(Received for publication, January 31, 1997, and in revised form, May 12, 1997)
From the Département de Microbiologie et Immunologie, Université de Montréal, Montréal, Québec H3C 3J7, Canada
Previous studies have shown that the reovirus
1 core protein harbors a putative nucleotide-binding motif and
exhibits an affinity for nucleic acids. In addition, a nucleoside
triphosphate phosphohydrolase activity present in reovirus cores has
been recently assigned to
1 using gene reassortment analysis. In
this study, it was demonstrated that the recombinant
1 protein,
expressed in the yeast Pichia pastoris, is able to
hydrolyze nucleoside 5
-triphosphates or deoxynucleoside
5
-triphosphates. This activity was absolutely dependent on the
presence of a divalent cation, Mg2+ or Mn2+.
The protein can also unwind double-stranded nucleic acid molecules in
the presence of a nucleoside 5
-triphosphate or deoxynucleoside 5
-triphosphate. These results provide the first biochemical evidence that the reovirus
1 protein is a nucleoside triphosphate
phosphohydrolase/helicase and strongly support the idea that
1
participates in transcription of the viral genome.
Mammalian reoviruses are members of the Reoviridae
family, and since their genome is made up of 10 segments of
double-stranded RNA (dsRNA)1 and replicates
in the cytoplasm, they must encode their own transcriptional and
replicative enzymes (1). During reovirus infection, the viral genome
remains in the inner capsid (core) of the virus, composed of two major
(1 and
2) and two minor (
3 and µ2) proteins. Gene
reassortment experiments have resulted in the assignment of
transcriptase activity to
3 (2). Although functions of other core
proteins have not been firmly established, it is suspected that
additional enzymatic functions are needed to achieve transcription and
replication of the viral genome. For example, it has been postulated,
by analogy with other viruses, that a helicase function could be
present in the viral core (1).
Nucleic acid helicases unwind double-stranded DNA and/or RNA, a process
energetically coupled to the hydrolysis of nucleoside 5-triphosphates
(NTPs) or deoxynucleoside 5
-triphosphates (dNTPs) (3, 4). Helicases
play a key role in nucleic acid replication, transcription, splicing,
translocation, recombination, and repair (5-8). Helicases of
prokaryotic, eukaryotic, and viral origins have been isolated and
classified into defined superfamilies (9-14). These proteins are
characterized by seven conserved motifs designated I, Ia, and II-VI
(15). Motifs I and II are very well conserved and correspond to the A
and B consensus sequences of a nucleotide-binding domain (16).
Superfamily II includes an expanding group of DNA and RNA helicases
that harbor a DEA(D/H) sequence in motif II (17). The sequences present
in motifs III-V are less strictly conserved, and their roles are not
clearly defined, whereas motif VI is supposed to be involved in nucleic
acid binding given its high content of positively charged amino acids
(13).
The 1 protein, a major component of the reovirus core, exhibits an
affinity for dsRNA and dsDNA in filter binding assays and can also bind
single-stranded RNA in gel retardation assays (18).2 Furthermore, analysis of gene
reassortment has recently resulted in the assignment of NTPase activity
present in reovirus cores to the L3 gene encoding
1 (19). In this
study, we present a biochemical characterization of the enzymatic
activities exhibited by the
1 protein encoded by the cloned L3 gene.
The protein was expressed in the yeast Pichia pastoris and
recovered by chromatography using the affinity of the protein for zinc,
which is conferred by its zinc-finger motif. This protein was able to
hydrolyze all NTPs and dNTPs with release of inorganic phosphate;
however, different NTPs and dNTPs are utilized with various
efficiencies. The protein can also unwind double-stranded nucleic acid
molecules; this reaction requires the presence of a NTP or dNTP at a
concentration consistent with a functional coupling between hydrolysis
of nucleotide and helicase activity. These findings strongly support
the idea that
1 participates as a helicase during transcription of
the viral genome.
The methylotrophic yeast
P. pastoris strain GS115 (his4) was used for the
expression of 1. All manipulations were performed according to the
manufacturer's instructions (Invitrogen). The 5
-end of the cloned L3
gene encoding
1 was first reconstructed by polymerase chain reaction
to remove homopolymers introduced in the original cloning procedure
(20). The
1 expression plasmid pHIL-L3 was then constructed by
inserting the complete L3 gene of reovirus serotype 3 at the unique
EcoRI site of the pHIL-D2 expression vector. The pHIL-D2
expression vector contains a HIS4 gene, and its unique
EcoRI site is flanked by the 5
- and 3
-regulatory sequences
of the methanol-inducible alcohol oxidase gene (AOX1) of
P. pastoris. Yeast cells were transformed with linear
(NotI-digested) DNA of pHIL-D2 or pHIL-L3 and selected on
histidine-deficient medium. Transformants were identified, liquid
cultures were prepared, and protein expression was induced by the
addition of 0.5% methanol for 20 h at 30 °C. The yeast cells
were recovered by centrifugation and disrupted with glass beads, and
the lysates (1 ml) were submitted to affinity chromatography on zinc
chelate affinity adsorbent columns (Boehringer Mannheim). The columns
were washed extensively (100 ml) with wash buffer (50 mM
Tris-HCl, pH 8.0, and 150 mM NaCl), and the bound proteins
were eluted sequentially with 1 ml of elution buffers, each made of 0.1 M sodium phosphate and 0.8 M NaCl and adjusted
at pH 7.5, 7.0, 6.5, 6.0, or 5.5 (21). Protein dosage was performed
using the Bradford protein assay (Bio-Rad).
Proteins were submitted to
electrophoresis on SDS-10% polyacrylamide gels. Following
electrophoresis, the proteins were transferred to a nitrocellulose
filter and probed with a 1:1000 dilution of monospecific anti-1
antibody (a generous gift of Dr. Michael R. Roner, Florida Atlantic
University). The antigen-antibody complexes were detected using a goat
anti-rabbit IgG conjugated to horseradish peroxidase and the Amersham
enhanced chemiluminescence Western blotting kit.
Standard reactions (22) were
performed in a buffer containing 30 mM Hepes-KOH, 2 mM MgCl2, and 0.2 pmol of
[-32P]ATP (4500 Ci/mmol; ICN). Approximately 18 fmol
(2.5 ng) of
1 protein were used in each 15-µl reaction when an
enriched
1 fraction recovered from zinc affinity chromatography was
used. The reactions were incubated at various temperatures and stopped
by the addition of 0.1 M EDTA at the times indicated. The
ATPase activity was detected by the release of
32Pi from [
-32P]ATP. Reaction
products (2 µl) were applied onto plastic-backed polyethyleneimine
cellulose sheets (Aldrich) and separated by ascending chromatography in
0.375 M potassium phosphate buffer, pH 3.5. The sheets were
then air-dried and subjected to autoradiography.
The Michaelis-Menten constant (Km) was determined by
the isotopic dilution method with unlabeled ATP. A similar experiment was performed with [-32P]GTP (4500 Ci/mmol; ICN) and
unlabeled GTP. The determination of Km for dATP and
dCTP was performed with 0.2 pmol of [
-32P]dATP (3000 Ci/mmol; ICN) or [
-32P]dCTP (3000 Ci/mmol; ICN) using
isotopic dilution with unlabeled dATP and dCTP, respectively. In these
two cases, deoxynucleoside triphosphate substrates were separated from
deoxynucleoside diphosphate reaction products by polyethyleneimine
cellulose chromatography in 0.8 M acetic acid and 0.9 M LiCl buffer.
The spots corresponding to the radiolabeled substrates or reaction products were identified following autoradiography; the corresponding regions were then excised from the polyethyleneimine cellulose sheets; and the radioactivity was measured by Cerenkov counting. For calculations, background values were first subtracted from product values by quantitation of radioactivity at the same level on chromatograms of control unincubated substrate. Thereafter, the ratio of generated products to total material was calculated by quantitation of both reaction products and residual substrate in each lane.
Synthesis of Helicase SubstratesThe substrate for the DNA
helicase assay was prepared by annealing the 17-base pair universal
primer (5-GTAAAACGACGGCCAGT-3
) to single-stranded M13 DNA. Chain
elongation was then performed with Sequenase (U. S. Biochemical Corp.)
for 30 min at 37 °C in the presence of dGTP, dTTP, and
[
-32P]dATP. The resulting substrate contained an
extended radiolabeled oligomer of 22 nucleotides hybridized to M13 DNA
and was purified by gel filtration (Sephadex G-25) to remove most of
the unincorporated nucleotides.
The substrate for RNA helicase activity was prepared as a partial RNA
duplex using pBluescript II SK+ (Stratagene), which
contains the T3 and T7 polymerase promoters. The first strand (59 nucleotides) was prepared by in vitro transcription with T3
RNA polymerase (Pharmacia Biotech Inc.) of the SmaI-digested plasmid. The labeled strand (89 nucleotides) was synthesized from the
XbaI-digested plasmid in the presence of
[-32P]UTP with T7 RNA polymerase (Pharmacia). The
original DNA templates were removed by DNase treatment (RQ1 DNase,
RNase-free), extracted in phenol/chloroform, and precipitated with
ethanol. The two RNA species were dissolved in a buffer containing 10 mM Tris-HCl, pH 7.5, and 200 mM NaCl. The RNAs
were then annealed by heating to 90 °C for 15 min and cooled to room
temperature to produce a tailed RNA molecule with a 17-base pair duplex
region.
In the standard assay (23), the DNA and RNA
substrates (0.3 pmol) were used in a total volume of 15 µl containing
25 mM Hepes-KOH, pH 7.5, 5 mM
MgCl2, 20 mM NaCl, 1 mM
dithiothreitol, and 5 mM ATP. Approximately 18 fmol (2.5 ng) of 1 protein were used in each 15-µl reaction when the
enriched
1 fraction recovered from zinc affinity chromatography was
used; incubation was for 30 min at 37 °C. Components were added or
removed in some reactions, as described below. The reactions were
stopped by adding 0.5% SDS, 50 mM EDTA, and 40% glycerol.
The reaction products were analyzed by electrophoresis on 15% (DNA) or
10% (RNA) nondenaturing Tris borate/EDTA-polyacrylamide gels. The gels
were dried under vacuum and exposed for autoradiography.
Sequence
comparisons revealed that 1 possesses two nucleotide-binding motifs
normally present in NTPase (Fig. 1): a PRKTKGKS sequence
(A site) in the N-terminal region and a DEAD motif (B site).
Furthermore,
1 harbors the characteristic motifs found in members of
the RNA/DNA helicase superfamily II, with slight variations occurring
in certain motifs. To study these putative enzymatic activities, the
1 protein encoded by the cloned L3 gene was expressed using the
P. pastoris yeast expression system as described under
"Experimental Procedures." Enrichment of the protein and removal of
endogenous yeast ATPases were then achieved by affinity chromatography,
taking advantage of the presence of a putative zinc-binding finger
motif on
1.
The expression level of 1 was too low for detection by either
Coomassie Blue or silver staining in the purified fractions (data not
shown). However, immunoblotting analysis, using a monospecific anti-
1 antibody, did reveal the enrichment of an immunoreactive protein in the eluted fractions; this protein comigrated with the
authentic
1 protein found in purified reovirions (Fig.
2A). The amount of
1 protein in these
fractions was estimated to be ~0.5 ng/µl by comparisons with the
signal obtained by immunoblotting using serial dilutions of purified
reovirus as a standard. Protein dosage also revealed that enrichment of
the protein was ~10-fold compared with the crude lysate (Table
I).
|
The 1 protein expressed in P. pastoris was first tested
for its ability to hydrolyze ATP by monitoring the amount of free radioactive inorganic phosphate liberated upon incubation of the protein fractions with [
-32P]ATP. As shown in Fig. 2
(B and C), ATPase activity endogenous to P. pastoris was found in the flow-through and first wash fractions of
the zinc column, but dropped to undetectable levels in later wash
fractions. Following elution at lower pH, ATPase activity was recovered
from cultures harboring the
1 expression vector (Fig.
2C), whereas equivalent fractions from control cultures harboring vector with no insert were devoid of ATPase activity (Fig.
2B). The same results were obtained from five independent negative cultures and two separate cultures expressing recombinant
1
(data not shown). A summary of the
1 enrichment procedure is
presented in Table I; the eluate fraction obtained at pH 7.5 was used
in the rest of this work.
In an effort to further rule out any possible contamination with a
copurifying yeast enzyme, thermoresistance of the NTPase activity was examined (Fig. 3). The approach used takes
advantage of the fact that reovirus core enzymes are active at high
temperature, whereas P. pastoris has an optimal growth
temperature of 30 °C, and its enzymes are thus expected to be
thermosensitive. Cell extracts from yeast cells harboring the control
vector (pHIL-D2) or 1 expression vector (pHIL-L3) were thus examined
for ATPase activity prior to or following heating at 42 °C for an
extended period (5 h). This treatment essentially abolished the yeast
endogenous ATPase activity (Fig. 3A), whereas part of the
activity was retained in lysates from
1-expressing yeast cells (Fig.
3B); the ATPase activity present in
1-enriched fractions
from zinc chelate affinity chromatography was completely resistant to
thermal inactivation (Fig. 3D), whereas no activity was
found in eluate fractions from yeast cells harboring the control vector
pHIL-D2 (Fig. 3C).
Characterization of
The
enzymatic activity of recombinant 1 found in elution fractions was
then further investigated. This activity was absolutely dependent on
the presence of the divalent cation Mg2+ or
Mn2+, whereas the Ca2+, Cu2+, and
Zn2+ cations were not effective cofactors (data not shown).
The ATPase activity increased sharply with MgCl2
concentration, reached a maximum at 2.5 mM, and was
constant up to 10 mM (Fig. 4A).
Similar results were obtained when Mn2+ was substituted for
Mg2+ (data not shown).
To gain additional insight into the 1-associated ATPase
activity, the effect of temperature on the reaction was examined (Fig.
4B). The temperature optimum of the reaction, as judged by
the maximum rate of ATP hydrolysis, was 42-50 °C; however, the
activity was rapidly lost upon incubation at 50 °C. A 1-h incubation
at 42, 37, and 30 °C resulted in similar rates of ATP hydrolysis,
whereas the enzyme was less active at 25 °C.
The kinetics parameters of the reaction were determined at 37 °C and 10 min, at which time the reaction is still proceeding at its initial rate. The ATPase reaction velocity reached a plateau at ~7 µM ATP and exhibited a Km of 1 µM as determined by a double-reciprocal plot (Fig. 4C). Identical results were obtained in three separate experiments.
To determine the substrate specificity of the 1 protein, the
hydrolysis of other ribonucleotides and deoxyribonucleotides was also
tested. The protein exhibited a strong preference for adenosine
nucleotides (ATP and dATP), which were hydrolyzed very efficiently. The
enzyme displayed some activity on other NTPs and dNTPs tested, although
the efficiency was different from one substrate to the other; the order
of preference was dATP > ATP > GTP > dCTP according
to the kcat values obtained with these substrates (Table II).
|
The ability of the
1 protein to unwind double-stranded nucleic acids was then
investigated. The RNA helicase reaction catalyzed by the reovirus
1
protein was demonstrated by strand displacement of a partial RNA duplex
as described under "Experimental Procedures." The
1-associated
RNA helicase activity required the presence of ATP, which could be
substituted by either one of the NTPs and dNTPs (Fig.
5A) (data not shown). The unwinding activity
was also dependent on the presence of a divalent cation
(Mg2+ or Mn2+), as was the NTPase/dNTPase
activity.
The helicase activity on a DNA substrate was also tested by the ability
of the protein to displace (unwind) a short oligonucleotide annealed to
single-stranded M13 DNA. As shown in Fig. 5B, the 1
protein is able to unwind the partial DNA duplex in the presence of
either ATP or dATP. This
1-associated DNA helicase activity could
also be stimulated by all NTPs and dNTPs (data not shown) and had the
same requirement for divalent cations as the RNA helicase activity.
Additional controls were also performed to ensure that the helicase
activity could not be due to a copurifying yeast protein. Crude lysate
from control yeast cells was examined for the presence of helicase
activity and was found to be negative. Furthermore, the enriched eluate
from the zinc affinity column was also negative in the control cells,
whereas the activity found in purified 1 eluate fractions was
thermoresistant (Fig. 5C), as was the NTPase activity.
Our understanding of the roles of individual core proteins
in reovirus replication is still incomplete, although genetic
information has been accumulated during the past few years (24, 25).
Investigation of the biochemical activity of the reovirus 1 protein
was inspired by sequence comparisons with other NTPase/helicase
proteins. Although the
1 protein does not show any significant
similarities to other proteins, it possesses the characteristic motifs
found in the DEAD subfamily of the RNA/DNA helicase superfamily II. All
these motifs are located in the amino-terminal third of
1; the
affinity of the protein for nucleic acids was previously assigned to
this region (18). Being one of the largest reovirus proteins, the
1
protein could not be easily expressed in bacterial expression systems.
The recently described P. pastoris yeast expression system was instead used as a convenient way to produce recombinant
1. The
expression levels obtained were low but sufficient for investigations of
1 catalytic activities. The reason for this low expression level
is unknown, but might reflect toxicity of NTPase activity when
overexpressed, thus resulting in the selection of cell clones expressing lower levels of the protein. Expression levels using P. pastoris have been reported to vary widely among proteins
(Invitrogen). Since a truncated form of
1 encompassing the
zinc-finger motif was previously expressed as a fusion protein in
bacterial cells (18) and found to attach to zinc chelate
columns,3 this procedure was thus adequate
for
1 recovery. The zinc affinity chromatography procedure allowed
us to efficiently reduce endogenous NTPase and helicase activities
below the detection level of our assays while enriching both
immunodetectable
1 and its associated NTPase activity by ~10-fold.
Thermoresistance was an additional proof that the NTPase activity
measured could not be due to a contaminating yeast protein; although it
is expected that NTPase activity could be abundant in yeast cells,
those endogenous activities were completely inactivated upon extensive
heat treatment.
All four NTPs and dNTPs tested were hydrolyzed by 1, and it is most
likely that all NTPs and dNTPs can be hydrolyzed since either one of
them can substitute for ATP in the helicase assay. The efficiency of
utilization appears to vary among the different NTPs and dNTPs, with
the highest rate of hydrolysis for dATP, followed by ATP and other
NTPs. However, it should be mentioned that even though all
Km values are in the micromolar range, they are
lower for NTPs than for dNTPs. At a low concentration of substrates, as
might be the case inside the viral core in vivo, ATP will
thus likely be the preferred energy source.
A NTPase activity has been previously found in purified reovirus cores
(19, 26, 27), and the activity has been recently assigned to 1 using
gene reassortment analysis (19). Interestingly, the core enzyme is
capable of hydrolyzing either NTPs or dNTPs in the presence of
Mg2+ or Mn2+ and also has an unusual
temperature optimum of 51 °C (27, 28). The observation that dATP was
more efficiently hydrolyzed relative to ATP at high substrate
concentration was also made in previous studies (19, 27). The catalytic
properties of the recombinant
1 protein analyzed in this study are
consistent with these previous findings and further support the idea
that
1 is the protein responsible for the NTPase activity observed
in reovirus cores.
The recombinant 1-associated NTPase/dNTPase had an apparent
temperature optimum of 50 °C, but the activity declined rapidly at
this high temperature, most likely due to thermal inactivation of the
enzyme. Recent studies indicated that
1 interacts with the
3
(polymerase) and
2 (guanylyltransferase) core proteins (29, 30);
such interactions might account for the higher stability of NTPase
activity found in purified viral cores compared with the isolated
recombinant
1 protein.
The Km values for ATP and GTP hydrolysis calculated
from Lineweaver-Burk plots were 1 and 2 µM, respectively.
These Km values are in the same range observed for
the well studied SV40 large T antigen, which is also a DNA and RNA
helicase (31). To date, only three other viral helicases have been
shown to possess both DNA and RNA helicase activities; they are SV40 large T antigen, vaccinia virus protein 18R, and hepatitis C virus protein NS3 (32-34). One cellular eukaryotic protein, nuclear DNA helicase II, also possesses the ability to unwind dsDNA and dsRNA substrates (35). The reovirus 1 protein shares with these four helicases the ability to hydrolyze all NTPs and dNTPs. However, SV40
large T antigen unwinds dsDNA when ATP is present, but unwinds dsRNA
with UTP, CTP, or GTP as cofactor (32). The bound nucleotide seems to
determine whether the T antigen acts as an RNA or DNA helicase. In
contrast, the reovirus
1 protein and hepatitis C virus protein NS3
can use the energy provided by the hydrolysis of any NTPs or dNTPs for
both their RNA and DNA helicase activities (34).
Many viral helicases have high basal NTPase activity in the absence of
added nucleic acids. This is the case of the
potyvirus-flavivirus-pestivirus proteins that have been recently
characterized (36-39). In this respect, they differ from most cellular
proteins of the NTPase/helicase superfamilies. For example, cellular
protein p68 and eukaryotic translation initiation factor 4A exhibit
almost no detectable ATPase activity in the absence of RNA (40, 41).
Our results demonstrated that the reovirus 1 protein, like many
viral NTPases, also exhibits NTPase and dNTPase activities in the
absence of added nucleic acids. Although it cannot be completely
excluded that nucleic acids can stimulate NTPase activity, the addition of various nucleic acids (single-stranded DNA, dsDNA, single-stranded RNA, and dsRNA) has failed to produce a significant change in kcat values for NTPase activity (data not
shown).
The helicase activity of 1 was not detected at an ATP concentration
of 0.1 µM and was maximal at 10 µM (data
not shown). This is consistent with the Km value of
1 µM for ATPase activity and with a probable functional
coupling between the two reactions where NTP or dNTP hydrolysis
generates the energy required for unwinding of nucleic acids.
The significance, if any, of the DNA helicase activity of the 1
protein is currently not understood, but a previous report already
demonstrated the
1 affinity for dsRNA and dsDNA in filter binding
assays (18). As previously mentioned, the NS3 protein from hepatitis C
virus is another example of an RNA virus protein with an associated DNA
helicase activity. Like the reovirus
1 protein, the role of the DNA
helicase activity in the multiplication cycle or pathogenesis of
hepatitis C virus is not yet established (34).
In addition to earlier genetic evidence that assigned transcriptase
activity to 3, more recent biochemical evidence has confirmed that
this protein possesses RNA polymerase activity (29). However, the
protein appears to be unable by itself to transcribe its natural dsRNA
substrate. This has led to the idea that additional proteins are
involved in transcription. The
1-associated RNA helicase activity
reported in this study strongly suggests an involvement of this protein
in the transcription of the double-stranded reovirus genome.
Although it is possible that normal transcription and replication of
the viral genome require the formation of a well structured core or the
participation of cellular protein(s), this report is the third case,
with the previous examples of 2 and
3 (29, 42), where an actual
enzymatic activity could be demonstrated on a solubilized reovirus
protein. To our knowledge, this report of
1 helicase activity is the
first demonstration of such an activity in a mammalian dsRNA virus.
The reovirus core also contains enzymes that modify the 5-end of newly
synthesized mRNAs by adding a cap structure similar to the one
present on cellular mRNAs (43, 44). An RNA triphosphatase activity
(polynucleotide phosphohydrolase) is involved in the formation of the
reovirus cap structure by releasing inorganic phosphate from the
5
-triphosphate end of the nascent mRNAs (1). The nature of the
core protein involved in this process is currently unknown, but the
1 protein, with its ability to release the terminal phosphate from
free nucleotides, is certainly an attractive candidate to exert a
similar activity on polynucleotides harboring a 5
-triphosphate end.
We thank Carole Danis for expert technical
assistance and Dr. Pierre Belhumeur for suggesting the use of the yeast
expression system. We also thank Dr. Michael R. Roner for the generous
gift of the anti-1 antibody.