Research Article: Selectivity of Enzymatic Conversion of Oligonucleotide Probes during Nucleotide Polymorphism Analysis of DNA

Date Published: April , 2010

Publisher: A.I. Gordeyev

Author(s): O.A. Vinogradova, D.V. Pyshnyi.



The analysis of DNA nucleotide polymorphisms is one of the main goals of DNA diagnostics.
DNA–dependent enzymes (DNA polymerases and DNA ligases) are widely used to enhance the
sensitivity and reliability of systems intended for the detection of point mutations in genetic
material. In this article, we have summarized the data on the selectiveness of
DNA–dependent enzymes and on the structural factors in enzymes and DNA which influence
the effectiveness of mismatch discrimination during enzymatic conversion of oligonucleotide
probes on a DNA template. The data presented characterize the sensitivity of a series of
DNA–dependent enzymes that are widely used in the detection of noncomplementary base
pairs in nucleic acid substrate complexes. We have analyzed the spatial properties of the
enzyme–substrate complexes. These properties are vital for the enzymatic reaction and the
recognition of perfect DNA–substrates. We also discuss relevant approaches to increasing
the selectivity of enzyme–dependent reactions. These approaches involve the use of
modified oligonucleotide probes which “disturb” the native structure of the
DNA–substrate complexes.

Partial Text

Single nucleotide polymorphism (SNP) is the most common form of genetic variations in the
genome. Currently, the number of known single nucleotide mutations in the human genome is in
excess of 9 million [1]. Such mutations are often
important genetic markers that can determine the phenotypic and physiological traits of an
individual and are also the molecular basis of certain diseases.

This review uses the term “enzyme selectivity,” which is the ability of an enzyme
to detect a non–complementary base pair in a substrate complex under certain conditions,
thus lowering the effectiveness of the enzymatic conversion of the imperfect complexes as
compared to perfect (fully complementary) ones. It is known that the ability of an enzyme to
identify a certain non–complementary base pair in a DNA–substrate depends on the
type of base pair, its nucleotide surroundings, and the location in regard to the site of the
enzymatic conversion. The selective activity of enzymes also depends on several external
factors, such as the buffering quality of the environment, temperature, and temporal
conditions, so an analysis of the literature does not lead to an easy establishment of the
general mechanisms of enzyme discrimination in some mismatches and tolerance towards others.
Some of the difficulties in the analysis and comparison of the effective detection of
mismatches are due to the different methods used for measuring the selectivity of enzymes in
various studies. Most often, the authors compare the following characteristics: yield of the
products of the enzymatic reaction, initial rates of product accumulation, and the ratio
between Vmax and Km. Usually, they consider the difference between the
values of the threshold cycle (ΔСt) during a real–time
PCR reaction for a perfect and imperfect template, or they analyze the
occurrence frequencies of the mismatch in the products of the enzymatic conversion of a random
oligonucleotide library paired into complexes with a DNA–template of known structure.

The catalytic cores of DNA–polymerases extracted from different organisms have varying
amino–acid sequences and belong to different families, but they still have a similar
structure and consist of three domains, which are assembled in a structure reminiscent in shape
of a half–open palm. The domains have appropriate names such as “palm,”
“thumb,” and other “fingers” (Fig.
2, А) [38–40]. The domains of the A–family DNA–polymerases consist of six
evolutionarily conserved motifs ( А , В , С , 1, 2 and 6), which are thought
to play the main role in the formation of the active site and the network of specific bonds
with the DNA–substrate [39–42]. The most conserved motifs are А , В and
С , two of which ( А and С ) are present in all the known DNA– and
RNA–polymerases. Motifs 1, 2 and 6 also have a fairly conservative spatial structure, but
they show more variety in terms of amino acid sequence. Compared to the highly conservative
А , В and С domains, these other domains are less involved in forming bonds
with DNA. To capture the dsDNA–substrate, the enzyme uses the “palm” (motifs
А , 2, 6) and the “thumb” (motif 1) domains. The “fingers” domain
closes above the “palm” forming a pocket (cavity) for the newly formed base pair.
This pocket is mainly made up of motif B amino acid residues. The fragments responsible for the
capture of the 3’–terminus of the primer, the inserted nucleotide, and the two
magnesium ions needed for the catalysis are all localized on the inner surface of the
“fingers” (motifs В , 6) and on the surface of the “palm” at the
base of the “fingers” (motifs А , С ). The polymerase active site,
which accomplishes the addition of nucleotides to the growing strand, is situated in the
“palm” domain [40, 42]. Some DNA–polymerases also have additional domains, which can, for
instance, add 3’ → 5’ exonuclease activity.

In order to perform effective catalysis, the molecules of DNA–processing enzymes undergo
conformational transitions. During a catalytic cycle, polymerases experience two main
structural changes (Fig. 4). The first is coupled with the
binding of the DNA–substrate, which enters the open crevice between the
“thumb” and the “palm” of the enzyme. The upper edge of the
“thumb” interacts with the substrate from the side of the minor groove of the
double helix, and thus it bends towards the surface of the palm. This causes the
“thumb” domain to form a hollow cylinder, which has a fragment of the
DNA–helix firmly lodged inside. Then, the second conformational change occurs; the
“fingers” turn towards the “palm,” which is coupled with the binding of
a nucleosidetriphosphate molecule in the polymerase’s active site. This change is called
the transition between the “open” and “closed” states of the enzyme,
and it is the final positioning and binding of the substrate in the enzyme’s active site.
This is the step when the bonds between the “fingers” domain and the inserted
nucleotide form, which allows to analyze the geometry of the transitional state, and thus the
complementarity of the forming base pair [49–51].

Changes in the structures of the enzyme and the substrate “tune” both of them to
each other, creating a whole network of protein–nucleic acid interactions based on
hydrogen and ionic bonds, as well as on Van–der–Waals interactions. This network of
bonds is highly specific, and the residues of the active site, which are the most conservative
ones, are often incorporated into this network. Unwinding of the DNA–duplex near the
enzyme’s active site increases the availability of various sites in the minor groove of
the double helix, which can in turn interact with the protein structure. For the most part,
these interactions are tight nonsequence–specific interactions, which are based on
hydrogen bonds between the centers present in any canonic base pair (electron acceptors, which
are in the N3 position of purine residues and in the O2 position of pyrimidine residues) and
the conserved amino acids in the protein [51, 52, 59, 60]. In turn, the induced A–form of the duplex is
stabilized by a network of Van–der–Waals interactions between the amino acid
residues and the carbohydrate fragments and/or heterocyclic bases of the nucleotide [43, 47, 51, 52].

There are many hypotheses on the factors affecting the sensitivity of DNA–dependent
enzymes towards noncanonic base pairs in the part of the DNA–substrate that is recognized
by the enzyme. These factors determine the selective activity of the enzymes including
DNA–ligases and DNA–polymerases. Several of these factors will be discussed

In the previous section, we reviewed the structure of DNA–dependent enzymes and the
complexes they form with the substrate in order to identify the factors affecting the
selectivity of enzymatic conversion. This section will review the structural traits of the
DNA–complex which can increase the selectivity of the enzymatic reaction. One of the
simplest ways to increase enzymatic ligation effectiveness is to use a method based on
“modified” probes, which consist of tandems of short oligonucleotides [83–85]. The
presence among the ligated components of mini–probes of penta– and even
tetra–nucleotides makes these composite complexes less effective as substrates, and their
enzymatic selectivity appears to be high [83, 85]. If a tetranucleotide is used as the central part of a
three–part tandem, the discrimination factor for any type of mismatch in the region of
the substrate complex is more than 300 when using the mesophilic Т 4 DNA–ligase
[85]. Such high selectivity of the enzyme is
unattainable if the DNA–duplex is formed by oligonucleotides, which are long enough to
provide optimal conditions for enzyme binding on the molecule (see [37] for an instance).

Another way to increase the selectivity of enzyme–dependent reactions not involving the
use of modified nucleotide analogs is based on the use of DNA–substrates with an
intentionally added mismatch next to the polymorphic site to be analyzed. The effectiveness of
such an approach was demonstrated for Taq DNA–polymerase [14, 86–90] and Tth DNA–ligase reactions [30]. This method involves placing the studied mismatch on the
3’–terminus of the elongated oligonucleotide or the ligated ОН
–component and the additional mismatch in proximity to the 3’–terminus,
specifically in the 2nd, 3rd or 4th positions [14, 30, 88–90], and in
some cases in the 5th or 7th [86].
In these cases, the “perfect” complex has a single planned noncomplementary pair,
while the complex with a mismatch contains two disruptions (Fig.

Currently, oligonucleotides carrying modified bases or with an altered
carbohydrate–phosphate backbone occupy a distinct niche in DNA hybridization probes
design. Some modifications (PNA peptydylnucleic acids [91;], LNA and ENA “locked”
nucleic acids [92, 93]) increase the stability of the modified complexes, which can be used to
increase the accuracy of DNA analysis at the level of hybrid complex formation. Other
modifications (N4–alkylcytosine [94],
5–methyl– and 5–(1–propargyl)uracyl [95]) can equalize the hybridizational characteristics of complexes with a
different nucleotide content, which is important during a parallel analysis of different DNA
sequences. It is worth noting that not all oligonucleotide modifications are compatible with
DNA–dependent enzymes, since the introduction of these modifications can disrupt the
protein–nucleic interactions which are needed for effective enzymatic catalysis.
Nevertheless, introduction of certain nucleotide analogs could become a method for increasing
the selectivity of enzymes towards mismatches in modified DNA–duplexes (Table 3).

One of the modifications used is a synthetic analog of a deoxyribonucleotide, which bears a
universal 3–nitropyrolle base. This base is named universal because it can form pairs
with all the natural bases thanks to its small size, which is comparable to that of the natural
bases, and the ability it retains to take part in stacking interactions. The effect of this
analog on the selectivity of the Tth DNA–ligase [30] was studied by introducing it into the 3rd position from the
pair to be analyzed, which was placed on the 3’–terminus of the ОН
–component. The choice of the position was based on data on increased selectivity upon
introduction of an additional mismatch, which was found to be optimal in the –3 position
from the enzymatic conversion site. The presence of the nucleotide analog caused a 9–fold
selectivity increase of the Tth DNA–ligase, which is 2.5–fold more
than the increased selectivity effect seen upon the introduction of an additional mismatch
based on canonic bases. Taq polymerase also exhibited decreased formation of
PCR products during the use of a single mismatch and a primer with a
3–nitropyrolle, as compared with a normal primer [96]. The unpredictable binding of oligonucleotides bearing such a modification
with the DNA–template is a disadvantage of this approach.

Such derivative oligonucleotides involve oligomers, which contain modified bicyclic
RNA–like monomers with 2’–O, 4’–C methylene and 2’–O,
4’–C ethylene links, LNA (Locked Nucleic Acid) [98–102], and
ENA (Ethylene Nucleic Acid) [103],

Increased selectivity of DNA–polymerases was demonstrated during the use of
oligonucleotide primers modified at the internucleotide phosphodiester residue [110–112].
Substitution of the native phosphate groups, located between the first and the second
nucleosides at the 3’–terminus of the primer, with thiophosphates increased the
3’–terminal mismatch discrimination efficiency of Vent and
Pfu DNA–polymerases [110–112]. Such modifications
did not alter the stability of lengthy DNA–complexes, but their presence increased the
discrimination of nucleotide mismatches, single or multiplex, and even those located at a
distance from the enzymatic conversion site. The DNA–polymerases showed no detectable
elongation of the modified oligonucleotide, even when the mismatches were located up to 8
nucleotides from the 3’–terminus of the primer [111]. Notably, according to the data presented above, DNA–polymerases
form tight interactions with the carbohydrate–phosphate backbone of the DNA helix up to
the above–mentioned position in the primer strand. However, the authors also noted that
for the conditions of the enzymatic reaction to be as stringent as possible, reactions using
phosphothioate analog oligonucleotides had to be performed with DNA–polymerases that
possesed exonuclease proofreading activity and the conditions for allele–specific
PCR had to be adjusted [112].

The data reviewed in this paper prove that the problem of achieving high selectivity in the
enzymatic conversion of oligonucleotide probes during nucleotide polymorphism analysis in DNA
is an issue depending on multiple factors. It is safe to assume that a universal analysis
scheme which allows an unequivocal discrimination of any nucleotide variation in DNA and which
uses the discussed analytic approaches has yet to be devised. The choice of a DNA analysis
scheme requires a complex design of the components of the analytical procedure, which factors
in the “two sides of the same medal.” Firstly, it is the specifics of substrate
complex recognition by the DNA–processing enzyme, and secondly, the structural
characteristics of the DNA–substrate which is formed by a molecular probe, based on an
oligonucleotide or its derivative. This review summarizes the most relevant facts that
characterize the peculiarities of nucleotide polymorphism analysis of DNA using
DNA–ligases and DNA–polymerases. The data presented reveal the fundamental
principles of selective oligonucleotide probe conversion during enzymatic DNA–analysis
and also point out the most promising recent developments in this field of research. Our
analysis of the available data shows that, despite the large amount of studies reviewed in this
paper, the problem of achieving selectivity in probe conversion remains unresolved and
undoubtedly requires further research.