ABSTRACT
The
Pk-rec
gene, encoding a RecA/RAD51 homologue from the hyperthermophilic archaeon
Pyrococcus
sp. KOD1, was expressed in
Escherichia coli
. The recombinant
Pk
-REC was purified to homogeneity and was shown to be in a dimeric form. A
striking property of the purified recombinant
Pk
-REC was the unusual DNase activity on both single- and double-stranded DNAs along with the ATPase activity. The reaction
product of this DNase activity was mononucleotides. The optimum temperature and
pH for the DNase activity were 60
o
C and 8-8.5, respectively. In addition, the metal ion requirement for DNase
activity was different from that for the ATPase activity. The protein exhibited
no DNase activity in the presence of Zn
2+
ion, which was one of the most preferable divalent cations for ATPase activity.
Another unique characteristic of the recombinant protein was that the reaction
product of ATPase activity was AMP instead of ADP.
Pk
-REC may represent a common prototype of the RecA family proteins with high
RecA-like activity.
Genetic recombination is one of the processes which allows for natural selection
in the evolution of organisms. Most of the genetic recombination studies have
been carried out using prokaryotic systems, particularly the bacterium
Escherichia coli
, as well as some lower eukaryotes. The bacterial RecA protein is known to play
an essential role in homologous recombination, DNA strand exchange, DNA repair
and coprotease activity in response to DNA damage resulting in the SOS
response, prophage induction and LexA cleavage (
1
-
4
). Purified RecA catalyzes the strand exchange reactions between a single-stranded DNA (ssDNA) and a double-stranded DNA (dsDNA) or between two dsDNAs, which is believed to be
the central mechanism of homologous genetic recombination. RecA also
demonstrates nucleoside triphosphatase activity. All of these activities have been extensively characterized
in vitro
and all require the presence of DNA and nucleoside triphosphate. In lower eukaryotes (
Saccharomyces cerevisiae
) two classes of genes are involved in genetic recombination. One class includes
the RAD52 epistasis group of genes (
5
-
7
) involved in mitotic DNA repair and meiotic recombination. Among these gene
products, RAD51 protein is structurally and functionally similar to RecA
protein of
E.coli
(
8
,
9
). The genes belonging to the other class are essential only for meiotic
recombination. Among these gene products, Dmc1 is structurally and evolutionary
related to RAD51 and RecA.
Genes for RecA/RAD51 protein homologues have recently been reported from the
third primary kingdom, archaea (
10
,
11
). It is of extreme interest to determine the enzymatic properties of
hyperthermophilic archaeal RecA/RAD51 homologues, because hyperthermophilic
archaea can grow in extreme environments which are probably similar to that of
the primitive earth and, therefore, it is expected that enzymes from these cells may represent prototypes of enzymes which belong to the same protein family. Since some
eukaryotic genes, such as the TATA-binding protein gene, have also been found in archaea (
12
-
15
), a similar transcriptional system seems to be shared by eukarya and archaea. Moreover phylogenetic trees based upon aminoacyl-tRNA synthetase (
16
), H
+
-ATPase (
17
) and elongation factors genes (
18
) show that archaea and eukarya share a more common ancestor than either does
with bacteria. We have previously cloned the
recA
/
rad51
homologue gene (
Pk-rec
) from the hyperthermophilic archaeon
Pyrococcus
sp. KOD1 and analyzed its sequence (
11
). It was found that
Pk
-REC lacks the N- and C-terminal domains as compared to other members of this protein
family. It has been suggested that for the
E.coli
RecA protein the N-terminal domain contributes to its self assembly (
19
) and the C-terminal domain contributes to regulate the binding affinities to both
single- and double-stranded DNAs (
20
). Nevertheless, recombinant
Pk
-REC complemented the UV light sensitivity of
E.coli recA
mutant strain (
11
), suggesting that
Pk
-REC is functional
in vivo
and its structure may be an ancestral prototype of bacterial and eukaryotic
enzymes. In this report, we have purified recombinant
Pk
-REC and analyzed its enzymatic properties. The unique biochemical
activities of this protein are discussed.
Pyrococcus
sp. KOD1 was isolated from a solfataric hot spring in Kagoshima, Japan (
21
).
Escherichia coli
strain JM109 was used as a host for subcloning the gene fragments and DNA
manipulations.
Escherichia coli
strain HMS174(DE3)pLysS. [F
-
recA
hsdR
(r
k12
-m
k12
+
)Rif
R
(DE3)pLysS(Cm
R
)] was used as a host to express the
Pk-rec
gene by using pET-8c expression vector (Novagen, Madison, WI).
L broth [10 g tryptone (Difco Lab., Detroit, MI), 5 g yeast extract (Difco Lab.)
and 5 g NaCl in 1 liter of deionized water, pH 7.2) was used for
E.coli
culture. NZCYM medium (10 g NZ amine, 5 g NaCl, 5 g yeast extract, 1 g casamino acids and 2 g MgSO
4
7H
2
O in 1 liter deionized water pH 7.0) was used for the expression of the gene by induction with 1 mM isopropyl-[beta]-D-thioglactopyranoside (IPTG) (Sigma, St Louis, MO).
Ampicillin (Sigma) was routinely used at a final concentration of 50 [mu]g/ml.
Restriction endonucleases were purchased from Takara Shuzo Co. (Kyoto, Japan),
DNA polymerase and T4 DNA ligase were purchased from Pharmacia Biochemicals Inc. (Uppsala, Sweden). Digestion of DNA
with restriction enzymes and analysis of DNA fragments by agarose gel electrophoresis were performed under standard
conditions (
22
). Transformation was performed by the calcium chloride procedure as described
by Sambrook
et al
. (
22
). Mini scale preparation of
E.coli
plasmid DNA was done by the alkaline lysis method (
22
) and large scale plasmid DNA preparation was performed by using Qiagen plasmid
Maxi kit (Qiagen Inc., Chatsworth, CA).
Alignment of amino acid sequences and estimation of the number of amino acid
substitutions per site or evolutionary distances between sequences of extant
species was measured by calculating the proportion of amino acid difference
between the sequences compared (
23
); positions where gaps are present in any of the aligned sequence were excluded
from the analysis. Based on the evolutionary distance matrix, a phylogenetic
tree was inferred by neighbour joining method. To obtain the reliability of the
inferred phylogenetic tree, the bootstrap method (
24
) was also applied. The bootstrap resamplings were repeated 1000 times, and for
each resampling a tree was inferred. The bootstrap probability that a
particular tree topology occurs during the resampling was evaluated. Amino acid
sequences in the conserved main central domain was used for the calculation
with
Pk
-REC using the ODEN program (National Institute of Genetics, Mishima, Japan) and Bioresearch/Sinca
(Fujitsu, Tokyo, Japan).
Escherichia coli
strain HMS174(DE3)pLysS carrying the expression plasmid containing
Pk-rec
gene was grown overnight at 37oC in NZCYM medium containing ampicillin. The preculture was inoculated (1%)
into fresh NZCYM medium and the cultivation was continued till the optical
density at 660 nm reached 0.35. The culture was then induced with 1 mM (final
concentration) IPTG and incubated for another 4 h at 37oC. Cells were harvested by centrifugation at 6000
g
for 10 min and washed with 50 mM sodium phosphate buffer (pH 7.0). The cell pellet was resuspended in the same buffer and cells were disrupted by sonication. Soluble and insoluble fractions were separated by centrifugation (15 000
g
for 30 min). The recombinant
Pk
-REC was recovered mainly in soluble fraction and purified by HiTrap Q,
Mono Q and HiTrap Heparin affinity columns (Pharmacia) with NaCl/50 mM sodium
phosphate buffer of increasing NaCl molarity for elution. All purification
steps were performed at 4oC. Purity of the protein was examined by SDS-PAGE using the Phast system and Phast gel (Pharmacia). A mixture of [rabbit muscle phosphorylase b (94 000), bovine serum albumin (67 000), egg white ovalbumin (43 000), bovine erythrocyte carbonic anhydrate (30 000), soya bean
trypsin inhibitor (20 100) and bovine milk [alpha]-lactalbumin (14 400)] was used for molecular weight standards. Protein concentration was
determined with bicinchoninic acid (BCA) protein assay kit (Pierce, Rockford, IL) according to the manufacturer's instructions using bovine serum albumin as a standard.
Figure
Figure
The molecular mass of the native enzyme was determined by analytical gel filtration chromatography using Superose 6 [HR 10/30] column
(Pharmacia). Molecular weight standards [bovine thyroglobulin (670 000), bovine [gamma]-globin (158 000), chicken ovalbumin (44 000), horse myoglobin (17
000) and vitamin B12 (1500)] were purchased from BioRad (BioRad Lab., Hercules, CA).
Figure
Deoxyribonuclease (DNase) reaction mixture (30 [mu]l), containing double ([Phi]X174 vector plasmid) or single (M13 mp19) stranded, linear or
circular DNA (5 [mu]g) and purified recombinant
Pk
-REC (0.1 [mu]g), were incubated at 60oC for 10 min in 30 mM Tris-HCl (pH 8.0) and 10 mM MgCl
2
. For each reaction sample, 5 [mu]l sample loading buffer [40 mM Tris-acetate (pH 8.0), 1 mM EDTA, 50% glycerol and a trace of bromophenol
blue] was added, and stored on ice to stop the reaction, and then analyzed by
1% agarose gel electrophoresis in Tris-acetate/EDTA buffer or 15% native polyacrylamide gel in Tris-borate/EDTA buffer (
22
). DNA was visualized under ultraviolet light by ethidium bromide staining. DNase activity was also determined by
measuring the liberated acid soluble oligo- or mononucleotides (
25
). For quantitative assay, the reaction mixture contained 50 [mu]g DNA, 1 [mu]g
Pk
-REC, 30 mM Tris-HCl (pH 8.0), 10 mM MgCl
2
, and sterilized water to make a total volume of mixture 60 [mu]l. The reaction was carried out at 60oC and stopped after 30 min by the addition of 60 [mu]l acid lanthanum reagent composed of 0.02 M La(NO
3
)
3
in 0.2 N HCl. The precipitates were removed by centrifugation, and the
absorbance of the supernatant solution was measured by a model DU 260
spectrophotometer (Beckman, Fullerton, CA) at 260 nm against a blank sample
incubated without enzyme. The reaction product of DNase activity was also
analyzed by HPLC using Wakopak WS-DNA (4.6 * 150mm, Wako, Osaka, Japan) column with increasing concentration of acetonitrile in 0.1 M triethylamine-acetate (TEAA) buffer at pH 7.0 (50/50 v/v). Deoxyadenosine monophosphate
(dAMP), deoxyadnyryl deoxyadenosine (dA-dA) and a randomly synthesized 27 nucleotide oligomer were used as size
standards. Absorbance was monitored at 260 nm.
ATPase activity of
Pk
-REC was assayed in a reaction mixture containing 20 mM Tris-HCl (pH 8.0), 2 mM DTT, 0.1 mM ATP, 2 mM MgCl
2
, 100 [mu]g/ml BSA, 0.1 [mu]g
Pk
-REC and water to make a final volume of 25 [mu]l. Each reaction contained 100 nCi [[alpha]-
32
P]ATP. Reaction mixtures were prepared on ice and incubations were at 55oC for 30 min followed by storage on ice: 1 [mu]l was then spotted directly onto Polygramr CEL 300 PEI thin layer chromatography plates (Macherey-Nagel, GmbH & Co., Duren, Germany). The substrate and products of the reaction were separated by one
dimensional chromatography using 1 M LiCl. The reaction product of ATPase
activity was also examined by HPLC using TSKgel DEAE-2SW (4.6 * 250 mm) column (Tosoh, Tokyo, Japan) using an increasing molarity
of sodium phosphate buffer (pH 3.0) in acetonitrile (80/20 v/v). Adenosine
triphosphate (ATP), adenosine diphosphate (ADP) and adenosine monophosphate
(AMP) were used as standards. Absorbance was measured at 260 nm.
The metal ions in the purified recombinant
Pk
-REC solution were removed by chelating with 20 mM EDTA and subsequent
dialysis for 36 h at 4oC against MilliQ water. DNase and ATPase activities of the dialyzed protein
were examined as described above with the slight modification that Mg
2+
was replaced by Ca
2+
, Cd
2+
, Co
2+
, Cu
2+
, Fe
2+
, Mn
2+
, Ni
2+
, Pb
2+
, Sr
2+
or Zn
2 +
for DNase activity, and by Co
2+
or Zn
2+
for ATPase activity. The salts used to analyze the effect of metal ions on the
enzymatic activities were CaCl
2
, CdCl
2
, CoCl
2
, CuSO
4
, FeSO
4
, MgCl
2
, MnCl
2
, NiCl
2
, PbCl
2
, SrCl
2
and ZnCl
2
.
The phylogenetic tree was constructed by comparing the amino acid sequence of
Pk
-REC with those of known RecA/RAD51 homologues from bacteria, eukaryotes
and other archaea. This phylogenetic tree, shown in Figure
1
, reveals that KOD1 obviously diverged from other organisms before eukaryotes,
bacteria and other archaea with an inclination for archaea. This result
suggests that
Pk-rec
gene is a common prototype of
recA
-like gene family.
Figure
Figure
In order to characterize the enzymatic properties of
Pk
-REC, the encoding gene was overexpressed by utilizing the T7 promoter
expression system of
E.coli
strain HMS174(DE3)pLysS. Gene expression was induced with 1 mM IPTG at 37oC and
Pk
-REC was successfully produced in soluble form. Most of the host proteins
were precipitated by heat treatment at 80oC for 15 min. The
Pk
-REC protein, in the soluble fraction after heat treatment, was purified by
ion exchange and affinity chromatography. The protein was eluted at 0.35 M NaCl
(pH 7.0) through MonoQ column and at 0.45 M NaCl (pH 7.0) through HiTrap
Heparin affinity column. The purified recombinant
Pk
-REC protein gave a single band on SDS-PAGE (Fig.
2
a). The molecular weight of the protein estimated from SDS-PAGE is ~28 kDa while the weight calculated from amino acid sequence is 24
kDa. In order to examine whether the protein exists in monomeric form, the
molecular weight of the protein in the native state was analyzed by gel
filtration chromatography, as well. It was determined to be 49 kDa (Fig.
2
b), suggesting that
Pk
-REC exists as a dimer.
When the purified recombinant
Pk
-REC was tested for its DNA binding activity, a common and basic
characteristic of RecA protein family, we found that it has DNA degrading
(DNase) activity. DNase activity was detected both for single- and double-stranded DNAs (Fig.
3
). It was also observed that DNase activity of
Pk
-REC is independent of ATP.
Pk
-REC has no detectable bias for either double-stranded or single-stranded DNA. When RNA, instead of DNA, was used as a
substrate, no nuclease activity was observed. Enzyme activity was further
assayed at various temperatures and 60oC proved to be the optimum (Fig.
4
a). Heat stability of the enzyme was examined at 60 and 80oC and it was found that the enzyme is rather heat labile having an
enzymatic half life of 20 min for both 60 and 80 oC (Fig.
4
b). The pH profile of
Pk
-REC activity showed a broad plateau extending from pH 7.5 to 9 with
maximal activity at pH 8-8.5 (Fig.
4
c). The final reaction product of
Pk
-REC DNase activity appeared to be mononucleotides. This was observed in
the reaction using a 27 bp double-stranded DNA fragment as a substrate and analyzing the reaction product on
15% native polyacrylamide gel (data not shown) as well as on HPLC (Fig.
5
). When a dinucleotide (dA-dA) was used as a substrate, no activity was observed. Concerning the
ionic strength, DNase activity of
Pk
-REC gradually decreased as the concentration of Tris-HCl in the reaction mixture increased beyond 60 mM. It was fully
inhibited in the presence of 200 mM Tris-HCl. When the enzyme activity was examined in the presence of Ca
2+
, Co
2+
, Cd
2+
, Cu
2+
, Fe
2+
, Mn
2+
, Mg
2+
, Ni
2+
, Pb
2+
, Sr
2+
and Zn
2+
,
Pk
-REC exhibited DNase activity to a variant degree in the presence of all
these tested metal ions except for Zn
2+
. No DNase activity, even no binding, was observed in the presence of Zn
2+
(Fig.
6
). In order to confirm that the observed DNase activity is not derived from any
undetectable contaminant from the host cells, the comparable amounts of host
cells carrying the vector only were treated in the same way as the cells
containing the
Pk-rec
gene. The crude lysate was treated for 15 min at 80oC and the resulting supernatant was assayed for DNase activity. No DNase
activity was observed.
ATP hydrolyzing activity by
Pk
-REC was compared in the presence and absence of DNA and no detectable
difference was observed. These results indicate that, unlike other RecA
proteins, the ATPase activity of
Pk
-REC is independent of both single-stranded and double-stranded DNAs. The reaction product of
Pk
-REC activity was migrated on TLC plates with a
R
f
value equivalent to AMP (Fig.
7
A). The reaction product was further analyzed by HPLC and there it also eluted
with a retention time completely matching AMP (Fig.
7
B). Some of the divalent metal cations (Co
2+
, Mg
2+
and Zn
2+
) were compared as a metal cofactor for
Pk
-REC ATPase activity. Of them Zn
2+
proved to be the most effective (data not shown).
The gene locus encoding
Pk
-REC was found adjacent to the 16S rRNA and ribosephosphate pyrophosphokinase genes (unpublished data) on the KOD1 genome.
Pk
-REC consists of only a central domain of proteins in this family from
bacteria, eukarya, and even from other archaea (
11
) and is almost equally apart from bacteria and eukarya with slight inclination
towards the latter. Putative proteins named RadA from acidophilic, halophilic and methanogenic hyperthermophilic archaea are inclined to a higher degree towards eukarya (Fig.
1
) (
10
) and are larger in size as compared to
Pk
-REC (
11
). The amino acid sequence comparison suggested
Pk
-REC to be a common prototype of RecA/RAD51 group proteins. RecA family
members usually exist in aggregates and N-terminal domain in RecA is reported to contribute for its self assembly (
19
,
26
). Since
Pk
-REC lacks the N-terminal domain of RecA, it is understandable that it does not
aggregate in a multimeric form but instead forms a dimer. Since the previous
work on the C-terminal truncated RecA protein showed enhanced DNA binding activities (
20
), the comparison of
Pk
-REC with RecA suggested it to have high RecA like activity. In fact, it
complemented
recA
defect of
E.coli in vivo
(
11
). When the purified recombinant
Pk
-REC protein was examined for its DNA binding activity, we found that it
had DNase activity and this activity was specific for deoxyribose sugar since
no activity was observed when RNA was used as a substrate. The optimum
temperature for enzyme activity was shown to be 60oC, although the optimum growth temperature of the strain KOD1 is 95oC. This suggests the existence of a protein stabilizing system in KOD1
cells. The reaction product of DNase activity was shown to be a mononucleotide, indicating that
Pk
-REC cleaves the phosphodiester bond non-specifically. When a dinucleotide (dA-dA) was used as a substrate, no cleavage was observed, probably
because the reaction rate was dramatically reduced. This phenomenon is a common characteristic of other non specific DNases.
It was found that Mg
2+
can be replaced by other divalent cations, such as Ca
2+
, Cd
2+
, Cu
2+
, Mn
2+
, Ni
2+
, Pb
2+
and Sr
2+
but DNase activity was reduced to some extent, however when Mg
2+
was substituted with Co
2+
or Fe
2+
no detectable change was observed. It should be noted that the protein exhibits
little DNase activity in the presence of Zn
2+
ion. Because the protein exhibits a very weak DNase activity even in the
absence of any metal ion, it seems likely that Zn
2+
ion binds to the Mg
2+
binding site of the protein and inhibits the DNase activity.
ATP hydrolysis is a common characteristic of the RecA protein family with ADP as
the reaction product and DNA is usually required as a cofactor (
27
). Moreover, it was shown that the ATP hydrolyzing reaction by
Pk
-REC is independent of DNA. Likewise, the DNase activity does not depend on
ATP. The analyses of metal ion requirement showed that Zn
2+
was not utilized as a metal cofactor for DNase activity but was the most
effective metal cofactor for the ATPase activity. This may suggest that the
protein possesses two different catalytic sites for DNase and ATPase
activities. Alternatively, these catalytic sites may be overlapped and the
protein possesses only one metal binding site. Preference of the enzyme for
metal ions may vary for different substrates. The DNase activity is not due to
contamination of nucleases from the
E.coli
cells, since the cell lysate prepared from
E.coli
HMS174(DE3)pLysS in the similar way as
Pk
-REC has no DNase activity. Furthermore, when double amino acid replacements were introduced in the ATPase active site (positions correspond to Lys
72 and Thr 73 in
E.coli
RecA), the mutated protein lost all ATPase activity and showed a dramatic
decrease in DNase activity (20-fold decrease in
V
max
) (personal communication). This result strongly demonstrates that
Pk
-REC has both ATPase and DNase activities and the active sites of these
activities seems closely related. Another unique characteristic of
Pk
-REC was its reaction product of ATPase activity. The reaction product of
RecA like ATPase activity from both bacteria and eukarya is ADP while the major
reaction product of
Pk
-REC ATPase activity is AMP. Since ADP is also detected as a minor product,
Pk
-REC may be able to cleave both sites of the [beta]-phosphate group in ATP. Previously,
Pk
-REC was shown to complement the UV sensitivity of an
E.coli recA
null mutant strain. All of these results demonstrate that
Pk
-REC is highly functional both
in vivo
and
in vitro
. Archaea in general and hyperthermophilic archaea in particular, are thought to
evolve from, and still live in environments very similar to that of primitive
earth. Cells at that time might have been exposed to harsh conditions including
fairly strong UV light. Under this stress it would be essential to maintain
effective DNA repair systems. From the perspective of evolution, the primitive
highly thermal environments might have given rise to this unique RecA like
characteristic followed by the transfer of DNase activity to other enzymes with the reduction of growth temperatures and the development of metabolic systems.
*To whom correspondence should be addressed. Tel: +81 6 879 7440; Fax: +81 6 879
7441; Email: t.imanaka@cell.bio.eng.osaka-u.ac.jp





REFERENCES
Return
