieur, D. (1993)
Pyrococcus
sp. Nov., a new hyperthermophilic archaeon iso-
abyssi
lated from a deep-sea hydrothermal vent.
Arch. Micro-
and hyperthermophiles.
22, 738–746.
Bioessays
160, 338–349.
biol.
Musgrave, D.R., Sandman, K., and Reeve, J.N. (1991)
DNA binding by the archaeal histone HMf results in
Fahrner, R.L., Cascio, D., Lake, J.A., and Slesarev, A.
(2001) An ancestral nuclear protein assembly: crystal
positive supercoiling.
10397–10401.
Musgrave, D.R., Sandman, K., Stroup, D., and Reeve, J.N.
(1992) DNA binding proteins and genome topology in
88,
Proc. Natl. Acad. Sci. USA
structure of the
histone.
Methanopyrus kandleri
10, 2002–2007.
Protein
Sci.
Forterre, P. (1995) Thermoreduction, a hypothesis for the
origin of prokaryotes. 318, 415–422.
C. R. Acad. Sci. III
Forterre, P. and Phillipe, H. (1999) Where is the root of
the universal tree of life? 21, 871–879.
thermophilic prokaryotes. In
Biocatalysis at Extreme
edited by M.W.W. Adams and R.M. Kelly,
Temperatures,
Bioessays
American Chemical Society, Washington, DC, pp.
174–188.
Gerard, E., Jolivet, E., Pieur, D., and Forterre, P. (2001)
DNA protection mechanisms are not involved in the ra-

253
ARCHAEAL GENOMIC STRESS RESPONSES
Musgrave, D.R., Forterre, P., and Slesarev, A. (2000) Neg- Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989)
Molec-
ative constrained supercoiling in archaeal nucleosomes
Cold Spring Harbor Laboratory Press,
.
ular Cloning,
Cold Spring Harbor, NY.
Nelson, K.E., Clayton, R.A., Gill, S.R., Gwinn, M.L., Haft, Sandman, K. and Reeve, J.N. (2001) Chromosome pack-
35, 341–349.
Mol. Microbiol.
D.H., Hickey, E.K., Peterson, J.D., Nelson, W.C.,
aging by archaeal histones.
50,
Adv. Appl. Microbiol.
Ketchum, K.A., McDonald, L., Utterback, T.R., Malek,
75–99.
J.A., Linher, K.D., Garret, M.M., Stewart, A.M., Cotton, Sandman, K., Krzycki J.A., Dobrinski, B., Lurz, R., and
M.D., Pratt, M.S., Phillips, C.A., Richardson, D., Hei-
delberg, J., Sutton, G.G., Fleischmann, R.D., Eisen, J.A.,
White, O., Salzberg, S.L., Smith, H.O., Venter, J.C., and
Fraser, C.M. (1999) Evidence for lateral gene transfer
Reeve, J.N. (1990)HMf, a DNA-binding protein isolated
from the hyperthermophilic archaeon
Methanothermus
is most closely related to histones.
fervidus,
Proc. Natl.
87, 5788–5791.
Acad. Sci. USA
between Archaea and Bacteria from genome sequence Santisteban, M.S., Kalashnikova, T., and Smith, M.M.
of
399, 323–329.
(2000) Histone H2A.Z regulates transcription and is
Thermotoga maritima. Nature
Nicholson, W.L., Setlow, B., and Setlow, P. (1991) Ultra-
violet radiation of DNA complexed with alpha/beta-
partially redundant with nucleosome remodelling com-
plexes.
103, 411–422.
Cell
type small, acid soluble spore proteins from spores of Woese, C.R. (1987) Bacterial evolution.
51,
Microbiol. Rev.
or
species make spore photoprod-
Clostridium
221–271.
Woese, C.R. (1998) The universal ancestor.
Bacillus
uct but not thymine dimers.
88, 8288–8292.
Proc. Natl. Acad. Sci. USA
Proc. Natl.
95, 6854–6859.
Acad. Sci. USA
Olsen, G.J. and Woese, C.R. (1997) Archaeal genomics: an Woese, C.R., Kandler, O., and. Wheelis, M.L. (1990) To-
overview. 89, 991–994. wards a natural system of organisms: proposal for the
domains Archaea, Bacteria, and Eucarya.
Cell
Peak, M.J., Robb, F.T., and Peak, J.G. (1995) Extreme re-
sistance to thermally induced DNA backbone breaks in
Proc. Natl.
87, 4576–4579.
Acad. Sci. USA
Wu, J.S. and Grunstein, M. (2000) 25 years after the nu-
cleosome model: chromatin modifications.
the hyperthermophilic Archaeaon
Pyrococcus furiosus. J.
177, 6316–6318.
Bacteriol.
Pugh, B.F., Schutte, B.C., and Cox, M.M. (1989) Extent of
duplex underwinding induced by recA protein bind- Wu, H.Y., Shyy, S.H., Wang, J.C., and Liu, L.F. (1988)
ing in the presence of ATP. 205, 487–492. Transcription generates positively and negatively su-
percoiled domains in the template. 53, 433–440.
Trends Bio-
25, 619–623.
chem. Sci.
J. Mol. Biol.
Reanney, D.C. (1974) On the origin of prokaryotes.
J.
Cell
48, 243–251.
Zhang, X. (2001) Examination of the stationary phase re-
sponse in the EuryArchaea [M.Phil. Thesis]. University
of Waikato, Hamilton, New Zealand.
Theor. Biol.
Rivera, M.C., Jain, R., Moore, J.E., and Lake, J.A. (1998)
Genomic evidence for two functionally distinct gene
classes.
95, 6239–6244.
Proc. Natl. Acad. Sci. USA
Ronimus, R.S. and Musgrave, D.R. (1996a) A gene, han1A
encoding an archaeal histone-like protein from the
species AN1: homology with eukaryal hi-
Thermococcus
stone consensus sequence and the implications for de-
Address reprint requests to:
Dr. David Musgrave
Department of Biological Sciences
University of Waikato
lineation of the histone fold.
1–7.
1307,
Biochim. Biophys. Acta
P.O. Box 3105
Hamilton, New Zealand
Ronimus, R.S. and Musgrave, D.R. (1996b) Purification
and characterisation of a histone-like protein from the
archaeal isolate AN1, a member of the
Thermococcales.
20, 77–86.
Mol. Microbiol.
musgrave@waikato.ac.nz
E-mail: