ABSTRACT
We describe here the establishment of a cell-free transcription system for the hyperthermophilic Archaeon
Pyrococcus
furiosus
using the cloned glutamate dehydrogenase (
gdh
) gene as template. The
in
vitro
system that operated up to a temperature of 85
o
C initiated transcription 23 bp downstream of a TATA box located 45 bp upstream
of the translational start codon of
gdh
mRNA, at the same site as in
Pyrococcus
cells. Mutational analyses revealed that this TATA box is essential for
in
vitro
initiation of transcription.
Pyrococcus
transcriptional components were separated into at least two distinct
transcription factor activities and RNA polymerase. One of these transcription
factors could be functionally replaced by
Methanococcus
aTFB and
Thermococcus
TATA binding protein (TBP). Immunochemical analyses demonstrated a structural
relationship between
Pyrococcus
aTFB and
Thermococcus
TBP. These findings indicate that a TATA box and a TBP are essential components
of the
Pyrococcus
transcriptional machinery.
In recent years, microorganisms growing optimally around the boiling point of
water have been described (
1
). Among these hyperthermophiles members of the genus
Pyrococcus
, which belongs to the order Thermococcales (
2
), appear most appropriate as model organisms to study principles of
thermoadaptation and gene expression at extreme temperatures.
Pyrococcus
strains grow fast to high cell densities (
3
) and can be grown on maltose or cellobiose as a single carbon and energy source
(
4
,
5
). They have a protein cell wall that can be easily lysed (
4
). In addition, pathways of sugar catabolism have been studied extensively in
this organism (
6
,
7
) and evidence for regulation of gene expression depending on carbon source for
growth and growth conditions has been obtained (
8
; de Vos and Antranikian, unpublished data).
Within Archaea, transcription has been studied in some detail in the methanogen
Methanococcus
thermolithotrophicus
(
9
-
11
) and the Crenarchaeon
Sulfolobus
shibatae
(
12
,
13
). In
Methanococcus
, initiation of transcription
in
vitro
is mediated by two archaeal transcription factors, aTFA and aTFB. aTFB binds to
the TATA box of archaeal promoters (
14
) and can be replaced by yeast and human TATA binding protein (TBP) in cell-free transcription reactions (
15
). These findings suggest that aTFB is homologous to eukaryotic TBPs. The nature
of aTFA has not yet been elucidated, but it seems to stabilize binding of aTFB
and of eukaryotic TBPs to the archaeal TATA box (
14
,
15
) and incubation of both aTFA and aTFB with archaeal promoter DNA results in
template commitment (
14
).
Recently, genes have been identified in
Pyrococcus
,
Thermococcus
and
Sulfolobus
genomes which show significant sequence similarity to the RNA polymerase II
transcription factors TFIIB and TBP, which is the DNA binding component of
transcription factor TFIID (
16
-
22
). The translation product of this putative TBP gene from
Pyrococcus woesei
has been shown to bind to an archaeal TATA box (
18
) and the translation product of the putative
Thermococcus celer
TBP gene is able to substitute for
Methanococcus
aTFB in cell-free transcription reactions (
23
). However, active transcription factors have not yet been purified from
Pyrococcus
cells and specific interaction of transcription factors from
Thermococcales
with
Pyrococcus
RNA polymerase has not yet been demonstrated. We describe here the
reconstitution of a specific cell-free transcription system from cellular fractions of
Pyrococcus furiosus
using the
Pyrococcus gdh
gene as template.
gdh
is one of the first genes cloned and sequenced from
Pyrococcus
(
24
) and its production is regulated by the growth conditions (de Vos and
Antranikian, unpublished data). The system described here may prove useful to
unravel the mechanisms of gene regulation in archaeal cells and to address
questions related to gene expression at temperatures close to the boiling point
of water.
Pyrococcus
furiosus
cells were grown at 90oC in a 100 l enamel-coated fermenter (Braun-Biotech) in a slightly modified medium used for cultivation of
Thermococcales (
25
) containing 1.0 g/l yeast extract (Difco), 5.0 g/l peptone (Difco), 5.0 g/l
soluble starch (Merck), 18.4 g/l NaCl, 12.6 g/l MgCl
2
[middot]6H
2
O, 0.16 g/l NaHCO
3
, 3.24 g/l Na
2
SO
4
, 2.38 g/l CaCl
2
[middot]2H
2
O, 0.56 g/l KCl, 0.5 g/l L-cysteine and 0.0001% resazurine. Trace minerals (80 mg/l KBr, 57 mg/l SrCl
2
[middot]6H
2
O, 22 mg/l H
3
BO
4
, 10 mg/l Na
2
HPO
4
, 4 mg/l sodium metasilicate[middot]5H
2
O, 2.4 mg/l NaF and 1.6 mg/l KNO
3
) were added from a stock solution. The pH was adjusted to 6.0 with sulphuric
acid. The fermenter was gassed with N
2
/CO
2
at a ratio of 80:20 (2 l/min) to remove hydrogen formed during growth of
Pyrococcus
. The agitation rate was 200 r.p.m. Cells were rapidly cooled before harvesting,
collected by centrifugation and stored at -70oC.
A cell extract from 22 g of cells (wet weight) of
P
.
furiosus
was prepared by polymin P and ammonium sulphate precipitation as described by
Zillig
et al
. (
26
,
27
). RNA polymerase and transcription factors were purified from the ammonium
sulphate precipitate as indicated in Figure
1
. Purification was performed in a buffer containing 50 mM Tris-HCl, pH 7.5, 1 mM EDTA, 50 mM KCl and 20% glycerol (v/v). The sizes of
DEAE-cellulose and heparin-Sepharose columns were 5 * 11 cm and 1.8 * 14 cm, respectively. Elution of bound proteins was
performed with linear salt gradients from 0.05 to 1 M KCl (10-fold column volumes each). Chromatography on Mono Q and Superdex 200
columns was performed as described previously (
10
). Protein concentrations were determined according to the procedure of Bradford
(
28
) using bovine serum albumin as standard.
DNA was purified by repeated centrifugation in CsCl density gradients as
described previously (
9
). Templates for
in
vitro
transcription studies were pUC19-based plasmids that were constructed by cloning PCR-amplified fragments containing the 5' and 3' regions of the
P.furiosus
gdh
gene (
24
). These fragments included part of the coding region and were fused in-frame at an artificial
Bam
HI site created by PCR at position +173 (Fig.
6
A). Plasmid pLUW400 consists of pUC19 carrying a 2.7 kb
Hin
dIII fragment including the complete
gdh
gene and its transcriptional terminator (
24
). To generate the 3' fragment, pLUW400 was used as template in a PCR amplification using the
reverse sequencing primer (primer b, Fig.
6
A; Pharmacia) and an oligonucleotide (primer a, Fig.
6
A) with the sequence CATA
GGATCC
AGAAGAAC-3' (
Bam
HI site in italics). As a PCR template for amplification of the 5' region, we constructed pLUW409 by inserting a 4.2 kb
Sph
I-
Nhe
I fragment of the chromosomal DNA of
P
.
furiosus
in pUC18. The configuration of pLUW409 was confirmed by restriction enzyme
digestion, hybridization and partial sequence analysis. In the PCR
amplification of the 5' region, primer c (Fig.
6
A) TTAC
GGATCC
GTCATCCATTTC-3' (
Bam
HI site in italics) was used in combination with the reverse sequencing primer
(primer b, Fig.
6
A). After subsequent digestion with
Xba
I,
Hin
fI or
Hin
dIII and
Bam
HI, fusion with the 3' fragment hydrolysed with
Bam
HI/
Ssp
I and ligation into pUC19 resulted in plasmids pLUW479, pLUW489 and pLUW477,
respectively (see Fig.
6
A). For the production of mutated promoter sequences, use was made of the
Hin
dIII site at position -33 in a PCR strategy by using primers CCGA
AAGCTT
TA
Reaction mixtures (100 [mu]l) for non-specific assays contained 40 mM HEPES, pH 7.3 (adjusted at 20oC), 250 mM KCl, 2.5 mM MgCl
2
, 0.1 mM EDTA, 1 mM DTT, 1 mM ATP, 0.1 mM UTP, 0.7 [mu]Ci [[alpha]-
32
P]UTP (3000 Ci/mmol), 2.5 [mu]g poly(dA[middot]dT)] and 5 or 2 [mu]l (Mono Q fractions) from various chromatographic purification
steps. Incubation was performed for 30 min at 70oC. Radioactivity insoluble in 5% trichloroacetic acid was collected on
glassfibre filters (Whatman GF/C) and quantitated by liquid scintillation
counting. Reaction mixtures (100 [mu]l) for specific assays used the same salt and buffer conditions as for non-specific assays but poly(dA[middot]dT)] was replaced by 2 [mu]g linearized template DNA (pLUW409 or mutated templates)
and 0.33 mM each of ATP, CTP and GTP, 0.002 mM UTP and 2 [mu]Ci [[alpha]-
32
P]UTP (3000 Ci/mmol) were added as substrates for synthesis of
in
vitro
RNA products. Five microlitres of the Superdex 200 fraction of RNA polymerase
and 10 [mu]l each of the Superdex 200 fraction of aTFB and the heparin-Sepharose fraction of aTFA were added to start transcription
reactions. Autoclaved mineral oil (Sigma) was added as an overlay to reaction
vessels to prevent evaporation and reactions were incubated for 30 min at 70 or
80oC. RNA was purified for electrophoresis in denaturing polyacrylamide gels
as described previously (
9
).
For analysis of the transcription start site
in
vitro
, transcription reactions were performed as described, but without labelled
precursors. In control reactions, nucleotides were omitted from cell-free transcription reactions. The single-stranded end-labelled DNA primer 5'-CTCCATATATTGGGCAGCTC-3' (primer 1, Fig.
2
A; 40 000 c.p.m.), complementary to nucleotides +75 to +94 of the
gdh
gene, was annealed with RNA from one transcription reaction, extended with
reverse transcriptase and cDNA analysed by electrophoresis on a 6% DNA
sequencing gel as described previously (
9
). As a modification, RNA extracted from one transcription reaction was co-precipitated with end-labelled primer (40 000 c.p.m.) and the pellet dissolved in either
annealing buffer (10 mM Tris-HCl, pH 7.5, 40 mM KCl, 0.5 mM EDTA) or 5* annealing buffer. When the primer was hybridized in 5* annealing buffer ~80% of cDNA terminated at the G and 20% at the C
residue of the initiator element (not shown). When annealing buffer was used,
both signals showed similar intensities (see Fig.
3
A). The DNA sequence of the
gdh
gene was analysed as a standard by the dideoxy chain termination method using
the same primer as for primer extension analysis. The transcription start site
in
vivo
was analysed using RNA isolated from
P
.
furiosus
and the end-labelled primer (primer 2, Fig.
2
A) with the sequence 5'-CTTATCTCCATATATTGGGC-3' essentially as described previously (
24
).
Immunochemical relationships of archaeal transcription factors were analysed by
challenging Western blots with antibodies directed against
Thermococcus
TBP as described previously (
23
). The IgG fraction was purified from rabbit serum by affinity chromatography on
protein A-Sepharose (Pharmacia).
A DNA fragment containing the coding region of
Thermococcus
TBP was cloned into the
Nde
I and
Eco
RI sites of the expression vector pET-17b (Novagen) as described previously (
23
). The resulting plasmid pTBPTc.17 was used for the expression of
Thermococcus
TBP in
Escherichia coli
BL21 (DE3) without N- or C-terminal fusion. Expression of the archaeal insert was induced by
adding isopropyl-1-thio-D-galactopyranoside (final concentration 1 mM) to a
growing culture (OD
600
= 0.7). The crude extract was heated for 15 min at 70oC. Precipitated material was collected by centrifugation (100 000
g
for 60 min) and the TBP purified from the supernatant by chromatography on Mono
Q and Superdex 200 columns as described (
23
).
To purify transcriptional components of
P
.
furiosus
a seven-step purification scheme was employed (Fig.
1
). As template for cell-free transcription reactions we used a recombinant plasmid linearized with
Nde
I that contains the
gdh
gene of
P.furiosus
(Figs
2
A and
6
A, pLUW409). When initiation of transcription occurs at the predicted site 22-24 bp downstream of the TATA box (
30
) of the putative
gdh
promoter, a run-off transcript of 327-329 nt should be expected. When the Superdex fraction of RNA
polymerase was incubated in cell-free transcription reactions with template DNA alone (Fig.
2
B, lane 1), with the
M
r
48 000 Superdex fraction, tentatively designated
Pyrococcus
aTFB (Fig.
2
B, lane 5) or the fraction eluted from heparin-Sepharose at 500 mM KCl, tentatively designated
Pyrococcus
aTFA (Fig.
1
, left side and Fig.
2
B, lane 4), distinct RNA products were not detected. However, incubation of the
enzyme in the presence of aTFB and aTFA resulted in the synthesis of a RNA
product of correct size (Fig.
2
B, lane 7). This finding indicates that at least two independent transcription
factors are required for transcription from the
gdh
promoter
in
vitro
.
Pyrococcus
transcriptional components were incubated with the
gdh
template at temperatures between 60 and 90oC. A transcript of correct size was synthesized over a broad range of
temperature (Fig.
2
C). RNA synthesis was optimal between 70 and 80oC (Fig.
2
C, lanes 3-5). The upper limit of cell-free transcription from a linearized template was 85oC (Fig.
2
C, lanes 6 and 7).
To define the transcription start site precisely, a primer extension protocol
was used. When 5'-labelled primer 1 or primer 2 (Fig.
2
A) were annealed with RNA extracted from
in
vitro
transcription reactions or from
Pyrococcus
cells, cDNAs of 95/94 (Fig.
3
A) and 100 (Fig.
3
B) nt were synthesized by reverse transcriptase. From these results two
transcription start sites can be inferred, which are located at the C and G
residues 22 and 23 nt downstream of the TATA box within the initiator element (
30
) of this promoter. However, the existence of the longer reverse transcriptase
product is strictly dependent on experimental conditions (see Materials and
Methods) and is most likely caused by the addition of a single nucleotide to
the 5'-end of the cDNA by reverse transcriptase (
31
). Hence, the G residue (labelled +1 in Fig.
2
A) appears to represent the correct transcription start site
in
vitro
and
in
vivo
.
Table
1
summarizes the results of purification of
Pyrococcus
transcriptional components. aTFB and RNA polymerase were enriched from the DEAE
fraction by a factor of 22 and 16, respectively, whereas aTFA was enriched from
the flow-through fraction of the DEAE-cellulose column only by a factor of 3.2. In agreement with these
data, analysis of the polypeptides in these fractions revealed that aTFB and
RNA polymerase are highly purified fractions, whereas the heparin-Sepharose fraction of aTFA is rather crude and might be composed of
independent transcription factors. Further purification of aTFA was not
possible due to its low stability.
Table 1
To investigate the relationship of
Pyrococcus
and
Methanococcus
transcription factors,
Pyrococcus
transcriptional components aTFA and aTFB were substituted by
Methanococcus
transcription factors in cell-free transcription reactions. When
Pyrococcus
aTFA was replaced by
Methanococcus
aTFA (Fig.
4
A, lane 9) or
Methanococcus
aTFB (Fig.
4
A, lane 7),
Pyrococcus
RNA polymerase was unable to synthesize distinct RNA products. It was not
possible to detect specific transcription when
Pyrococcus
aTFB was replaced by
Methanococcus
aTFA (Fig.
4
A, lane 8). In contrast, when
Pyrococcus
aTFB was replaced by
Methanococcus
aTFB, an RNA product of correct size was observed (Fig.
4
A, lane 6, compare with lane 1). These findings indicate that aTFB, in contrast
to aTFA, is functionally interchangeable between
Pyrococcus
and
Methanococcus
and suggests that the
Pyrococcus
aTFB fraction contains an archaeal TATA binding protein.
To identify
Pyrococcus
promoter elements in a functional assay a set of upstream DNA deletions as well
as a deletion and single point mutation in the TATA box of this template were
constructed and used as templates for cell-free transcription reactions. The template used for the experiments shown
in Figures
2
B and
6
B, lane 1 includes 2.4 kbp of
gdh
DNA upstream from the gene (pLUW409, Fig.
6
A). For analyses of mutated DNAs, DNA linearized with
Bam
HI was used. When transcription initiates accurately at these templates an RNA
product of 173 nt was expected. When the DNA region from -2400 to -201 was deleted, a strong signal corresponding to an RNA product
of correct size was observed (Fig.
6
B, lane 2), indicating that the DNA region upstream of position -201 is not essential for cell-free transcription. Deletion of the DNA sequence from position -201 to -86 also did not affect template activity (Fig.
6
B, lane 3). However, when the DNA region between positions -86 and -33 was deleted, transcriptional activity was only 6% of wild-type levels (Fig.
6
A and B, lane 4). Transcription initiation was completely inhibited when 6 nt of
the TATA box were deleted or when the T residue at position 5 of the TATA box
was replaced by a G residue (Fig.
6
B, lanes 5 and 6). These findings indicate that the TATA box is essential for
transcription by this
in
vitro
system. A second control region may exist in the DNA segment between positions -86 and -33.
We describe here a specific cell-free transcription system for a hyperthermophilic Euryarchaeon (
32
) using as template the gene coding for an abundant
Pyrococcus
protein, glutamate dehydrogenase. Cell-free transcription in
Methanococcus
and
Sulfolobus
has been shown to initiate at the purine of a pyrimidine/purine dinucleotide at
a distance of 18-26 bp downstream of the TATA box (
30
,
33
). Analysis of run-off transcripts (Figs
2
B and
4
B) and primer extension experiments (Fig.
3
A and B) demonstrate that
Pyrococcus
RNA polymerase initiates
in
vitro
transcription at the G of a CG dinucleotide of the
gdh
gene at a distance of 23 bp downstream of an AT-rich sequence which is in sequence identical to the consensus TATA box of
archaeal promoters (
34
,
33
). These findings suggest that the crucial elements of
Methanococcus
and
Sulfolobus
promoters are also conserved in regulatory sequences of
Pyrococcus
DNA. The
in vivo
transcription start site identified in this study was located 16 nt upstream of
that determined earlier (
24
). Most likely this can be attributed to the better quality of the RNA batch
used and the inferred transcription start site corresponds to that observed in
the
in
vitro
system. The observation that transcription
in
vitro
initiates at the same site as
gdh
mRNA in
Pyrococcus
cells indicates that this cell-free system is capable of recognizing promoters correctly
in
vitro
. Analyses of upstream DNA deletions as well as deletions and a single point
mutant of the TATA box revealed that the TATA box is essential for specific
transcription of the
gdh
template by the
Pyrococcus
cell-free system (Fig.
6
). Moreover, the heterologous tRNA
Val
gene of
Methanococcus
is also transcribed with high activity (Fig.
4
B) by the
Pyrococcus
-derived system, indicating a high degree of conservation of transcription
signals and of the corresponding
trans
-acting components between methanogenic Archaea and Thermococcales. The
evidence presented here suggests that promoters of hyperthermophiles contain
similar elements as promoters from mesophilic and extreme thermophilic Archaea.
Additional mutational analyses are required to verify the existence of a second
putative control signal in the DNA region between positions -86 and -33 upstream of the
gdh
template.
Similar to
Methanococcus
(
11
), at least two transcription factors are necessary for cell-free transcription. One factor, like
Methanococcus
aTFB (
10
), co-elutes during several chromatographic steps with RNA polymerase and is
separated from the enzyme by chromatography on a Superdex 200 column (Fig.
1
). Analysis of the molecular mass of this factor by Superdex chromatography
revealed a molecular mass of 48 000, which is also similar to that of
Methanococcus
aTFB (54 000) analysed by the same procedure. The factor separating from the
RNA polymerase during Superdex 200 chromatography (Fig.
1
) can be functionally replaced by
Methanococcus
aTFB (Fig.
4
A, lane 6) and by bacterially produced
Thermococcus
TBP (Fig.
5
B, lane 6). Moreover, the
Pyrococcus
aTFB fraction contains a polypeptide which is serologically related to
Thermococcus
TBP (Fig.
5
A). The molecular mass of this polypeptide determined by SDS-PAGE is identical (Fig.
5
A) to that of the TBP from
T.celer
and
P.woesei
, which is 21 300 (
19
,
18
), and approximately half of the value estimated for native aTFB (48 000),
suggesting that
Pyrococcus
aTFB is, similarly to
Methanococcus
aTFB (
10
), a dimer in solution. The relationship of
P.furiosus
aTFB to
Thermococcus
TBP in structure and function suggests that these polypeptides are homologous.
The second factor is not functionally interchangeable with
Methanococcus
aTFA. Further purification of this fraction and cloning of the gene(s) encoding
this activity is required to elucidate the nature of this second factor and its
function in transcription.
Pyrococcus furiosus
thrives at temperatures between 70 and 103oC (
4
) and shows optimal growth rate at 100oC. Thus far unknown mechanisms protect the cellular DNA from thermal
degradation in
Pyrococcus
cells. Our results show that the
Pyrococcus
transcriptional machinery works
in
vitro
up to a temperature of 85oC (Fig.
2
C). Similarly to the cell-free system established for the hyperthermophilic bacterium
Thermotoga
(
35
), activity of the
Pyrococcus
system decreases rapidly at temperatures higher than 85oC. As linearized templates were used for transcription experiments with
Thermotoga
and
Pyrococcus
RNA polymerase, lack of activity at temperatures above 85oC seems to be caused by denaturation of the template and not by
inactivation of transcriptional components. Supercoiled DNA was shown to be
resistant to thermal denaturation at temperatures up to 107oC (
36
). Initial experiments in our laboratory showed that supercoiled
gdh
template can be accurately transcribed
in
vitro
up to a temperature of 95oC, but analysis of transcripts from supercoiled templates was hampered by
the inability of the
Pyrococcus
system to terminate transcription correctly
in
vitro
. However, the cell-free system described here still initiates transcription at temperatures
close to the growth optimum of
P
.
furiosus
. It therefore may prove useful to investigate specific problems of
transcription at high temperatures, like the effects of reverse gyration of DNA
(
37
,
38
) on transcription, to unravel mechanisms of gene regulation in archaeal cells
and to investigate in more detail the function of the putative transcription
factors identified in the genomes of hyperthermophiles (
16
-
22
).
This work was supported by the European Community Biotechnology Program, a grant
from the Deutsche Forschungsgemeinschaft and the Fonds der Chemischen
Industrie.
Purification step
Total protein
Total activity
Specific activity
(mg)
(U)
(U/mg)
aTFA
AS precipitate
389
2484
6.4
DEAE-cellulose
29
11 045
380.4
Heparin-Sepharose
4.3
5330
1245.4
aTFB
AS precipitate
389
3600
9.3
DEAE-cellulose
261
13 757
52.6
Heparin-Sepharose
22.4
5761
257.4
Mono Q
13.5
3636
268.5
Superdex 200
1.5
1730
1153.3
RNA polymerase
AS precipitate
389
4914
12.6
DEAE-cellulose
261
14 851
56.8
Heparin-Sepharose
22.4
12 820
572.9
Mono Q
13.5
10 414
771.4
Superdex 200
10.1
9594
887.7
REFERENCES
Return