The inhibitory effect of citrus flavonoids naringenin and hesperetin against purine nucleoside phosphorylase: Spectroscopic, atomic force microscopy and molecular modeling studies

Article abstract of DOI:10.1016/j.molliq.2020.112954

In this work, the inhibitory effect of two citrus flavonoids naringenin and hesperetin on human purine nucleoside phosphorylase (hPNP) and their binding mechanism were evaluated. Results from enzymatic kinetics revealed that naringenin and hesperetin reversibly inhibited hPNP via a mixed-type manner with IC₅₀ values of 4.83 × 10^?4 M and 5.32 × 10^?4 M, respectively. Analysis of molecular modeling revealed that both naringenin and hesperetin bound directly into the active site by generating multiple forces including hydrogen bonding, π–π and π-Alkyl interactions with His64, Glu201, Ser220, His257, Phe200 and Val217 residues of hPNP, which caused the inhibition of hPNP activity. Moreover, conformational analysis by three-dimension fluorescence, circular dichroism and atomic force microscopy revealed that the binding of naringenin and hesperetin to hPNP induced changes in the microenvironment, secondary structure and morphology of hPNP. These results suggested that occupying the active site and enzymatic conformational perturbation induced by naringenin and hesperetin are the main reasons for reducing the inhibition of hPNP activity, which would be helpful in understanding the inhibitory mechanism of naringenin and hesperetin against hPNP.

Full text of DOI:10.1016/j.molliq.2020.112954

Journal of Molecular Liquids 307 (2020) 112954
Contents lists available at ScienceDirect
Journal of Molecular Liquids
journal homepage: www.elsevier.com/locate/molliq
The inhibitory effect of citrus flavonoids naringenin and hesperetin
against purine nucleoside phosphorylase: Spectroscopic, atomic force
microscopy and molecular modeling studies
b
c
Xingang Lv a, Qilei Wang a, Lang-Hong Wang a, , Er-Fang Ren , Deming Gong
⁎
a
College of Food Science and Technology, Northwest University, Xi'an 710069, China
b
Guangxi Subtropical Crops Research Institute, Nanning 530001, China
c
Department of Biomedicine, New Zealand Institute of Natural Medicine Research, Auckland 2104, New Zealand
a r t i c l e i n f o
a b s t r a c t
Article history:
In this work, the inhibitory effect of two citrus flavonoids naringenin and hesperetin on human purine nucleoside
phosphorylase (hPNP) and their binding mechanism were evaluated. Results from enzymatic kinetics revealed
that naringenin and hesperetin reversibly inhibited hPNP via a mixed-type manner with IC50 values of
4.83 × 10−4 M and 5.32 × 10−4 M, respectively. Analysis of molecular modeling revealed that both naringenin
and hesperetin bound directly into the active site by generating multiple forces including hydrogen bonding,
π–π and π-Alkyl interactions with His64, Glu201, Ser220, His257, Phe200 and Val217 residues of hPNP, which
caused the inhibition of hPNP activity. Moreover, conformational analysis by three-dimension fluorescence, cir-
cular dichroism and atomic force microscopy revealed that the binding of naringenin and hesperetin to hPNP in-
duced changes in the microenvironment, secondary structure and morphology of hPNP. These results suggested
that occupying the active site and enzymatic conformational perturbation induced by naringenin and hesperetin
are the main reasons for reducing the inhibition of hPNP activity, which would be helpful in understanding the
inhibitory mechanism of naringenin and hesperetin against hPNP.
Received 3 March 2020
Accepted 20 March 2020
Available online 21 March 2020
Keywords:
Citrus flavonoid
Purine nucleoside phosphorylase
Enzymatic kinetics
Molecular modeling
Atomic force microscopy
© 2020 Elsevier B.V. All rights reserved.
1. Introduction
could be a potential therapeutic target and consequently becoming of in-
terest in molecular biology and clinical medicine. Currently, there are
Purine nucleoside phosphorylase (PNP) is a crucial enzyme in the
metabolism of purines with reversible phosphorolysis of guanine, hypo-
xanthine, and many of nucleoside analogues [1]. Structural analysis has
revealed that PNPs found in mammals are homotrimers that specific to
cleave the glycosidic bond of 6-oxopurine nucleosides while that from
prokaryotes are homohexamers vary in their specificity, accepting 6-
amino purine nucleosides as well [2]. PNP allows the cells to form nucle-
otides by using purine bases, to avoid the energy-wasting ex-novo syn-
thesis. Therefore, it is essential for cell survival and function, and being
involved in various diseases under abnormal activity. In humans, the de-
ficiency of PNP leads to T-lymphocytopenia, a kind of metabolic disorder
having immunodeficiency and depletion of T-cells [3]. However, the in-
crease of expression levels of PNP mRNA and biological activity was
found to have a relationship with T-cell cancers and in autoimmune dis-
eases. For example, higher PNP activity was found with different types of
tumors, including pancreatic ductal adenocarcinoma, human colon carci-
noma, leukemias and lymphomas [4–6]. These findings suggest that PNP
only two oral PNP inhibitors (forodesine hydrochloride and ulodesine)
in clinical trials for the treatment of T-cell malignancies [7]. However,
these inhibitors impose efficacy and side effects with high price and
complex production technology, and consequently, new inhibitors with
high efficiency and minor side effects are certainly warranted.
Flavonoids, a group of bioactive substances naturally occur in various
vegetables and fruits having a broad range of pharmaceutical activities
[8–10]. Naringenin (NAR, Fig. 1A) and hesperetin (HES, Fig. 1B) are two
of major bioactive flavonoids in grapefruit and sweet orange, respectively.
They have been well acclaimed for antioxidant, anti-inflammatory, anti-
ulcer and anti-carcinogenic activity, which are presumed to beneficial
for reducing the risk of cardiovascular disease and cancers. Moreover,
NAR and HES are regarded to have the potency in inhibiting a variety of
enzymes that play crucial roles in promoting human health and regulat-
ing physiological functions. For example, a report conducted by Priscilla
et al., suggesting NAR is an effective inhibitor of intestinal α-glucosidase
to reduce the postprandial blood glucose levels [11]. Luigi and coworkers
found that HES is of potential clinical relevance that inhibits the activity of
cytochrome P450 (CPY2C9) at low micromolar concentrations [12].
Moreover, both NAR and HES also have been revealed to have abilities
⁎
Corresponding author.
E-mail address: wlhong@nwu.edu.cn (L.-H. Wang).
https://doi.org/10.1016/j.molliq.2020.112954
0167-7322/© 2020 Elsevier B.V. All rights reserved.

2
X. Lv et al. / Journal of Molecular Liquids 307 (2020) 112954
Co. (St. Louis, MO, USA). Stock solutions of hPNP (2.60 × 10−4 M) and
inosine (2.84 × 10−2 M) were prepared in buffered aqueous solution
(0.01 M potassium phosphate buffer with 1 mM MgCl2, pH 7.4). NAR
(purity, 99.0%) and HES (purity, 99.0%) were purchased from Aladdin
Chemical Co. (Shanghai, China), and their stock solution
(1.5 × 10−2 M) were prepared in dimethylsulfoxide (DMSO) under
sterile conditions which was then stored in the dark at 4 °C. The final
concentrations of DMSO in the experiments were kept at b2.0%, which
shows no effect on the structure and activity of the enzyme. All stock so-
lutions were stored at 0–4 °C. All chemicals were of analytical reagent
grade. Ultrapure water was used in the preparation of the samples
and throughout the whole experiment.
2.2. hPNP activity and type inhibition analysis
The activity of hPNP was determined the phosphorolysis of inosine
to hypoxanthine based on a published method [2]. Briefly, 1.0 mL
hPNP (1.30 × 10−7 M) was incubated with varying concentrations of
NAR/HES in potassium phosphate buffer solution for 3 h at 37 °C with
gentle mixing. Then, the catalytic reaction was initiated by adding ino-
sine (final concentration was 1.42 × 10−4 M). After 1 min, the reaction
was stopped by heating the mixture in boiling water (100 °C) for 3 min
to completely extinguish catalytic activity. After denatured enzymes
were removed by centrifugation (15,000 ×g, 10 min), aliquots of
500 μL of the supernatants were filtered using 0.22 μm syringe filters
and collected for liquid chromatography. The determination of hypo-
xanthine was carried out by loading 20 μL each sample on an XSelect®-
HSS T3 (4.6 × 250 mm, 5 μm, Waters, Milford, MA, USA) with the
absorbance detection at 254 nm. The C18 column (column temperature,
30 °C) was used with 10 mM ammonium formate (pH 3.6) as the mobile
phase in the flow rate of 1 mL/min.
The relative enzymatic activity was expressed as E (%), based on the
Eq. (1):
Eð%Þ ¼ C1=C0 ꢀ 100%
ð1Þ
where E, C1 and C0 represent the relative activity of hPNP, the concentra-
tion of hypoxanthine in the absence and presence of flavonoids (NAR or
HES), respectively.
In order to characterize the kinetic mechanism of inhibition of hPNP
by NAR and HES, the formation of hypoxanthine was catalyzed by hPNP
was measured in the above-specified incubation mixture containing in-
creasing concentrations of inosine in the absence or presence of four
concentrations of NAR or HES (0, 1.25, 2.50, 3.7 × 10−4 M). Based on
the experiment above, the constant of inhibition (Km) and the inhibi-
tory type could be determined by the Lineweaver−Burk plot.
Fig. 1. The molecular structure of NAR (A) and HES (B).
in inhibiting tyrosine kinase activity in vitro by occupying the ATP binding
site of the enzyme [13]. However, no information is available about their
inhibitory activity against hPNP. Since hPNP is considered as a therapeutic
target for the therapy of T-cell cancers and in autoimmune diseases, find-
ing safe and high-effective hPNP inhibitors can help to extend different
ways for the treatment of these diseases.
Therefore, in this work, the inhibitory effects of NAR and HES on hPNP
and their interactions including binding constant, site and hPNP confor-
mational changes were investigated by high-performance liquid chro-
matography (HPLC), fluorescence and 3D-fluorescence, circular
dichroism (CD), atomic force microscopy (AFM), and computational mo-
lecular dockings. This study was expected to provide useful information
for potential applications of NAR and HES as hPNP inhibitors in the clinic.
2.3. Fluorescence studies
The fluorescence spectra of hPNP in the presence of different
amounts of NAR/HES were performed under 37 °C using a model F-
7000 fluorescence spectrophotometer (Hitachi High-Technologies Co.,
Tokyo, Japan) with an excitation wavelength of 280 nm. Both the exci-
tation and emission bandwidths were set at 5.0 nm. Appropriate blanks
corresponding to the buffer were subtracted to correct background fluo-
rescence, and the inner filter effects according to a previous method
[14].
The three-dimensional fluorescence spectra of hPNP in the absence
and presence of NAR/HES were recorded at excitation and emission
wavelengths from 200 to 450 nm using 5.0/5.0 nm slit widths.
2. Materials and methods
2.4. Circular dichroism (CD) studies
2.1. Reagents
The CD signal of hPNP was recorded in a BioLogic MOS 450 CD spec-
trometer (Bio-Logic, Claix, France) using a 0.1 cm path length cuvette
and baseline corrected for the buffer signal. The concentration of hPNP
Human purine nucleoside phosphorylase (hPNP, EC 2.4.2.1,
303.2 U mL−1) and inosine (purity ≥99%) were from Sigma–Aldrich

X. Lv et al. / Journal of Molecular Liquids 307 (2020) 112954
3
was kept at 5.4 μM, and the molar ratios of the mixture ([NAR or HES]/
Based on the Michaelis−Menten kinetics, the Lineweaver-Burk
[hPNP]) were varied as 0:1, 2:1 and 4:1.
equation for mixed type inhibition is as follows [19]:
ꢀ
ꢁ
ꢀ
ꢁ
1
v
Km
Vmax½Sꢁ
1
½Iꢁ
1
Vmax
½Iꢁ
¼
1 þ
þ
1 þ
ð2Þ
2.5. Molecular simulation
Ki
Kis
Molecular docking of the optimized ligands (NAR and HES) with
hPNP (PDB: 3K8O) was performed in AutoDockTools-1.5.6rc3 with
Autodock 4.2 program package using Lamarckian genetic algorithm
(LGA) as the search engine. The default grid spacing of 0.7 Å with a
grid-box 120 Å × 120 Å × 120 Å was applied by AutoGrid program.
For each ligand, 100 possible conformations were generated, and
based on glide score and conformation clusters, the optimal conforma-
tion with the lowest energy was selected for further analysis.
Secondary plots can be built as:
Km½Iꢁ
Km
Vmax
Slop ¼
Ki þ
ð3Þ
ð4Þ
Vmax
1
1
1
Y−intercept ¼
¼
þ
½Iꢁ
app
Vmax KisVmax
V
max
where ν is the enzyme reaction rate in the absence and presence of NAR
or HES. Km is the Michaelis–Menten constant. [I] is the concentration of
NAR or HES, and [S] is the concentration of inosine; Ki and Kis are the
equilibrium constant for binding with free enzyme and the enzyme-
substrate complex, respectively.
2.6. Atomic force microscopy (AFM) measurements
AFM measurements were performed using an MFP-3D-S AFM
(Santa Barbara, CA) in tapping mode as described previously [15].
Briefly, 1.0 mL hPNP (1.30 × 10−7 M) with NAR/HES (molar ratio, 1:0
and 1:5) was incubated for 3.0 h at 25 °C. Then, 15 μL of each sample
was air dried on a freshly cleaved mica plate with standing for 0.5 h,
and followed by washed gently with 300 μL ultrapure water. After suit-
able drying and calibration, the desired AFM images were obtained.
The secondary plot of slope or Y-intercept vs. [NAR] was linear
(Fig. 3C and D inset), suggesting that NAR and HES had a single inhibi-
tion site or a single class of inhibition sites on hPNP [20]. Based on sec-
ondary plots, the values of Ki for NAR and HES were obtained to be
(3.48 0.1) × 10−4 M−1 and (3.94 0.1) × 10−4 M−1, respectively.
Accordingly, the Kis of inhibitors with respect to inosine–hPNP com-
plexes of NAR and HES were determined to be (1.53
0.3) × 10−3 M−1 and (1.27 0.3) × 10−3 M−1. The value of Kis was ob-
viously greater than that of Ki, indicating that the affinity of NAR and
HES to hPNP may be stronger than that of the inosine–hPNP complex
[21]. Similar results were reported by Wen and coworkers, who also ob-
served that flavonoids including chrysin, baicalein and apigenin per-
formed as mixed-type inhibitors of hPNP [2].
2.7. Statistical analysis
Results were expressed as means SD. Data were analyzed using
OriginPro 8.0 (OriginLab, Northampton, MA, USA).
3. Results and discussion
3.3. Fluorescence quenching of hPNP by Nar and HES
3.1. The inhibitory effects on hPNP
The inhibitory activity of NAR and HES on hPNP demonstrates that
there is a direct interaction between these two flavonoids and hPNP.
Therefore, fluorescence quenching of hPNP was induced by NAR and
HES to investigate their interactions. As shown in Fig. 4A and B, the max-
imum fluorescence emission peak of hPNP at 338 nm decreased with an
apparent red-shift after gradually adding of NAR and HES (0 to
5.73 × 10−4 M), indicating that the binding of NAR and HES to hPNP in-
deed happened, and their interactions resulted in some perturbations in
the conformation of hPNP [22].
Fig. 2A and B exhibit the effects of NAR and HES in the range of con-
centrations from 0 to 6.88 × 10−4 M on the hypoxanthine formation, re-
spectively. Obviously, the production of hypoxanthine was gradually
decreased with the increasing concentration of NAR and HES, indicating
that both they decreased hypoxanthine formation in a concentration-
dependent manner. These results were well consistent with the loss rel-
ative activity of hPNP (Fig. 2C). The concentration of NAR and HES that
resulted in a loss of 50% enzyme activity (IC50) was estimated to be
4.83 × 10−4 M and 5.32 × 10−4 M, respectively, implying that NAR
and HES might have a potential to inhibit the activity of hPNP.
For fluorescence quenching, an improved method was used to calcu-
late the binding constant (Ka) and the number of binding sites (n) [23].
3.2. Inhibition mechanism of NAR and HES on the activity of hPNP
F0−F
F
1
log
¼ n logKa−n log
½Qtꢁ−
ð5Þ
F0−F
½Ptꢁ
To confirm the reversibility of these two citrus flavonoids inhibition
toward hPNP, the plots of the v vs [hPNP] (enzyme concentration) in the
presence of increasing concentrations of NAR and HES were con-
structed. As shown in Fig. 3A and B, all data were fitted and gave a family
of straight lines with intercepts essentially equal to 0. Moreover, in-
creasing concentrations of NAR and HES led to a decrease in the slopes
of the lines, suggesting that these two flavonoids reversibly inhibited
hPNP activity, that led to the decrease of efficiency in term of the pro-
duction of hypoxanthine [16].
The kinetic behavior of hPNP during phosphorolysis of inosine in the
presence of NAR and HES was evaluated using double-reciprocal
Lineweaver-Burk plots at different concentrations of inosine and NAR/
HES. As can be seen in Fig. 3C and D, the Lineweaver-Burk plots gave a
family of lines with different slopes and different Y-axis intercepts indi-
cating that NAR and HES are typically mixed-type hPNP inhibitors
[17,18]. This means that NAR and HES not only bind with the free
hPNP but also with the inosine–hPNP complex resulting in an increase
F
where [Pt] and [Qt] denote the total concentrations of the hPNP and
NAR/HES, respectively. F0 and F are the corrected fluorescence intensi-
ties of hPNP in the absence and presence of NAR/HES, respectively,
based on a reported method [24]. Based on the Eq. (5) and Fig. 4C, the
calculated Ka and n values for NAR−hPNP and HES−hPNP were
(4.73 0.03) × 104 M−1 and (3.94 0.04) × 104 M−1, and 1.25 and
1.36, indicating a strong static interaction of these two flavonoids with
hPNP [25]. The estimated Ka values for NAR−hPNP was larger than
HES−hPNP, suggesting that the stability of NAR–hPNP N HES–hPNP,
which is well consistent with the Ki values from inhibition kinetic
results.
3.4. Three-dimensional fluorescence analysis
To study the possible hPNP conformational change, 3-D fluorescence
spectra of hPNP and the NAR/HES−hPNP complex were measured,
of Km and a decrease of Vmax
.

4
X. Lv et al. / Journal of Molecular Liquids 307 (2020) 112954
respectively. As shown in Fig. 5A, Peak 1 (λex = 280 nm, λem = 335 nm)
mainly reflects the spectral behavior of Trp and Tyr, which is mainly
caused by the transition of n → π* of aromatic amino acids in hPNP,
and Peak 2 (λex = 230 nm, λem = 335 nm) exhibits the fluorescence
properties of polypeptide backbone structure of hPNP [26]. A direct
comparison of Fig. 5A with B and C clearly reveal the 3D fluorescence
characteristics of hPNP as a result of interaction with NAR and HES in
terms of emission intensity reduction and position of both peak 1 and
peak 2. Specifically, the fluorescence intensity of Peak 1 was decreased
from 2363 to 815 and 1026 with the addition of NAR and HES, while
Peak 2 was decreased from 1085 to 86.1 and 103, respectively. More-
over, both Peak 1 and 2 have an apparent red-shift in that λem at
335 nm moved to 350 nm. Since the two peaks in hPNP originate from
different types of residues, these perturbations suggested the microen-
vironment of tryptophan residues and conformation of the binding of
NAR and HES to hPNP induced some conformational and micro-
environmental changes [27,28].
3.5. CD spectra analysis
CD spectroscopy is a practical and sensitive in analyzing the confor-
mation (secondary and tertiary structure) changes of proteins [29].
Generally, CD spectra of proteins could be divided into two wavelength
regions, namely far-UV CD and near-UV CD spectrum, which located in
the range of 190–250 nm and 250–350 nm, respectively. Therefore, this
method was used to study the alterations in the secondary and tertiary
structure of hPNP after being exposed to NAR and HES.
Fig. 6A and C shows far UV-CD measurements (190–250 nm) of free
hPNP and NAR/HES–hPNP complexes that revealed the structural
changes of hPNP. The free hPNP displayed a broad negative band around
210–230 nm, which is in good agreement with the CD spectra of human
erythrocytic PNP in the literature [30]. After binding with NAR, the neg-
ative ellipticities of hPNP were decreased, indicating the noticeable
change in the secondary structures induced by the binding of NAR.
The relative content (%) of the secondary structure, including α-helix,
β-sheet, β-turns and random coils was calculated by the DichroWeb on-
line program (http://dichroweb.cryst.bbk.ac.uk), and listed in Table 1.
This table clearly shows that the percentage of the α-helix and β-turn
in the presence of certain concentrations of NAR was decreased, while
β-sheet was increased. Specially, the contents of the α-helix and β-
turn decreased from 26.5 to 25.0, 21.6% and from 21.1 to 19.4, 20.0%, re-
spectively, while the contents of the β-sheet increased from 18.1 to 20.3,
23.7% at the increasing molar ratios ([hPNP/NAR]) from 1:0 to 1:2 and
1:4. Similarly, the β-sheet increased to 20.6% and 24.3%, while the α-
helix reduced to 24.3% and 20.9% as well as the β-turn mildly decreased
to 20.2% and 19.8% that were observed after binding with HES at molar
ratios of 1:2 and 1:4. These results implied the structural compactness of
hPNP, which might lead to the contraction of the hydrophobic
pocket along with decreasing the landing of the substrate and finally
causing an inhibition against hPNP.
The near-UV CD spectra in the region of 250–350 nm was attributed
to the existence of aromatic residues (Phe, Tyr, and Trp) and disulfide
bonds, which could be used to investigate the tertiary structural
changes of hPNP [31,32]. With the gradual addition of NAR or HES to
hPNP solution, a significant increase in the molar ellipticity was ob-
served in the wavelength range of 280–350 nm (Fig. 6B and D), indicat-
ing alterations of the environments of Trp residues, which related to the
drastic perturbations in the tertiary structure of hPNP after binding with
Fig. 2. Effects of various concentrations of NAR (A) or HES (B) on the production of
hypoxanthine by hPNP. The concentration of inosine was 1.42 × 10−4 M; (C) Effect of
NAR and HES on the relative activity of hPNP; the concentration of NAR/HES was 0, 0.63,
1.25, 1.88, 2.50, 3.13, …, 6.25 and 6.88 × 10−4 M for curves 1 → 12. Data are presented
as the means (n = 3).

X. Lv et al. / Journal of Molecular Liquids 307 (2020) 112954
5
Fig. 3. Plots of ν vs. [hPNP] for NAR (A) and HES (B). The concentration of NAR and HES were 0, 1.25, 2.50, 3.75 × 10−4 M for curves 1 → 4, respectively; the final concentration of inosine
was 1.42 × 10−4 M and c(hPNP) = 0, 0.65, 1.30, 2.60, 3.90 and 5.20 × 10–7 M. (C) and (D) Lineweaver-Burk plots for NAR and HES, respectively. The final hPNP concentration was
1.30 × 10−7 M. The secondary plots of slope (E) and Y-intercept (F) vs. [NAR]/[HES]. NAR/HES concentrations were 0, 1.25, 2.50, 3.75 and 5.0 × 10−4 M for curves 1 → 5, respectively.
these two flavonoids. The above-mentioned results indicated that NAR
and HES could affect the secondary structure of hPNP and result in the
loss of catalytic function of hPNP, which further confirmed their inhibi-
tory activities [33].
3.6. Surface morphological changes in hPNP
The morphological changes including distributions of shapes and
size of hPNP before and after NAR/HES bound were observed by AFM

6
X. Lv et al. / Journal of Molecular Liquids 307 (2020) 112954
Fig. 5. (A) 3-D fluorescence spectra of hPNP (A), NAR–hPNP system (B) and HES–hPNP
system (C). c(hPNP) = 8.67 × 10−7 M and c(NAR/HES) = 5.21 × 10−6 M.
Fig. 4. The fluorescence emission spectra of PNP in the presence of increasing amounts of
NAR (A) and HES (B). Fluorescence emission spectra were recorded over a wavelength
range of 300–450 nm at an excitation wavelength of 280 nm. The concentrations of
particles of hPNP became swollen and the mean height of hPNP reached
to 12.7 and 14.2 nm, respectively. The results from AFM microscopy
suggested that NAR and HES increased the hydrophobicity of hPNP sur-
rounding microenvironment by forming a flavonoid–hPNP complex
that induced both morphological changes and aggregation of hPNP mol-
ecules through protein-protein hydrophobic interaction. In this situa-
tion, the contact of hPNP molecules with water on the surface area
were mostly minimized in the stable complex structures [34]. Therefore,
these results obtained from AFM indicated that the formation of the
NAR−hPNP or HES−hPNP complex was accompanied by a serious de-
stabilization of the native conformation of the protein [35,36].
NAR/HES concentrations were 0, 0.52, 1.04, 1.56, 2.08, 2.61, …, 5.21 and 5.73 × 10−4
M
for curves 1 → 11. The hPNP concentration was 8.67 × 10−7 M. (C) the plots for the
fluorescence quenching of PNP by NAR and HES based on the Eq. (5).
(Fig. 7). Fig. 7A and B are the two- and two-dimensional topographic
images of hPNP alone. The mean heights of hPNP in the absence of
NAR/HES were about 3.8 nm, which well consistent with the size
(3.2 nm) previously determined by Wen et al. [2]. Upon addition of
NAR (Fig. 7C and D) and HES (Fig. 7E and F), the small-bright circular

X. Lv et al. / Journal of Molecular Liquids 307 (2020) 112954
7
Table 1
3.7. Computational docking of NAR and HES on hPNP
The contents of secondary structures in hPNP in the absence and presence of NAR and HES.
The docking analysis was performed using AutoDock 4.2 program to
investigate the binding sites of NAR– and HES–hPNP interaction at a
molecular level. As shown in Fig. 8A and B, the minimal binding energy
of hPNP with NAR and HES were −6.8 and −6.1 kcal/mol with the
highest populated clusters of 19 and 14 out of 100 (the blue bar), re-
spectively. Based on the energetically favorable results, the lowest ener-
gies were selected to visualize the docking postures of NAR– and HES–
hPNP interaction using Discovery Studio Visualizer 4.0. Results from
the 2D and 3D schematic views shown that both NAR and HES docked
well in the hydrophobic pocket (active site) of hPNP and surrounded
by the amino acid residues, including Ser33, Val61, His64, His86,
Arg84, Tyr88, Asn115, Ala116, Ala117, Gly118, Phe200, Glu201,
Val217, Ser220, Met219, Asn243, Val245, His257, etc., suggesting that
NAR and HES shared a common binding region with the substrate (ino-
sine) (Fig. 8C and D). Among them, Val61, Tyr88, Ala116, Ala117,
Tyr192, Val217, Met219, Val245, Val260 are hydrophobic residues, it
could be speculated that the hydrophobic interactions played an impor-
tant role in the binding mechanism of NAR and HES with hPNP. More-
over, the effects of hydrogen bonds, π–π (Pi-Pi) and π-Alkyl
interactions could not be neglected for stabilizing NAR– and HES–
Molar ratio
hPNP/NAR
α-Helix
β-Sheet
β-Turn
Random coil
1:0
1:2
1:4
1:0
1:2
1:4
26.5
25.0
21.6
25.6
24.3
20.9
18.1
20.3
23.7
19.0
20.6
24.3
21.1
19.4
20.0
20.5
20.2
19.8
34.3
35.3
34.7
34.9
34.9
35.0
hPNP/HES
hPNP postures. For NAR–hPNP interaction, NAR formed four hydrogen
bonds with His64, Glu201, Ser220 and His257 of hPNP residues, and
to produce π–π (Pi-Pi) and π-Alkyl forces with the amino acids
Phe200, Val217 and Met219 (Fig. 8C). Similarly, in the docking of HES
with hPNP, the same hydrogen bonds were found between the hydroxyl
group of HES with the residues His64, Glu201, Ser220 and His257, and
generate π–π (Pi-Pi) and π-Alkyl forces with Phe200 and Val217. Conse-
quently, it can be concluded that His64, Glu201, Ser220, His257, Phe200
and Val217 around the reactive site, and played a crucial role in com-
plexation process of NAR and HES with hPNP.
Fig. 6. The CD spectra of hPNP in the presence of increasing amounts of NAR: (A) Far-UV CD spectra and (B) Near-UV spectra; The CD spectra of hPNP in the presence of increasing amounts
of HES: (C) Far-UV CD spectra and (D) Near-UV spectra; c(hPNP) = 5.4 μM, the molar ratios ([hPNP]/[Flavonoids]) are 1:0, 1:2 and 1:4, respectively.

8
X. Lv et al. / Journal of Molecular Liquids 307 (2020) 112954
Fig. 7. AFM topography image of PNP (A), NAR–hPNP complex (C), HES–hPNP complex (E) adsorbed onto mica with tapping mode in air. The scan size of the image was 5.0 μm × 5.0 μm. c
(PNP) = 1.30 × 10−7 mol L−1, the molar ratios of flavonoids to PNP were 5:1. B, D and F are the two-dimensional graphs for A, C and E, respectively.

X. Lv et al. / Journal of Molecular Liquids 307 (2020) 112954
9
Fig. 8. Cluster analysis of the AutoDock docking runs of NAR (A) and HES (B) with hPNP; Three-dimensional (3D) and two-dimensional (2D) interaction patternsof hPNP with NAR (C) and
HES (D) were generated using Discovery Studio Visualizer v4.0.

10
X. Lv et al. / Journal of Molecular Liquids 307 (2020) 112954
cytochrome P450 2C9-mediated drug metabolism in vitro, Drug Metab.
4. Conclusions
Pharmacokinet. 25 (2010) 466–476.
[13] B.B. Chandrika, M. Steephan, T.S. Kumar, A. Sabu, M. Haridas, Hesperetin and
naringenin sensitize HER2 positive cancer cells to death by serving as HER2 tyrosine
kinase inhibitors, Life Sci. 160 (2016) 47–56.
[14] S. Shahraki, H.S. Delarami, M. Saeidifar, Catalase inhibition by two Schiff base deriv-
atives. Kinetics, thermodynamic and molecular docking studies, J. Mol. Liq. 287
(2019), 111003.
[15] R.M. Hjelm, K.E. Garcia, S. Babanova, K. Artyushkova, I. Matanovic, S. Banta, P.
Atanassov, Functional interfaces for biomimetic energy harvesting: CNTs-DNA ma-
trix for enzyme assembly, Biochim. Biophys. Acta Bioenerg. 1857 (2016) 612–620.
[16] Y. Wang, G. Zhang, J. Pan, D. Gong, Novel insights into the inhibitory mechanism of
kaempferol on xanthine oxidase, J. Agric. Food Chem. 63 (2015) 526–534.
[17] M. Fan, H. Ding, G. Zhang, X. Hu, D. Gong, Relationships of dietary flavonoid struc-
ture with its tyrosinase inhibitory activity and affinity, LWT Food Sci. Technol. 107
(2019) 25–34.
[18] X. Wu, M. Hu, X. Hu, H. Ding, D. Gong, G. Zhang, Inhibitory mechanism of epicate-
chin gallate on α-amylase and α-glucosidase and its combinational effect with
acarbose or epigallocatechin gallate, J. Mol. Liq. 290 (2019), 111202.
[19] A. Cornish-Bowden, R. Eisenthal, Statistical considerations in the estimation of en-
zyme kinetic parameters by the direct linear plot and other methods, Biochem. J.
139 (1974) 721–730.
hPNP is one of the most important targets for therapy of T-cell malig-
nancies, therefore finding significant inhibitors, especially from natural
compounds is of importance. In the work, the inhibitory effects of two
citrus flavonoids including NAR and HES on hPNP and interactions be-
tween them were investigated by a combination of differential experi-
mental and computational methods. The results revealed that both
NAR and HES could bind directly into the active cavity, and in this pro-
cess induce the alterations in the microenvironment, secondary struc-
ture and morphology of hPNP, which resulted in the reduced catalytic
activity and fluorescence quenching of hPNP. All these experimental re-
sults and theoretical data have provided some information for the inter-
actions between NAR and HES and hPNP, and it would be valuable for
understand the inhibitory mechanism of flavonoid compounds on
hPNP, and search for other natural PNP inhibitors.
[20] R. Yekta, G. Dehghan, S. Rashtbari, R. Ghadari, A.A. Moosavi-Movahedi, The inhibi-
tory effect of farnesiferol C against catalase; kinetics, interaction mechanism and
molecular docking simulation, Int. J. Biol. Macromol. 113 (2018) 1258–1265.
[21] D. Kalita, D.G. Holm, D.V. LaBarbera, J.M. Petrash, S.S. Jayanty, Inhibition of α-
glucosidase, α-amylase, and aldose reductase by potato polyphenolic compounds,
PLoS One 13 (2018).
[22] H.J. Zeng, Y.P. Wang, R. Yang, J. You, L.B. Qu, Inhibitory effects of daidzein and genis-
tein on trypsin: insights from spectroscopic and molecular docking studies, Int. J.
Biol. Macromol. 89 (2016) 336–343.
Declaration of competing interest
The authors hereby declare that there is no conflict of interest.
Acknowledgments
[23] L.H. Wang, M.S. Wang, X.A. Zeng, D.M. Gong, Y.B. Huang, An in vitro investigation of
the inhibitory mechanism of β-galactosidase by cinnamaldehyde alone and in com-
bination with carvacrol and thymol, Biochim. Biophys. Acta, Gen. Subj. 1861 (
(2017) 3189–3198.
[24] R. Wang, X. Hu, J. Pan, G. Zhang, D. Gong, Interaction of isoeugenol with calf thymus
DNA and its protective effect on DNA oxidative damage, J. Mol. Liq. 282 (2019)
356–365.
This research was supported by the National Natural Science Foun-
dation of China (21706213) and Scientific Research Staring Foundation
from the Northwest University, China. Financial support from the Chi-
nese Postdoctoral Science Foundation (2016M602853) is also gratefully
acknowledged.
[25] X. Peng, G. Zhang, Y. Liao, D. Gong, Inhibitory kinetics and mechanism of kaempferol
References
on α-glucosidase, Food Chem. 190 (2016) 207–215.
[26] W. Wang, N. Gan, Q. Sun, D. Wu, R. Gan, M. Zhang, P. Tang, H. Li, Study on the inter-
action of ertugliflozin with human serum albumin in vitro by multispectroscopic
methods, molecular docking, and molecular dynamics simulation, Spectrochim.
Acta A 219 (2019) 83–90.
[27] H. Yang, G. Hou, L. Zhang, L. Ju, C. Liu, Exploring the effect of bisphenol S on sludge
hydrolysis and mechanism of the interaction between bisphenol S and α-amylase
through spectrophotometric methods, J. Photochem. Photobiol. B 167 (2017)
128–135.
[1] P. Giuliani, M. Zuccarini, S. Buccella, M. Rossini, I. D’Alimonte, R. Ciccarelli, M. Marzo,
A. Marzo, P. Di Iorio, F. Caciagli, Development of a new HPLC method using fluores-
cence detection without derivatization for determining purine nucleoside phos-
phorylase activity in human plasma, J. Chromatogr. B 1009 (2016) 114–121.
[2] Q.H. Wen, L.H. Wang, X.A. Zeng, D.B. Niu, M.S. Wang, Hydroxyl-related differences
for three dietary flavonoids as inhibitors of human purine nucleoside phosphory-
lase, Int. J. Biol. Macromol. 118 (2018) 588–598.
[3] M. Camici, V. Micheli, P.L. Ipata, M.G. Tozzi, Pediatric neurological syndromes and in-
born errors of purine metabolism, Neurochem. Int. 56 (2010) 367–378.
[4] S. Kojima, T. Chiyomaru, K. Kawakami, H. Yoshino, H. Enokida, N. Nohata, M. Fuse, T.
Ichikawa, Y. Naya, M. Nakagawa, Tumour suppressors miR-1 and miR-133a target
the oncogenic function of purine nucleoside phosphorylase (PNP) in prostate can-
cer, Br. J. Cancer 106 (2012) 405–413.
[5] O. Sanfilippo, M. Camici, M.G. Tozzi, M. Turriani, A. Faranda, P. Ipata, R. Silvestrini,
Relationship between the levels of purine salvage pathway enzymes and clinical/bi-
ological aggressiveness of human colon carcinoma, Cancer Biochem. Biophys. 14
(1994) 57–66.
[6] F. Ravandi, V. Gandhi, Novel purine nucleoside analogues for T-cell-lineage acute
lymphoblastic leukaemia and lymphoma, Expert Opin. Investig. Drugs 15 (2006)
1601–1613.
[7] M.S. Castilho, M.P. Postigo, C.B. de Paula, C.A. Montanari, G. Oliva, A.D. Andricopulo,
Two-and three-dimensional quantitative structure–activity relationships for a series
of purine nucleoside phosphorylase inhibitors, Bioorg. Med. Chem. 14 (2006)
516–527.
[28] M.Z. Kabir, Z. Benbekhti, N.F.W. Ridzwan, R. Merrouche, N. Bouras, S.B. Mohamad, S.
Tayyab, Biophysical and in silico investigations of the molecular association between
a potent RNA polymerase inhibitor, thiolutin and human serum albumin, J. Mol. Liq.
303 (2020) 1–11, https://doi.org/10.1016/j.molliq.2020.112648.
[29] S. Tayyab, M.M. Izzudin, M.Z. Kabir, S.R. Feroz, W.-V. Tee, S.B. Mohamad, Z. Alias,
Binding of an anticancer drug, axitinib to human serum albumin: fluorescence
quenching and molecular docking study, J. Photochem. Photobiol. B 162 (2016)
386–394.
[30] J.D. Stoeckler, R.P. Agarwal, K.C. Agarwal, K. Schmid, R.E. Parks Jr., Purine nucleoside
phosphorylase from human erythrocytes: physicochemical properties of the crystal-
line enzyme, Biochemistry 17 (1978) 278–283.
[31] S. Lin, G. Zhang, Y. Liao, D. Gong, The inhibitory kinetics and mechanism of dietary
vitamins D3 and B2 on xanthine oxidase, Food Funct. 7 (2016) 2849–2861.
[32] S. Farhadian, B. Shareghi, S. Asgharzadeh, M. Rajabi, H. Asadi, Structural characteri-
zation of α-chymotrypsin after binding to curcumin: spectroscopic and computa-
tional analysis of their binding mechanism, J. Mol. Liq. 289 (2019), 111111.
[33] C.P. Sun, J.K. Yan, J. Yi, X.Y. Zhang, Z.L. Yu, X.K. Huo, J.H. Liang, J. Ning, L. Feng, C.
Wang, The study of inhibitory effect of natural flavonoids toward β-glucuronidase
and interaction of flavonoids with β-glucuronidase, Int. J. Biol. Macromol. 143
(2020) 349–358.
[8] S. Das, P. Ghosh, S. Koley, A.S. Roy, Binding of naringin and naringenin with hen egg
white lysozyme: a spectroscopic investigation and molecular docking study,
Spectrochim. Acta A 192 (2018) 211–221.
[9] L. Chen, H. Teng, Z. Jia, M. Battino, A. Miron, Z. Yu, H. Cao, J. Xiao, Intracellular signal-
ing pathways of inflammation modulated by dietary flavonoids: the most recent ev-
idence, Crit. Rev. Food Sci. Nutr. 58 (2018) 2908–2924.
[34] Z. Xiong, W. Liu, L. Zhou, L. Zou, J. Chen, Mushroom (Agaricus bisporus)
polyphenoloxidase inhibited by apigenin: multi-spectroscopic analyses and compu-
tational docking simulation, Food Chem. 203 (2016) 430–439.
[10] J. Xiao, Dietary flavonoid aglycones and their glycosides: which show better biolog-
[35] Y. Zheng, W. Yang, W. Sun, S. Chen, D. Liu, X. Kong, J. Tian, X. Ye, Inhibition of porcine
pancreatic α-amylase activity by chlorogenic acid, J. Funct. Foods 64 (2020),
103587.
[36] Z. Chen, Y. Chen, Z. Xue, X. Gao, Y. Jia, Y. Wang, Y. Lu, J. Zhang, M. Zhang, H. Chen,
Insight into the inactivation mechanism of soybean Bowman-Birk trypsin inhibitor
(BBTI) induced by epigallocatechin gallate and epigallocatechin: fluorescence, ther-
modynamics and docking studies, Food Chem. 303 (2020), 125380.
ical significance? Crit. Rev. Food Sci. Nutr. 57 (2017) 1874–1905.
[11] D.H. Priscilla, D. Roy, A. Suresh, V. Kumar, K. Thirumurugan, Naringenin inhibits α-
glucosidase activity: a promising strategy for the regulation of postprandial hyper-
glycemia in high fat diet fed streptozotocin induced diabetic rats, Chem. Biol. Inter-
act. 210 (2014) 77–85.
[12] L. Quintieri, S. Bortolozzo, S. Stragliotto, S. Moro, M. Pavanetto, A. Nassi, P. Palatini,
M. Floreani, Flavonoids diosmetin and hesperetin are potent inhibitors of