Published online January 2, 2004
Nucleic Acids Research, 2004, Vol. 32, No. 1 e2
DOI: 10.1093/nar/gnh011
Ligase-mediated construction of branched DNA
strands: a novel DNA joining activity catalyzed by T4
DNA ligase
Maritha Mendel-Hartvig1,2, Anil Kumar1 and Ulf Landegren1,*
1
Rudbeck Laboratory, Department of Genetics and Pathology, Se-75185 Uppsala, Sweden and
Biosciences, Uppsala, Sweden
2
Amersham
Received September 15, 2003; Revised November 6, 2003; Accepted November 17, 2003
ABSTRACT
Branched nucleic acid strands exist as intermediates in certain biological reactions, and bifurcating
DNA also presents interesting opportunities in biotechnological applications. We describe here how
T4 DNA ligase can be used for ef®cient construction
of DNA molecules having one 5¢ end but two distinct
3¢ ends that extend from the 2¢ and 3¢ carbons,
respectively, of an internal nucleotide. The nature of
the reaction products is investigated, and optimal
reaction conditions are reported for the construction of branched oligonucleotides. We discuss the
utility of these branched DNA nanostructures for
gene detection.
INTRODUCTION
Single DNA strands are typically linear strings of nucleotides
with one 5¢ and one 3¢ extreme end, but there are examples of
nucleic acid molecules with segments coupled in parallel from
a branching point rather than in series. The well known lariatshaped intermediate formed in the process of RNA splicing
arises through the covalent attachment of the 5¢ end of a
transcribed and cleaved intron to the 2¢ hydroxyl of an
adenosine residue, internal to the intron sequence (1). This
nucleotide thus represents a branched position in the RNA
molecule, connected to downstream sequences via 2¢,5¢- and
3¢,5¢-phosphodiester linkages. Also retroelements that exist in
some populations of Escherichia coli contain a branched
position, an internal G ribonucleotide residue where the 2¢
hydroxyl has been used to prime a polymerization reaction
using the same RNA strand as a template (2).
Branched nucleic acid strands have been chemically
synthesized using branching nucleotides that give rise to
fork-like structures (3±5). Such branching oligoribonucleotides have been useful to study messenger RNA splicing and to
investigate the biological role of RNA lariats and forks in vivo.
Chemically synthesized branched nucleic acid strands have
also been used for gene detection by attaching one hybridizing
DNA segment to several identical detectable sequences in a
forked structure, in order to increase sensitivity in hybridization-based detection of, for example, viral sequences (6). Socalled DNA dendrimers have also been used for enhanced
detection. These DNA molecules branch dichotomically in
several generations and can be constructed by introducing a
modi®ed residue having two blocking residues at intervals
during solid-phase chemical synthesis for subsequent extension of two separate, but identical 5¢ end sequences (7). Braich
and Damha developed an interesting strategy to start synthesis
of a second 3¢ strand during chemical synthesis by deprotecting a 3¢ carbon of an internal residue (8). Dif®culties of
selectively deprotecting this residue, and of performing 5¢-to3¢ synthesis from this position, may limit the generality of the
procedure.
We were interested in developing enzymatic means to
construct DNA strands that branch into two distinct 3¢ end
sequences from a given sugar residue along a DNA strand.
DNA ligases serve the purpose of sealing nicks in doublestranded DNA molecules (9±11), but they have also been
employed in a much wider range of reactions, such as the
joining of duplex DNA molecules with either cohesive or
blunt ends (12), or of ligation substrates where one or more of
the strands are RNA (13±15). Under appropriate conditions
ligases have also been shown to join oligodeoxyribonucleotides containing abasic sites, gaps or mismatched bases at the
ligation junction (16,17). We show here that T4 DNA ligase
can also be used to add an extra 3¢ strand to an internal
nucleotide residue in a single-stranded DNA molecule,
connected to the downstream sequence via a 2¢,5¢ phosphodiester bond. This leaves the 3¢ hydroxyl at this internal
residue free to be enzymatically joined to the 5¢ phosphate of
the extra DNA strand if the ligation substrate is positioned
through hybridization to a template strand.
MATERIALS AND METHODS
Oligonucleotides
Oligonucleotides were synthesized on an Applied Biosystems
380B DNA Synthesizer using phosphoramidite chemistry or
*To whom correspondence should be addressed. Tel: +46 18 471 4910; Fax: +46 18 471 4808; Email: [email protected]
Present address:
Anil Kumar, Solexa Limited, Chesterford Research Park, Little Chesterford, Nr Saffron Walden, Essex CB10 1XL, UK
The authors wish it to be known that, in their opinion, the ®rst two authors should be regarded as joint First Authors
Nucleic Acids Research, Vol. 32 No. 1 ã Oxford University Press 2004; all rights reserved
e2 Nucleic Acids Research, 2004, Vol. 32, No. 1
PAGE 2 OF 7
Figure 1. Structure of a substrate for construction of branched DNA by templated enzymatic ligation. The outer box illustrates that an internal nucleotide, the
boxed C residue, is connected to the downstream sequence via the 2¢ carbon, and that the 3¢ hydroxyl of this nucleotide is juxtaposed to the 5¢ phosphate of a
separate oligonucleotide by hybridization to a third, template oligonucleotide. H and D indicate cleavage sites for the restriction enzymes HinfI and DdeI,
respectively (see Figure 3B).
ordered from Interactiva or Eurogentech. We purchased the
5¢DMT, 3¢ TBDMS cytidine (N-Bz) phosphoramidite nucleoside reagent from ChemGenes, catalog number ANP-5682.
This reagent is suitable to introduce a ribonucleotide connected via a 2¢,5¢ phosphodiester bond to the preceding
nucleotide. After complete synthesis, deblocking and HPLC
puri®cation, this reagent yields a residue with a free 3¢
hydroxyl function in the nucleotide chain, referred to herein as
a 3¢ hydroxy ribonucleotide residue. All sequences are written
in the 5¢±3¢ direction (see also Fig. 1). The sequences of
oligonucleotides ON1a, ON1b and ON1c were: GACTGCTCGGACTCAGCCTACACTAGCGACATTGATGGACCTTCCGTCTCGTTCT, where the underlined C represents an
unconventional cytosine 3¢ hydroxy ribonucleotide in ON1a, a
regular cytosine ribonucleotide in ON1b and a cytosine
deoxyribonucleotide in ON1c. Oligonucleotide ON1d
(GACTGCTCGGACTCAGCCTACACTAGCGAC)
was
interrupted after the same C residue. Oligonucleotides
ON2a, (p-TCATCCAAGGTGCCTTTGCTGAGAC), ON2b
(p-TCATCCAAGGTGCCTTTGCTGAGACGTAATAGCGTTGAC) and ON2c (p-TCATCCAAGGTGCCTTTGCTGAGACTTTTTTTTTT-biotin), were used to form 3¢ branch
sequences. The following reagents were used as templates for
branch ligation: ON3a (CAAAGGCACCTTGGATGAGTCGCTAGTGTAGGC), ON3b (GTCTCAGCAAAGGCACCTTGGATGAGTCGCTAGTGTAGGCTGAGTCCGAGCAGTC), ON3c (CAGAACAGTCGTCTCAGCAAAGGCACCTTGGATGAGTCGCTAGTGTAGGCTGAGTCCGAG),
and ON3d (CAGAACAGTCGTCTCAGCAAAGGCA-
CCTTGGATGAGTCGCTAGTGTAGGCTGAGTCCGAGCAGTCGGCTCAGCAC).
Finally, oligonucleotides ON4 (AGAACGAGACGGAAGGTCC) and ON5 (GTCTCAGCAAAGGCACCTTG) were
used to prime extension reactions from the 2¢±5¢ and 3¢±5¢
branches, respectively, of the branched product.
Radiolabeling
Ten picomole oligonucleotide was 5¢ radiolabeled by the
addition of a 5¢ phosphate using 30 units of T4 polynucleotide
kinase (Amersham Biosciences) in 50 ml of 10 mM TrisOAc
pH 7.5, 10 mM MgAc2, 50 mM KAc and 30 mCi [l-32P]ATP
(3000 mCi/mmol, Amersham Biosciences) at 37°C for 30 min.
For 3¢ radiolabeling, 10 pmol oligonucleotide was incubated
with 18 units of terminal deoxynucleotidyl transferase
(Amersham Biosciences) in the same buffer containing
50 mCi [a-32P]dCTP (3000 mCi/mmol, Amersham
Biosciences) at 37°C for 30 min. After the reactions the
enzymes were heat-inactivated at 65°C for 10 min, and
oligonucleotides were puri®ed on MicroSpinÔ G-50 columns
(Amersham Biosciences).
Branch ligation
Unless otherwise indicated, 1 pmol radiolabeled ON2 and
3 pmol of both ON1 and ON3 were incubated at 65°C for
10 min in 20 ml of reaction buffer containing 10 mM TrisOAc
pH 7.5, 10 mM Mg(Ac)2, 50 mM KAc, 0.1 mg/ml bovine
serum albumin (BSA), 10 mM ATP, and allowed to cool to
room temperature for at least 10 min. Thereafter 2.8 Weiss
PAGE 3 OF 7
Nucleic Acids Research, 2004, Vol. 32, No. 1 e2
buffer was added [25 mM Tris-borate pH 8.3, 10 mM EDTA,
75% formamide, 0.05% (w/v) xylene cyanol FF, 0.05% (w/v)
bromophenol blue], and the samples were separated in a
denaturing 10% polyacrylamide gel and analyzed on a
PhosphorImager (Amersham Biosciences).
Restriction enzyme digestion
Radiolabeled ligation product (25 fmol) was digested with
5 units of HinfI or 4 units of DdeI restriction enzyme
(Amersham Biosciences) in 5 ml of 20 mM TrisOAc pH 7.5,
20 mM Mg(Ac)2, 100 mM KAc, at 37°C overnight.
Solid-phase replication of branched oligonucleotides
Ten microliter branch ligation reactions with 0.25 pmol 5¢
radiolabeled and 3¢ biotinylated ON2c, were immobilized to
20 ml of streptavidin-coated Dynabeads (Dynal), denatured,
and washed according to recommended procedures. The
extension primers, 0.25 pmol of ON4 or ON5, were incubated
together with the immobilized branched DNA at 65°C for
10 min in 5 ml of extension buffer. They were allowed to cool
to room temperature for at least 10 min. The 10 ml extension
reactions contained either 200 ng f29 DNA polymerase (a
kind gift of Dr Margarita Salas) in 50 mM Tris±HCl pH 7.5,
10 mM MgCl2, 20 mM (NH4)2SO4, 1 mM of 200 mM dNTP,
0.15 mg/ml BSA, 1 mM dithiothreitol, or 7.5 units of T7
Sequenase version 2.0 DNA polymerase (Amersham
Biosciences) in 40 mM Tris±HCl pH 7.5, 20 mM MgCl2,
50 mM NaCl, 200 mM dNTP, 0.15 mg/ml BSA. The reactions
were incubated at 37°C for 1 h, and stopped by adding an
equal volume of loading buffer. The samples were analyzed on
a denaturing 10% polyacrylamide gel.
RESULTS AND DISCUSSION
Construction of branched DNA by ligation
Figure 2. Formation of ligation products involving an oligonucleotide with
a 2¢,5¢ linkage. (A) Dependence of ligation product size on the length of the
labeled 5¢ phosphorylated ON2a (25 nt) and ON2b (39 nt) and independence
of template lengths. M, molecular weight markers. (B) Oligonucleotides
with a 2¢,5¢ linkage, a regular 3¢±5¢ DNA linkage or a 3¢±5¢ linkage
connecting a single ribonucleotide residue to the downstream sequence were
combined with templates of the indicated lengths. Reaction products were
analyzed on a denaturing 10% polyacrylamide gel.
units of T4 DNA ligase (Amersham Biosciences) was added
and the reactions were incubated at 37°C overnight, followed
by heat inactivation at 65°C for 10 min.
Alkaline treatment
One picomol branched 3¢-labeled oligonucleotide in 6.5 ml of
ligation reaction was incubated at 37°C overnight after the
addition of 6.5 ml of 0.6 M NaOH, 2 mM EDTA (18).
Thereafter 2.2 ml of 2 M HAc was added for neutralization,
and the nucleic acid was precipitated with ethanol and
dissolved in 6.5 ml of H2O. An equal volume of loading
Ligation of DNA strands by DNA ligases typically involves
the joining of 5¢ phosphates and 3¢ hydroxyls of adjoining
DNA ends, hybridizing to a common template DNA strand
(9±11,19). In order to attempt ligase-mediated construction of
branched DNA, we synthesized an oligonucleotide with an
internal 3¢ hydroxy ribonucleotide residue. This nucleotide
was thus connected to the downstream sequence via a 2¢,5¢
phosphodiester linkage, leaving a free 3¢ hydroxyl at this
internal position. The 3¢ hydroxyl group was brought close to
the 5¢ phosphate of a separate oligonucleotide by hybridizing
the strands to be joined to a common template oligonucleotide.
The question we asked was whether a DNA ligase could join a
5¢ phosphate of one strand to an internal 3¢ hydroxyl along a
different DNA strand in a DNA-templated reaction, despite
the DNA strand extending from the adjacent 2¢ carbon.
Figure 1 illustrates the structure of the oligonucleotide
reagents and of the complex serving as substrate for the
ligation reaction.
Characterization of the ligation product
We investigated the structure of the ligation products by
denaturing polyacrylamide gel electrophoresis. Of the three
oligonucleotides in the ligation substrates, only the branch
oligonucleotides, ON2a or ON2b, had 5¢ phosphatesÐthe 32P
added for labeling. Therefore only ligation to the 5¢ end of
e2 Nucleic Acids Research, 2004, Vol. 32, No. 1
PAGE 4 OF 7
Figure 3. Analysis of the structure of the ligation products formed in branch-ligation reactions. (A) Investigation of the susceptibility of 3¢-labeled molecules
with a 2¢,5¢ linkage to alkali treatment before and after branch ligation. (B) Regular nick- and branch-ligated products, radiolabeled via a 5¢ phosphate, were
investigated for their susceptibility to cleavage by the restriction enzymes HinfI and DdeI. Cleavage sites are indicated in Figure 1. The samples were analyzed on a denaturing 10% polyacrylamide gel. M, molecular weight markers.
ON2 oligonucleotides could be detected. Figure 2A demonstrates that the migration of the ligation products was
dependent on the sizes of the labeled 3¢ branch oligonucleotides, con®rming that these reagents participated in the
ligation products. Western and Rose have shown that under
certain conditions T4 DNA ligase is able to form an
intramolecular loop by joining the end of a single DNA strand
to that of an oligonucleotide hybridizing to the DNA strand,
without apparent Watson±Crick basepairing on one side of the
ligation site (20). A similar reaction could have joined the
branch oligonucleotides (ON2) to the 3¢ ends of the template
oligonucleotides (ON3). No such effect was seen under the
conditions used here, however, as the different sizes of the
three different ON3 oligonucleotides that were used to
template the ligation reactions clearly did not in¯uence the
migration of the ligation products (Fig. 2A).
Ligation products were obtained in high yields, ranging
from 70±78% of the labeled strands in the illustrated reactions.
Only strands containing an internal 2¢,5¢ linkage with a free 3¢
hydroxyl were substrates for the ligation reaction. No ligation
was observed with oligonucleotides having a regular 3¢±5¢
linkage in the corresponding position, whether or not there
was a free 2¢ hydroxyl present (Fig. 2B).
DNA strands containing internal RNA residues are susceptible to cleavage by treatment with alkali. The 3¢ hydroxy
ribonucleotide residue, having an internal 3¢ hydroxyl, is also
expected to render the strands sensitive to alkaline hydrolysis.
By contrast, successful ligation products that have this internal
3¢ carbon attached to an extraneous DNA strand should no
longer be sensitive to alkali. In the experiment shown in
Figure 3A the 3¢ radiolabeled oligonucleotide with a 3¢
hydroxy ribonucleotide residue was shown to be sensitive to
alkaline cleavage as expected, whereas the ligation product
was not. The pattern of cleavage of the ligation products with
restriction enzymes HinfI and DdeI is also consistent with a
branched structure of the ligation products. We compared
restriction cleavage of a conventional nick-ligated and a
branch-ligated product (Fig. 3B). The branched products
migrated distinctly from the corresponding linear fragments of
the same size, and the migration of the restriction digested
PAGE 5 OF 7
Nucleic Acids Research, 2004, Vol. 32, No. 1 e2
products was consistent with the expected cleavage pattern. A
HinfI site situated at the branch point could be cleaved in the
nick-ligated duplex but not in the branch-ligation reaction
product, indicating that the branched structure inhibited the
restriction enzyme. We conclude that the ligation products
indeed represent branched structures, created by ligation of the
5¢ phosphorylated strands to other strands having internal 2¢,5¢
linkages.
Optimization of ligation conditions
Ligation reactions proceed in four steps. First the enzyme is
activated by covalent coupling of ATP, followed by binding to
the substrate. The adenylyl group is then transferred to the 5¢
phosphate on the nicked DNA, and ®nally the activated
phosphate is attacked by a nearby 3¢ hydroxyl group, forming
a phosphodiester bond between the two adjacent strands and
releasing AMP (10,11,19). Attempts to use T4 DNA ligase to
join dif®cult substrates, such as blunt ends of duplex DNA
molecules or DNA strands hybridized to an RNA template,
tend to result in the interruption of the ligation process after
the cofactor has been transferred to the 5¢ end of the ligation
substrate. This leads to the accumulation of 5¢-adenylated
DNA strands. This tendency can be reduced, however, and
ligation can be allowed to proceed by using substantially
lower amounts of the cofactor ATP (21,22). Adenylated
intermediates accumulate also in branch ligation. At 1 mM
ATP, the concentration commonly used in nick ligation,
adenylation dominated over ligation (Fig. 4A and B). By
lowering ATP concentrations to 1±10 mM, the ligation product
yield increased to levels similar to those of nick ligation.
Similar yields of branched products were observed at
temperatures of 5, 16 or 37°C (results not shown).
Compared to the joining of conventional nicked ligation
substrates, considerably more ligase is required for ef®cient
branch ligation, and branched ligation products accumulate at
a far slower rate, reaching yields of ~70±80% only after
overnight incubation (Fig. 4A). Maximal ligation yields were
observed at the highest concentration of ligase used, corresponding to an estimated four ligase molecules per substrate
(Fig. 5). Blunt-end ligation requires 10±1003 more ligase
than cohesive end ligation, and both RNA templated ligation
as well as the looped ligation investigated by Western and
Rose require a molar excess of ligase over substrate
molecules. The reason for the large requirement of ligase in
the branch-ligation reaction is not understood. It is possible
that the enzyme remains bound to the substrate after the ligase
has covalently joined the oligonucleotides, unable to go on to
ligate new substrates. However, no band shift was obtained
when the ligation reaction products were analyzed for
electrophoretic mobility shift according to Rossi et al. (21),
indicating that the ligase leaves the substrate after ligation
(data not shown).
Figure 4. Comparison of the effect of ATP concentration on the yield of
(A) ligation and (B) adenylation products of regular nick-ligated (open
symbols) and branch-ligated (®lled symbols) products as a function of time.
The large and small symbols represent reactions performed with 1 mM and
1 mM ATP, respectively.
ribonucleotide in ON1b, although paused products were
observed (results not shown). By contrast, both f29 DNA
polymerase and T7 Sequenase generated products interrupted
around the branch-point of branched templates, although a
faint band, somewhat smaller than the expected full-length
product, can also be seen on the gel when extension was
primed from the 3¢ branch (see arrows in Fig. 6). We conclude
that the branched site served as a block for replication.
The branched site blocks replication
Concluding remarks
It was of interest to determine if the branched oligonucleotides
formed by enzymatic ligation could serve as templates for
DNA polymerases. Lorsch et al. have shown that reverse
transcriptase is able to read through a 2¢,5¢ linkage in a
template (23). We tested several DNA polymerases and found
that most were able to read through the 3¢ hydroxy
ribonucleotide residue in the linear ON1a and the 3¢±5¢
Among the DNA ligases, the T4 DNA ligase enzyme is
unusually adept at joining modi®ed nucleic acid substrates. To
our knowledge it is the only ligase capable of ef®cient bluntend ligation (11). By using high concentration of T4 DNA
ligase and low ATP concentration the yield of the branched
ligation product can reach levels similar to those for regular
nick ligation, although the reaction is quite slow.
e2 Nucleic Acids Research, 2004, Vol. 32, No. 1
PAGE 6 OF 7
Figure 5. The yield of branched product was investigated as a function of
ligase concentration. Circles represent branched ligation products and
squares represent adenylation products.
Templated branch ligation allows the construction of DNA
strands having two different 3¢ ends. This further extends the
already rich scope for DNA nano-scale structures, composed
of rigid duplex and ¯exible single-stranded segments (24). In
gene detection applications, branched DNA strands could
extend the range of functionalities of probe molecules. For
example, it is possible to construct padlock probes (25) that
after circularization and ®rm binding to a target molecule can
still extend a 3¢ end that can be used to prime rolling-circle
replication of a separately added circular DNA strand for
enhanced detection (26). Branched oligonucleotides could
also be used as bi-allelic padlock probes having two different
3¢ ends, speci®c for alternate alleles at a given locus, and
separately detectable.
ACKNOWLEDGEMENTS
This work was supported by Amersham Biosciences, the
Swedish Medical and Technological Research Councils of
Sweden, the Wallenberg Foundation, and Polysaccharide
Research AB Uppsala. A.K. was supported by a stipend
from the Swedish Institute. Dr Margarita Salas kindly
provided the f29 polymerase. We thank Drs Marek
Kwiatkowski, Mats Gullberg and Mats Nilsson for providing
valuable comments on the manuscript.
REFERENCES
1. Padgett,R.A., Konarska,M.M., Grabowski,P.J., Hardy,S.F. and
Sharp,P.A. (1984) Lariat RNAs as intermediates and products in the
splicing of messenger RNA precursors. Science, 225, 898±903.
2. Yee,T., Furuichi,T., Inouye,S. and Inouye,M. (1984) Multicopy singlestranded DNA isolated from a Gram-negative bacterium, Myxococcus
xanthus. Cell, 38, 203±209.
3. Horn,T. and Urdea,M.S. (1989) Forks and combs and DNA: the synthesis
of branched oligodeoxyribonucleotides. Nucleic Acids Res., 17,
6959±6967.
4. Sund,C., Foldesi,A., Yamakage,S. and Chattopadhyaya,J. (1991)
Synthesis of branched nona- and deca-RNA modelling the lariat formed
in pre-mRNA processing reaction (splicing). Nucleic Acids Symp. Ser.,
24, 9±12.
5. Damha,M.J., Ganeshan,K., Hudson,R.H.E. and Zabarylo,S.V. (1992)
Solid-phase synthesis of branched oligoribonucleotides related to
messenger RNA splicing intermediates. Nucleic Acids Res., 20,
6565±6573.
Figure 6. Attempted replication of immobilized branched 5¢-radiolabeled
oligonucleotides across the 2¢±5 and the 3¢±5¢ linkages at the branch point.
Comparison between f29 DNA polymerase and T7 Sequenase versus
2.0 DNA polymerase. Reaction products were analyzed on a denaturing
10% polyacrylamide gel. The extension reactions were primed from the 2¢
branch in lanes 1, 2, 5 and 6, and from the 3¢±5¢ branch in lanes 3, 4, 7
and 8. M, molecular weight markers.
6. Urdea,M.S. (1994) Branched DNA signal ampli®cation. Bio/Technol.,
12, 926±928.
7. Horn,T., Chang,C.-A., Fultz,T.J., Ahle,D., Hamren,S.J. and Urdea,M.S.
(1991) Branched oligodeoxyribonucleotides for use as ampli®cation
multimers in bioassays: chemical synthesis and characterization. Nucleic
Acids Symp. Ser., 24, 201±202.
8. Braich,R.S. and Damha,M.J. (1997) Regiospeci®c solid-phase synthesis
of branched oligonucleotides. Effect of vicinal 2¢,5¢- (or 2¢,3¢-) and
3¢,5¢-phosphodiester linkages on the formation of hairpin DNA.
Bioconjugate Chem., 8, 370±377.
9. Lehman,I.R. (1974) DNA ligase: structure, mechanism, and function.
Science, 186, 790±797.
10. Higgins,N.P. and Cozzarelli,N.R. (1979) DNA-joining enzymes: a
review. Methods Enzymol., 68, 50±71.
11. Engler,M.J. and Richardson,C.C. (1982) DNA ligases. The Enzymes, XV,
3±29.
12. Sgaramella,V. (1972) The covalent joining of DNA molecules at their
base-paired ends. Farmaco [Sci.], 27, 809±817.
13. Kleppe,K., van de Sande,J.H. and Khorona,H.G. (1970) Polynucleotide
ligase-catalyzed joining of deoxyribo-oligonucleotides on
ribopolynucleotide templates and of ribo-oligonucleotides on
deoxyribopolynucleotide templates. Proc. Natl Acad. Sci. USA, 67,
68±73.
14. Fareed,G.C., Wilt,E.M. and Richardson,C.C. (1971) Enzymatic breakage
and joining of deoxyribonucleic acid. 8. Hybrids of ribo- and
deoxyribonucleotide homopolymers as substrates for polynucleotide
ligase of bacteriophage T4. J. Biol. Chem., 246, 925±932.
15. Nilsson,M., Barbany,G., Antson,D.O., Gertow,K. and Landegren,U.
(2000) Enhanced detection and distinction of RNA by enzymatic probe
ligation. Nat. Biotechnol., 18, 791±793.
16. Wu,D.Y. and Wallace,R.B. (1989) The ligation ampli®cation reaction
(LAR)±ampli®cation of speci®c DNA sequences using sequential rounds
of template-dependent ligation. Genomics, 4, 560±569.
PAGE 7 OF 7
17. Landegren,U., Kaiser,R., Sanders,J. and Hood,L. (1988) A ligasemediated gene detection technique. Science, 241, 1077±1080.
18. Sriskanda,V. and Shuman,S. (1998) Speci®city and ®delity of strand
joining by Chlorella virus DNA ligase. Nucleic Acids Res., 26,
3536±3541.
19. Weiss,B., Jacquemin-Sablon,A., Live,T.R., Fareed,G.C. and
Richardson,C.C. (1968) Enzymatic breakage and joining of
deoxyribonucleic acid. VI. Further puri®cation and properties of
polynucleotide ligase from Escherichia coli infected with bacteriophage
T4. J. Biol. Chem., 243, 4543±4555.
20. Western,L.M. and Rose,S.J. (1991) A novel DNA joining activity
catalyzed by T4 DNA ligase. Nucleic Acids Res., 19, 809±813.
21. Rossi,R., Montecucco,A., Ciarrocchi,G. and Biamonti,G. (1997)
Functional characterization of the T4 DNA ligase: a new insight into the
mechanism of action. Nucleic Acids Res., 25, 2106±2113.
Nucleic Acids Research, 2004, Vol. 32, No. 1 e2
22. Nilsson,M., Antson,D.O., Barbany,G. and Landegren,U. (2001)
DNA-templated DNA ligation for transcript analysis. Nucleic Acids Res.,
29, 578±581.
23. Lorsch,J.R., Bartel,D.P. and Szostak,J.W. (1995) Reverse transcriptase
reads through a 2¢-5¢ linkage and a 2¢-thiophosphate in a template.
Nucleic Acids Res., 23, 2811±2814.
24. Seeman,N.C. (2003) DNA in a material world. Nature, 421, 427±431.
25. Nilsson,M., Malmgren,H., Samiotaki,M., Kwiatkowski,M.,
Chowdhary,B.P. and Landegren,U. (1994) Padlock probes: circularizing
oligonucleotides for localized DNA detection. Science, 265, 2085±2088.
26. Lizardi,P.M., Huang,X., Zhu,Z., Bray-Ward,P., Thomas,D.C. and
Ward,D.C. (1998) Mutation detection and single-molecule counting
using isothermal rolling-circle ampli®cation. Nature Genet., 19,
225±232.