Substrate specificity of E. coli uridine phosphorylase. Further evidences of high-syn conformation of the substrate in uridine phosphorolysis

Article abstract of DOI:10.1080/15257770.2016.1223306

Twenty five uridine analogues have been tested and compared with uridine with respect to their potency to bind to E. coli uridine phosphorylase. The kinetic constants of the phosphorolysis reaction of uridine derivatives modified at 2′-, 3′- and 5′-positions of the sugar moiety and 2-, 4-, 5- and 6-positions of the heterocyclic base were determined. The absence of the 2′- or 5′-hydroxyl group is not crucial for the successful binding and phosphorolysis. On the other hand, the absence of both the 2′- and 5′-hydroxyl groups leads to the loss of substrate binding to the enzyme. The same effect was observed when the 3′-hydroxyl group is absent, thus underlining the key role of this group. Our data shed some light on the mechanism of ribo- and 2′-deoxyribonucleoside discrimination by E. coli uridine phosphorylase and E. coli thymidine phosphorylase. A comparison of the kinetic results obtained in the present study with the available X-ray structures and analysis of hydrogen bonding in the enzyme-substrate complex demonstrates that uridine adopts an unusual high-syn conformation in the active site of uridine phosphorylase.

Full text of DOI:10.1080/15257770.2016.1223306

Nucleosides, Nucleotides and Nucleic Acids
ISSN: 1525-7770 (Print) 1532-2335 (Online) Journal homepage: http://www.tandfonline.com/loi/lncn20
Substrate Specificity of E. Coli Uridine
Phosphorylase. Further Evidences of High-
syn Conformation of the Substrate in Uridine
Phosphorolysis
C. S. Alexeev, G. G. Sivets, T. N. Safonova & S. N. Mikhailov
To cite this article: C. S. Alexeev, G. G. Sivets, T. N. Safonova & S. N. Mikhailov (2016): Substrate
Specificity of E. Coli Uridine Phosphorylase. Further Evidences of High-syn Conformation of
the Substrate in Uridine Phosphorolysis, Nucleosides, Nucleotides and Nucleic Acids, DOI:
10.1080/15257770.2016.1223306
To link to this article: http://dx.doi.org/10.1080/15257770.2016.1223306
Published online: 15 Nov 2016.
Submit your article to this journal
Article views: 2
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=lncn20
Download by: [University of Otago]
Date: 18 November 2016, At: 01:47

NUCLEOSIDES, NUCLEOTIDES AND NUCLEIC ACIDS
, VOL. , NO. , –

http://dx.doi.org/./..
Substrate Specificity of E. Coli Uridine Phosphorylase.
Further Evidences of High-syn Conformation of the
Substrate in Uridine Phosphorolysis
a
b
c
a
C. S. Alexeev , G. G. Sivets , T. N. Safonova , and S. N. Mikhailov
a
b
Engelhardt Institute of Molecular Biology, Russian Academy of Sciences, Moscow, Russia; Institute of
c
Bioorganic Chemistry, National Academy of Sciences of Belarus, Minsk, Belarus; Bach Institute of
Biochemistry, Russian Academy of Sciences, Moscow, Russia
ABSTRACT
ARTICLE HISTORY
Received  April 
Accepted  August 
Twenty five uridine analogues have been tested and compared
with uridine with respect to their potency to bind to E. coli uri-
dine phosphorylase. The kinetic constants of the phosphorolysis
reaction of uridine derivatives modified at 2 -, 3 - and 5 -positions
of the sugar moiety and 2-, 4-, 5- and 6-positions of the hetero-
cyclic base were determined. The absence of the 2 - or 5 -hydroxyl
group is not crucial for the successful binding and phosphoroly-
KEYWORDS
ꢀ
ꢀ
ꢀ
Uridine phosphorylase;
substrate specificity;
modified uridines;
conformation of uridine
ꢀ
ꢀ
ꢀ
ꢀ
sis. On the other hand, the absence of both the 2 - and 5 -hydroxyl
groups leads to the loss of substrate binding to the enzyme. The
ꢀ
same effect was observed when the 3 -hydroxyl group is absent,
thus underlining the key role of this group. Our data shed some
ꢀ
light on the mechanism of ribo- and 2 -deoxyribonucleoside dis-
crimination by E. coli uridine phosphorylase and E. coli thymidine
phosphorylase. A comparison of the kinetic results obtained in
the present study with the available X-ray structures and analysis
of hydrogen bonding in the enzyme-substrate complex demon-
strates that uridine adopts an unusual high-syn conformation in
the active site of uridine phosphorylase.
Introduction
Nucleoside phosphorylases are involved in salvage pathways of nucleoside biosyn-
thesis and catalyze the reversible phosphorolysis of nucleoside to a free heterocyclic
base with the formation of α-D-ribose-1-phosphate (Sug-p), and the equilibrium
[
1, 2]
of this reaction is shifted towards the nucleoside.
(
Thus, uridine phosphorylase
[2]
UP) catalyzes the following reaction depicted in Scheme 1. This class of enzymes
includes thymidine phosphorylase (TP; EC 2.4.2.4), UP (EC 2.4.2.3), and purine
nucleoside phosphorylase (PNP; EC 2.4.2.1). These enzymes were found in virtually
all organisms and share similar primary structures. The most specific enzyme is TP,
whose substrate is thymidine. UP catalyzes the phosphorolysis of both thymidine
CONTACTS S. N. Mikhailov
Sciences, Vavilov str. , Moscow , Russia.
 Taylor & Francis Group, LLC
smikh@eimb.ru
Engelhardt Institute of Molecular Biology, Russian Academy of
©

2
C. S. ALEXEEV ET AL.
Scheme . Reversible uridine phosphorolysis catalyzed by UP.
and uridine, whereas cytidine is not a substrate for both enzymes. PNP may cleave
ꢀ
[1,2]
the N-glycosidic bond in purine ribo(and 2 -deoxyribo)nucleosides.
UP activity is usually elevated in various tumor tissues, including breast can-
cer, compared to matched normal tissues and this induction appears to contribute
to the therapeutic efficacy of fluoropyrimidines in cancer patients.[3] Bacterial uri-
dine phosphorylase functions as a dimer composed of identical subunits that have
a molecular mass most commonly reported equal to 27.5 kDa.[
1]
Nucleoside phosphorylases are used in industry for the synthesis of drugs
[4–7]
(
Cladribin, Fludarabine, and Nelarabin) and practically important nucleosides.
ꢀ
The ability of UP and PNP to cleave 2 -deoxynucleosides, β-D-arabinofuranosyl
nucleosides and to catalyze their formation[ is widely used. In these cases, enzy-
matic methods for creating the N-glycosidic bond can compete with the chemical
synthesis. In general, this transglycosylation reaction is shown in Scheme 2.
The knowledge of the substrate specificity and equilibrium constants will help
6,8]
2
in predicting and optimizing the yield of the desired Nuc . For example, the use of
2
a lower concentration of inorganic phosphate and an excess of Base will shift the
2
equilibrium towards the Nuc formation. In these reactions one nucleoside phos-
2
phorylase may be used to catalyze the formation of pyrimidine or purine Nuc
1
2
starting from the appropriate Nuc and pyrimidine or purine (Base ). In the sec-
ond variant both UP and PNP may be used to transform pyrimidine Nuc to
purine Nuc . This approach is very popular because the chemical transformation
1
2
of the sugar moiety of pyrimidine nucleosides, for example, the synthesis of 1-β-
D-arabinofuranosyluracil is, in general, more efficient and less costly compared to
purine nucleosides. Moreover, the equilibrium of PNP phosphorolysis is shifted to
the nucleoside to a much higher extent compared with UP and may be considered
as a driving force for the total transglycosylation process. Some other approaches
were employed to improve the yields of transglycosylation reactions. For exam-
ple, α-D-arabinose-1-phosphate was successfully used for the preparation of β-D-
[8]
arabinofuranosyl nucleosides. 7-Methylguanosine derivatives as a source of the
ribose moiety have important advantages because their phospholysis with PNP is
[9–12]
shifted towards the formation of α-D-ribose-1-phosphate.
Scheme. Nucleosidetransglycosylationreaction(transferofthesugarmoietyfromoneheterocyclic
base to another via α-D-sugar--phosphate, Sug-p).

NUCLEOSIDES, NUCLEOTIDES AND NUCLEIC ACIDS
3
Bacterial nucleoside phosphorylases are rather thermostable and may be utilized
at elevated temperatures (50°–60°C), thus providing higher solubility of substrates
and, consequently, their higher concentration in the reaction mixture. This is espe-
cially important for purine heterocyclic bases and nucleosides poorly soluble in
aqueous buffers. The particular conditions of the transglycosylation reaction should
be optimized to ensure the maximum yields.
The substrate specificity of these enzymes has not been fully investigated, and
the role of the functional groups of nucleosides in the formation of the produc-
tive enzyme-substrate complex is unknown. This is due to the lack of convenient
methods for the determination of kinetic parameters of the phosphorolysis reaction.
Therefore, knowledge of the substrate specificity and the architecture of the active
site of these enzymes is of considerable interest. The investigation of the substrate
properties and determination of the kinetic parameters is an important step towards
the creation of enzymes with improved characteristics. It should be noted that the
kinetic parameters of this enzyme have been measured predominantly for natural
ꢀ
[13, 14]
ribo- and 2 -deoxynucleosides
and studies of the properties of modified nucle-
[15–17]
osides were limited to the determination of their inhibitory properties.
The aim of this work is to study the substrate and inhibitory properties of uridine
derivatives in the reactions catalyzed by E. coli UP in order to shed some light on
the conformation of the substrate in the productive complex with the enzyme.
Results and discussion
We studied the E. coli UP-catalyzed phosphorolysis of uridine and its 25 deriva-
tives modified in the heterocyclic base and the sugar moiety. The kinetic constants
(
6
K , K , k ) of the phosphorolysis reaction of uridines modified at 2-, 4-, 5- and
-positions of the heterocyclic base and 2 -, 3 - and 5 -positions of the sugar moiety
M I cat
ꢀ
ꢀ
ꢀ
were determined. These data are summarized in Table 1.
-Methyluridine is slowly converted into 6-methyluracil, and the substrate bind-
ing is significantly weaker compared to the 5-methyl derivative. These results are
6
[
18]
in agreement with those reported previously.
The replacement of oxygen with
[19]
˚
˚
sulfur (the van der Waals radii are O, 1.40 A; S, 1.85 A ) in position four has no
appreciable effect on the substrate affinity for the enzyme (Table 1). By contrast, 2-
thiouridine is not a substrate for UP. The substrate properties of UP are summarized
in Figure 1.
ꢀ
ꢀ
2
-Deoxyuridine and 5 -deoxyuridine show rather good substrate properties,
as follows from their kinetic constants (Table 1). Therefore, we can assume that
ꢀ
ꢀ
the absence of the 2 - or 5 -hydroxyl group is not crucial for successful bind-
ꢀ
ꢀ
ing and phosphorolysis. On the other hand, the absence of both the 2 - and 5 -
hydroxyl groups results in the inability of the substrate to bind to the enzyme,
as in the case of 5 -deoxythymidine, 5 -deoxy-5 -chlorothymidine and other 5 -
deoxy derivatives. The lack of substrate properties of 3 -fluoro-3 -deoxyuridine and
-deoxythymidine clearly demonstrates the key role of the substrate 3 -hydroxyl
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
3
group. Our data provide evidence that hydroxyl groups of the ribose moiety of

4
C. S. ALEXEEV ET AL.
Table . Kinetic constants and equilibrium constants (Keq) for uridine and its analogs in phosphorol-
ysis reaction catalyzed by E. coli UP at pH . and °C.
kcat/KM
kcat/KM
−

−
KM (KI) kcat
sec
KM (KI) kcat
sec
−
−
Substrate
Urd
μM
sec
/μM
Substrate
Thd
μM
sec
/μM


.


.
∗
∗∗



.
—

-thiouridine
-thiouridine


.
.


.
-methyluridine
-methyluridine



.


.
ꢀ

-deoxyuridine
-β-D-arabino-
furanosyluracil



.


.
ꢀ
ꢀ

-homouridine
 -C-methyluridine (D-allo)
∗
∗∗



.
—
ꢀ
ꢀ

-C-methyluridine (L-talo)
 -homothymidine
(
Continued on next page)

NUCLEOSIDES, NUCLEOTIDES AND NUCLEIC ACIDS
5
Table . (Continued).
Substrate
KM (KI) kcat kcat/KM
−
Substrate
KM (KI) kcat
kcat/KM
−

−
−
μM
sec
sec
μM
sec
sec
/
μM
/μM
(
)
—
—

.
.
ꢀ
ꢀ

-C-methylthymidine (D-
ribo)
 -C-methylthymidine (L-
lyxo)
∗
∗∗



.
—
—
—
—
ꢀ
ꢀ

-deoxyuridine
 -deoxythymidine



.
—
—
ꢀ
ꢀ
ꢀ

-deoxy- -
 -deoxythymidine
methyleneuridine
∗
∗∗
∗
∗∗
—
ꢀ
ꢀ
ꢀ ꢀ
 -deoxy- -azidothymidine

-deoxy- -chlorothymidine
∗
∗∗
∗
∗∗
—
ꢀ
ꢀ
ꢀ
ꢀ

-deoxy- -bromothymidine
 -deoxy- -iodothymidine
(
)
—
—
()
—
—
ꢀ
ꢀ
ꢀ

-fluoro- -deoxyuridine
, -O-anhydrouridine
(
Continued on next page)

6
C. S. ALEXEEV ET AL.
Table . (Continued).
Substrate
KM (KI) kcat kcat/KM
Substrate
KM (KI) kcat kcat/KM
−
−
−
−
sec
μM
sec
sec
μM
sec
/
μM
/μM
∗
∗∗
∗
∗∗
—
—
ꢀ
ꢀ
 -C-methyluridine

-C-methyluridine
a
The accuracy of the estimation is better than % for Km and kcat.
Km> μM; the reaction rate was smaller than .% of that for the natural substrate; For the preliminary
∗
∗∗
account of this work see Ref [].
the substrate are involved in hydrogen bonding, thus providing its binding to the
enzyme to form the UP substrate complex.
Our data shed some light on the mechanism of discrimination between ribo- and
ꢀ
2
-deoxyribonucleosides by UP and TP. The most specific enzyme is TP, whose sub-
ꢀ
strates are thymidine and 2 -deoxyuridine. UP catalyzes the phosphorolysis of both
thymidine and uridine. Uridine and 5-methyluridine are not E. coli TP substrates,
but they effectively inhibit the enzyme activity.[ On the other hand, UP catalyzes
the phosphorolysis of uridine 20 times more effectively than 5-methyluridine (cf.
20]
ꢀ
the k /K values). This difference between 2 -deoxyuridine and thymidine is sub-
cat
M
stantially smaller (5 times) (Table 1). Meanwhile, UP catalyzes the phosphorolysis
of uridine 60 times more effectively compared to thymidine. The above pairwise
ꢀ
comparison provides evidence that the presence of the 2 -hydroxyl group with the
simultaneous absence of the bulky 5-methyl group is crucial for the formation of the
productive UP substrate complex.
The efficiency of phosphorolysis of 1-β-D-arabinofuranosyluracil is 300 times
ꢀ
lower than that of uridine but only 25 times lower as compared to 2 -deoxyuridine.
ꢀ
Meanwhile, 2,2 -O-anhydrouridine, a nucleoside with a fixed orientation around the
glycosidic bond, binds to E. coli UP more effectively than uridine. The same situation
has been observed earlier for UPs of different origin.[
15–17]
Let us consider the effect of substituents on the nucleoside conformation. It is
generally believed that nucleosides are rather flexible molecules, resulting in their
Figure . Tolerated (left) and untolerated (right) modifications of uridine.

NUCLEOSIDES, NUCLEOTIDES AND NUCLEIC ACIDS
7
Figure . Anti and syn conformations of uridine (top) and the steric contacts in high-anti and high-syn
ꢀ
ꢀ
conformations with the  -proton and the  -C-methyl group (bottom).
existence in a manifold of different conformational states separated by low-energy
barriers, which can be overcome at room temperature. The rotation around the gly-
cosidic linkage in natural uridine is hindered by steric encumbrance between the
ꢀ
base H6 proton and the ribose H2 proton (high-anti conformation) and between
ꢀ
the C2 carbonyl group and the H2 proton (high-syn conformation), resulting
in two energy barriers. The former barrier is relatively low, whereas the latter
one is much higher. The anti and syn conformations of uridine are shown in
Figure 2. In pyrimidine derivatives the barriers to hindered rotation around the
N-glycosidic bond, resulting from steric encumbrance between the C2 carbonyl
ꢀ
group and the H2 proton, were estimated by Wagner and Jochims using NMR
[22]
spectroscopy. Derivatives of 1-(β-D-ribofuranosyl)isocyanuric acids (Figure 3)
were used as model compounds. Two methyl groups of l,3-dimethyl-5-(2,3,5-tri-O-
acetyl-β-D-ribofurunosyl)-s-triazine-2,4,6(1H,3H,5H)-trione have different chem-
ical shifts due to different environments. Two singlets at low temperatures coalesce
to one singlet at elevated temperature. The Gibbs free energies of activation of the
rotation around the glycosidic bond in this model compound were estimated at 43–
−
1 [22]
4
6 kJmoI .
.
Figure . Structure of -(β-D-ribofuranosyl)isocyanuric acid.

8
C. S. ALEXEEV ET AL.
In nucleosides the least favorable conformations of the ribose moiety in terms
ꢀ
ꢀ
of steric interactions are W (04 -exo) and E (04 -endo) conformations, in which the
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
substituents at C2 and C3 are eclipsed with the nearly zero C1 -C2 -C3 -C4 torsion
−
1 [23]
angle. The activation energy barrier was estimated at 20–23 (± 0.9) kJ mol .
Functionally competent analogues of nucleosides and nucleotides, namely C’-
methylnucleosides and their phosphoric esters, are widely used for the study of dif-
ferent enzymes catalyzing the synthesis and decomposition of nucleic acids.[24–26]
The presence of bulky methyl groups can give rise to intramolecular and inter-
molecular steric contacts when fixing the substrate in a particular conformation of
the enzyme-substrate complex, thus preventing substrate binding. Conformational
maps (the influence of the pseudorotation angle P and the glycoside angle χ on
the potential energy E) were calculated for nucleosides and C’-methylnucleosides
in order to detect intramolecular contacts between the methyl group and the het-
[24–26]
ꢀ
erocyclic base.
Energetically forbidden conformations were observed in 3 -
C-methylnucleosides only in the case of N-conformers. The restricted conforma-
ꢀ
tions of 5 -C-methylnucleosides were observed only if the methyl group lies above
ꢀ
the furanose ring. The introduction of the bulky methyl group in the 2 -position
leads to a substantial increase in the energy barrier to the syn-anti conversion due
to intramolecular contacts of this group with the 2-keto group and the proton
[24–26]
in 6-position of pyrimidines.
. These steric interactions should be taken into
account in the discussion of the substrate properties of C’-methylnucleosides.
ꢀ
ꢀ
ꢀ
The replacement of protons with a methyl group at the 2 , 3 or 5 positions of
the nucleoside gives nucleoside analogs containing all functional groups of natural
compounds, for example, all possible binding sites for the complex formation with
the enzyme. As a result, comparable binding constants for analogs and natural
compounds should be observed. Two extreme cases of the applicability of the
[
24–26]
proposed
approach may be considered: (1) the enzyme efficiently binds and
is fulfilled.
transforms the analog; therefore, the condition E(analog)∼E
(natural substance)
(
2) the enzyme does not bind the analog; the substrate conformation may be located
in the regions where E(analog)>> E(natural substance) provided that the introduction of
a bulky methyl group results in intramolecular rather than intermolecular steric
hindrance. This concept can explain the conformation of the substrate during its
[
24–26]
transformation by several enzymes of nucleic acid biosynthesis.
ꢀ
Both 5 -C-methyluridines are substrates for UP, and the efficiency of phos-
phorolysis of 5 -C in the L-talo configuration is nearly the same as for uri-
ꢀ
dine. Another diastereomer in the D-allo configuration reacts with phosphate an
order of magnitude slower. The same configurational preferences may be seen
ꢀ
ꢀ
for 5 -C-methylthymidine (L-talo) but the phosphorolysis is much slower. 2 -C-
ꢀ
Methyluridine and 3 -C-methyluridine are not substrates for UP and virtually do
not bind to the enzyme.
ꢀ
ꢀ
Molecular modeling of the binding of 2 -C-methyluridine and 3 -C-
methyluridine to UP was performed based on the X-ray structure of the complex
[27]
ꢀ
of uridine with UP. When complexed with UP, 3 -C-methyluridine forms a close
ꢀ
˚
steric contact (2.45 A) between the 3 -C-methyl group and sulfur of methionine

NUCLEOSIDES, NUCLEOTIDES AND NUCLEIC ACIDS
9
ꢀ
[]
Figure . High-syn and C -exo conformation of uridine complexed with a sulfate ion and UP. The
views from different sides.
ꢀ
1
94. By contrast, the additional methyl group of 2 -C-methyluridine in the complex
with UP does not form contacts with the enzyme but causes intramolecular steric
hindrance.
The obtained kinetic results may be best explained by the binding of UP to uri-
dine in the high-syn conformation. There is the following evidence for this unusual
conformation of the uridine complexed with UP when the carbonyl is located near
ꢀ
the 2 -proton. The efficiency of phosphorolysis of 1-β-D-arabinofuranosyluracil is
ꢀ
2
5 times lower as compared to 2 -deoxyuridine evidently due to the steric hindrance
ꢀ
caused by the 2 -hydroxyl group and the 2-keto group. The tendency to lose sub-
strate properties is observed with an increase in the van der Waals radius of the 2 -
substituent (H, 1.30A; O, 1.40A; Me, 2.00A ). Meanwhile, 2,2 -O-anhydrouridine
that models the high-syn conformation of uridine binds to UP better than uridine,
despite the absence of a proton in the 3-position of the heterocyclic base, thus play-
ing an important role in the substrate binding.[1] In this case, the orientation of the
ꢀ
[19]
ꢀ
˚
˚
˚
heterocycle substantially outweighs the contribution of hydrogen bonding by the
ꢀ
2
2
-hydroxyl and 3-NH groups in the active site of UP. This strongly suggests that the
ꢀ
-keto group is located quite close to 2 -H in the substrate-UP complex. This is in
line with the lack of substrate properties of 2-thiouridine.
It is interesting to compare the obtained results with the X-ray diffraction data
for the complexes of uridine with UP. The high-resolution structure of Shewanella
oneidensis MR-1 uridine phosphorylase complexed with uridine and sulfate[ is
best suited for this purpose because the active site region of the enzyme in this com-
plex is well resolved and all contacts (hydrogen bonds and hydrophobic contacts)
are reliably identified. Besides, the presence of sulfate, which mimics phosphate, is
crucial for the formation of the active site.
27]
This enzyme is homologous to E. coli UP and the active sites of both enzymes
are virtually identical. The equilibrium constants and the kinetic characteristics of
these enzymes with respect to uridine, thymidine and 5-methyluridine are nearly
[28]
the same.
The unusual high-syn conformation of the substrate was unequivocally estab-
lished by X-ray crystallographic analysis of Shewanella oneidensis UP complexed
[27]
˚
with sulfate and uridine (1.75 A resolution).
This structure is characterized by
the torsion angle C2-N1-C1´-C2´= 13.5° and the N1-C1´ glycosidic bond lengths
ꢀ
˚
of 1.49 A (Figure 4). The ribose residue adopts the least favorable C4 -exo confor-
ꢀ
ꢀ
mation in terms of steric interactions, in which the C2 and C3 hydroxyl groups are

1
0
C. S. ALEXEEV ET AL.
Figure . The active site containing uridine and a sulfate ion. His belongs to another subunit. Hydro-
gen bonds are shown as orange broken lines.
eclipsed and both the CH OH group and the heterocyclic base (bulky substituents)
2
are maximally axial (Figure 4). Additionally, the ribose ring is highly flattened, as
ꢀ
ꢀ
ꢀ
ꢀ
˚
can be seen from the torsion angles (C1 -C2 -C3 -C4 is −2°). There is also a very
ꢀ
ꢀ
short contact between the O2 and O3 atoms (2.25 A). The distance between the
ꢀ
˚
carbonyl O2 atom and the C2 atom of the ribose is 2.92 A (the distance between
ꢀ
[27]
˚
H2 and O2 can be estimated at 1.8 A).
All functional groups of the uridine molecule are involved in an extensive hydro-
gen bond network with UP (see Figure 5). Thus, the O4 atom of uridine forms a
hydrogen bond with the NH2 atom of Arg165, the O2 atom forms a hydrogen bond
with Gln163, and the latter residue forms a hydrogen bond with the N3 atom of
ꢀ
ꢀ
uridine. The ribose residue forms several hydrogen bonds: O4 and Thr91, O2 and
Glu195, O3 and Glu195, O5 and His5.
ꢀ
ꢀ
[27]
Conclusions
The combined kinetic and structural data provide clear evidence that UP binds uri-
dine in the most energetically unfavorable conformation, which, to the best of our
knowledge, has no precedents in the enzymes of nucleic acid metabolism. This is
possible due to multiple interactions between the substrate and the protein envi-
ronment (active site residues) mainly through hydrogen bonds. These results are
important for understanding the mechanism of action of this class of enzymes.
An analysis of the conformations of nucleosides in solution and rotational bar-
riers suggests that the energy difference between the ground state of uridine and
−
1
uridine complexed with UP may be as high as 63–69 kJ mol . The binding in
a high-energy conformation results in the weakening of the glycosidic bond. The

NUCLEOSIDES, NUCLEOTIDES AND NUCLEIC ACIDS
11
observed conformation of uridine complexed with sulfate (mimetic of phosphate)
may be very similar to its conformation in the transient state.
The ability of E. coli UP to cleave and catalyze the formation of β-D-ribo-(2-
deoxyribo- and arabino-)furanosyl nucleosides is widely used in enzymatic trans-
glycosylation reactions, in which a sugar residue is transferred from a pyrimidine
nucleoside to the purine heterocyclic base.
Experimental part
The following reagents were used in this work: thymidine, uridine, 5-methyluridine,
ꢀ
ꢀ
2
-deoxyuridine, 1-β-D-arabinofuranosyluracil, 2,2 -O-anhydrouridine, (Sigma,
USA). Kinetic constants were measured using a Cary 300 spectrophotometer (Var-
ian, Australia). High performance liquid chromatography was performed on a
Gilson chromatographer (USA) using a 4.6 × 150 mm column (5 μm, Nucleosil
C18).
The following nucleosides were obtained according to the previously pub-
[
29]
ꢀ
[30]
ꢀ
lished protocols: 4-thiouridine , 5 -homouridine , 5 -C-methyluridines (D-
[31,32]
ꢀ
[33]
ꢀ
allo and L-talo)
, 5 -homothymidine , 5 -C-methylthymidines (D-ribo
and L-lyxo) , 5 -deoxyuridine , 5 -deoxythymidine , 1-(5,6-dideoxy-
β-D-ribo-hex-5-enofuranosyl)adenine
[33]
ꢀ
[34]
ꢀ
[35]
ꢀ
ꢀ
[36]
[39]
ꢀ
(5 -deoxy-5 -methyleneuridine)
,
.
ꢀ
ꢀ
[37]
ꢀ
[38]
ꢀ
3
2
-fluoro-3 -deoxyuridine
,
3 -C-methyluridine
,
2 -C-methyluridine
ꢀ
-Thiouridine and 6-methyluridine were the gift of Dr. A.Holy. 5 -Chloro-5 -
deoxythymidine, 5 -bromo-5 -deoxythymidine, 5 -iodo-5 -deoxythymidine, 5 -
azido-5 -deoxythymidine were the gift of Dr. A.M. Kritzyn. (Chemical structures
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
of modified nucleosides are shown in Table 1.)
Enzymic activity assays for E. coli UP
The UP activity was assayed spectrophotometrically at 25°C and estimated from
the difference in A280 for the nucleoside and the heterocyclic base resulting
from its enzymatic cleavage. Calculations were performed assuming ꢀε
=
280
−
1
−1
2
030 M ·cm for the uridine/uracil pair. Measurements were taken in 50 mM
potassium-phosphate (pH 7.5). The reaction mixture (1 mL) contained 0.5 μmol
of uridine and 0.18 units of UP. E. coli UP was obtained according to the literature
[40]
procedure
and was supplied by Dr. R.S. Esipov from Shemyakin–Ovchinnikov
Institute of Bioorganic Chemistry of the Russian Academy of Sciences.
Estimation of the Michaelis constant (K ) for phosphorolysis of uridine and its
M
derivatives
The constant KM of phosphorolysis of uridine by UP was determined from the
spectrophotometric estimation of the reaction rate at 280 nm. The substrate con-
centration varied in the range of 10−200 μM for uridine derivatives at a constant
phosphate concentration (50 mM), pH 7.5, at 25°C. The reactions were initiated by
adding 0.03 units of the enzyme; the rate was measured for 5 min. To obtain K , we
M

1
2
C. S. ALEXEEV ET AL.
used nonlinear regression analysis of the Michaelis–Menten and Lineweaver–Burk
plots.
Estimation of the inhibition constant (KI)
K of uridine derivatives were measured spectrophotometrically[29]. An enzyme
I
was preincubated at various concentrations of an inhibitor at 25°C for 15 min,
and the initial rate was measured for phosphorolysis of the natural substrate or its
thio derivative at 280 and 365 nm in the case of uridine and 4-thiouridine, respec-
tively. Substrates were used at 0.5 or 1 K . The estimation of K followed the Dixon
M
I
method.
Conflict of interest
The authors confirm that this article content has no conflict of interest.
Author contributions
Conceived and designed the experiments: S.N. Mikhailov. Performed the kinetic experiments:
C.S. Alexeev. Analyzed the data: S.N. Mikhailov. Analysis of crystallographic data: T. N. Safonova.
ꢀ
ꢀ
ꢀ
Synthesis of 3 -deoxythymidine and 3 -fluoro-3 -deoxyuridine: G.G.Sivets. Wrote the paper: S.N.
Mikhailov.
Acknowledgments
The research was supported by the Russian Science Foundation, grant No 16-14-00178. T. N.
Safonova was supported by the Russian Science Foundation (RSF; grant No. 14-24-00172).
The authors are grateful to Dr. K.M. Polyakov and Dr. N.G. Panova for help and fruitful dis-
cussions.
ORCID
S. N. Mikhailov http://orcid.org/0000-0002-1363-9706
References
1
2
3
4
5
. Pugmire, M.; Ealick, S. Structural analyses reveal two distinct families of nucleoside phos-
phorylases. Biochem. J. 2002, 361, 1–25.
. Lewkowicz, E.S.; Iribarren, A.M. Nucleoside Phosphorylases. Curr. Org. Chem. 2006, 10,
1
197–1215.
. Yan, R.; Wan, L.; Pizzorno, G.; Cao, D. Uridine phosphorylase in breast cancer: a new prog-
nostic factor. Front Biosci. 2006, 11, 2759–2766.
. Utagawa, T. Enzymatic preparation of nucleoside antibiotics. J. Mol. Catal. B: Enzym. 1999,
6
, 215–222.
. Zaks, A. Industrial biocatalysis. Curr. Opin. Chem. Biol. 2001, 5, 130–136

NUCLEOSIDES, NUCLEOTIDES AND NUCLEIC ACIDS
13
6
7
. Mikhailopulo, I.A. Biotechnology of nucleic acid constituents-State of the art and perspec-
tives. Curr. Org. Chem. 2007, 11, 317–335.
. Iglesias, L.E.; Lewkowicz, E.S.; Medici, R.; Bianchi, P.; Iribarren, A.M. Biocatalytic
approaches applied to the synthesis of nucleoside prodrugs. Biotechnol. Adv. 2015, 33,
4
12–434.
8
9
. Konstantinova, I.D.; Antonov, K.V.; Fateev, I.V.; Miroshnikov, A.I.; Stepchenko, V.A.; Bara-
novsky, A.V.; Mikhailopulo, I.A. A Chemo-Enzymatic Synthesis of β;-D-Arabinofuranosyl
Purine Nucleosides. Synthesis 2011, 1555–1560.
. Kulikowska, E.; Bzowska, A.; Wierzchowski, J.; Shugar, D. Properties of two unusual,
and fluorescent, substrates of purine-nucleoside phosphorylase: 7-methylguanosine and 7-
methylinosine. Biochimica et Biophys. Acta (BBA)-Protein Struct. Mol. Enzymology. 1986,
8
74, 355–363.
1
0. Hennen, W.J.; Wong, C.H. A new method for the enzymic synthesis of nucleosides using
purine nucleoside phosphorylase. J. Org. Chem. 1989, 54 (19), 4692–4695.
1
1. Nixon, A.E.; Hunter, J.L.; Bonifacio, G.; Eccleston, J.F.; Webb, M.R. Purine nucleoside phos-
phorylase: its use in a spectroscopic assay for inorganic phosphate and for removing inor-
ganic phosphate with the aid of phosphodeoxyribomutase. Anal. Biochem. 1998, 265, 299–
3
07.
1
2. Ubiali, D.; Morelli, C.F.; Rabuffetti, M.; Cattaneo, G.; Serra, I.; Bavaro, T.; Speranza, G. Sub-
strate Specificity of a Purine Nucleoside Phosphorylase from Aeromonas hydrophila Toward
6
-Substituted Purines and its Use as a Biocatalyst in the Synthesis of the Corresponding
Ribonucleosides. Curr. Org. Chem. 2015, 19, 2220–2225.
13. Krenitsky, T.A. Uridine phosphorylase from Escherichia coli: Kinetic properties and mecha-
nism. Biochimica et Biophys. Acta (BBA)-Enzymology 1976, 429 (2), 352–358.
14. Leer, J.C.; Hammer-Jespersen, K.; Schwartz, M. Uridine Phosphorylase from Escherichia coli.
Physical and Chemical Characterization. Eur. J. Biochem. 1977, 75, 217–224.
15. Niedzwicki, J.G.; El Kouni, M.H.; Chu, S.H.; Cha, S. Structure-activity relationship of
ligands of the pyrimidine nucleoside phosphorylases. Biochem. Pharmacol. 1983, 32,
3
99–415.
1
6. el Kouni, M.H.; Naguib, F.N.; Chu, S.H.; Cha, S.M.; Ueda, T.; Gosselin, G.; Imbach, J.-L.;
Sheally, F.; Otter, B.A. Effect of the N-glycosidic bond conformation and modifications in the
pentose moiety on the binding of nucleoside ligands to uridine phosphorylase. Mol. Phar-
macol. 1988, 34, 104–110.
1
7. el Kouni, M.H.; Naguib, F.N.; Panzica, R.P.; Otter, B.A.; Chu, S.H.; Gosselin, G.; Chu, C.K.;
Schinazi, R.F.; Shealy, Y.F.; Goudgaon, N.; Ozerov, A.A.; Ueda, T.; Iltzsch, M.H. Effects of
modifications in the pentose moiety and conformational changes on the binding of nucle-
oside ligands to uridine phosphorylase fromToxoplasma gondii. Biochem. Pharmacol. 1996,
5
1, 1687–1700.
1
8. Krajewska, E.; Shugar, D. Pyrimidine ribonucleoside phosphorylase activity vs 5-and/or 6-
substituted uracil and uridine analogues, including conformational aspects. Biochem. Phar-
macol. 1982, 31(6), 1097–1102.
1
9. GordonA.J.; FordR.A. In: The Chemist’s Companion: A Handbook of Practical Data, Tech-
niques, and References. 1973, New-York: Wiley-Interscience.
2
0. Panova, N.G.; Alexeev, C.S.; Kuzmichov, A.S.; Shcheveleva, E.V.; Gavryushov, S.A.; Polyakov,
K.M.; Kritzyn, A.M.; Mikhailov, S.N.; Esipov, R.S.; Miroshnikov, A.I. Substrate specificity of
E. coli thymidine phosphorylase. Biochemistry (Moscow) 2007, 72(1), 21–28.
1. Alexeev, C.S.; Polyakov, K.M.; Sivets, G.G.; Safonova, T.N.; Mikhailov, S.N. Substrate speci-
ficity of E. coli uridine phosphorylase. Evidence of high-syn conformation of substrate. Col-
lect. Symp. Ser. 2014, 14, 213–216.
2
2
2. Wagner, G.; Jochims, J.C. Barrieren der behinderten Rotation urn die N-glycosidische
Bindung. IV. 1-(β;-D-Ribofuranosy1)isocyanursauren. Chem. Ber. 1979, 112, 1941–1947.

14
C. S. ALEXEEV ET AL.
2
2
3. Thibaudeau, C.; Acharya, P.; Chattopadhyaya, J. In: Stereoelectronic effects in nucleosides and
nucleotides and their structural implications. 2005, Sweden: Uppsala University Press.
4. Moiseyev, G.P.; Yakovlev, G.I.; Lysov, Yu.P.; Chernyi, A.A.; Polyakov, K.M.; Oivanen,
M.; Lonnberg, H.; Beigelman, L.N.; Efimtseva, E.V.; Mikhailov, S.N. Determination of
the nucleotide conformation in the productive enzyme-substrate complexes of RNA-
depolymerases. FEBS Lett. 1997, 404, 169–172.
2
5. Tunitskaya, V.L.; Rusakova, E.E.; Padyukova, N.Sh.; Ermolinsky, B.S.; Chernyi, A.A.;
Kochetkov, S.N.; Lysov, Yu.P.; Mikhailov, S.N. Substrate properties of C’-methyl UTP deriva-
tives in T7 RNA polymerase reactions. Evidence for N-type NTP conformation. FEBS Lett.
1
997, 400, 263–266.
2
2
6. Mikhailov, S.N.; Lysov, Yu.P.; YakovlevG.I. Use of functionally competent nucleosides and
nucleotide analogs in studying enzyme-substrate interactions (review). Mol. Biol. (Russia)
1
7. Safonova, T.N.; Mikhailov, S.N.; Veiko, V.P.; Mordkovich, N.N.; Manuvera, V.A.; Alekseev,
C.S.; Kovalchuk, M.V.; Popov, V.O.; PolyakovK.M. High-syn conformation of uridine and
asymmetry of the hexameric molecule revealed in the high-resolution structures of She-
wanella oneidensis MR-1 uridine phosphorylase in the free form and in complex with uridine.
Acta Crystallogr. D. 2014, D70, 3310–3319.
999, 33, 340–354.
2
2
3
8. Mordkovich, N.N.; Safonova, T.N.; Manuvera, V.A.; Veiko, V.P.; Polyakov, K.M.; Alek-
seev, K.S.; Mikhailov, S.N.; PopovV.O. Physicochemical Characterization of Uridine Phos-
phorylase from Shewanella oneidensis MR-1. Dokl. Biochem. Biophys. (Russia) 2013, 451,
1
87–189.
9. Panova, N.G.; Shcheveleva, E.V.; Alexeev, C.S.; Mukhortov, V.G.; Zuev, A.N.; Mikhailov,
S.N.; Esipov, R.S.; Chuvikovsky, D.V.; Miroshnikov, A.I. Using of 4-thiouridine and 4-
thiothymidine for pyrimidine nucleoside phosphorylase studying. Mol. Biol. (Russia) 2004,
3
8, 770–776.
0. Mikhailov, S.N.; Padyukova, N.S.; Karpeiskii, M.Y.; Kolobushkina, L.I.; Beigelman, L.N. ().
ꢀ
Use of 5-deoxy-ribo-hexofuranose derivatives for the preparation of 5 -nucleotide phos-
phonates and homoribonucleosides. Collect. Czech. Chem. Commun. 1989, 54, 1055–
1
066.
ꢀ
3
3
1. Mikhailov, S.N.; Karpeisky, M.Ya. Nucleotide conformational analogs. V. Synthesis of 5 -C-
methylnucleosides on the basis of 6-deoxy-D-allose. Bioorgan. Khim. (USSR) 1979, 5, 895–
9
2. Mikhailov, S.N.; Padyukova, N.Sh.; Struchkova, M.I.; Yarotsky, S.V. Synthesis and NMR spec-
tra of some derivatives of methyl 6-deoxy-α;- and β;-D-allofuranosides and methyl 6-deoxy-
α;- and β;-L-talofuranosides. Bioorgan. Khim. (USSR) 1982, 8, 926–931.
05.
3
3
3. Padyukova, N.Sh.; Fomitcheva, M.V.; Mikhailov, S.N.; Janta-Lipinsky, M. Synthesis of thymi-
dine 5 -derivatives. Bioorgan. Khim. (USSR) 1990, 16 (5), 668–673.
4. Mikhailov, S.N.; Grishko, N.B. Acyclic analogs of nucleosides. Synthesis of chiral 1,5-
dihydroxy-4-methyl-3-oxapent-2-yl derivatives of uracil. Chem. Heterocyclic Compd. 1988,
ꢀ
2
4(1), 75–78.
3
3
3
5. Wang, Y.; Hogenkamp, H.P.; Long, R.A.; Revankar, G.R.; Robins, R.K. A convenient synthesis
of 5 -deoxyribonucleosides. Carbohydr. Res. 1977, 59, 449–457.
6. Lerner, L.M. Preparation of the enantiomeric forms of 9-(5, 6-dideoxy-. beta.-D-ribo-hex-
5
7. Mikhailopulo, I.A.; Pricota, T.I.; Poopeiko, N.E.; Sivets, G.G.; Kvasyuk, E.I.; Sviryaeva, T.V.;
Savochkina, L.P.; Beabealashvilli, R.S. 3 -Fluoro-3 -deoxyribonucleoside 5 -triphosphates:
ꢀ
-enofuranosyl) adenine. J. Org. Chem. 1978, 43(12), 2469–2473.
ꢀ
ꢀ
ꢀ
Synthesis and use as terminators of RNA biosynthesis. FEBS Lett. 1989, 250, 139–
1
41.
3
8. Beigelman, L.N.; Gurskaya, G.V.; Tsapkina, E.N.; Mikhailov, S.N.().Epimerization during the
acetolysis of 3-O-acetyl-5-O-benzoyl-1,2-O-isopropylidene-3-C-methyl-α-d-ribofuranose.

NUCLEOSIDES, NUCLEOTIDES AND NUCLEIC ACIDS
15
ꢀ
Synthesis of 3 -C-methylnucleosides with theβ-d-ribo-and α-d-arabino configurations. Car-
bohydr. Res. 1988, 181, 77–88.
3
9. Beigelman, L.N.; Ermolinsky, B.S.; Gurskaya, G.V.; Tsapkina, E.N.; Karpeisky, M.Y.;
ꢀ
Mikhailov, S.N. New syntheses of 2 -C-methylnucleosides starting from D-glucose and D-
ribose. Carbohydr. Res. 1987, 166, 219–232.
4
0. Esipov, R.S.; Gurevich, A.I.; Chuvikovsky, D.V.; Chupova, L.A.; Muravyova, T.I.; Mirosh-
nikov, A.I. Overexpression of Escherichia coli genes encoding nucleoside phosphorylases in
the pET/Bl21(DE3) system yields active recombinant enzymes. Protein Expres. Purif. 2002,
2
4, 56–60.