Characterization of an ATP-dependent DNA ligase encoded by Chlorella virus PBCV-1

Abstract

We report that Chlorella virus PBCV-1 encodes a 298-amino-acid ATP-dependent DNA ligase. The PBCV-1 enzyme is the smallest member of the covalent nucleotidyl transferase superfamily, which includes the ATP-dependent polynucleotide ligases and the GTP-dependent RNA capping enzymes. The specificity of PBCV-1 DNA ligase was investigated by using purified recombinant protein. The enzyme catalyzed efficient strand joining on a singly nicked DNA in the presence of magnesium and ATP (Km, 75 microM). Other nucleoside triphosphates or deoxynucleoside triphosphates could not substitute for ATP. PBCV-1 ligase was unable to ligate across a 2-nucleotide gap and ligated poorly across a 1-nucleotide gap. A native gel mobility shift assay showed that PBCV-1 DNA ligase discriminated between nicked and gapped DNAs at the substrate-binding step. These findings underscore the importance of a properly positioned 3' OH acceptor terminus in substrate recognition and reaction chemistry.

Full text

JOURNAL OF VIROLOGY,
0022-538X/97/$04.0010
Mar. 1997, p. 1931–1937 Vol. 71, No. 3
Copyright q1997, American Society for Microbiology
Characterization of an ATP-Dependent DNA Ligase Encoded
by Chlorella Virus PBCV-1
C. KIONG HO,
1
JAMES L. VAN ETTEN,
2
AND STEWART SHUMAN
1
*
Molecular Biology Program, Sloan-Kettering Institute, New York, New York 10021,
1
and Department of
Plant Pathology, University of Nebraska, Lincoln, Nebraska 68583-0722
2
Received 4 November 1996/Accepted 10 December 1996
We report that Chlorella virus PBCV-1 encodes a 298-amino-acid ATP-dependent DNA ligase. The PBCV-1
enzyme is the smallest member of the covalent nucleotidyl transferase superfamily, which includes the
ATP-dependent polynucleotide ligases and the GTP-dependent RNA capping enzymes. The specificity of
PBCV-1 DNA ligase was investigated by using purified recombinant protein. The enzyme catalyzed efficient
strand joining on a singly nicked DNA in the presence of magnesium and ATP (K
m
,75mM). Other nucleoside
triphosphates or deoxynucleoside triphosphates could not substitute for ATP. PBCV-1 ligase was unable to
ligate across a 2-nucleotide gap and ligated poorly across a 1-nucleotide gap. A native gel mobility shift assay
showed that PBCV-1 DNA ligase discriminated between nicked and gapped DNAs at the substrate-binding
step. These findings underscore the importance of a properly positioned 3*OH acceptor terminus in substrate
recognition and reaction chemistry.
The ATP-dependent DNA ligases catalyze the joining of 59
phosphate-terminated donor strands to 39hydroxyl-terminated
acceptor strands via three sequential nucleotidyl transfer reac-
tions (10). In the first step, nucleophilic attack on the a-phos-
phate of ATP by ligase results in liberation of pyrophosphate
(PP
i
) and formation of a covalent intermediate in which AMP
is linked to the ε-amino group of a lysine. The nucleotide is
then transferred to the 59end of the donor polynucleotide to
form DNA-adenylate—an inverted (59)-(59) bridge structure,
AppN. Attack by the 39OH of the acceptor strand on the
DNA-adenylate joins the two polynucleotides and liberates
AMP.
Animal cells contain multiple DNA ligase isozymes encoded
by separate genes (1, 28, 29). ATP-dependent DNA ligases are
also encoded by animal DNA viruses, e.g., the poxviruses and
African swine fever virus (ASFV), by the T-odd and T-even
bacteriophages (T4, T6, T3, and T7), by yeasts, plants, and
archaea (7). The ATP-dependent DNA ligases belong to a
superfamily of covalent nucleotidyl transferases that includes
the GTP-dependent eukaryotic mRNA capping enzymes (20,
22). The ligase-capping enzyme superfamily is defined by a set
of six short motifs (Fig. 1). The lysine within motif I (KxDG) is
the active site of AMP transfer by the ligases (6, 25, 28) and
GMP transfer by the capping enzymes (2, 4, 15, 18). Conserved
residues within motifs I, III, IV, and V are critical for covalent
nucleotidyl transfer, as shown by mutational analyses (2–4, 6, 8,
18, 20, 21). The recently reported crystal structure of T7 DNA
ligase shows that the ATP-binding site is made up of conserved
motifs I, III, IIIa, IV, and V (24).
The bacteriophage T7 and T3 enzymes (359- and 346-ami-
no-acid polypeptides, respectively) are the smallest ATP-de-
pendent ligases described to date (7). Cellular DNA ligases are
much larger; for example, human ligases I, III, and IV are 919-,
922-, and 844-amino-acid polypeptides, respectively (1, 29).
Vaccinia virus and ASFV ligases are intermediate in size (552
and 419 amino acids, respectively) (5, 23). Sequence compar-
isons of cellular and virus-encoded proteins suggest that a
catalytic domain common to all ATP-dependent ligases is em-
bellished by additional isozyme-specific protein segments situ-
ated at their amino or carboxyl termini. Virus-encoded en-
zymes, by virtue of their small size, may well define the catalytic
core of the ligase family.
Here, we report a new viral DNA ligase encoded by Para-
mecium bursaria Chlorella virus 1 (PBCV-1). PBCV-1 is the
prototype of a family of large, polyhedral DNA viruses that
replicate in unicellular eukaryotic Chlorella-like green algae
(27). The PBCV-1 genome, like the genomes of the poxviruses
and ASFV, is a linear, double-strand DNA molecule with in-
verted terminal repeats and covalently closed hairpin telo-
meres. The sequence of the 330-kbp PBCV-1 genome has been
determined (9, 11–13); PBCV-1 encodes ;380 polypeptides.
An open reading frame encoding a ligase-like protein was
encountered between nucleotide coordinates 264,000 and
265,000 of the PBCV-1 genome (11a; Genbank accession num-
ber is U77663). The predicted gene product includes the six
motifs shared among the cellular and DNA virus-encoded
ATP-dependent DNA ligases (Fig. 2). The order and spacing
of these motifs in the PBCV-1 ligase-like protein are similar to
those seen in other ligase family members (Fig. 1). The
PBCV-1 polypeptide, at 298 amino acids, is smaller than any
known ligase or capping enzyme. We have expressed the
PBCV-1 protein in Escherichia coli and purified it to apparent
homogeneity. We report that the recombinant protein is in-
deed an ATP-dependent DNA ligase. A biochemical charac-
terization of the PBCV-1 ligase is presented.
MATERIALS AND METHODS
T7-based vector for expression of PBCV-1 DNA ligase. Oligonucleotide prim-
ers complementary to the 59and 39ends of the putative PBCV-1 ligase gene were
used to amplify the 298-amino-acid open reading frame. Total PBCV-1 genomic
DNA was used as the template for a PCR catalyzed by Pfu DNA polymerase
(Stratagene). The sequence of the 59flanking primer was 59-CATGAAGTTAC
GTGTGTCATATGGCAATCACAAAGCC; that of the 39flanking primer was
59-CAAGACTTCGTAAAAACGGATCCTACATGGGATGA. These primers
were designed to introduce NdeI and BamHI restriction sites at the 59and 39
ends, respectively, of the ligase gene. The 0.9-kbp PCR product was digested with
NdeI and BamHI and then cloned into the NdeI and BamHI sites of T7-based
expression plasmid pET3c (Novagen). The resulting plasmid, pET-ChV-ligase,
was transformed into Escherichia coli BL21(DE3). Dideoxy sequencing of the
entire insert of pET-ChV-ligase confirmed that no alterations of the genomic
DNA sequence were introduced during PCR amplification and cloning of the
ligase gene.
* Corresponding author.
1931

Expression and purification of recombinant PBCV-1 ligase. A 500-ml culture
of E. coli BL21(DE3)/pET-ChV-ligase was grown at 378C in Luria-Bertani me-
dium containing 0.1 mg of ampicillin per ml until the A
600
reached 0.5. The
culture was adjusted to 0.4 mM isopropyl-b-D-thiogalactopyranoside (IPTG),
and incubation was continued at 378C for 2 h. Cells were harvested by centrif-
ugation, and the pellet was stored at 2808C. All subsequent procedures were
performed at 48C. Thawed bacteria were resuspended in 50 ml of buffer A (50
mM Tris HCl [pH 7.5], 2 mM dithiothreitol [DTT], 1 mM EDTA, 10% sucrose)
containing 0.2 M NaCl. Lysozyme was added to a final concentration of 2 mg/ml,
and the sample was sonicated for 30 s. Triton X-100 was added to a 0.1% final
concentration. The suspension was frozen on dry ice and then allowed to thaw at
48C. Sonication was repeated. Insoluble material was removed by centrifugation
for 45 min at 18,000 rpm in a Sorvall SS34 rotor. The soluble extract (60 mg of
protein) was adjusted to 50 mM NaCl by addition of 3 volumes of buffer A, and
this material was applied to a 25-ml column of DEAE-cellulose that had been
equilibrated with buffer A containing 50 mM NaCl. The flowthrough fraction (21
mg of protein) was applied to a 10-ml column of phosphocellulose that had been
equilibrated in buffer A containing 50 mM NaCl. The column was washed with
the same buffer and then eluted stepwise with buffer B (50 mM Tris HCl [pH
8.0], 2 mM DTT, 10% glycerol) containing 0.1, 0.15, 0.2, 0.3, 0.5, and 1.0 M NaCl.
The polypeptide composition of the column fractions was monitored by sodium
dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE). The recom-
binant PBCV-1 protein was retained on the column and was recovered predom-
inantly in the 0.3 M fraction (7 mg of protein). An aliquot of the 0.3 M phos-
phocellulose fraction was applied to a 4.8-ml 15 to 30% glycerol gradient
containing 50 mM Tris-HCl (pH 8.0), 2 mM DTT, 0.5 M NaCl, and 0.1% Triton
X-100. The gradient was centrifuged at 50,000 rpm for 24 h at 48C in a Beckman
SW50 rotor. Fractions were collected from the bottom of the tube. Marker
proteins—bovine serum albumin, soybean trypsin inhibitor, and cytochrome
c—were sedimented in a parallel gradient. Enzyme fractions were stored at
2808C and thawed on ice just prior to use. Protein concentrations were deter-
mined by the Bio-Rad dye-binding assay with bovine serum albumin as the
standard.
Enzyme-AMP complex formation. Standard reaction mixtures (20 ml) contain-
ing 50 mM Tris HCl (pH 7.5), 5 mM DTT, 5 mM MgCl
2
,10mM[a-
32
P]ATP, and
enzyme were incubated for 5 min at 378C and then halted by adding SDS to a 1%
final concentration. The samples were electrophoresed through a 12% polyacryl-
amide gel containing 0.1% SDS. Label transfer to the 34-kDa PBCV-1 polypep-
tide was visualized by autoradiographic exposure of the dried gel and was quan-
titated by scanning the gel with a FUJIX BAS1000 Bio-Imaging Analyzer.
Ligase substrate. The standard substrate used in ligase assays was a 36-bp
DNA duplex containing a centrally placed nick (see Fig. 5). This DNA was
formed by annealing two 18-mer oligonucleotides to a complementary 36-mer
strand. The 18-mer constituting the donor strand was 59
32
P labeled and gel
purified as previously described (19, 21). The labeled donor was annealed to the
complementary 36-mer in the presence of a 39OH-terminated acceptor strand in
0.2 M NaCl by heating at 658C for 2 min, followed by slow cooling to room
temperature. The molar ratio of the 18-mer donor to 36-mer complement to
18-mer acceptor strands in the hybridization mixture was typically 1:3:6.
DNA ligation. Reaction mixtures (20 ml) containing 50 mM Tris HCl (pH 7.5),
5 mM DTT, 10 mM MgCl
2
, 1 mM ATP, 1 pmol of 59
32
P-labeled DNA substrate,
and enzyme were incubated at 228C. Reactions were initiated by enzyme addition
and halted by addition of 1 ml of 0.5 M EDTA and 5 ml of formamide. The
samples were heated at 958C for 5 min and then electrophoresed through a
12.5% polyacrylamide gel containing 7 M urea in 0.53TBE (45 mM Tris borate,
1 mM EDTA). The labeled 36-mer ligation product was well resolved from the
59-labeled 18-mer donor strand. The extent of ligation [36-mer/(18-mer 136-
mer)] was determined by scanning the gel with a PhosphorImager.
RESULTS
Expression of the PBCV-1 ligase-like protein in bacteria.
The PBCV-1 open reading frame encoding a ligase-like pro-
tein was PCR amplified from viral genomic DNA and cloned
into a T7 RNA polymerase-based bacterial expression vector.
FIG. 1. Conserved sequence elements define a superfamily of covalent nucleotidyl transferases. Six colinear sequence elements, designated motifs I, III, IIIa, IV,
V, and VI, are conserved in polynucleotide ligases and mRNA capping enzymes as shown. The aligned amino acid sequences are those of the DNA ligases (DNA)
encoded by Chlorella virus PBCV-1 (ChV), vaccinia virus (VAC), Shope fibroma virus (SFV), fowlpox virus (FPV), Saccharomyces cerevisiae (Sc), Schizosaccharomyces
pombe (Sp), Arabidopsis thaliana (At), human ligase I (Hu1), human ligase 3 (Hu3), human ligase 4 (Hu4), Xenopus laevis (Xe), Caenorhabditis elegans (Ce), Candida
albicans (Ca), ASFV (ASF), Desulfurolobus ambivalens (Dam), Methanococcus jannaschii (Mja), and bacteriophages T4, T3, and T7. Also included is T4 RNA ligase.
Grouped below the ligases are the aligned sequences of capping enzymes (CE) encoded by Chlorella virus PBCV-1 (ChV), ASFV (ASF), S. cerevisiae (Sc), S. pombe
(Sp), vaccinia virus (VAC), Shope fibroma virus (SFV), and molluscum contagiosum virus (MCV). The number of intervening amino acid residues separating the motifs
is indicated between the motifs. Residues in the vaccinia virus DNA ligase that were found by mutational analysis to be essential for activity (21) are indicated by asterisks.
FIG. 2. Amino acid sequence of a putative PBCV-1 DNA ligase. Conserved
motifs I, III, IIIa, IV, V, and VI are boxed.
1932 HO ET AL. J. VIROL.

The pET-ChV-ligase plasmid was introduced into E. coli
BL21(DE3), a strain that contains the T7 RNA polymerase
gene under the control of a lacUV5 promoter. A prominent
34-kDa polypeptide was detectable by SDS-PAGE in whole-
cell extracts of IPTG-induced bacteria (Fig. 3A, lane 1). This
polypeptide was not present when bacteria containing the pET
vector alone were induced with IPTG (data not shown). After
centrifugal separation of the crude lysate, the PBCV-1 protein
was recovered in the soluble supernatant fraction (Fig. 3A,
lane 2).
Recombinant PBCV-1 34-kDa protein forms a covalent en-
zyme-adenylate complex. The initial step in DNA ligation in-
volves formation of a covalent enzyme-adenylate intermediate,
EpA. EpA formation by DNA ligases can be detected with
high sensitivity and specificity by label transfer from
[a-
32
P]ATP to the enzyme. To assay the adenylyltransferase
activity of the expressed PBCV-1 protein, we incubated either
whole-cell or soluble extracts of IPTG-induced BL21(DE3)/
pET-ChV-ligase cells in the presence of [a-
32
P]ATP and a
divalent cation. This resulted in the formation of a nucleotidyl-
protein adduct that migrated as a single 34-kDa species during
SDS-PAGE (Fig. 3B, lanes 1 and 2). Labeling of this polypep-
tide was not detected in extracts prepared from bacteria that
lacked the PBCV-1 gene (data not shown). We concluded that
the expressed 34-kDa PBCV-1 protein is active in nucleotidyl
transfer.
Purification of recombinant PBCV-1 ligase. The 34-kDa
polypeptide was purified from the soluble bacterial extract by
sequential DEAE-cellulose and phosphocellulose column
chromatography steps. The recombinant protein did not bind
to DEAE-cellulose at low ionic strength (50 mM NaCl). SDS-
PAGE analysis of the DEAE flowthrough fraction (Fig. 3A,
lane 3) showed that many of the bacterial polypeptides were
eliminated during this step. The PBCV-1 protein adsorbed to
phosphocellulose and was recovered during step elution with
0.3 M NaCl (Fig. 3A, lane 8). The phosphocellulose prepara-
tion was highly enriched with respect to the 34-kDa polypep-
tide; approximately 7 mg was obtained from a 500-ml culture
of IPTG-induced bacteria. The adenylyltransferase activity co-
incided with the 34-kDa polypeptide during column chroma-
tography (Fig. 3B and other data not shown).
When the phosphocellulose fraction was centrifuged
through a 15 to 30% glycerol gradient in 0.5 M NaCl, a single
peak of adenylyltransferase activity was detected that coin-
cided with the 34-kDa polypeptide (Fig. 4). We estimated a
sedimentation coefficient of 3.1S relative to marker proteins
sedimented in a parallel gradient. This suggested that the
PBCV-1 adenylyltransferase is a monomer of the 34-kDa pro-
tein.
DNA ligation. We assayed the ability of the recombinant
PBCV-1 protein to seal a 36-mer synthetic duplex DNA sub-
strate containing a single nick. The structure of the substrate is
shown in Fig. 5. Ligase activity was evinced by conversion of
the 59
32
P-labeled 18-mer donor strand into an internally la-
beled 36-mer product. The DNA ligase activity profile across
the glycerol gradient paralleled that of enzyme-adenylate com-
plex formation (Fig. 4B). These results demonstrate that the
34-kDa PBCV-1 protein is indeed a DNA ligase. Further char-
acterization of the PBCV-1 ligase was performed by using the
glycerol gradient preparation (peak fraction 19).
The extent of ligation of the nicked duplex (added at a 50
nM concentration with respect to the 59-labeled donor strand)
during a 10-min incubation in the presence of 1 mM ATP
increased linearly with the concentration of PBCV-1 ligase
from 50 to 500 pM (Fig. 6A). In the linear range of enzyme
dependence in this experiment, the recombinant ligase joined
about 100 to 120 fmol of DNA ends per fmol of enzyme. To
estimate the ratio of product to enzyme, the enzyme molarity
was calculated based on total protein concentration, assuming
enzyme homogeneity. It was also assumed that all of the en-
zyme molecules in the preparation were catalytically active.
The reaction saturated at .500 pM enzyme with 85% of the
labeled donor strand converted to 36-mer in 10 min. This
upper limit of ligation probably reflected incomplete annealing
of all three component strands to form the nicked substrate.
Ligation could be detected in the absence of added ATP, but
only at high levels of input enzyme (Fig. 6B). ATP-indepen-
dent ligation was attributed to preadenylylated ligase in the
enzyme preparation. The linear dependence of ATP-indepen-
dent strand joining on enzyme indicated that about 0.9 mol of
FIG. 3. Expression, purification, and adenylyltransferase activity of recombi-
nant PBCV-1 ligase. (A) The polypeptide compositions of recombinant PBCV-1
protein at sequential stages of purification were analyzed by SDS-PAGE, as
follows: whole-cell lysate of IPTG-induced BL21(DE3)/pET-ChV-ligase (lane
1); soluble lysate fraction (lane 2), DEAE-cellulose flowthrough fraction (lane
3), phosphocellulose flowthrough fraction (lane 4), and 0.1 M NaCl (lane 5), 0.15
M NaCl (lanes 6), 0.2 M NaCl (lane 7), 0.3 M NaCl (lane 8), 0.5 M NaCl (lane
9), and 1.0 M NaCl (lane 10) phosphocellulose eluates. The gel was fixed and
stained with Coomassie blue dye. The positions and sizes (kilodaltons) of co-
electrophoresed marker polypeptides are shown on the left. (B) Enzyme-adeny-
late complex formation. Reaction mixtures (20 ml) contained 50 mM Tris HCl
(pH 7.5), 5 mM DTT, 5 mM MgCl
2
,10mM[a-
32
P]ATP, and recombinant
PBCV-1 protein at the following stages of purification: whole-cell lysate (lane 1),
soluble lysate fraction (lane 2), DEAE-cellulose flowthrough fraction (lane 3),
phosphocellulose flowthrough fraction (lane 4), and 0.1 M NaCl (lane 5), 0.15 M
NaCl (lanes 6), 0.2 M NaCl (lane 7), 0.3 M NaCl (lane 8), 0.5 M NaCl (lane 9),
and 1.0 M NaCl (lane 10) phosphocellulose eluates. The reaction products were
resolved by SDS-PAGE. An autoradiograph of the dried gel is shown. The
positions and sizes (in kilodaltons) of prestained marker polypeptides are indi-
cated on the left.
VOL. 71, 1997 CHLORELLA VIRUS DNA LIGASE 1933

ends was sealed per mol of ligase (Fig. 6B), implying that 90%
of the enzyme molecules had AMP bound at the active site.
Kinetics, ATP dependence, and nucleotide cofactor specific-
ity of ligation. The kinetics of ligation were examined in DNA
excess (50 nM) in the presence of 1 mM ATP and 10 mM
MgCl
2
. The initial rate was proportional to enzyme concentra-
tion in the range of 125 to 500 pM (Fig. 7A). In subsequent
experiments, ligation assay mixtures contained 250 pM enzyme
and activity was determined after a 2-min incubation at 228C
(unless otherwise specified). Under these conditions, ligase
activity in 50 mM Tris-HCl buffer was optimal between pHs 7.0
and 7.5. Activity at pH 9.5 was ;50% of that at pH 7.5 (data
not shown). The rate of ligation increased with ATP concen-
tration from 10 to 200 mM and leveled off at 0.5 to 2 mM (Fig.
7B). A K
m
of 75 mM ATP was calculated from a double-
reciprocal plot of the data. Other ribonucleoside triphosphates
or deoxynucleoside triphosphates at a 0.5 mM concentration
could not substitute for ATP (Fig. 8).
Divalent cation dependence and specificity. Ligation de-
pended on a divalent cation in excess of the input ATP; activity
was enhanced steadily as Mg was increased from 2 to 20 mM
(Fig. 9B). The divalent cation requirement was satisfied by Mn
and, to a lesser extent, by Co but not by Ca, Cu, or Zn (Fig. 9A
and B).
DNA substrate specificity—nicks versus gaps. The structure
of the ligation substrate was altered such that the 39hydroxyl-
terminated acceptor strand was separated from the 59phos-
phate donor terminus by a 2- or 1-nucleotide (nt) gap (Fig. 5).
The specific activity of PBCV-1 ligase on a 1-nt gap substrate
was 1% of the activity of a nicked duplex DNA (Fig. 10).
PBCV-1 ligase was incapable of joining strands across the 2-nt
gap. The implication is that the 39OH must be positioned fairly
precisely relative to the 59phosphate donor terminus for liga-
tion to occur.
Specificity of ligase binding to DNA—nick versus gap. A
native gel mobility shift assay was employed to examine the
binding of PBCV-1 ligase to the
32
P-labeled, nicked duplex
DNA (19). Binding reactions were performed in the absence of
magnesium and ATP to preclude conversion of substrate to
product during the incubation. Control experiments verified
that ATP-independent ligation of the nicked DNA substrate by
stoichiometric amounts of PBCV-1 ligase required a divalent
cation. No strand joining occurred under the reaction condi-
tions employed in our gel shift experiments (data not shown).
Mixing the ligase with nicked substrate resulted in the forma-
tion of a discrete protein-DNA complex that migrated more
slowly than the free DNA during electrophoresis through a 6%
FIG. 4. Glycerol gradient sedimentation. (A) The phosphocellulose ligase
preparation was sedimented in a 15 to 30% glycerol gradient as described in
Materials and Methods. Aliquots (20 ml) of the phosphocellulose fraction (P-
Cell) and the indicated glycerol gradient fractions were analyzed by SDS-PAGE.
Polypeptides were visualized by staining with Coomassie blue dye. Marker pro-
teins (bovine serum albumin [BSA], soybean trypsin inhibitor [STI], and cyto-
chrome c[cytC]) were centrifuged in a parallel gradient; the marker peaks are
indicated below the gel. (B) Aliquots of the glycerol gradient fraction were
assayed for enzyme-adenylate complex formation and DNA strand-joining ac-
tivities as described in Materials and Methods. The DNA ligation reaction
mixtures contained 1 pmol of nicked duplex DNA and 0.1 ml(1ml of a 1:10
dilution) of each of the indicated gradient fractions. Incubation was for 10 min
at 228C. Adenylyltransferase reaction mixtures contained 10 mM[a-
32
P]ATP and
1ml of the indicated fractions; incubation was for 5 min at 378C. Adenylyltrans-
ferase activity was gauged by the signal intensity of the radiolabeled ligase
polypeptide. PSL, photostimulatable luminescence.
FIG. 5. Ligase substrates. Duplex substrates for PBCV-1 ligase were pre-
pared by annealing a 59-end-labeled 18-mer donor strand to a 36-mer comple-
mentary strand and an 18-mer acceptor strand. The structure of the standard
nicked duplex substrate is shown at the top; the position of the 59-end-labeled
nucleotide is indicated by the dots. The structures of 1- and 2-nt-gapped duplexes
are shown below that of the nicked duplex.
FIG. 6. ATP-dependent and ATP-independent ligation of duplex DNA con-
taining a single nick. (A) Complete reaction mixtures (20 ml) containing 50 mM
Tris HCl (pH 7.5), 5 mM DTT, 10 mM MgCl
2
, 1 mM ATP, 1 pmol of nicked
DNA substrate, and the indicated amounts of purified PBCV-1 ligase (glycerol
gradient fraction) were incubated for 10 min at 228C. (B) ATP was omitted from
the reaction mixtures. Extent of ligation is plotted as a function of input ligase.
1934 HO ET AL. J. VIROL.

native polyacrylamide gel (Fig. 11). The abundance of this
complex increased in proportion to the amount of input ligase.
To estimate binding affinity, the gel was scanned with a
PhosphorImager; the apparent dissociation constant, calcu-
lated as described by Riggs et al. (17), was 15 nM.
Little or no specific complex was detected when PBCV-1
ligase was incubated with 1- or 2-nt gap DNA (Fig. 11). A
diffuse smear of shifted material was detected at a 200 nM
ligase concentration. Thus, PBCV-1 DNA ligase bound specif-
ically at a DNA nick and was capable of discriminating be-
tween nicked and 2-nt-gapped DNAs at the substrate-binding
step. This affirms the importance of the 39OH acceptor strand
in substrate recognition.
Analysis of enzyme-AMP complex formation. The PBCV-1
ligase reacted specifically with [a-
32
P]ATP. The amount of
enzyme-AMP complex formed during a 5-min incubation at
378C in the presence of 10 mM[a-
32
P]ATP was proportional to
the amount of added protein (data not shown). EpA formation
increased as a function of ATP concentration and reached
saturation near 20 mM[a-
32
P]ATP (Fig. 12A). Half saturation
was achieved at ;5mM ATP. [a-
32
P]dATP was an extremely
poor substrate for EpA formation. We estimated from the
NTP titration experiment in Fig. 12A that ATP was 100-fold
more effective than dATP in EpA formation. Hence, PBCV-1
ligase, which utilized ATP, but not dATP, as a cofactor in the
strand-joining reaction, discriminated between ribose and de-
oxyribose sugars at the step of EpA formation. No protein-
nucleotide complex was formed with [a-
32
P]GTP (data not
shown).
We calculated, based on the molar amount of AMP label
transfer versus the molar amount of ligase added, that ;5% of
the input protein was converted to ligase-[
32
P]adenylate. This
was consistent with the estimate (based on ATP-independent
ligation activity) that 90% of the enzyme molecules in the
preparation were preadenylated and, hence, incapable of EpA
formation. We observed a fivefold increase in the extent of
label transfer from [a-
32
P]ATP to protein when a nicked DNA
substrate was added to the reaction mixture (data not shown).
This would be expected if transfer of unlabeled AMP from
preadenylated enzyme to the 59phosphate of the donor strand
(step 2 of the ligase reaction) regenerates unadenylated en-
zyme that can then react with [a-
32
P]ATP.
EpA formation depended on a divalent cation cofactor. This
requirement was satisfied by either 5 mM magnesium or 5 mM
manganese and, to lesser extent, by 5 mM cobalt (Fig. 12B).
Calcium, copper, and zinc were inactive at this concentration
(Fig. 12B). The yield of EpA was proportional to the magne-
FIG. 7. Kinetics and ATP concentration dependence of strand joining. (A)
Kinetics. Reaction mixtures (100 ml) contained 50 mM Tris HCl (pH 7.5), 5 mM
DTT, 10 mM MgCl
2
, 1 mM ATP, 5 pmol of nicked DNA, and either 500, 250,
or 125 pM purified PBCV-1 ligase. The reaction was initiated by enzyme addi-
tion. Aliquots (10 ml) were withdrawn at the times indicated, and the reaction
was quenched immediately. Extent of ligation is plotted as a function of incu-
bation time. (B) ATP dependence. Reaction mixtures (20 ml) containing 50 mM
Tris HCl (pH 7.5), 5 mM DTT, 10 mM MgCl
2
, 1 pmol of nicked DNA, 5 fmol
of PBCV-1 ligase, and ATP as indicated were incubated for 2 min at 228C. Extent
of ligation is plotted as a function of ATP concentration.
FIG. 8. Nucleotide specificity. Reaction mixtures (20 ml) containing 50 mM
Tris HCl (pH 7.5), 5 mM DTT, 10 mM MgCl
2
, 1 pmol of nicked DNA, 5 fmol
of PBCV-1 ligase, and 0.5 mM nucleoside triphosphate or deoxynucleoside
triphosphate, as indicated, were incubated for 2 min at 228C. Nucleotide was
omitted from a control reaction (—). The reaction products were electropho-
resed through a 12% polyacrylamide gel containing 7 M urea in 0.53TBE. An
autoradiograph of the dried gel is shown. The positions of the 59-end
32
P-labeled
18-mer donor strand and the 36-mer ligation product are indicated on the left.
FIG. 9. Divalent cation specificity of strand joining. (A) Reaction mixtures
(20 ml) containing 50 mM Tris HCl (pH 7.5), 1 mM ATP, 1 pmol of nicked DNA,
5 fmol of PBCV-1 ligase, and the indicated divalent cation at 10 mM were
incubated for 2 min at 228C. Divalent cation was omitted from a control reaction.
Mg, Mn, Ca, and Co were added as chloride salts; Cu and Zn were added as
sulfates. (B) Reaction mixtures (20 ml) containing 50 mM Tris HCl (pH 7.5), 5
mM DTT, 1 mM ATP, 1 pmol of nicked DNA, 5 fmol of PBCV-1 ligase, and
MgCl
2
or MnCl
2
, as indicated, were incubated for 2 min at 228C. Extent of
ligation is plotted as a function of divalent cation concentration.
VOL. 71, 1997 CHLORELLA VIRUS DNA LIGASE 1935

sium concentration from 0.1 to 1 mM and plateaued at 2 to 10
mM (Fig. 12C). Manganese was a more effective cofactor than
magnesium at 0.5 mM but was progressively less active at
higher concentrations (Fig. 12C).
DISCUSSION
AChlorella virus PBCV-1 gene encoding a putative DNA
ligase was identified during sequencing of the viral DNA ge-
nome. We show that the 298-amino-acid gene product is an
ATP-dependent DNA ligase. This was achieved by expressing
the PBCV-1 protein in bacteria, purifying the protein to ho-
mogeneity, and characterizing its enzymatic properties.
PBCV-1 ligase, like other cellular and virus-encoded DNA
ligases, catalyzes strand joining via an enzyme-AMP interme-
diate. The PBCV-1 enzyme displays strict specificity for ATP
as the nucleotide cofactor; dATP is inactive. PBCV-1 ligase
thus resembles the T4, vaccinia virus, and cellular type I en-
zymes in its discrimination of the nucleoside triphosphate
sugar moiety (14, 16, 19, 21, 26).
The high efficiency of PBCV-1 ligase in strand joining across
a nick in duplex DNA contrasts sharply with the low efficiency
of ligation across a 1-nt gap and the inability to seal strands
across a 2-nt gap. Vaccinia virus ligase and yeast CDC9 ligase
display similar properties. The latter two enzymes synthesize
substantial levels of the DNA-adenylate intermediate on a 1-nt
gap DNA substrate (19, 26). Similar results were obtained
when stoichiometric levels of PBCV-1 ligase were reacted with
the 1-nt gap DNA (6a).
A native gel mobility shift assay was used to demonstrate
that formation of a stable complex between PBCV-1 ligase and
DNA depends on a DNA nick. The PBCV-1 enzyme discrim-
inates clearly at the DNA-binding step between nicked DNA
molecules that can be sealed versus 2-nt-gapped molecules that
cannot. Even a 1-nt gap significantly reduces the affinity of
ligase for the DNA. This implies that the protein initially
contacts both the donor and acceptor DNA strands on either
side of the nick prior to any covalent modification of the DNA
substrate. Similar specificity for binding to DNA nicks has been
documented for the vaccinia virus ligase (19). Thus, both of
these virus-encoded enzymes have an intrinsic ability to bind
FIG. 10. DNA substrate specificity. Reaction mixtures (20 ml) containing 50
mM Tris HCl (pH 7.5), 5 mM DTT, 1 mM ATP, 1 pmol of either nicked DNA,
1- or 2-nt gap DNA substrate, and PBCV-1 ligase, as indicated, were incubated
for 10 min at 228C. Extent of ligation is plotted as a function of input enzyme.
FIG. 11. Nucleic acid binding specificity. Reaction mixtures (10 ml) con-
tained 50 mM Tris HCl (pH 7.5); 5 mM DTT; 5% glycerol; 50 nM
32
P-labeled,
nicked DNA (left), 1-nt gap DNA (center), or 2-nt gap DNA (right); and 0, 25,
50, 100, or 200 nM purified PBCV-1 ligase (proceeding from left to right within
each titration series). After incubation for 10 min at 228C, the samples were
electrophoresed at 100 V through a native 6% polyacrylamide gel in 0.253TBE
(22.5 mM Tris-borate, 0.6 mM EDTA). An autoradiogram of the dried gel is
shown. Labeled species corresponding to free DNA and the ligase-DNA complex
are indicated at the left.
FIG. 12. Analysis of enzyme-AMP complex formation. (A) ATP dependence. Reaction mixtures (20 ml) containing 50 mM Tris HCl (pH 7.5), 5 mM MgCl
2
,5mM
DTT, 2 pmol of PBCV-1 ligase (glycerol gradient fraction), and [a-
32
P]ATP or [a-
32
P]dATP at the concentrations indicated were incubated for 5 min at 378C. Extent
of EpA formation is plotted as a function of NTP concentration. (B) Divalent cation specificity. Reaction mixtures (20 ml) containing 50 mM Tris HCl (pH 7.5), 10
mM[a-
32
P]ATP, 2 pmol of PBCV-1 ligase, and the indicated divalent cation at 5 mM were incubated for 5 min at 378C. Mg, Mn, Ca, and Co were added as chloride
salts; Cu and Zn were added as sulfates. (C) Divalent cation dependence. Reaction mixtures (20 ml) containing 50 mM Tris HCl (pH 7.5), 5 mM DTT, 10 mM
[a-
32
P]ATP, 2 pmol of PBCV-1 ligase, and MgCl
2
or MnCl
2
at the concentrations indicated were incubated for 5 min at 378C. Extent of EpA formation is plotted as
a function of divalent cation concentration.
1936 HO ET AL. J. VIROL.

preferentially to DNA sites where their action is required.
Electrophoretic resolution of the ligase-DNA complex from
unbound DNA will permit the use of chemical and enzymatic
footprinting techniques to delineate the interface between
PBCV-1 ligase and the DNA substrate, a subject about which
almost nothing is known for any DNA ligase. Of particular
interest is identification of the region(s) of the ligase polypep-
tide that mediates nick recognition.
Because the Chlorella viruses are not amenable to genetic
manipulation, it is not possible to determine the biological
function of the DNA ligase during the PBCV-1 replication
cycle. A role during viral DNA replication, repair, or recom-
bination is plausible, although the molecular mechanisms of
these transactions are largely unexplored for PBCV-1. Avail-
able insights come largely from analysis of the genomic DNA
sequence, which identifies by sequence homology several po-
tential replication and repair proteins. These include, in addi-
tion to the DNA ligase, a DNA polymerase, a polymerase
processivity factor, a helicase, a DNA glycosylase-apyrimidinic
lyase, and a type II topoisomerase (9, 11–13). The DNA ligase
is the first instance in which it has been shown that a purified
recombinant PBCV-1 DNA replication protein actually has the
enzymatic properties attributed to it on the basis of sequence
homology.
The PBCV-1 ligase is the smallest DNA ligase described to
date. It may well represent the minimum catalytic unit of an
ATP-dependent ligase. Insofar as PBCV-1 ligase is also
smaller than any known mRNA capping enzyme, it may con-
stitute the catalytic core of the nucleotidyl transferase super-
family. PBCV-1 ligase includes the six conserved motifs that
define the family but contains no additional sequence at the
carboxyl terminus downstream of motif VI. It contains only 26
amino acids N terminal to the presumptive active site, Lys-27,
within motif I. In this light, the PBCV-1 ligase emerges as an
excellent model for further structural and functional studies of
a eukaryotic ligase. The fact that the recombinant PBCV-1
ligase is purified in high yield as ligase-adenylate offers an
opportunity to solve the structure of the covalent reaction
intermediate. Realization of this goal would extend and com-
plement the insights provided by the crystal structure of the
bacteriophage T7 enzyme bound noncovalently to ATP (24).
ACKNOWLEDGMENTS
We thank Yu Li for help in sequencing the PBCV-1 DNA ligase
gene.
This investigation was supported by NIH grants GM42498 (S.S.) and
GM32441 (J.V.E.) and ACS grant FRA-432 (S.S.).
REFERENCES
1. Barnes, D. E., L. H. Johnston, K. Kodama, A. E. Tomkinson, D. D. Lasko,
and T. Lindahl. 1990. Human DNA ligase I cDNA: cloning and functional
expression in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 87:6679–
6683.
2. Cong, P., and S. Shuman. 1993. Covalent catalysis in nucleotidyl transfer: a
KTDG motif essential for enzyme-GMP complex formation by mRNA cap-
ping enzyme is conserved at the active sites of RNA and DNA ligases. J. Biol.
Chem. 268:7256–7260.
3. Cong, P., and S. Shuman. 1995. Mutational analysis of mRNA capping
enzyme identifies amino acids involved in GTP binding, enzyme-guanylate
formation, and GMP transfer to RNA. Mol. Cell. Biol. 15:6222–6231.
4. Fresco, L. D., and S. Buratowski. 1994. Active site of the mRNA-capping
enzyme guanylyltransferase from Saccharomyces cerevisiae: similarity to the
nucleotidyl attachment motif of DNA and RNA ligases. Proc. Natl. Acad.
Sci. USA 91:6624–6628.
5. Hammond, J. M., S. M. Kerr, G. L. Smith, and L. K. Dixon. 1992. An African
swine fever virus gene with homology to DNA ligases. Nucleic Acids Res.
20:2667–2671.
6. Heaphy, S., M. Singh, and M. J. Gait. 1987. Effect of single amino acid
changes in the region of the adenylylation site of T4 RNA ligase. Biochem-
istry 26:1688–1696.
6a.Ho, C. K., and S. Shuman. Unpublished data.
7. Kletzin, A. 1992. Molecular characterization of a DNA ligase gene of the
extremely thermophilic archaeon Desulfurolobus ambivalens shows close phy-
logenetic relationship to eukaryotic ligases. Nucleic Acids Res. 20:5389–
5396.
8. Kodama, K., D. E. Barnes, and T. Lindahl. 1991. In vitro mutagenesis and
functional expression in Escherichia coli of a cDNA encoding the catalytic
domain of human DNA ligase I. Nucleic Acids Res. 19:6093–6099.
9. Kutish, G. F., Y. Li, Z. Lu, D. L. Furuta, D. L. Rock, and J. L. Van Etten.
1996. Analysis of 76 kb of the Chlorella virus PBCV-1 330-kb genome: map
positions 182 to 258. Virology 223:303–317.
10. Lehman, I. R. 1974. DNA ligase: structure, mechanism, and function. Sci-
ence 186:790–797.
11. Li, Y., Z. Lu, D. E. Burbank, G. F. Kutish, D. L. Rock, and J. L. Van Etten.
1995. Analysis of 43 kb of the Chlorella virus PBCV-1 330-kb genome: map
positions 45 to 88. Virology 212:134–150.
11a.Li, Y., Z. Lu, S. Ropp, L. Sun, G. F. Kutish, D. L. Rock, and J. L. Van Etten.
Analysis of 72 kb of DNA at the right end of the Chlorella virus PBCV-1
genome. Unpublished data.
12. Lu, Z., Y. Li, Q. Que, G. F. Kutish, D. L. Rock, and J. L. Van Etten. 1996.
Analysis of 94 kb of the Chlorella virus PBCV-1 330-kb genome: map posi-
tions 88 to 182. Virology 216:102–123.
13. Lu, Z., Y. Li, Y. Zhang, G. F. Kutish, D. L. Rock, and J. L. Van Etten. 1995.
Analysis of 45 kb of DNA located at the left end of the Chlorella virus
PBCV-1 genome. Virology 206:339–352.
14. Montecucco, A., M. Lestingi, G. Pedrali-Noy, S. Spadari, and G. Ciarrocchi.
1990. Use of ATP, dATP and their alpha-thio derivatives to study DNA
ligase adenylation. Biochem. J. 271:265–268.
15. Niles, E. G., and L. Christen. 1993. Identification of the vaccinia virus
mRNA guanylyltransferase active site lysine. J. Biol. Chem. 268:
24986–24989.
16. Raae, A. J., and K. Kleppe. 1978. Effect of ATP analogues on T4 polynu-
cleotide ligase. Biochem. Biophys. Res. Commun. 81:24–27.
17. Riggs, A. D., H. Suzuki, and S. Bourgeois. 1970. lac repressor-operator
interactions: equilibrium studies. J. Mol. Biol. 48:67–83.
18. Schwer, B., and S. Shuman. 1994. Mutational analysis of yeast mRNA
capping enzyme. Proc. Natl. Acad. Sci. USA 91:4328–4332.
19. Shuman, S. 1995. Vaccinia virus DNA ligase: specificity, fidelity, and inhi-
bition. Biochemistry 34:16138–16147.
20. Shuman, S., Y. Liu, and B. Schwer. 1994. Covalent catalysis in nucleotidyl
transfer reactions: essential motifs in Saccharomyces cerevisiae RNA capping
enzyme are conserved in Schizosaccharomyces pombe and viral capping en-
zymes and among polynucleotide ligases. Proc. Natl. Acad. Sci. USA 91:
12046–12050.
21. Shuman, S., and X. M. Ru. 1995. Mutational analysis of vaccinia DNA ligase
defines residues essential for covalent catalysis. Virology 211:73–83.
22. Shuman, S., and B. Schwer. 1995. RNA capping enzyme and DNA ligase: a
superfamily of covalent nucleotidyl transferases. Mol. Microbiol. 17:405–410.
23. Smith, G. L., Y. S. Chan, and S. M. Kerr. 1989. Transcriptional mapping and
nucleotide sequence of a vaccinia virus gene encoding a polypeptide with
extensive homology to DNA ligases. Nucleic Acids Res. 17:9051–9062.
24. Subramanya, H. S., A. J. Doherty, S. R. Ashford, and D. B. Wigley. 1996.
Crystal structure of an ATP-dependent DNA ligase from bacteriophage T7.
Cell 85:607–615.
25. Tomkinson, A. E., N. F. Totty, M. Ginsburg, and T. Lindahl. 1991. Location
of the active site for enzyme-adenylate formation in DNA ligases. Proc. Natl.
Acad. Sci. USA 88:400–404.
26. Tomkinson, A. E., N. J. Tappe, and E. C. Friedberg. 1992. DNA ligase I from
Saccharomyces cerevisiae: physical and biochemical characterization of the
CDC9 gene product. Biochemistry 31:11762–11771.
27. Van Etten, J. L., L. C. Lane, and R. H. Meints. 1991. Virus and virus-like
particles of the eukaryotic algae. Microbiol. Rev. 55:586–620.
28. Wang, Y. C., W. A. Burkhart, Z. B. Mackey, M. B. Moyer, W. Ramos, I.
Husain, J. Chen, J. M. Besterman, and A. E. Tomkinson. 1994. Mammalian
DNA ligase II is highly homologous with vaccinia DNA ligase. Identification
of the DNA ligase II active site for enzyme-adenylate formation. J. Biol.
Chem. 269:31923–31928.
29. Wei, Y. F., P. Robins, K. Carter, K. Caldecott, D. J. Pappin, G. L. Yu, R. P.
Wang, B. K. Shell, R. A. Nash, P. Schar, et al. 1995. Molecular cloning and
expression of human cDNAs encoding a novel DNA ligase IV and DNA
ligase III, an enzyme active in DNA repair and recombination. Mol. Cell.
Biol. 1555:3206–3216.
VOL. 71, 1997 CHLORELLA VIRUS DNA LIGASE 1937