114
inhibitors a priori bind into appropriate active site in the pol structure and compete
with natural enzyme substrates. Consequently, the structure of polymerase complex
with nucleoside inhibitor can be easy reconstructed in silico with high accuracy and
without any additional experimental data. On the contrary, the structural analysis
of DNA polymerase interaction with non-nucleotide compounds requires the preliminary identification of appropriate binding site(-s), which is a very sophisticated
task without additional input. Direct in silico identification of the interactive site,
based on structures of protein and ligand only (blind docking) in many cases is not
completely exact. So, it is not surprising that one usually uses a combination of
computational and instrumental approaches to identify non-nucleoside compounds’
sites on the DNA polymerase surface.
Mizushina et al. [82, 114] identified for the first time the mode of interaction
between the human DNA pol β and the lithocholic acid (LCA) (see Table 4.2). The
39-kDa pol β was separated proteolytically into two fragments corresponding to
the template-primer binding domain (8 kDa) and the catalytic domain (31 kDa).
It was shown that LCA bound tightly to the 8-kDa fragment but not to the 31-kDa
fragment. In
1
H–
15
N HMQC NMR analysis of pol β with LCA, the 8-kDa domain
bound to LCA as a 1:1 complex with a dissociation constant (K D ) of 1.56 mM. The
chemical shifts were observed only in residues mainly in helix-3, helix-4, and the
79–87 turn of the same face. No significant shifts were observed for helix-1, helix-2, and other loops of the 8-kDa domain [82]. The maximal shift was observed
for three amino acid residues—Lys60, Leu77, and Thr79 of pol β on the LCA. Obtained data were used for the further docking of LCA molecule into the appropriate
region on the pol β surface and interaction interface reconstruction [114].
Later, the same group reported about reconstruction pol β complexes with LCA
derivates—3-alpha-methoxy-5-beta-cholan-24-oic acid (compound 2) and 3-alphaO-lauroyl-5-beta-cholan-24-oic acid (compound 9) [115]. The docking was carried
out using a fixed docking procedure in the Affinity module within Insight II
3
modeling software (Accerlys Inc., San Diego, CA). The calculations used a CVFF force
field in the Discovery module and the Monte Carlo strategy in the Affinity module
of Insight II. Each energy-minimized final-docking position of LCA derivatives
was evaluated using interactive score function in the Ludi module. The Ludi score
includes the contribution of the loss of translational and rotational entropy of the
fragment, number and quality of hydrogen bonds, and contributions from ionic and
lipophilic interactions to the binding energy. According to the data obtained, compound 2 shares the same binding site with LCA. Compound 9, containing long fatty
acid moiety, is one of the strongest pol β inhibitors (K D = 1.7 nM) and binds into a
different site in the surface 8-kDa pol β domain. A critical role in the compound
9 binding belongs to amino acid residues Leu11, Lys35, His51, and Thr79 [115].
Second site is also able to bind another pol β inhibitor—nervonic acid (NA) [115].
Inhibitor of different eukaryotic DNA polymerase (pol α, pol δ, pol ε, pol γ,
pol ι, pol κ and TdT) sulfoquinovosylmonoacylglycerol (SQMG) also was found
to identify its binding mode with pol β [104]. According to the data of the NMR
3
Now Accerlys Inc. has discounted a development and support of Insight II. Functionality of this
software is transferred to Discovery Studio and Pipeline Pilot program suites.
A. Yu. Nyporko
inhibitors a priori bind into appropriate active site in the pol structure and compete
with natural enzyme substrates. Consequently, the structure of polymerase complex
with nucleoside inhibitor can be easy reconstructed in silico with high accuracy and
without any additional experimental data. On the contrary, the structural analysis
of DNA polymerase interaction with non-nucleotide compounds requires the preliminary identification of appropriate binding site(-s), which is a very sophisticated
task without additional input. Direct in silico identification of the interactive site,
based on structures of protein and ligand only (blind docking) in many cases is not
completely exact. So, it is not surprising that one usually uses a combination of
computational and instrumental approaches to identify non-nucleoside compounds’
sites on the DNA polymerase surface.
Mizushina et al. [82, 114] identified for the first time the mode of interaction
between the human DNA pol β and the lithocholic acid (LCA) (see Table 4.2). The
39-kDa pol β was separated proteolytically into two fragments corresponding to
the template-primer binding domain (8 kDa) and the catalytic domain (31 kDa).
It was shown that LCA bound tightly to the 8-kDa fragment but not to the 31-kDa
fragment. In
1
H–
15
N HMQC NMR analysis of pol β with LCA, the 8-kDa domain
bound to LCA as a 1:1 complex with a dissociation constant (K D ) of 1.56 mM. The
chemical shifts were observed only in residues mainly in helix-3, helix-4, and the
79–87 turn of the same face. No significant shifts were observed for helix-1, helix-2, and other loops of the 8-kDa domain [82]. The maximal shift was observed
for three amino acid residues—Lys60, Leu77, and Thr79 of pol β on the LCA. Obtained data were used for the further docking of LCA molecule into the appropriate
region on the pol β surface and interaction interface reconstruction [114].
Later, the same group reported about reconstruction pol β complexes with LCA
derivates—3-alpha-methoxy-5-beta-cholan-24-oic acid (compound 2) and 3-alphaO-lauroyl-5-beta-cholan-24-oic acid (compound 9) [115]. The docking was carried
out using a fixed docking procedure in the Affinity module within Insight II
3
modeling software (Accerlys Inc., San Diego, CA). The calculations used a CVFF force
field in the Discovery module and the Monte Carlo strategy in the Affinity module
of Insight II. Each energy-minimized final-docking position of LCA derivatives
was evaluated using interactive score function in the Ludi module. The Ludi score
includes the contribution of the loss of translational and rotational entropy of the
fragment, number and quality of hydrogen bonds, and contributions from ionic and
lipophilic interactions to the binding energy. According to the data obtained, compound 2 shares the same binding site with LCA. Compound 9, containing long fatty
acid moiety, is one of the strongest pol β inhibitors (K D = 1.7 nM) and binds into a
different site in the surface 8-kDa pol β domain. A critical role in the compound
9 binding belongs to amino acid residues Leu11, Lys35, His51, and Thr79 [115].
Second site is also able to bind another pol β inhibitor—nervonic acid (NA) [115].
Inhibitor of different eukaryotic DNA polymerase (pol α, pol δ, pol ε, pol γ,
pol ι, pol κ and TdT) sulfoquinovosylmonoacylglycerol (SQMG) also was found
to identify its binding mode with pol β [104]. According to the data of the NMR
3
Now Accerlys Inc. has discounted a development and support of Insight II. Functionality of this
software is transferred to Discovery Studio and Pipeline Pilot program suites.
A. Yu. Nyporko