Download DNA replication in thermophiles

236
Biochemical Society Transactions (2004) Volume 32, part 2
DNA replication in thermophiles
A.I. Majernı́k, E.R. Jenkinson and J.P.J. Chong1
Centre for Extremophile Research, Department of Biology and Biochemistry, University of Bath, Claverton Down, Bath BA2 7AY, U.K.
Abstract
DNA replication enzymes in the thermophilic Archaea have previously attracted attention due to their
obvious use in methods such as PCR. The proofreading ability of the Pyrococcus furiosus DNA polymerase
has resulted in a commercially successful product (Pfu polymerase). One of the many notable features of
the Archaea is the fact that their DNA processing enzymes appear on the whole to be more like those found
in eukaryotes than bacteria. These proteins also appear to be simpler versions of those found in eukaryotes.
For these reasons, archaeal organisms make potentially interesting model systems to explore the molecular
mechanisms of processes such as DNA replication, repair and recombination. Why archaeal DNA-manipulation
systems were adopted over bacterial systems by eukaryotic cells remains a most interesting question that
we suggest may be linked to thermophily.
The ‘replicon’ model revisited
The process of DNA replication appears to take place in a
number of steps that were originally proposed in a model
for understanding the regulation of DNA replication in bacteria. The replicon model proposed by Jacob, Brenner and
Cuzin [1] has a number of features that can be obviously
related to the situation found in both bacteria and eukaryotes,
where the study of DNA replication is well advanced.
It seems likely that these features will also be present in
the archaeal paradigm, which is discussed further below. The
salient features of the replicon model were that (i) an
initiator protein could control DNA synthesis by acting on
a replicator element and (ii) the activated replicator would
allow replication of the DNA attached to the replicator (the
replicon) to occur (Figure 1A).
These features can be identified in bacteria such as
Escherichia coli where DnaA acts as an initiator protein by
specifically recognizing and binding to the origin sequence,
oriC [2] (Figure 1B). The process whereby DNA replication
is allowed to occur is then completed through a number of
steps: additional DnaA proteins are recruited to the origin,
where binding to both DNA elements and other DnaA
molecules occurs [3]. The multimerization of DnaA while
bound to DNA causes torsional stresses in nearby sequences,
and a localized unwinding of an A/T-rich DNA element.
The DnaA multimer together with this single-stranded DNA
allow the recruitment of a DnaB–DnaC complex, which
results in the loading of a hexameric DnaB complex on
to the unwound DNA [4]. DnaB in this multimeric form
displays DNA helicase (unwinding) activity that precedes
the replication fork proteins required to duplicate the DNA
Key words: Archaea, Cdc6, cell cycle, DNA replication, mini-chromosome maintenance (MCM),
origin recognition complex (ORC).
Abbreviations used: ORC, origin recognition complex; MCM, mini-chromosome maintenance.
1
To whom correspondence should be addressed, at the present address: Department
of Biology (Area 5), University of York, P.O. Box 373, York Y010 5YW, U.K. (e-mail jpjc1@
york.ac.uk).
C 2004
Biochemical Society
duplex. The same steps must take place in eukaryotic
DNA replication, although different proteins are involved
in the process. In this case the ORC (origin recognition
complex) binds to specific DNA sequences known as origins
of replication (or ‘origins’) [5]. Whether localized duplex
unwinding occurs is currently not clear. The ORC then
appears to act as a landing pad for other proteins involved
in DNA replication [6], most notably the recruitment of
the MCM (mini-chromosome maintenance) proteins to the
DNA via the Cdc6 and Cdt1 proteins [7,8]. MCM proteins
have been shown to form hexameric complexes and are
thought to provide the DNA helicase activity required for
eukaryotic DNA replication (Figure 1C) [9]. Initial parallels
in the replication processes of bacteria and eukaryotes can
therefore be drawn thus: both mechanisms show recognition
of a specific DNA element through the recruitment of a
multimeric protein complex. This complex acts to recruit
other proteins, and in particular the DNA helicase. In
both cases the helicase is hexameric. Since Archaea are the
remaining major evolutionary ‘domain’ of life, consideration
of the components likely to be involved in archaeal DNA
replication allows us to consider the universal applicability of
the replicon model to DNA replication.
DNA replication in Archaea: roles for Cdc6
Analysis of whole archaeal genome sequences has allowed
the identification of sequences related to eukaryotic Cdc6
and MCM proteins but not to DnaA/B/C [10]. As is often
the case with the Archaea, it is difficult to make all but the
broadest of generalizations concerning this process. In most
cases, Archaea contain one or more Cdc6-like sequence and
one MCM sequence (see [11] for specific deviations from this
situation). It has been recently shown that the archaeal Cdc6
protein from Pyrobaculum aerophilum and a thermophilic
bacterial DnaA protein from Aquifex aeolicus have a very
high level of structural similarity although they show no
Thermophiles 2003
Figure 1 The replicon model updated
(A) Simplified schematic of the replicon model. A replicator protein binds to the DNA initiator sequence and causes the
replicon to be duplicated. DNA replication initiates at an origin. (B) Extended replicon scheme applied to E. coli. The replicator
protein DnaA binds to the initiator oriC sequence. Multimerization of DnaA leads to localized unwinding of the DNA at an
AT-rich origin and recruitment of the hexameric DnaB helicase via a DnaB–DnaC intermediate. (C) The replicon scheme
in eukaryotes. Autonomously replicating sequences (ARS) containing an origin of replication and an ORC-binding site are
occupied by the ORC protein complex, which may cause localized duplex unwinding. Hexameric MCM complexes are recruited
to the origin by the loading factors Cdc6 and Cdt1. Loaded MCMs act as a replication licensing factor, allowing the origin of
fire once per cell cycle, ensuring replciation of one complete genomes’-worth of DNA.
similarity at the level of amino acid sequence [12,13]. It is
therefore very tempting to speculate that archaeal Cdc6 may
be acting as the initiator protein and binding specifically to
archaeal origin DNA sequences. This is supported by data
from Forterre and co-workers who have shown not only
that the loci of a number of Cdc6-like genes are adjacent to
putative and actual origins of replication (a situation also seen
in bacteria with DnaA genes), but also physical interaction of
the Pyrococcus abyssi Cdc6 protein with DNA containing
specific origin sequences [14–16]. It is of further note that in
eukaryotes Cdc6 and a number of the ORC subunits have
significant sequence similarity [17,18]. Thus the appealing
scenario presented is that the single Cdc6-like protein found
in some Archaea may multimerize on origins in a similar
way to bacterial DnaA, causing similar torsional stresses
and localized DNA melting, and ultimately that the Cdc6
sequences found in Archaea may have evolved into the sixsubunit ORC complex of eukaryotes. However, it could also
be the case that the Cdc6-like homologues are ‘guilty by
association’, just as is the case in eukaryotic systems where
Cdc6 can also be detected at origins [19], and the Archaea
may use a completely different protein to fulfil the role of
ORC/DnaA.
An alternative role for the archaeal Cdc6 proteins is as
the helicase loading proteins. In eukaryotic DNA replication
initiation the presence of both Cdc6 and Cdt1 is required
for ‘loading’ of the MCMs on to DNA [7,8]. Again this
possibility is complicated by the lack of homogeneity
between species. Archaeal genomes with between 1 and 13
Cdc6-like genes have been reported. Biochemical studies
of the two Cdc6-like proteins in Methanothermobacter
thermautotrophicus have shown that the Cdc6-like proteins
have ATPase activity and can autophosphorylate [20]. It has
also been shown that ATP-bound Cdc6 can modulate the
DNA helicase activity of MCM complexes [21]. Functional
differences between the two Cdc6 proteins in this organism
have yet to be demonstrated. Thus in species where more
than one Cdc6-like gene exists, the simplest explanation is
that one gene plays the initiator role of ORC and DnaA,
while another acts as the helicase loader (equivalent to
eukaryotic cdc6 and cdt1 or bacterial DnaC). The role of
the additional Cdc6-like sequences found in species such as
C 2004
Biochemical Society
237
238
Biochemical Society Transactions (2004) Volume 32, part 2
Sulfolobus and Halobacterium then arises. One possibility is
that these organisms may replicate their genomes from more
than one origin of replication, and that these proteins show
different specificities for particular DNA sequences. Another
possibility is that these multiple Cdc6-like proteins are
snapshots in the evolution of the eukaryotic ORC, and that a
heterologous Cdc6 complex more reminiscent of the current
eukaryotic ORC is formed in these organisms. In species that
have only a single Cdc6-like sequence (notably Pyrococcus
species), it will be interesting to determine whether this
protein plays a dual role as origin marker and helicase
loader, or whether another, non-homologous protein plays
one of these roles. A final mystery is the apparent lack of a
Cdc6 gene in the genomes of Methanococcus jannaschii and
Methanopyrus kandleri.
Archaeal MCM: a model protein
A number of biochemical studies have characterized archaeal
MCMs as producing hexamers or double hexamers [22–24].
These proteins have provided a good system for initial forays
into understanding the molecular mechanism by which the
MCMs may act as a DNA helicase. The homomeric nature
of these multimeric complexes provides a much simpler
model for both structural and functional studies. A recent
crystal structure of the N-terminal 286 amino acids of the
M. thermautotrophicus protein has provided the impetus for
further interest in the way in which these proteins might cause
duplex DNA to be unwound [25]. This structure reinforces
the idea that the MCM complex is hexameric (which contrasts
with one study suggesting that heptamers may form [26]),
and also provides data to support the notion that the central
channel in this complex is sufficiently large to accommodate
double-stranded DNA. It may therefore still be the case that
the helicase activity of the MCMs is provided in the form
of a rotary pump as suggested by Laskey and Madine [27].
What is certainly clear is that the archaeal system provides a
great asset to our understanding of what is a key protein in
eukaryotic DNA replication.
Why make things more complicated?
The massive increase in the amount of DNA contained
in eukaryotic cells compared with prokaryotic ones is undoubtedly a major factor in the increased complexity found
in the eukaryotic DNA replication process. The replicon was
originally defined as “a genetic element such as an episome or
a chromosome (of a bacterium or of a phage) . . . [that] . . . can
only replicate as a whole” [1]. While this has been shown to
be the case in Bacteria and some Archaea, where replication
of a circular genomic DNA molecule is initiated from a single
origin of replication by a pair of replication forks that travel
in opposite directions around the chromosome, the replicon
is a far more malleable entity in eukaryotes. Eukaryotic
genomes are contained on multiple chromosomes that
replicate from thousands of different origins simultaneously.
This necessitates not only a requirement for a ‘replication
C 2004
Biochemical Society
licensing factor’ – a marker of un-duplicated DNA in the
replicating cell [28] – but also more sophisticated regulation
of the process of cell division: DNA replication should
certainly not be re-initiated until sister chromatids have
been safely segregated into separate compartments, and
cytokinesis should not be completed until the genomes
have been partitioned. This contrasts with the situation
in Bacteria, where multiple initiation events may result in
several replicons being duplicated simultaneously [29,30].
The bacterial mode of replication appears to be the case in
at least some Archaea. Archaeal genomes are encoded on
circular chromosomes. Replication studies in Pyrococcus have
identified a single replication origin that is replicated by a pair
of replication forks [15]. Flow cytometry studies in a number of archaeal species also indicate that several genomes’worth of DNA may be contained in a single cell [31,32].
Whether these observations are generalizable over the whole
domain remains to be determined. As mentioned before,
Pyrococcus species have only a single Cdc6-like sequence,
and therefore a single origin may be expected. The situation
with regards to functional origins in the species containing
multiple Cdc6-like sequences has yet to be experimentally
determined. In the same way, only a few species of Archaea
have been examined by flow cytometry. It is interesting that
flow cytometric analysis of the euryarchaeon Methanococcus
jannaschii (which appears not to have a Cdc6-like sequence,
but has multiple MCM sequences) also indicates the presence
of multiple genomes of DNA in individual cells [31].
Whatever the actual situation may be, and however conserved it is between individual species whether Crenarchaeota
or Euryarchaeota, it is clear that the mechanisms governing
cell-cycle control and division are controlled by novel means.
The sophisticated cell cycle of eukaryotes is controlled
by a number of different kinase activities, most notably
those of the cyclin-dependent kinases. The expansion of
the complement of related proteins found in eukaryotic
replication initiation may also arise from this necessarily
complex co-ordination of events – multiple ORC and MCM
subunits allow a more subtle control of DNA replication
events. No homologues to the cyclin-dependent kinases are
found in Archaea. Although in some cases there are FtsZ
homologues (a key protein in bacterial cell-cycle regulation),
interacting proteins appear to be non-homologous, and
therefore novel (reviewed in [33]). The Archaea therefore
present an interesting paradigm of replication proteins that
most likely use mechanisms of great relevance to eukaryotic
DNA replication studies, but which are controlled by novel
means. The identification of cell-cycle control proteins in
the Archaea will undoubtedly be of great interest to a wide
audience and yield greater understanding not only of archaeal
but also eukaryotic organisms.
Archaeal DNA replication as the prototype
for eukaryotic DNA replication
One of the most interesting aspects of the Archaea is their
position in evolution. The broadest of comparisons of the
Thermophiles 2003
three domains of life gives us the previously mentioned
observation that the proteins involved in eukaryotic DNA
metabolism appear to be more related to those in Archaea than
in Bacteria. Given that Bacteria are clearly the more ancient
organisms, the question must arise as to why eukaryotic cells
have adopted the archaeal DNA-processing machinery over
the bacterial system? Why did the Archaea need to solve the
problem of DNA replication a second time unless they coevolved with the Bacteria? Obviously producing new machinery for such a fundamental process is fraught with hazards.
The archaeal proteins do not appear to have arisen out of
a gradual evolution of bacterial proteins, but the wholesale
adoption of a different system – for example, the DNA
helicase in the bacterial system (DnaB) is processive in the
opposite direction (5 to 3 ) to that of the archaeal MCM (3
to 5 ). What force could be exerted to effect such a change?
One potential influence might be the presence of increased
temperature. In order to become a thermophilic organism,
a number of changes must be put in place to maintain the
integrity of the genome [34]. DNA will breathe and even melt
at higher temperatures and these elevated temperatures will
also increase the frequency of DNA damage (undoubtedly a
contributing factor to the presence of eukaryotic-like DNA
repair pathways in many of the Archaea) [35,36].
One solution to the melting problem is to bind compacting
proteins to the DNA. For example, the binding of histones
to DNA increases the temperature at which a given sequence
will melt. However, the addition of histones introduces
further problems to the DNA replication machinery, which
must cope with these additional obstacles before the duplex
can be melted. Therefore, one possible explanation for the
wholesale adoption of a different DNA replication system
may be a need for the ancient nascent thermophile to be
able to unwind histone-bound DNA. Having evolved DNA
replication and repair systems able to deal with chromatin
rather than ‘naked’ DNA allows further sophistication of
the cell. As well as conferring thermophily, histones compact
the DNA, allowing more genetic material to be stored in
a given volume. More genetic material allows further complexity to arise and may therefore be the explanation for why
upgraded archaeal DNA manipulation systems are found
in modern-day eukaryotes. This line of reasoning would
support the hypothesis that eukaryotic cells arose from a
thermophile, and would suggest that rather than describing
archaeal Cdc6 and MCM as eukaryotic-like, it might be
more accurate to describe eukaryotic Cdc6 and MCMs as
‘archaeon-like’.
Note added in proof (received 13 February 2004)
A considerable understanding of the role of Cdc6 and
potential multiple replication origins in the Archaea has been
afforded recently [37].
We thank Dr Alex Jeffries for a critical reading of this review. J.P.J.C.
is a BBSRC David Phillips Research Fellow.
References
1 Jacob, F., Brenner, S. and Cuzin, F. (1963) Cold Spring Harbor Symp.
Quant. Biol. 28, 329–348
2 Bramhill, D. and Kornberg, A. (1988) Cell 54, 915–918
3 Donachie, W.D. (1993) Annu. Rev. Microbiol. 47, 199–230
4 Messer, W., Blaesing, F., Jakimowicz, D., Krause, M., Majka, J., Nardmann,
J., Schaper, S., Seitz, H., Speck, C., Weigel, C. et al. (2001) Biochimie 83,
5–12
5 Bell, S.P. and Stillman, B. (1992) Nature (London) 357, 128–134
6 Stillman, B. (1996) Science 274, 1659–1664
7 Nishitani, H., Lygerou, Z., Nishimoto, T. and Nurse, P. (2000)
Nature (London) 404, 625–628
8 Maiorano, D., Moreau, J. and Mechali, M. (2000) Nature (London) 404,
622–625
9 Labib, K. and Diffley, J.F. (2001) Curr. Opin. Genet. Dev. 11, 64–70
10 Edgell, D.R. and Doolittle, W.F. (1997) Cell 89, 995–998
11 Jenkinson, E.R. and Chong, J.P. (2003) Biochem. Soc. Trans. 31, 669–673
12 Liu, J., Smith, C.L., DeRyckere, D., DeAngelis, K., Martin, G.S. and Berger,
J.M. (2000) Mol. Cell 6, 637–648
13 Erzberger, J.P., Pirruccello, M.M. and Berger, J.M. (2002) EMBO J. 21,
4763–4773
14 Lopez, P., Philippe, H., Myllykallio, H. and Forterre, P. (1999)
Mol. Microbiol. 32, 883–886
15 Myllykallio, H., Lopez, P., Lopez-Garcia, P., Heilig, R., Saurin, W., Zivanovic,
Y., Philippe, H. and Forterre, P. (2000) Science 288, 2212–2215
16 Matsunaga, F., Forterre, P., Ishino, Y. and Myllykallio, H. (2001)
Proc. Natl. Acad. Sci. U.S.A. 98, 11152–11157
17 Bell, S.P., Mitchell, J., Leber, J., Kobayashi, R. and Stillman, B. (1995) Cell
83, 563–568
18 Neuwald, A.F., Aravind, L., Spouge, J.L. and Koonin, E.V. (1999)
Genome Res. 9, 27–43
19 Donovan, S., Harwood, J., Drury, L.S. and Diffley, J.F. (1997) Proc. Natl.
Acad. Sci. U.S.A. 94, 5611–5616
20 Grabowski, B. and Kelman, Z. (2001) J. Bacteriol. 183, 5459–5464
21 Shin, J.H., Grabowski, B., Kasiviswanathan, R., Bell, S.D. and Kelman, Z.
(2003) J. Biol. Chem. 278, 38059–38067
22 Kelman, Z., Lee, J.K. and Hurwitz, J. (1999) Proc. Natl. Acad. Sci. U.S.A.
96, 14783–14788
23 Chong, J.P., Hayashi, M.K., Simon, M.N., Xu, R.M. and Stillman, B. (2000)
Proc. Natl. Acad. Sci. U.S.A. 97, 1530–1535
24 Shechter, D.F., Ying, C.Y. and Gautier, J. (2000) J. Biol. Chem. 275,
15049–15059
25 Fletcher, R.J., Bishop, B.E., Leon, R.P., Sclafani, R.A., Ogata, C.M. and
Chen, X.S. (2003) Nat. Struct. Biol. 10, 160–167
26 Yu, X., VanLoock, M.S., Poplawski, A., Kelman, Z., Xiang, T., Tye, B.K. and
Egelman, E.H. (2002) EMBO Rep. 3, 792–797
27 Laskey, R.A. and Madine, M.A. (2003) EMBO Rep. 4, 26–30
28 Blow, J.J. and Laskey, R.A. (1988) Nature (London) 332, 546–548
29 Lewis, P.J. (2001) Microbiology 147, 519–526
30 Sawitzke, J. and Austin, S. (2001) Mol. Microbiol. 40, 786–794
31 Malandrin, L., Huber, H. and Bernander, R. (1999) Genetics 152,
1315–1323
32 Maisnier-Patin, S., Malandrin, L., Birkeland, N.K. and Bernander, R.
(2002) Mol. Microbiol. 45, 1443–1450
33 Bernander, R. (2003) Mol. Microbiol. 48, 599–604
34 Lopez-Garcia, P. (1999) J. Mol. Evol. 49, 439–452
35 Grogan, D.W. (1998) Mol. Microbiol. 28, 1043–1049
36 Grogan, D.W. (2000) Trends Microbiol. 8, 180–185
37 Robinson, N.P., Dionne, I., Lundgren, M., Marsh, V.L., Bernander, R. and
Bell, S.D. (2004) Cell 116, 25–38
Received 19 September 2003
C 2004
Biochemical Society
239