Molecular structure, in vitro anticancer study and molecular docking of new phosphate derivatives of betulin

Article abstract of DOI:10.3390/molecules26030737

A series of 30-diethylphosphate derivatives of betulin were synthesized and evaluated for their in vitro cytotoxic activity against human cancer cell lines, such as amelanotic melanoma (C-32), glioblastoma (SNB-19), and two lines of breast cancer (T47D, MDA-MB-231). The molecular structure and activities of the new compounds were also compared with their 29-phosphonate analogs. Compounds 7a and 7b showed the highest activity against C-32 and SNB-19 cell lines. The IC₅₀ values for 7a were 2.15 and 0.91 μM, and, for 7b, they were 0.76 and 0.8 μM for the C-32 and SNB-19 lines, respectively. The most potent compounds, 7a and 7b, were tested for their effects on markers of apoptosis, such as H3, TP53, BAX, and BCL-2. For the whole series of phosphate derivatives, a lipophilicity study was performed, and the ADME parameters were calculated. The most active products were docked to the active site of the EGFR protein. The relative binding affinity of selected phosphate betulin derivatives toward EGFR was compared with standard erlotinib on the basis of ChemScore and K_DEEP score. Positively, all derivatives docked inside the cavity and showed significant interactions. Moreover, a molecular dynamics study also reveals that ligands 7a,b form stable complexes and the plateau phase started after 7 ns.

Full text of DOI:10.3390/molecules26030737

molecules
Article
Molecular Structure, In Vitro Anticancer Study and Molecular
Docking of New Phosphate Derivatives of Betulin
1
,
2
1
1
1
Elwira Chrobak *, Maria Jastrz e˛ bska , Ewa B e˛ benek , Monika Kadela-Tomanek , Krzysztof Marciniec
,
3
2
4
1
Małgorzata Latocha , Roman Wrzalik , Joachim Kusz and Stanisław Boryczka
1
Department of Organic Chemistry, Faculty of Pharmaceutical Sciences in Sosnowiec,
Medical University of Silesia, Katowice, 4 Jagiello n´ ska Str., 41-200 Sosnowiec, Poland;
ebebenek@sum.edu.pl (E.B.); mkadela@sum.edu.pl (M.K.-T.); kmarciniec@sum.edu.pl (K.M.);
boryczka@sum.edu.pl (S.B.)
2
Silesian Center for Education and Interdisciplinary Research, Institute of Physics, University of Silesia,
5Pułku Piechoty 1a, 41-500 Chorzów, Poland; maria.jastrzebska@us.edu.pl (M.J.);
7
roman.wrzalik@us.edu.pl (R.W.)
3
4
Department of Cell Biology, Faculty of Pharmaceutical Sciences in Sosnowiec, Medical University of Silesia,
Katowice, 8 Jedno s´ ci Str., 41-200 Sosnowiec, Poland; mlatocha@sum.edu.pl
Institute of Physics, University of Silesia, 75 Pułku Piechoty Str.1a, 41-500 Chorzów, Poland;
joachim.kusz@us.edu.pl
*
Correspondence: echrobak@sum.edu.pl
Abstract: A series of 30-diethylphosphate derivatives of betulin were synthesized and evaluated
for their in vitro cytotoxic activity against human cancer cell lines, such as amelanotic melanoma
(
C-32), glioblastoma (SNB-19), and two lines of breast cancer (T47D, MDA-MB-231). The molecular
ꢀ
ꢁꢂꢀꢃꢄꢅꢆꢇ
structure and activities of the new compounds were also compared with their 29-phosphonate
ꢀ
ꢁꢂꢃꢄꢅꢆ
analogs. Compounds 7a and 7b showed the highest activity against C-32 and SNB-19 cell lines. The
IC50 values for 7a were 2.15 and 0.91 µM, and, for 7b, they were 0.76 and 0.8 µM for the C-32 and
Citation: Chrobak, E.; Jastrz e˛ bska,
M.; B e˛ benek, E.; Kadela-Tomanek, M.;
Marciniec, K.; Latocha, M.; Wrzalik,
R.; Kusz, J.; Boryczka, S. Molecular
Structure, In Vitro Anticancer Study
and Molecular Docking of New
SNB-19 lines, respectively. The most potent compounds, 7a and 7b, were tested for their effects
on markers of apoptosis, such as H3, TP53, BAX, and BCL-2. For the whole series of phosphate
derivatives, a lipophilicity study was performed, and the ADME parameters were calculated. The
most active products were docked to the active site of the EGFR protein. The relative binding affinity
of selected phosphate betulin derivatives toward EGFR was compared with standard erlotinib on the
basis of ChemScore and KDEEP score. Positively, all derivatives docked inside the cavity and showed
Phosphate Derivatives of Betulin.
Molecules 2021, 26, 737. https://doi.
org/10.3390/molecules26030737
significant interactions. Moreover, a molecular dynamics study also reveals that ligands 7a,b form
stable complexes and the plateau phase started after 7 ns.
Academic Editor: Silvie Rimpelová
Received: 26 November 2020
Accepted: 28 January 2021
Keywords: natural compounds; betulin; anticancer study; molecular docking; EGFR
Published: 31 January 2021
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
1
. Introduction
Cancer is a large group of diseases in which the mechanisms of cell growth control
have been lost. Today, it is one of the leading causes of death in people all over the world
approximately 30% of all premature deaths among adults aged 30–69 years not caused by
infectious diseases) [ ]. Chemotherapy with synthetic drugs is burdened with numerous
(
1
side effects. Therefore, substances of natural origin are being considered a potential source
of new anticancer drugs.
Triterpenes are an important class of compounds synthesized by various plants, which
show a broad spectrum of biological activity [2,3]. Betulin and betulinic acid are among
the triterpenes of the lupane type. The structure of these compounds contains substituents
such as isopropenyl moiety and hydroxyl and/or carboxyl groups that can be easily
functionalized. This allows for the continuous expansion of the library of derivatives
(terpenoids), and among these compounds, new substances with potential biological
activity are sought. In the group of semi-synthetic derivatives of betulin, numerous
Copyright: © 2021 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4
.0/).
Molecules 2021, 26, 737. https://doi.org/10.3390/molecules26030737
https://www.mdpi.com/journal/molecules

Molecules 2021, 26, 737
2 of 30
compounds resulting from modification in the isopropenyl group have also been described.
The products obtained in this way were tested for various activities, including, for example,
anti-inflammatory, antiviral, and anticancer properties [4–6].
As a result of introducing a function containing phosphorus atoms into isopropenyl
moiety of betulin diesters, compounds with antitumor (diethylphosphonate derivatives)
and antischistosomal (triphenylphosphonium derivatives) activity were obtained [7,8].
Anti-HIV-1 studies performed with the series of phosphorus analogs of bevirimat showed
that the 30-diethylphosphate derivative has comparable activity to bevirimat, but with
higher selectivity [9].
The carbon–carbon triple bond occurs in drugs used in the pharmacotherapy of vari-
ous diseases and is still the subject of many research studies related to the development of
new pharmaceuticals [10]. Examples include the antiretroviral efavirenz, the contraceptive
norethynodrel, oxybutynin for the treatment of overactive bladder, pargyline with the
antidepressant effect, the antitumor erlotinib, and the antifungal terbinafine [11].
The literature review shows that the presence of an alkynyl moiety in a compound
molecule may favorably influence its antitumor activity [12–14]. This is also confirmed
by the studies conducted earlier by our team [15 19]. It, therefore, seemed interesting to
–
obtain new betulin derivatives containing diethylphosphate group in position C30, and an
alkynyl substituent in position C3 and/or C28 (Figure 1).
Figure 1. Schematic structure of target compounds.
The aim of the present study was the synthesis of new betulin derivatives, analysis
of the structure of these compounds, and evaluation of their in vitro cytotoxic activity.
Human glioma (SNB-19), melanoma (C-32), and two breast cancer (T47D and MDA-MB-
231) cell lines were used in the research. In addition, in silico techniques such as ADMET
parameter prediction and molecular docking were used to more fully characterize novel
substances with potential anticancer effects. The comparison of the structure and activity
of phosphonate and phosphate derivatives carried out in the paper was to answer the
question of how the replacement of the phosphonate with the phosphate group in the
isopropenyl substituent affects the cytotoxic activity of betulin alkynyl derivative.
2. Results and Discussion
2.1. Synthesis
In order to modify the betulin molecule or its derivatives at position C30, first the
functionalization of the allylic hydrogen is carried out, introducing a substituent of greater
reactivity into it. The most common method for this purpose is bromination at the allyl
position with N-bromosuccinimide (NBS) [20]. The 30-bromo- derivatives thus obtained
are a convenient substrate for further transformations. They have been used in reactions
with thiolates, aliphatic and alicyclic alcohols, potassium isocyanate, sodium azide, acid
anhydrides, dimethylaminopyridine (DMAP), triphenylphosphine, and trialkyl phosphite.
As a result of these reactions, new compounds were obtained that were evaluated for
biological activity [6–8,20].

Molecules 2021, 26, 737
3 of 30
Another route for the preparation of the 30-substituted derivatives is the introduction
of a hydroxyl group by oxidation of the isopropenyl moiety with m-chloroperoxybenzoic
acid (m-CPBA) [21].
3
,28-Diacetyl-30-hydroxybetulin
this work. This compound has a free hydroxyl group at position 30, which can be converted
into a phosphate substituent (derivative ). The phosphorylation reaction was carried out
3 (Scheme 1) was used as the starting compound for
4
with diethyl chlorophosphate in a pyridine solution. The next step was the hydrolysis of
acetyl groups at C3 and C28 leading to deprotection of the hydroxyl groups, which enabled
their further functionalization. Hydrolysis of the acetyl group is most often performed in
an alkaline medium [22]. First, the substituent is unblocked at the C28 position (primary
hydroxyl group), then at C3 (secondary hydroxyl group). Unfortunately, under these
conditions, the C-O-P(O)(OEt) ester bond can also be hydrolyzed. Hydrolysis was carried
2
out with sodium hydroxide in a water/tetrahydrofuran/methanol mixture in a volume
ratio of 1:2:1. The course of the reaction was monitored by thin layer chromatography, and
thus the optimal duration of 5 h was determined. Analysis of the spectra of the hydrolysis
product 5 showed that no removal of the phosphate group had occurred under the reaction
31
conditions used. This fact is confirmed by the presence in the P spectrum of the signal of
the phosphorus atom at −0.91 ppm.
Scheme 1. Synthesis of 30-diethoxyphosphoryloxybetulin
5 and its alkynyl derivatives 6a–6e and
7a–
7e. Reaction conditions: (i) (CH CO) O, pyridine, room temp.; (ii) m-CPBA, CHCl , reaction at
3
2
3
boiling temp.; (iii) (C H O) P(O)Cl, DMAP, pyridine, room temp.; (iv) NaOH, THF/MeOH/H O,
2
5
2
2
room temp.; (v) for A—DCC, DMAP, CH Cl , from
benzene, from −5 to room temp.
−
10 to room temperature, for B—pyridine,
2
2

Molecules 2021, 26, 737
4 of 30
As already mentioned, the introduction of a substituent containing a triple bond to
the betulin molecule may have a beneficial effect on the cytotoxic activity of the compound.
Reactions of 30-diethoxyphosphoryloxybetulin
using free hydroxyl groups. Acids with different structures were selected for the synthesis,
i.e., propiolic acid with a terminal triple bond and acids with a C C bond in the middle of
5 with four alkynyl acids were performed
≡
the carbon chain, i.e., 2-butynoic, cyclopropylpropiolic, and phenylpropiolic acid.
Reactions with acids were performed using the Steglich method, which is suitable for
substrates requiring mild conditions and a non-acidic environment [23]. The use of a 17%
excess of alkynyl acid guarantees complete conversion of the hydroxyl group at the C28
position of the starting substrate
7d that were isolated in 71–82% yields. Additionally, small amounts of diesters 6a
are formed in the reaction that were obtained in 5–12% yields.
The reaction of with propargyl chloroformate was carried out in benzene in the
5, leading to the formation of monosubstituted derivatives
7a–
–6d
5
presence of pyridine to give the alkynyloxycarbonyl derivatives 7e and 6e in yields 60%
and 15%, respectively.
All products were purified by column chromatography. The structures of compounds
1
13
31
3–5, 6a–6e, and 7a–e were confirmed based on the analysis of H, C, and P NMR spectra
and the mass spectrometry method (HR-MS).
In general, there is a wealth of information about the antitumor activity of betulinic
acid. Taking this into account, phosphate derivatives of betulonic
8 and betulinic acid 9
were also included in this study (Scheme 2). In our earlier work, we described the synthesis
and testing of anti-HIV-1 activity of these compounds, but so far they have not been tested
as anticancer agents [9].
C H O
C H O
C2H5O
C2H5O
2
5
O
2
5
O
O
P
P
P
C H O
C H O
2 5
2
5
O
O
O
OH
[O]
COOH [H]
COOH
HO
O
HO
5
8
9
Scheme 2. Synthesis of 30-phosphate derivatives of betulonic (
8
) and betulinic acid (9) from 30-
diethoxyphosphoryloxybetulin (5).
2.2. Structural Studies
One of the steps in the synthesis of derivatives containing a phosphate group at
the C30 position is the hydrolysis of acetyl groups at the C3 and C28 positions. This
reaction is carried out in an alkaline medium. As previously described, for 3,28-diacetyl-
3
0-diethylphosphonobetulin, this reaction proceeds with allyl–vinyl isomerization in the
isopropenyl group [ ]. The compound contains a phosphate group in the allyl position
C30), so the substituent is connected via an oxygen atom. Based on H and P NMR
spectra of the hydrolysis product (compound ), isomerization was not observed. The
7
4
1
31
(
5
31
structure of product
δ −0.91 ppm), and also H NMR (signals
5
was confirmed based on the analysis of P NMR spectra (singlet at
1
δ = 4.15 ppm and δ = 1.37 ppm), which prove the
presence of the ethyl groups of the phosphate substituent (Figure S2a,c). The position of
the H3 proton signals ( = 3.21 ppm) and H28 ( = 3.80 ppm and = 3.32 ppm) corresponds
to the chemical shift of these protons in the spectra of unsubstituted betulin containing free
hydroxyl groups. In the 13C NMR spectrum of compound
(Figure S2b), the signals of
the C3 and C28 carbon atoms are shifted by about 3 ppm (δ = 77.9 ppm and δ = 59.2 ppm,
respectively) compared to the position in the substrate spectrum ( = 80.9 ppm and
δ
δ
δ
5
4
δ
δ = 62.4 ppm) including acetyl groups (Figure S1b). This shift is due to the absence of the
adjacent electron-accepting carbonyl group, which had a revealing effect.
Using the X-ray structural study of a single-crystal and Raman spectra, the structure
of new 30-phosphate derivatives was analyzed. Additionally, a comparison was made

Molecules 2021, 26, 737
5 of 30
with selected 29-phosphonate analogs in terms of the position of the double bond in the
isopropenyl group.
2.2.1. Raman Spectra
The experimental Raman spectra for phosphate and phosphonate derivatives were
−
1
measured in the frequency range of 200–2000 cm using a confocal alpha 300R (WITec,
Germany) spectrometer.
Compounds for which experimental Raman spectra were made were grouped into
three sets depending on the type of substituents, i.e., (5, 10), (7a, 11), and (7c, 13) (Figure 2).
O
P
30
20
2
9
O
O
O
P
20
5, 10 R = H
a, 11 R = C(O)C CH
7c, 13 R = C(O)C
O
O
30
29
O
7
OR
OR
C
HO
HO
5
, 7a,7c
Figure 2. Structure of allyl substituted compounds
0, 11, and 13 (phosphonate derivatives).
10, 11, 13
5, 7a, and 7c (phosphate derivatives) and vinyl
1
The two compounds in each set differ in the position of the C=C double bond and
formation of the O=P-O-C (allyl substituted compounds 7a, and 7c) or O=P-C=C (vinyl
substituted compounds 10 11, and 13) groups, respectively (Figure 2). In phosphate
5,
,
derivatives, this group is attached through an oxygen atom, and the distances of the
phosphoryl group (P=O) from the C=C double bond are different.
Experimental frequencies and band assignments are collected in Table 1. The theoreti-
cal Raman spectra for selected compounds, i.e., 5 and 10, were calculated using the DFT
method (see Materials and Methods). In Table S1, the calculated band frequencies and
assignments for 5 and 10 are collected and included in the Supplementary Materials.
Table 1. Experimental band frequencies and assignments of Raman spectra for compounds 5, 7a, 7c, 10, 11, and 13.
Experimental Spectra (cm 1
−
)
5
7a
7c
10
-
11
13
-
1719
1652
1704
1650
1715
1627
1710
1629
C=O stretching
1
664
1625
C=C stretching in allyl/vinyl chains
CH , CH , CH bending in side chain and
3
2
1
465–1450 1465–1450
1458
1465–1450 1465–1450
1458
cycloalkane ring
-
1350
1325
1367
1326
1370
1325
-
1360
1334
P=O stretching, CH , CH , CH bending in side chain
3
2
1
1
324
1322
284
-
1265
-
1275
-
-
-
-
-
P-O stretching in O=P-O-C moiety
P-C stretching in O=P-C=C moiety
1260
1261
P=O, CH , CH , CH bending in side chain and
3
2
1
234
1235
1235
1237
1234
1234
cycloalkane ring
-
1170
1141
1097
1036
921
-
1187
1145
1098
1044
922
-
1170
-
1173
-
1186 s
1141
1101
1040
-
1
1
1
143
105
038
P-O stretching, CH , CH CH bending and C-C
3 2,
stretching in cycloalkane ring, C-O stretching
1100
1039
-
1100
1039
-
9
21
-
P-O bending in O=P-O-C moiety
P-C bending in O=P-C=C moiety
790
795
792
O-C=O stretching, CH , CH CH bending in
3
2,
-
755
766
-
752
752
cycloalkane ring
7
36
04
745
708
732
707
735
704
735
704
735
707
CH , CH CH bending in cycloalkane ring
3
2,
7
CH , CH in side chain, P-O stretching
3 2

Molecules 2021, 26, 737
6 of 30
Table 1. Cont.
Experimental Spectra (cm 1
−
)
5
7a
7c
657
614
-
10
11
13
662
620
619
560 w
533 s
462
413
-
6
6
6
5
5
4
4
3
2
62
14
15
60
35
60
14
88
32
665
618
595
558
535
459
410
389
234
661
619
593
561
531
560
414
386
233
661
618
595
560
530
460
410
-
CH , CH , CH bending in cycloalkane ring, in side
3
2
chains, stretching C-C in cycloalkane ring
-
530
470
402
-
stretching C-C in cycloalkane ring, C-H bending
OH bending
240
-
236
vw—very weak, w—weak, s—strong.
The band assignment of experimental spectra was performed by comparison with
those calculated as well as using the literature databases of the IR and Raman spectra [24,25]
.
Comparing the intensities and positions of the individual theoretical and experimental
peaks (Table 1 and Table S1), we can notice quite significant differences, which, however, do
not affect the quality of band assignment to individual vibrations. This difference is mainly
due to the approximations of the calculation model (the size of the base wave functions
and harmonic vibrations approximation) and the fact that the calculations were performed
for a single molecule and the interactions of molecules were not taken into account (see
Materials and Methods). Despite the limitations of the model, the spectral calculations
provided helpful additional information for the interpretation of the experimental data.
Raman spectra of all investigated compounds show intense bands in the range of
−
1
1
664–1625 cm assigned to the stretching vibration of the C=C double bound (Figure 3)
.
According to literature data, the C=C stretching band in alkyl-substituted ethylenes is
−
1
usually located in the neighborhood of 1650 cm . The position of the C=C stretching
vibration band is influenced by the adjacent functional groups, so in the molecule , one of
the carbon atoms is bound to two hydrogen atoms and the other one to hydrogen and the
CH - group connected to the phosphate residue. In compound 10, however, one carbon
5
-
2
atom is connected to a hydrogen atom and a methyl group, and the other to a hydrogen
atom and a phosphorus atom of the phosphonate group (Figure 2).
The frequencies of the individual C=C stretching bands
pounds are summarized in Table 2.
νc=c for the studied com-
Table 2. Bond length and selected group Raman frequencies for the double bond C20–C29 experi-
mental and calculated. Band shift (∆νC=C) was calculated as a difference between band frequencies
for following sets of compounds: 5–10, 7a–11, and 7c–13.
Raman Spectra
Bond Length
C20 = C29
Experimental
Shiftexp
Calculated Band
Frequency
Shiftcal
(∆ν
Band Frequency
(∆νC=C)exp
)
C=C cal
−1
−
1
−1
−1
[cm ]
(
νC=C)exp [cm
]
[cm
]
(ν
)
[cm
]
C=C cal
5
1.326(5)
1664
1677
3
2
2
9
68
1
0
a
1
1.338(3)
1625
1609
7
1
-
-
-
-
1652
-
-
-
-
-
5
1627
-
-
-
7
c
1650
1
1
3
1629

Molecules 2021, 26, 737
7 of 30
−
1
Figure 3. Experimental Raman spectra in the range 200–2000 cm for three sets of compounds:
(
5, 10) (A), (7a, 11) (B), and (7c, 13) (C).
Within each set of the compounds there is a band shift (∆νc=c)exp due to different
positions of the C=C bond in the side chain. For compounds
5 and 10, the change in

Molecules 2021, 26, 737
8 of 30
location of the C=C bond is accompanied by shortening of the bond length from 1.338(3)
Å to 1.326(5) Å and causes the band shift towards larger wavelength by the value of
−
1
(
∆νc=c)exp = 39 cm . The calculated band shift (∆νc=c)cal for these compounds is much
−
1
higher and equal to 68 cm (see Table 2). For sets of compounds 7a–11 and 7c–13, the
bond shifts (∆νc=c)exp are equal to 25 and 21 cm , respectively.
−
1
−
1
As shown in Table 1, the bands at 1284 and 1275 cm relate to stretching vibration of
the P-O bond in the O=P-O-C side chain, which appear in compounds and 7c. In turn, the
P-C stretching vibration in the O=P-C=C side chain at 1260 and 1261 cm can be observed
for compounds 10 and 13, respectively. It can be noticed that there are some deviations
from this scheme, i.e., the P-O and P-C stretching bands do not appear clearly in the spectra
of compounds 7a and 11, as shown in Figure 2.
5
−
1
Similarly, the bending vibrations of the P-O and P-C bonds are observed in ranges
−
1
9
21–926 and 790–795 cm , for 5, 7a, and 7c and 10, 11, and 13, respectively (see Table 1).
For compounds 7a 7c 11, and 13, the single peak of C=O stretching vibration in the
,
,
−
1
range 1704–1719 cm can be observed. The C-H bending vibration in the side chain and
in the cycloalkane ring as well as the C-C cycloalkane ring vibrations appear mainly at low
−
1
frequencies, e.g., about 700–500 cm , and they are confirmed by theoretical spectra (see
Table S1).
2.2.2. Crystal Structure
The single crystal suitable for the X-ray diffraction analysis was grown by slow
evaporation at room temperature technique from acetonitrile solution. The molecular
structure of the compound with atom numbering is shown in Figure 4. Additional
information about the crystal can be found in Supplementary Materials, Tables S2 and S3.
5
Figure 4. Molecular structure of the compound
5 showing the atom numbering scheme. Displacement
ellipsoids are drawn at the 50% probability level.
Compound
5 crystallizes in a monoclinic P2 space group. There are two molecules
1
per asymmetric unit. The molecular structure of
5
is stabilized by hydrogen bonds as
well as weak intermolecular C-H· · · O bonds (Table S4). In the pentacyclic scaffold of
betulin, the six-membered rings adopt chair conformations and are joined similarly, as
described previously for other betulin derivatives [7]. The five-membered ring assumes the
conformation of the twisted envelope. Structure analysis confirms that in the derivative
5

Molecules 2021, 26, 737
9 of 30
the diethyl phosphate substituent is connected to the betulin molecule via an ester bond
3
attached to a sp hybridized carbon (C30) atom. The length of the C20–C30 bond for
compound 5 is 1.490 Å, and it is longer than the C20–C29 bond equal 1.326 Å (Table 3).
Table 3. Selected bond lengths (Å) for the crystal structure of 30-diethoxyphosphoryloxybetulin
and 28-cyclopropylpropynoyl-29-diethoxyphosphorylbetulin 13.
5
Lengths of Selected Bonds [Å]
Compound
C20–C29
C20–C30
C20–C19
3
0-diethoxyphosphoryloxybetulin 5
8-cyclopropylpropynoyl-
9-diethoxyphosphorylbetulin 13 [7]
1.326(5)
1.490(5)
1.503(5)
2
1
.338(3)
1.515(3)
1.514(3)
2
The comparison of the structure of an isopropenyl group containing a phosphonate
13) and a phosphate (5) substituent is shown in Figure 5.
(
Figure 5. The structure of the substituted isopropenyl group: (A)—for compound 5, (B)—for reference
phosphonate derivative 13.
2
.3. Anticancer Activity
The anticancer activity has been tested for various alkynyl derivatives of triterpenes,
and the results obtained were promising [13 26 27]. On the other hand, the literature also
,
,
contains information that the introduction of a dialkylphosphate group into a molecule
of a known drug substance may improve its activity or even change it towards having an
antitumor effect [28
could create a derivative with higher antitumor activity. Thus, the phosphate derivatives
synthesized in this study 6a 6e, and 7a 7e were assessed for in vitro activity against
,29]. The combination of these two moieties in the betulin molecule
4
,
5,
–
–
four human tumor cell lines such as amelanotic melanoma (C-32), glioblastoma (SNB-19),
and two lines of breast cancer (Luminal A class—T47D; Claudin—low class; a triple
negative breast cancer—MDA-MB-231).
2.3.1. Cell Viability Studies
Evaluation of the effect of new betulin derivatives on the viability of the cultured
cells was performed using the WST-1(water soluble tetrazolium salt) assay. Betulin
and cisplatin were used as the reference compounds. The results, expressed as half the
maximum inhibitory concentration (IC ), are shown in Table 4.
1
50

Molecules 2021, 26, 737
10 of 30
Table 4. Effect of 30-phosphate derivatives of betulin and reference compounds (betulin and cisplatin)
on the viability of cells from four cancer cell lines studied.
Human Cell Line/IC50 ± SD [µM]
Compound
C-32
SNB-19
5.01 ± 0.78
10.93 ± 1.33
—
T47D
1.3 ± 0.13
1.04 ± 0.14
9.50 ± 0.74
2.61 ± 0.42
NA
MDA-MB-231
4
5
33.86 ± 1.59
43.46 ± 2.50
—
14.48 ± 1.37
0.40 ± 0.02
6
a
a
-
7
2.15 ± 0.22
NA
0.91±0.09
4.65 ± 0.51
0.76 ± 0.14
NA
6.20 ± 0.22
21.18 ± 1.24
4.93 ± 0.44
6
7
b
b
0.80 ± 0.03
NA
15.40 ± 0.51
NA
6
7
c
c
NA
11.90 ± 0.52
31.46 ± 1.56
6.90 ± 0.40
11.43 ± 0.76
3.97 ± 0.86
8.80 ± 0.17
0.60 ± 0.08
7.70 ± 0.11
6.99±1.07
9.00 ± 0.53
15.61 ± 1.11
4.29 ± 0.65
6.29 ± 0.12
11.22 ± 0.28
4.38 ± 0.04
9.23 ± 0.57
1.59 ± 0.09
10.20 ± 0.69
0.43 ± 0.14
11.67 ± 0.44
—
NA
9.11 ± 0.39
6
7
d
d
NA
NA
9.47 ± 0.03
NA
NA
6
7
e
e
—
NA
—
1
1
1
1
1
0 *
1 *
2 *
3 *
4 *
10.71 ± 0.86
0.70 ± 0.13
14.94 ± 0.90
12.2±0.84
NA
—
—
—
—
—
—
Betulin
9.93 ± 0.25
5.29 ± 0.12
NA
7.76 ± 1.87
5.47 ± 0.05
Cisplatin
6.26 ± 1.24
NA means IC50 > 50 µM. * Phosphonate derivatives of betulin; synthesis and the results of the anticancer activity
test were described previously [7].
For all cell lines tested, the 3,28-diester derivatives 6a
-
e
were less active (IC50 range
4.65–31.5 µM) than the 28-monoester derivatives 7a–7e (IC50 range 0.76–15.40 µM). Among
the monoester compounds, 7b and 7a were the most active. They contain, respectively, 2-
butynoyl and propynoyl substituents. Monoester derivatives of acids with a more extensive
structure, 7c (cyclopropylpropynoyl) and 7d (phenylpropynoyl), were significantly less
active, but their activity was higher compared to betulin 1.
It can be seen that among the cell lines tested, cells of both breast tumors were
characterized by the lowest sensitivity to alkynyl derivatives of betulin phosphate. The
highest activity for the T47D (IC50 = 1.04
determined for the 30-phosphate of betulin
the 28-propynoyl derivative 7a showed a significant effect on T47D cells (IC50 = 2.61
µ
M) and MDA-MB-231 (IC50 = 0.40
. In the group of phosphate derivatives, only
M).
µM) lines was
5
µ
As previously described, compounds containing a phosphonate group at position C29,
analogous betulin derivatives, were active against this cell line; however, the unsubstituted
compound 10 (IC50 = 1.55 µM) was less active than the 28-propynoyl 11 (IC50 = 0.70 µM)
analog [Table 4].
The C-32 and SNB-19 cells seem to be more sensitive to 30-phosphate betulin deriva-
tives. Compound 7b showed the strongest activity for both lines (IC50 = 0.76 and 0.8 µM
,
respectively).
The derivative 7a with the terminal triple bond was slightly weaker in relation to
SNB-19 cells (IC50 = 0.91 M) and more than two times less active towards the C-32
line (IC50 = 2.15 M). Among betulin 29-phosphonates, the activity of the correspond-
ing propynoyl derivative 11 (IC50 = 0.6 M) was significantly higher than that of the
-butynoyl derivative 12 (IC50 = 7.70 M). A meaningful difference in the cytotoxicity
µ
µ
µ
2
µ

Molecules 2021, 26, 737
11 of 30
of 28-propargyloxycarbonyloxy derivatives was also observed. The structure–activity
relationships observed in the 30-phosphate monoesters 7a–7e indicate that the order of
the substituents in the cytotoxic effect on the SNB-19 cell line is as follows: 2-butynoyl >
propynoyl > propargyloxycarbonyl > 3-cyclopropyl-2-propynoyl > 3-phenyl-2-propynoyl.
Comparing the activities of compound 7a with those determined for 28-O-propynoylbetulin
in relation to appropriate tumor cell lines (IC50 values 16.7, 4.2, and 110.8 µM for C-32,
SNB-19, and T47D, respectively) [27], it seems that the introduction of a diethylphosphate
substituent to the isopropenyl group may exert a beneficial effect on the cytotoxic activity
of the obtained derivatives.
The phosphate derivatives of betulonic (8) and betulinic (9) acids showed no activity
in relation to the examined breast cancer lines (T47D and MDA-MB-231). Activity against
C-32 and SNB-19 cells determined for these compounds as well as the references (29 and
30 phosphonate derivatives and betulinic acid) are summarized in Table 5.
Table 5. Effect of 30-phosphate derivatives of betulonic acid 8, betulinic acid 9, and reference compounds on the viability of
cells from two cancer cell lines studied (C-32 and SNB-19).
R
30-OP(O)(OC H )
29-P(O)(OC H ) *
30-P(O)(OC H ) *
H
2
5 2
2
5 2
2
5 2
O=
Comp. 8
OH
Comp. 9
R1
O=
OH
CH C(O)O
OH
OH
3
C-32
45.74 ± 6.85
45.75 ± 5.27
NA
45.52 ± 3.72
1.64 ± 0.18
47.46 ± 2.96
7.67 ± 1.57
30.74 ± 2.17
1.29 ± 0.15
NA
NA
8.63 ± 0.90
20.09 ± 1.26
IC50
[
µM]
SNB-19
5.98 ± 0.43
*
—Phosphonate derivatives of betulonic and betulinic acids; synthesis and the results of the anti-HIV-1 activity test were described
previously [9].NA Means IC50 > 50 µM .
Initial studies of the anticancer activity of betulinic acid were focused on human
melanoma cells, and then its effects on numerous other cancer cell lines, including glioblas-
toma, have been reported [30 31]. The evaluation of the activity of phosphorus derivatives
of betulonic acid and betulinic acid, carried out in our study, showed that these com-
pounds are much less active against C-32 melanoma cells (IC50 range 30.74–47.46 M)
than betulinic acid (IC50 = 8.63 M; R=H, R =OH). The tested acid derivatives showed
a much greater effect on SNB-19 cells. The strongest activity (IC50 = 1.29
sented in the 30-phosphonate derivative of 3-acetylbetulinic acid (R=30-P(O)(OC H ) ,
,
µ
µ
1
µM) was pre-
2
5 2
R =CH C(O)O). The IC value determined for 30-phosphate of betulinic acid
9 was more
1
3
50
than 4.5-fold higher.
2.3.2. Analysis of the Expression of Genes Related to the Apoptosis Process
In the fight against cancer, it is especially important to understand the mechanisms
involved in cell death. The effectiveness of anticancer therapy results not only from the
inhibition of proliferation, but also from the ability to induce apoptosis in cancer cells.
Apoptosis, also known as programmed cell death, is a fundamental physiological process
that plays a key role in the development and maintenance of tissue and body homeostasis.
A cell dying in the process of apoptosis undergoes both morphological changes (chromatin
condensation, changes in the structure of the cell nucleus, and formation of apoptotic bod-
ies) and biochemical changes (activation of many proteins classified as both transcription
factors and enzymes and changes in the structure of the cell membrane) [32]. Changes in
apoptotic cells are determined by the expression of relevant genes that are involved in this
complex process and act both pro-and anti-apoptotically [33–35].

Molecules 2021, 26, 737
12 of 30
In our work, we assessed the influence of the tested compounds on the expression of
selected genes related to apoptosis. The two most active derivatives 7a and 7b were selected
for the study. We examined their influence on the transcriptional activity of genes encoding
a proliferation marker (histone H3), a cell cycle regulator (p53), and proteins related to the
mitochondrial apoptosis pathway (BCL-2 and BAX). The study was performed after 24-h
exposure of the cultured cells to the test compounds (half of IC ).
50
The structural organization of DNA into chromatin is essential for maintaining cellular
identity. In the S phase of the cell cycle, the chromatin structure is multiplied (related to,
among others, the synthesis of histones), which is coupled with DNA replication. Histone
H3, being one of the components of chromatin, undergoes several post-translational mod-
ifications, influencing gene expression. According to the literature data, the level of H3
mRNA quantified by QRT-PCR can serve as a marker in the assessment of proliferative
activity and the development of brain glioma, for example [34,35].
In our study, compounds 7a and 7b reduced the number of mRNA copies of the
histone H3 gene in the cells of both tumor lines compared to the level determined for the
control (Figure 6). The antiproliferative effect of these compounds may therefore be due to
the ability to reduce the histone H3 genes at the mRNA level.
Figure 6. The effect of compounds 7a and 7b on transcriptional activity of H3 for SNB-19 (
exposed to compound 7a and 7b for 24 h at a concentration equal to half of IC50 determined for SNB-19 (0.3 and 0.25
respectively) and for C-32 (0.7 and 0.25 g/mL, respectively). For the reference compound, cisplatin, the appropriate
A
) and C-32 (B) cells. Cells were
µg/mL,
µ
concentrations were 0.8 µg/mL (SNB-19 line) and 0.6 µg/mL (C32 line). * p < 0.05 compared with the untreated control.
Most cancers have a mutation in the TP53 suppressor gene, which encodes the P53
protein. This leads to an inhibition of cell apoptosis and tumor growth. In apoptosis
induced by DNA damage, the p53 protein is involved in various activation steps in both
the internal and external pathways. It is related to its function as a transcription factor of
pro-apoptotic genes and its direct interaction with anti-apoptotic proteins, which results in
their blockade [36].
The results of our research presented in Figure 7 indicate that the tested compounds
increase the number of mRNA copies of TP53. Compound 7b upregulated TP53 expression
levels, but not as significantly as was observed for 7a.
There is a strong genetic and biochemical link between the tumor suppressor p53 and
the oncogene BCL-2 that is fundamental to cancer biology. The BCL-2 family of proteins
includes both pro-apoptotic (e.g., BAX) and anti-apoptotic (e.g., BCL-2) proteins. The task
of anti-apoptotic proteins from the BCL-2 family is to bind to the active form of BAX, which
prevents it from penetrating into the outer membrane of the mitochondria and blocks
apoptosis [33,37,38]. The effectiveness of anti-cancer therapies depends on the mechanisms
of BCL-2/BAX-dependent cancer cell killing [38].

Molecules 2021, 26, 737
13 of 30
Figure 7. The effect of compounds 7a and 7b on transcriptional activity of TP53 for SNB-19 (
exposed to compound 7a and 7b for 24 h at a concentration equal to half of IC50 determined for SNB-19 (0.3 and 0.25
respectively) and for C-32 (0.7and 0.25 g/mL, respectively). For the reference compound, cisplatin, the appropriate
concentrations were 0.8 µg/mL (SNB-19 line) and 0.6 µg/mL (C32 line). * p < 0.05 compared with the untreated control.
A) and C-32 (
B) cells. Cells were
µg/mL,
µ
As shown in Figure 8, the tested compounds increased the mRNA copy number to
BAX and BCL-2 ratio. Given this result, it can be assumed that the derivatives 7a and 7b
affect apoptosis via the mitochondrial route.
Figure 8. The effect of compounds 7a and 7b on BAX/BCL-2 for SNB-19 (
to compound 7a and 7b for 24 h at a concentration equal to half of IC50 determined for SNB-19 (0.3 and 0.25
respectively) and for C-32 (0.7 and 0.25 g/mL, respectively). For the reference compound, cisplatin, the appropriate
concentrations were 0.8 µg/mL (SNB-19 line) and 0.6 µg/mL (C32 line). * p < 0.05 compared with the untreated control.
A
) and C-32 (
B) cells. Cells were exposed
µg/mL,
µ
2.4. Physicochemical Descriptors (Lipophilicity, ADME Properties)
For new derivatives, log PTLC parameters were determined, and correlations between
the values of lipophilicity parameters determined experimentally and calculated with the
use of computer programs were ascertained.
Lipophilicity is an important physicochemical parameter related to the biological
activity of compounds. This descriptor, contained in Lipi n´ ski’s rule of five, concerns the
transport of a compound through the biological membrane system and the formation
of a compound–receptor complex [39]. To assess the lipophilic nature of triterpenoids
3–
5,
6a–6e, and 7a–7e, the reverse phase chromatography (RP-TLC) method was used
(
chromatographic procedure by RP-TLC is included in the Supplementary Materials). The
theoretical characteristics of the chromatographic parameters (RM0, log PTLC) of the tested
derivatives were also performed. The results are presented in Table 6.

Molecules 2021, 26, 737
14 of 30
Table 6. The experimental values of lipophilicity for tested compounds 3–5, 6a–6e, and 7a–7e.
Compound
RM0
-b
r
logPTLC
3
4
5
5.655
5.740
3.719
6.893
5.318
6.884
5.431
6.419
5.587
7.877
6.145
6.490
5.302
0.064
0.064
0.046
0.080
0.062
0.078
0.063
0.071
0.064
0.086
0.070
0.076
0.063
0.997
0.998
0.984
0.977
0.988
0.994
0.994
0.998
0.985
0.998
0.992
0.994
0.993
7.362
7.470
4.914
8.929
6.937
8.917
7.079
8.330
7.277
10.174
7.982
8.419
6.916
6
7
a
a
6
7
6
7
b
b
c
c
6
7
6
7
d
d
e
e
The lowest value of experimental lipophilicity (log PTLC = 4.914) was shown for
derivative with two hydroxyl groups in positions C3 and C28. The conversion of these
substituents into ester functions increased the lipophilicity of derivatives 6a 6e and 7a 7e.
5
–
–
The highest increase in the log PTLC value in the group of tested triterpenoids was recorded
for the 6d diester, which is related to the presence of two phenylpropynoyl groups.
The 7a–7e monoesters are characterized by lower log PTLC values than the compounds
6a–6e containing two ester groups. Depending on the structure of the substituent, the
monoesters can be ranked in the following order of increasing lipophilicity propargyloxy-
carbonyl < propynoyl < 2-butynoyl < cyclopropylpropynoyl < phenylpropynoyl. The
derivatives 7a and 7e with a terminal triple bond possess the lowest lipophilicity (log PTLC
.937 and 6.916, respectively) compared to the other alkynyl-substituted triterpenoids.
The experimentally determined lipophilicity of the tested compounds was compared
-
6
with the theoretical values. Additionally, the partition coefficients (log P) were calculated
by use of the seven theoretical programs such as ALOGPs, AC logP, miLogP, ALOGP,
MLOGP, XLOGP2, and XLOGP3 [40]. The results (Table S5, Supplementary Materials)
show that compound
value. In the case of the most lipophilic compound 6d, the same relationship was found.
The calculated theoretical values of log P of compounds 6a 6e, and 7a 7e are in
5 has the lowest lipophilicity, which is consistent with its log PTLC
3–5
,
–
–
the range of 5.15–13.40. The highest log P values were obtained with the XLOGP3 and the
lowest with the ALOGPs.
The correlation matrix for theoretical lipophilicity parameters of derivatives
3–5, 6a–6e,
and 7a 7e are presented in Table S6 (Supplementary Materials). The calculated correlation
–
coefficients are in the range 0.812–0.987, and the highest value of correlation is found in the
cases of XLOGP2 and XLOGP3 (0.987). These programs calculate the value of the log P
parameter using the same method based on the additive atom/group model.
The comparison of the experimental lipophilicity values (log PTLC) for the derivatives
3–5, 6a–6e, and 7a–7e, with the corresponding values obtained by different calculation
methods, is depicted in Figure 9. The highest correlation between theoretically and experi-
mentally received lipophilicity values were obtained in the case of the XLOGP2 program
(
0.861), while the lowest was in the case of MLOGP (0.763).
The Log PTLC values of the most active compounds 7a and 7b range from 6.937 to
7.982. This experimental lipophilicity does not comply with Lipi n´ ski’s rule (log P < 5). This
means that lipophilicity is not the only possible factor influencing the biological activity of
a compound.

Molecules 2021, 26, 737
15 of 30
1
1
1
4
2
0
8
6
4
2
0
LogP
3
4
5
6a
7a
6b
7b
6c
7c
6d
7d
6e
7e
Compounds
ALOGPs
AC logP
miLogP
ALOGP
MLOGP
XLOGP2
XLOGP3
LogPTLC
Figure 9. The values of lipophilicity of triterpenoids 3–5, 6a–6e, and 7a–7e.
For active monoesters 7a–7e, good correlation between antiproliferative activity
against the C-32 and SNB-19 cell lines and lipophilicity could be observed. The obtained
relationships described by the equations (A and B) are presented in Table 7.
Table 7. The correlation between the cytotoxicity (IC ) and the log P values for monoesters 7a–7e.
5
0
Compounds
Cell Line
Equation
r
2
7
7
a–7e
a–7e
C-32
SNB-19
IC50 = −21.887LogP + 331.21LogP—1242.1 (A)
0.720
0.826
2
IC50 = −12.262LogP + 186.35LogP—701.71 (B)
In silico pharmacokinetic analysis is an important method for verifying and simulating
the potential biological activity of a future drug substance. The main goal of early-phase
pharmacokinetic studies is the preselection of newly synthesized chemical compounds and
qualifying structures with optimal properties for further study. The bioavailability studies
of compounds, in addition to the lipophilicity parameter, also include the determination
of such descriptors as molecular weight (M), topological polar surface of the molecule
(
tPSA), log BB (Brain/blood partition coefficient), number of rotatable bonds (nROTB),
and number of donor (nHBD) and acceptor (nHBA) sites for hydrogen bonds [41]. All
ADME parameters for derivatives 6a 6e, and 7a 7e were calculated using ACD/iLab
software and tabulated in Table 8 [42].
3–5
,
–
–
Table 8. Prediction of ADME properties of compounds 3–5, 6a–6e, and 7a–7e.
Compound
M
logBB
nHBD
nHBA
nROTB
tPSA
3
4
5
542.80
678.89
594.81
698.88
646.85
726.93
660.87
779.01
686.91
851.07
722.94
758.93
676.87
0.07
1
0
2
0
1
0
1
0
1
0
1
0
1
5
8
6
8
7
8
7
7
7
7
7
10
8
7
13
9
72.83
107.17
95.03
107.17
101.10
107.17
101.10
90.10
101.10
90.10
101.10
125.63
110.33
−0.55
−0.32
−1.70
−0.85
−1.40
−0.79
−1.70
−0.85
−2.00
−1.40
−1.70
−0.85
6
7
a
a
15
12
15
12
17
13
17
13
19
14
6
7
6
7
b
b
c
c
6
7
6
7
d
d
e
e

Molecules 2021, 26, 737
16 of 30
2
The values of nHBA (
bioavailability of the 30-diethylphosphate derivatives of betulin.
≤
10), nHBD ( 5), and tPSA (<140 Å ) determined the oral
≤
Essentially, the determination of the brain/blood barrier (BBB) permeability parameter
has two aspects. The first is to identify neurotoxic substances that damage the brain.
In addition, it may be helpful in designing new drugs that can achieve the goal in the
CNS [43]. The presence of BBB makes the treatment of glioblastoma very difficult, as the
chemotherapeutic agents used cannot be delivered to the target tumor cells. The triterpenes
3–5
,
6a–
6e, and 7a
–7e have the log BB values that fit within an optimal range of the values
for drug-likeness (log BB
−
3 to 1.2) [44]. On the other hand, the log BB values obtained for
the tested phosphate derivatives (active in vitro against SNB-19 cells) are characteristic of
drugs that are inactive in the CNS. However, only a small number of low molecular weight
fat-soluble compounds have the ability to pass through the BBB. Other molecules need
a special transport system for this. The use of such forms as liposomes, dendrimers, or
polymeric micelles allows chemotherapeutic agents to overcome this barrier and reach the
goal of action [45].
It is also known that in the cancer process, some primary tumor cancer cells may
metastasize. Such cells have the ability to migrate and can colonize even distant tissues [46].
Metastasis is a significant problem in the treatment of neoplastic diseases and contributes
to the worsening of the patient’s prognosis. Metastasis of glioblastoma beyond the nervous
system is determined at the level of about 2% of cases, but the number of this type of
metastasis is increasing. This is due to the greater effectiveness of the primary tumor
treatment, which extends the survival time. Glioma metastases are most often diagnosed
in the lungs and pleura (60%) [47].
2.5. In Silico Study
As already mentioned, abnormalities in the expression of genes regulating the cell
cycle play an important role in the process of cancer development. Both the initiation and
development of the neoplastic process are related not only to such factors as mutations
of proto-oncogenes and suppressor genes, disturbances in the DNA repair process, and
mutations of anti- and pro-apoptotic genes. Increased expression of growth factors and
angiogenesis also plays an important role in these processes. The epidermal growth factor
receptor (EGFR) has been observed to be elevated in stomach, brain, and breast tumors.
This receptor belongs to the receptor tyrosine kinase (RTK), the pathway of which is the
most disrupted signaling cascade in glioblastoma [48,49].
According to our previous article on 3-substituted derivatives of betulin as potential
EGFR inhibitors [19], molecular modeling was carried out to predict the possible mode of
action of the newly synthesized compounds 7a and 7b. EGFR structure was obtained from
PDB (PDB ID: 1M17). The co-crystallized ligand for 1M17 was a known EGFR inhibitor
(
i.e., erlotinib) and was taken as the reference drug (Figure 10).
Figure 10. Structure of reference drug erlotinib.
As illustrated in Figure 11, the crystal structure of 1M17 shows that N1 of the erlotinib
quinazoline moiety receives a hydrogen bond from the amide nitrogen of Met769. Moreover,
the water molecule (W10) essentially forms a hydrogen-bonding network, which involves
nitrogen N3 of the erlotinib quinazoline moiety and the Thr766 side chain.

Molecules 2021, 26, 737
17 of 30
Figure 11. Three-dimensional (3D) binding interaction of erlotinib with EGFR (PDB ID: 1M17).
Genetic Optimization for Ligand Docking (GOLD) ver. 2020.1 [50] was used for the
hydrated docking analysis. GOLD is an automated ligand docking program that relies
on the use of a genetic algorithm to study the full range of conformational flexibility of
ligands with partial protein flexibility (flexibility of receptor hydrogens) [51]. The region
of interest used for GOLD docking was defined as all the EGFR protein residues within
the 10 Å of the reference ligand. For the remaining water molecules, the “Toggle” option
was on (having GOLD decide whether the water should be present or absent, i.e., bound
or displaced by the ligand during docking). The results of docking compounds 7a and 7b
to the active site of EGFR tyrosine kinase obtained in the GOLD program are presented
in Table 9. Ligand matches are presented in the arbitrary units [a.u.] of GOLD computer
program. The higher the calculated value, the better the binding affinities of the ligand for
its protein. The obtained results show that compound 7a has a slightly lower docking score
compared to the reference erlotinib (30.87 a.u. and 31.06 a.u., respectively). In contrast,
compound 7b shows a better binding affinity for EGFR compared to erlotinib (docking
score 31.33 a.u.).
Table 9. The top-ranked ChemScore fitness score function obtained with GOLD and binding energy
obtained with KDEEP for betulin derivatives 7a and 7b.
Compound
GOLD [a.u.]
KDEEP ∆G [kcal/mol]
7
a
30.87
31.33
31.06
−10.29
−11.48
7
b
erlotinib
−9.63 (−8.93 *)
*
Calculating for crystal structure of 1M17.
Binding energy is an important parameter for assessing the stability of the ligand-
protein complex. It is released when a ligand molecule binds to a protein, which lowers
the total energy of the complex. Thus, the greater the energy released when binding a
ligand to a protein, the greater the tendency of the ligand to associate with that protein.
For this reason, a binding energy calculation has been made using the KDEEP predictor.
KDEEP calculates binding energy
DCNNs) model [52].
The KDEEP calculations for the crystal structure of the complex of erlotinib with
∆G [kcal/mol] using deep convolutional neural networks
(
EGFR (1M17) and the complex obtained by docking erlotinib to EGFR tyrosine kinase
domain showed that both of these complexes achieved significantly higher binding energy
values compared to compounds 7a and 7b (Table 9). According to the obtained in KDEEP
results, compounds 7a,b show a better binding affinity to the active site of EGFR compared
to erlotinib.

Molecules 2021, 26, 737
18 of 30
Docked poses of ligand 7a and their interactions with amino acid residues are pre-
sented in Figures 12 and 13.
Figure 12. Docking pose of EGFR tyrosine kinase complex with 7a (A) and 7b (B).
The top docked pose of compound 7a in the binding site showed hydrophobic inter-
actions with Gly690, Thr830 (carbon hydrogen bond), Phe699 (π–alkyl), Ala698, Lys721,
Cys751, Ala719, Leu820, Val702, and Cys773 (alkyl–alkyl). Moreover, hydroxy groups of
triterpene in positions C3 were involved in hydrogen bonds to carboxyl oxygen atom of
Thr766 (Figure 13).
Figure 13. Two-dimensional (2D) binding interaction of 7a with EGFR.
The pose of compound 7b in the binding site showed hydrophobic interactions with 21
amino acids residues including interactions with Asp831, Phe832, and Gly833 (DFG motif).

Molecules 2021, 26, 737
19 of 30
According to the obtained literature data, interactions with DFG motif play a significant
role in tyrosine kinase activity [53]. Moreover, hydroxy groups of triterpene in positions
C3 were involved in hydrogen bonds to carboxyl oxygen atom of Thr766, stabilizing the
obtained complex (Figure 14).
Figure 14. Two-dimensional (2D) binding interaction of 7b with EGFR.
Molecular dynamics (MD) calculation provide a better understanding of structure–
function relationships in motion and other conformational changes by the proteins. There-
fore, an MD study was conducted in this paper to evaluate the dynamic behavior of the
top-scoring complexes for a period of 10 ns. The molecular mechanics/generalized born
surface area (MM/GBSA) method was selected for rescoring complexes, because it is
the fastest force-field based method that computes the free energy of binding, as com-
pared to the other computational free energy methods, such as free energy perturbation
(FEP) or thermodynamic integration (TI) methods [54]. Comparison studies have also
shown that MM/GBSA outperforms molecular mechanics/Poisson Boltzmann surface area
MM/PBSA [55]. The MM/GBSA calculation was performed using MolAICal software [56].
MolAICal provides a way to calculate the MM/GBSA based on the output results of molec-
ular dynamics simulations that are carried out by NAMD [57]. The calculated binding free
energies are shown in Table 10.
Table 10. Calculated binding free energies of tested compounds.
Compound
∆E(internal)
∆E(electrostatic) + ∆G(sol)
∆E(VDW) + ∆G(sol
)
∆G(bind)
7
a
−4.96
−5.91
−7.40
20.25
24.88
11.80
−61.82
−62.48
−35.18
−46.53
−43.51
−23.38
7
b
erlotinib
The binding free energies of 7a and 7b range from
are much smaller than those of erlotinib. According to the calculations performed, betulin
−
46.53 to
−
43.51 kcal/mol, which
derivatives 7a and 7b are better binders than the reference compound.

Molecules 2021, 26, 737
20 of 30
Analyzing the complexes ligand root-mean-square deviation (RMSD) and hydrogen
bonding can provide insights into a structural conformation faced by the protein and ligand
during simulation (Figures 15 and 16).
Figure 15. Graphical representation of the plots showing proteins CαRMSD (Å) and ligands atoms versus time (10 ns).
Figure 16. Graphical representation of the plots showing proteins RMSF [Å] versus residue index number of protein for 7a
and 7b complexes of EGFR tyrosine kinase.

Molecules 2021, 26, 737
21 of 30
The plots for EGFR proteins C
α
versus time for the simulation with compounds 7a
and 7b are shown in Figure 15. The visual analysis of the trajectory confirms the stability
of the structure of protein at around 6 ns. It should be mentioned that the RMSD values
obtained for the complexes were lower compared to the apo protein.
The plot for ligands RMSD versus time shows a plateau value of 1 Å and 1.3 Å for the
ligands 7a and 7b, respectively, at around 0.3 ns. There was no a significant shift in RMSD
values for the ligands backbone. The deviation value observed was within limits, which is
quite stable for binding of the ligands with the EGFR tyrosine kinase. The higher value of
RMSD for the 7b compound may be due to the butynyl group at carboxylate moiety.
The residue root mean square fluctuation (RMSF) is a measure of the flexibility of a
residue. It is typically calculated for the C atom of each residue. The plot for protein
α
RMSF versus residue number index shown in Figure 16 describes fluctuations in the range
of 0.5–2.5 Å for complexes EGFR with 7a and 7b. The fluctuations of the C-terminals ends
were higher. The catalytic residues are no longer higher than 0.5 Å for both complexes, and
are suitable for stable binding.
The data presented in Figure 17 show the results of analysis of hydrogen bond contacts
in the studied complexes. Compound 7a (Figure 17A) forms three hydrogen bonds, and
the binding with Thr766 residue occurs in the complex at the 28% time of measurement.
Ligand 7b forms two hydrogen bonds during simulation time, including a stable H-bond
with Thr766 (Figure 17B).
Figure 17. Pictorial representation of the number of H-bond contacts formed by ligand 7a
tyrosine kinase.
(A) and 7b (B) with EGFR
From the discussion, it can be summarized that the results obtained in silico are
consistent with the experimental data, and compounds 7a and 7b have an efficient binding
inside the cavity of EGFR tyrosine kinase.
3. Materials and Methods
3.1. General Experimental Procedures
Materials and methods are included in the Supplementary Materials.
3
,28-Diacetyl-30-hydroxybetulin
3
was obtained in 52% yield by oxidation of 3,28-
1
diacetylbetulin
2
with m-chloroperbenzoic acid in chloroform. Melting point and H and
13
C NMR spectra for compound 3 were consistent with the literature data [58].
Phosphate derivatives of betulinic
to the described method. Spectroscopic ( H, C, and P NMR) and melting point data
were consistent with the literature values [9].
8
and betulonic
9
acid were synthesized according
1
13
31

Molecules 2021, 26, 737
22 of 30
3.2. Procedures for Synthesizing Compounds
3.2.1. Synthesis of 3,28-Diacetoxy-30-diethoxyphosphoryloxybetulin 4
A flask with a solution of compound
3 (0.54 g, 1 mmol) and 12 mg (0.1 mmol) 4-
dimethylaminopyridine (DMAP) in 6 mL pyridine was placed in an ice-water bath. Then,
diethyl chlorophosphate (0.28 mL, 1.9 mmol) was added dropwise to the cooled mixture,
and stirring was continued for 24 h under an inert gas (argon) atmosphere after the
cooling bath was removed. The course of the reaction was monitored using thin layer
chromatography (TLC). After completion of the reaction, the pyridine was evaporated
under reduced pressure. Dichloromethane (30 mL) was added to the residue, and the
resulting solution was washed with 10% HCl (2
×
10 mL), saturated NaHCO (12 mL), and
3
then with water (2 10 mL). The organic phase was dried with anhydrous Na SO and then
×
2
4
concentrated. The product was isolated by column chromatography (SiO , hexane:ethyl
2
acetate, 3:2, v/v).
Yield: 72%, m. p. 61–63 C.
◦
1
TLC (dichloromethane:ethanol, 15:1, v/v): R = 0.48. H-NMR (600 MHz, CDCl )
δ (ppm):
f
3
5
.06 (s, 1H, H-29), 5.00 (s, 1H, H-29), 4.51 (m, 2H, H-30), 4.48 (m, 1H, H-3), 4.26 (d,
J = 10.8 Hz, 1H, H-28), 4.16 (m, 4H, 2 OCH CH ), 3.84 (d, J = 10.8 Hz, 1H, H-28), 2.39 (m,
×
2
3
1
6
0
H, H-19), 2.06 (s, 3H, C(O)CH ), 2.09 (s, 3H, C(O)CH ), 0.85–2.00 (m, CH, CH ), 1.38 (m,
3 3 2
H, 2
×
OCH CH ), 1.08 (s, 3H, CH ), 1.01 (s, 3H, CH ), 0.87 (s, 3H, CH ), 0.86 (s, 3H, CH ),
2 3 3 3 3 3
1
3
.85 (s, 3H, CH ), 0.80 (m, 1H, H-5). C-NMR (150 MHz, CDCl3)
δ
(ppm): 171.6; 171.0;
3
149.0; 110.5; 80.9; 63.8; 63.8; 62.4; 55.4; 50.2; 49.5; 46.3; 43.2; 42.7; 40.9; 38.4; 37.8; 37.5; 37.1;
3
4.3; 34.1; 31.3; 29.7; 28.0; 27.0; 26.6; 23.7; 21.3; 21.0; 20.9; 18.2; 16.2; 16.2; 16.1; 16.0; 14.9.
31
−
P-NMR (243 MHz, CDCl3)
δ
(ppm):
−
0.91. HR-MS (APCI) m/z: C H O P [(M–H) ],
38
62
8
Calc. 677.4182; Found 677.4180.
3
.2.2. Synthesis of 30-Diethoxyphosphoryloxybetulin 5
A solution of 730 mg sodium hydroxide in 10 mL of methanol, 20 mL of tetrahy-
drofuran, and 10 mL of water was prepared in the flask. Then, 680 mg of compound
(1 mmol) was added. The reaction was carried out at room temperature while stir-
4
ring with a magnetic stirrer. The progress of the reaction was monitored by TLC (SiO2,
dichloromethane:ethanol, 15:1, v/v). After 5 h, the reaction mixture was evaporated to
dryness. The residue was dissolved in dichloromethane and washed with 10 mL of hy-
drochloric acid and then three times with 30 mL of water. The organic phase was dried
with anhydrous Na SO , and the product was purified using column chromatography
2
4
(
SiO , dichloromethane:ethanol, 15:1, v/v).
2
◦
Yield: 85%, m. p. 194–195 C.
1
TLC (dichloromethane:ethanol, 15:1, v/v): R = 0.36. H-NMR (600 MHz, CDCl )
5
δ
(ppm):
.49 (d, J = 1.2 Hz, 1H, H-29); 4.99 (s, 1H, H-29), 4.51 (m, 2H, H-30), 4.15 (m, 4H, 2
OCH CH ), 3.80 (d, J = 10.8 Hz, 1H, H-28), 3.32 (d, J = 10.8 Hz, 1H, H-28), 3.21 (m, 1H, H-3),
f
3
×
2
3
2
.35 (m, 1H, H-19), 0.90–2.00 (m, CH, CH ), 1.37 (m, 6H, 2
×
OCH CH ), 1.07 (s, 3H, CH ),
2
2 3 3
1
.04 (s, 3H, CH ), 1.00 (s, 3H, CH ), 0.84 (s, 3H, CH ), 0.78 (s, 3H, CH ), 0.70 (m, 1H, H-5).
3
3
3
3
1
3
C-NMR (150 MHz, CDCl3)
δ
(ppm): 148.3; 109.2; 77.9; 68.0; 2
×
62.8; 59.2; 54.2; 49.3; 48.5;
46.7; 42.1; 41.6; 39.9; 37.8; 37.7; 36.2; 36.1; 33.2; 32.7; 30.5; 28.2; 27.0; 26.3; 26.0; 25.7; 19.9; 17.3;
31
2
×
15.2; 15.0; 14.3, 13.7. P-NMR (243 MHz, CDCl3)
δ
(ppm):
−
0.91. HR-MS (APCI) m/z:
−
C H O P [(M–H) ], Calc. 593.3971; Found 593.3959.
3
4
58
6
3.2.3. General Procedure for Synthesis of Alkynyl Derivatives 6a–6d and 7a–7d
A solution of 595 mg of 30-diethoxyphosphoryloxybetulin
5 (1 mmol) in 5 mL of
◦
dichloromethane was cooled in an ice–salt bath to
−
10 C, and 1.17 mmol of the corre-
sponding alkynyl acid was added. Then, a solution of 240 mg of DCC and 10 mg of DMAP
in 1 mL of dichloromethane was slowly added dropwise. The reaction was carried out
under an argon atmosphere while stirring with a magnetic stirrer. The flask with the
reaction mixture was kept in an ice–salt bath for the first 5 h, after which the reaction
was carried out at room temperature. After 24 h, the precipitate was filtered off and the

Molecules 2021, 26, 737
23 of 30
filtrate was concentrated to dryness in a vacuum evaporator. Subsequently, the reaction
products were purified using column chromatography (SiO , dichloromethane:ethanol,
2
15:1 or 40:1, v/v).
3
0-Diethoxyphosphoryloxy-3,28-dipropynoylbetulin 6a
◦
Yield: 5%, m. p. 127–130 C.
1
TLC (dichloromethane:ethanol, 15:1, v/v): R = 0.75. H-NMR (600 MHz, CDCl )
5
δ
(ppm):
.07 (s, 1H, H-29), 5.00 (s, 1H, H-29), 4.62 (m, 1H, H-3), 4.51 (m, 2H, H-30), 4.40 (d,
J = 10.8 Hz, 1H, H-28), 4.15 (m, 4H, 2 OCH CH ), 3.98 (d, J = 10.8 Hz, 1H, H-28), 2.92 (s,
H, C CH), 2.88 (s, 1H, C CH), 2.38 (m, 1H, H-19), 0.93–2.10 (m, CH, CH ), 1.37 (m, 6H,
OCH CH ), 1.05 (s, 3H, CH ), 1.00 (s, 3H, CH ), 0.91 (s, 3H, CH ), 0.90 (s, 3H, CH ),
f
3
×
2
3
1
2
0
≡
≡
2
×
2
3
3
3
3
δ
3
13
.88 (s, 3H, CH ), 0.74 (m, 1H, H-5). C-NMR (150 MHz, CDCl3)
(ppm): 153.2; 152.8;
68.9; 64.4; 63.9; 63.9; 55.3; 50.2; 49.5; 46.4; 43.1; 42.7; 40.9;
8.3; 37.9; 37.7; 37.5; 37.0; 34.2; 34.1; 31.2; 29.6; 27.9; 26.9; 26.6; 23.5; 20.9; 18.1; 16.5; 2 16.2;
0.93. HR-MS (APCI) m/z: C H O P
3
1
3
48.8; 110.7; 83.6; 75.0; 74.6; 2
×
×
31
1
6.1; 14.7. P-NMR (243 MHz, CDCl3)
δ
(ppm):
−
40
58
8
−
[
(M–H) ], Calc. 697.3869; Found 697.3856.
3
,28-Di(2-butynoyl)-30-diethoxyphosphoryloxybetulin 6b
◦
Yield: 10%, m. p. 66–70 C.
1
TLC (dichloromethane:ethanol, 40:1, v/v): R = 0.73. H-NMR (600 MHz, CDCl )
δ
(ppm):
.06 (s, 1H, H-29), 4.99 (s, 1H, H-29), 4.59 (m, 1H, H-3), 4.50 (m, 2H, H-30), 4.34 (d,
J = 10.8 Hz, 1H, H-28), 4.16 (m, 4H, 2 OCH CH3), 3.93 (d, J = 10.8 Hz, 1H, H-28), 2.37
f
3
5
×
2
(m, 1H, H-19), 2.02 (s, 3H, CH C
≡
C), 2.01 (s, 3H, C
≡
CCH ), 0.93–2.10 (m, CH, CH ), 1.36
3
3 2
(
m, 6H, 2
×
OCH CH ), 1.04 (s, 3H, CH ), 0.99 (s, 3H, CH ), 0.90 (s, 3H, CH ), 0.89 (s, 3H,
2
3
3
3
3
1
3
CH ), 0.86 (s, 3H, CH ), 0.80 (m, 1H, H-5). C-NMR (150 MHz, CDCl3)
δ
(ppm): 154.3;
53.9; 149.1; 111.4; 85.7; 85.0; 82.8; 72.5; 72.5; 69.0; 63.9; 63.8; 55.4; 50.2; 46.4; 45.5; 42.7; 40.9;
8.4; 37.9; 37.5; 37.0; 34.2; 34.1; 30.6; 29.6; 27.9; 27.3; 27.0; 23.6; 20.9; 19.7; 18.1; 16.5; 2 16.2;
0.90. HR-MS (APCI) m/z:
3
3
1
3
1
×
31
6.1; 16.0; 14.7; 3.9; 3.9. P-NMR (243 MHz, CDCl3)
δ
(ppm):
−
−
C H O P [(M-H) ], Calc. 725.4182; Found 725.4162.
42
62
8
3
,28-Di(2-cyclopropylpropynoyl)-30-diethoxyphosphoryloxybetulin 6c
◦
Yield: 8%, m. p. 67–70 C.
1
TLC (dichloromethane:ethanol, 15:1, v/v): R = 0.70. H-NMR (600 MHz, CDCl )
δ
(ppm):
.05 (s, 1H, H-29), 4.98 (s, 1H, H-29), 4.58 (m, 1H, H-3), 4.50 (m, 2H, H-30), 4.31 (d,
J = 10.8 Hz, 1H, H-28), 4.14 (m, 4H, 2 OCH CH ), 3.91 (d, J = 10.8 Hz, 1H, H-28), 2.37 (m,
f
3
5
×
2
3
1
3
0
H, H-19), 0.93–2.10 (m, CH, CH ), 1.36 (m, 6H, 2
×
OCH CH ), 1.02 (s, 3H, CH ), 0.97 (s,
2
2
3
3
H, CH ), 0.95 (m, 2
×
2H, CH C
≡
CH), 0.93 (m, 2
×
2H, CH C
≡
CH), 0.88 (s, 3H, CH3),
3
2
2
1
3
.87 (s, 3H, CH ), 0.85 (s, 3H, CH ), 0.77 (m, 1H, H-5). C-NMR (150 MHz, CDCl )
δ
(ppm):
64.2; 55.9; 50.7; 50.0; 47.0;
3.2; 41.4; 38.5; 38.3; 38.0; 37.5; 34.8; 34.6; 30.1; 28.4; 27.5; 27.1; 24.1; 21.4; 18.7; 17.1; 16.7; 16.7;
3
3
3
154.9; 154.5; 149.4; 111.1; 94.0; 93.3; 83.1; 69.4; 2
×
69.0; 64.4; 2
×
4
31
1
6.5; 15.2; 2
×
9.7; 0.0. P-NMR (243 MHz, CDCl3)
δ
(ppm):
−
0.91. HR-MS (APCI) m/z:
−
C H O P [(M–H) ], Calc. 629.3971; Found 629.3965.
37
58
6
3
0-Diethoxyphosphoryloxy-3,28-di(2-phenylpropynoyl)betulin 6d
◦
Yield: 12%, m. p. 89–91 C.
1
TLC (dichloromethane:ethanol, 40:1, v/v): R = 0.77. H-NMR (600 MHz, CDCl )
δ
(ppm):
.62 (m, 4H, Harom), 7.47 (m, 2H, Harom), 7.40 (m, 4H, Harom), 5.07 (s, 1H, H-29), 5.01 (s, 1H,
H-29), 4.68 (m, 1H, H-3), 4.52 (m, 2H, H-30), 4.43 (d, J = 10.8 Hz, 1H, H-28), 4.17 (m, 4H,
f
3
7
2
×
OCH CH ), 4.02 (d, J = 10.8Hz, 1H, H-28), 2.41 (m, 1H, H-19), 0.93–2.10 (m, CH, CH ),
2
3
2
1.38 (m, 6H, 2
×
OCH CH ), 1.08 (s, 3H, CH ), 1.02 (s, 3H, CH ), 0.95 (s, 3H, CH ), 0.94 (s,
2
3
3
3
3
13
3
H, CH ), 0.90 (s, 3H, CH ), 0.84 (m, 1H, H-5). C-NMR (150 MHz, CDCl3)
δ
(ppm): 154.7;
3
3
1
8
2
54.2; 149.0; 2
0.7; 68.9; 64.1; 2
8.0; 17.3; 27.0; 26.6; 23.6; 20.9; 18.2; 16.6; 2
×
133.0; 2
×
130.5; 2
×
129.7; 2
×
128.5; 2
×
119.9; 110.7; 86.5; 85.8; 83.2; 81.1;
×
63.8; 55.4; 50.2; 46.5; 42.7; 40.9; 38.4; 38.0; 37.6; 37.1; 34.3; 34.1; 31.2; 29.7;
31
×
16.2; 16.1; 14.8. P-NMR (243 MHz, CDCl3)
6
δ
−
(ppm):
−
0.88. HR-MS (APCI) m/z: C H O P [(M–H) ], Calc. 849.4495; Found 849.4483.
0-Diethoxyphosphoryloxy-28-propynoylbetulin 7a
37
58
3
◦
Yield: 72%, m. p. 177–180 C.

Molecules 2021, 26, 737
24 of 30
1
TLC (dichloromethane:ethanol, 15:1, v/v): R = 0.51. H-NMR (600 MHz, CDCl )
δ
(ppm):
f
3
5
4
C
.07 (s, 1H, H-29), 5.00 (s, 1H, H-29), 4.51 (m, 2H, H-30), 4.40 (d, J = 10.8Hz, 1H, H-28),
.15 (m, 4H, 2 OCH CH ), 3.99 (d, J = 10.8 Hz, 1H, H-28), 3.20 (m, 1H, H-3), 2.92 (s, 1H,
CH), 2.38 (m, 1H, H-19), 0.93–2.10 (m, CH, CH ), 1.37 (m, 6H, 2
×
2
3
≡
×
OCH CH ), 1.04 (s,
2 3
2
3
1
6
2
H, CH3), 1.01 (s, 3H, CH ), 0.99 (s, 3H, CH ), 0.84 (s, 3H, CH ), 0.79 (s, 3H, CH ), 0.69 (m,
3 3 3 3
H, H-5). 13C-NMR (150 MHz, CDCl3)
4.5; 2 63.9; 55.3; 50.3; 49.5; 46.4; 43.1; 42.7; 40.9; 38.9; 38.7; 37.6; 37.1; 34.2; 33.6; 31.2; 29.6;
8.0; 27.4; 27.0; 26.7; 20.8; 18.4; 18.3; 2
δ (ppm): 153.2; 148.8; 110.7; 78.9; 74.8; 74.7; 68.9;
×
3
1
×
16.2; 16.0; 15.4; 14.8. P-NMR (243 MHz, CDCl3)
δ
−
(ppm):
−
0.94. HR-MS (APCI) m/z: C H O P [(M–H) ], Calc. 645.3920; Found 645.3918.
3
7
58
7
28-(2-Butynoyl)-30-diethoxyphosphoryloxybetulin 7b
◦
Yield: 81%, m. p. 133–136 C.
TLC (dichloromethane:ethanol, 15:1, v/v): R = 0.52. H-NMR (600 MHz, CDCl )
1
δ
(ppm):
f
3
5
4
.06 (s, 1H, H-29), 4.99 (s, 1H, H-29), 4.50 (m, 2H, H-30), 4.35 (d, J = 10.8Hz, 1H, H-28),
.15 (m, 4H, 2 OCH CH ), 3.94 (d, J = 10.8 Hz, 1H, H-28), 3.20 (m, 1H, H-3), 2.38 (m, 1H,
×
2
3
H-19), 2.01 (s, 3H, C
≡
CCH ), 0.93–2.10 (m, CH, CH ), 1.37 (m, 6H, 2
×
OCH CH ), 1.04 (s,
3
2
2 3
3
H, CH ), 0.99 (s, 3H, CH ), 0.98 (s, 3H, CH ), 0.83 (s, 3H, CH ), 0.78 (s, 3H, CH ), 0.70 (m,
3
3
3
3
3
13
1
H, H-5). C-NMR (150 MHz, CDCl3)
δ (ppm): 154.3; 148.9; 110.5; 85.5; 78.9; 72.5; 68.8;
6
2
3.8; 63.8; 55.3; 50.3; 49.5; 46.4; 43.1; 42.7; 40.9; 38.9; 38.7; 37.5; 37.1; 34.2; 31.2; 29.6; 28.0; 27.4;
7.0; 26.6; 20.8; 18.3; 2
0.91. HR-MS (APCI) m/z: C H O P [(M–H)-], Calc. 659.4077; Found 659.4079.
×
16.2; 16.1; 15.4; 14.8, 3.9. 31P-NMR (243 MHz, CDCl3)
δ
(ppm):
−
38
60
7
2
8-(2-Cyclopropylpropynoyl)-30-diethoxyphosphoryloxybetulin 7c
◦
Yield: 82%, m. p. 133–134 C.
TLC (dichloromethane:ethanol, 15:1, v/v): R = 0.54. H-NMR (600 MHz, CDCl )
1
δ
(ppm):
f
3
5
4
.05 (s, 1H, H-29), 4.95 (s, 1H, H-29), 4.50 (m, 2H, H-30), 4.32 (d, J = 10.8 Hz, 1H, H-28),
.15 (m, 4H, 2 OCH CH ), 3.92 (d, J = 10.8 Hz, 1H, H-28), 3.20 (m, 1H, H-3), 2.38 (m, 1H,
×
2
3
H-19), 0.93–2.10 (m, CH, CH ), 1.36 (m, 6H, 2
CH ), 0.98 (s, 3H, CH ), 0.95 (s, 2H, C
×
OCH CH ), 1.03 (s, 3H, CH ), 0.99 (s, 3H,
2
2
3
3
≡
CCHCH ), 0.93 (s, 2H, C
≡
CCHCH ), 0.83 (s, 3H,
3
3
2
2
1
3
CH ), 0.77 (s, 3H, CH ), 0.69 (m, 1H, H-5). C-NMR (150 MHz, CDCl )
δ
(ppm): 155.0;
64.4; 64.2; 55.8; 50.8; 50.0; 43.5; 43.2; 41.4; 39.4; 39.2;
9.0; 38.1; 37.7; 34.8; 34.7; 34.4; 31.7; 30.2; 28.5; 27.9; 27.5; 26.0; 21.4; 18.8; 2 16.7; 16.6; 15.9;
0.93. HR-MS (APCI) m/z: C H O P
3
3
3
1
3
49.4; 111.0; 94.0; 79.5; 69.3; 69.0; 2
×
×
31
1
5.3; 9.75. P-NMR (243 MHz, CDCl3)
δ
(ppm):
−
40
62
7
−
[
(M–H) ], Calc. 685.4233; Found 685.4229.
3
0-Diethoxyphosphoryloxy-28-(2-phenylpropynoyl)betulin 7d
◦
Yield: 71%, m. p. 131–134 C.
1
TLC (dichloromethane:ethanol, 40:1, v/v): R = 0.52. H-NMR (600 MHz, CDCl )
δ
(ppm):
.62 (m, 2H, Harom), 7.47 (m, 1H, Harom), 7.40 (m, 2H, Harom), 5.07 (s, 1H, H-29), 5.01 (s, 1H,
H-29), 4.51 (m, 2H, H-30), 4.42 (d, J = 10.8 Hz, 1H, H-28), 4.16 (m, 4H, 2 OCH CH ), 4.03
f
3
7
×
2
3
(d, J = 10.8Hz, 1H, H-28), 3.20 (m, 1H, H-3), 2.41 (m, 1H, H-19), 0.93–2.10 (m, CH, CH ),
2
1
3
.37 (m, 6H, 2
×
OCH CH ), 1.06 (s, 3H, CH ), 1.01 (s, 3H, CH ), 0.99 (s, 3H, CH ), 0.84 (s,
2
3
3
3
3
13
H, CH ), 0.78 (s, 3H, CH ), 0.69 (m, 1H, H-5). C-NMR (150 MHz, CDCl3)
δ
(ppm): 154.6;
3
3
1
4
1
48.9; 133, 130.6; 129.7; 128.6; 119.7; 110.6; 86.4; 80.7; 78.9; 68.8; 64.1; 2
×
63.8; 55.3; 50.3; 49.5;
6.5; 43.2; 42.7; 40.9; 38.9; 38.2; 37.6; 37.1; 34.3; 34.2; 31.62; 29.7; 28.0; 27.4; 27.0; 26.7; 20.9;
31
9.7; 18.3; 2
×
16.2; 16.1; 15.4; 14.8. P-NMR (243 MHz, CDCl3)
δ
(ppm):
−
0.91. HR-MS
−
(
APCI) m/z: C H O P [(M–H) ], Calc. 721.4233; Found 721.4226.
43
62
7
3.2.4. Synthesis of Alkynyl Derivatives 6e and 7e
A solution of compound
pyridine was cooled to about
benzene was gradually added dropwise while stirring with a magnetic stirrer. Stirring was
continued for 18 h at room temperature. The volatiles were evaporated, 6 mL of chloroform
was added to the residue, and the resulting solution was washed with 10% sulfuric acid
and then with water. After drying over anhydrous sodium sulfate, the chloroform was
5
(595 mg, 1mmol) in 4.5 mL of dry benzene and 3.75 mL of
◦
−
5 C. Then, 1.5 mmol of propargyl chloroformate in 4 mL of
removed in a vacuum evaporator. The mono- and disubstituted derivatives, 6e and 7e
,
were isolated using column chromatography (SiO , dichloromethane:ethanol 15: 1, v/v).
2

Molecules 2021, 26, 737
25 of 30
3
0-Diethoxyphosphoryloxy-3,28-dipropargyloxycarbonyloxybetulin 6e
◦
Yield: 15%, m. p. 129–131 C.
TLC (dichloromethane:ethanol, 15:1, v/v): R = 0.82. H-NMR (600 MHz, CDCl )
1
δ
(ppm):
f
3
4
4
4
.98 (s, 1H, H-29), 4.91 (s, 1H, H-29), 4.67 (m, 2H, OCH -C
≡
C), 4.65 (m, 2H, OCH -C C),
≡
2
2
.29 (d, J = 6 Hz, 2H, H-30), 4.30 (d, J = 10.8 Hz, 1H, H-28), 4.27 (m, 1H, H-3), 4.07 (m,
H, 2 OCH CH ), 3.85 (d, J = 10.8 Hz, 1H, H-28), 2.47 (t, J = 2.4 Hz, 1H, C CH), 2.45
CH), 2.28 (m, 1H, H-19), 1.0–2.10 (m, CH, CH ), 1.29 (m, 6H, 2
×
≡
2
3
(t, J = 2.4 Hz, 1H, C
≡
×
2
OCH CH ), 0.96 (s, 3H, CH ), 0.91 (s, 3H, CH ), 0.85 (s, 3H, CH ), 0.78 (s, 3H, CH ), 0.77
2
3
3
3
3
3
13
(s, 3H, CH ), 0.70 (m, 1H, H-5). C-NMR (150 MHz, CDCl3)
δ
(ppm): 155.1; 154.6; 148.8;
3
1
10.7; 86.2; 75.7; 75.5; 68.9; 66.9; 63.9; 63.8; 2
8.1; 37.5; 37.0; 34.1; 31.2; 29.5; 27.6; 26.9; 26.6; 23.6; 20.9; 18.1; 16.4; 2
×
55.3; 55.0; 53.5; 50.2; 49.5; 46.6; 42.7; 40.9; 38.3;
3
×
16.2; 16.1; 16.0; 14.8.
62 10
31
−
P-NMR (243 MHz, CDCl3)
δ
(ppm):
−
0.9. HR-MS (APCI) m/z: C H O P [(M–H) ],
42
Calc. 757.4081; Found 757.4075.
3
0-Diethoxyphosphoryloxy-28-propargyloxycarbonyloxybetulin 7e
◦
Yield: 60%, m. p. 172–174 C.
TLC (dichloromethane:ethanol, 15:1, v/v): R = 0.64. H-NMR (600 MHz, CDCl )
1
δ
(ppm):
C), 4.42
OCH CH ), 3.85
f
3
4.97 (s, 1H, H-29), 4.91 (s, 1H, H-29), 4.67 (dd, J = 1.2 Hz, J = 2.4 Hz, 2H, OCH -C
≡
2
(
(
d, J = 6 Hz, 2H, H-30), 4.30 (d, J = 10.8 Hz, 1H, H-28), 4.07 (m, 4H, 2
d, J = 10.8 Hz, 1H, H-28), 3.11 (m, 1H, H-3), 2.47 (t, J = 2.4 Hz, 1H, C
×
2
3
≡
CH), 2.47 (m, 1H,
H-19), 0.93–2.10 (m, CH, CH ), 1.28 (m, 6H, 2
CH ), 0.90 (s, 3H, CH ), 0.75 (s, 3H, CH ), 0.69 (s, 3H, CH ), 0.61 (m, 1H, H-5). C-NMR
×
OCH CH ), 0.96 (s, 3H, CH ), 0.91 (s, 3H,
2
2
3
3
1
3
3
3
3
3
(150 MHz, CDCl3) δ (ppm): 155.1; 148.9; 110.6; 78.9; 75.7; 68.9; 66.9; 63.9; 63.8; 55.3; 55.2;
5
0.3; 49.5; 46.6; 42.7; 40.9; 38.9; 38.7; 37.5; 37.1; 34.2; 34.1; 31.2; 29.5; 28.0; 27.4; 27.0; 26.7; 20.8;
31
1
8.3; 2
×
16.2; 16.1; 16.0; 15.4; 14.8. P-NMR (243 MHz, CDCl3)
δ
(ppm):
−
0.91. HR-MS
−
(
APCI) m/z: C H O P [(M–H) ], Calc. 675.4026; Found 675.4019.
38
60
8
3
.3. Raman Spectroscopy
Raman spectra were measured using the Raman confocal microscope Alfa 300R
WITec Instruments Corp., Germany) equipped with an Olympus 50 /0.50 long working
distance objective and 532 nm wavelength diode laser. Averaged spectra were accumulated
(
×
−
00 times with a 2 s integration time, at the range 200–3800 cm and with a spectral
1
1
−
1
−1
resolution about 1.5 cm (at 2000 cm ). The power of the laser beam on the sample did
not exceed 10 mW, which prevented heating of the samples. OriginPro 9.1 software was
used to analyze the spectra. The spectra were smoothed (Savitzky–Golay method, seven
points) and background subtracted. The band positions were determined by the curve
−
1
fitting (Gaussian function shape) with accuracy of about 1 cm
.
The theoretical vibrational spectra of compounds and 10 were calculated using
5
Gaussin09 software. Structure optimization and spectra calculations were performed with
DFT B3LYP hybrid functional and 6–31G(d) basis set. No imaginary frequencies were
found for the optimized structures, which means that at least local energy minima were
obtained. All calculated frequencies were scaled by a scaling factor 0.960 in order to correct
the calculated value to experimentally observable.
3.4. Crystal Structure
The methodology for X-ray analysis [59–61] and additional information about the
crystal are included in the Supplementary Materials. Crystal structure was deposited at
the Cambridge Crystallographic Data Center, with deposit number CCDC 2,039,827, and is
available free of charge via www.ccdc.cam.ac.uk/data_request/cif.
3.5. Biological Activity
3.5.1. Cell Culture and Medium
The synthesized compounds and references were tested for cytotoxic activity in vitro
against human cancer cell lines such as amelanotic melanoma (C-32, ATCC, Rockville,
MD, USA), glioblastoma (SNB-19, DSMZ, Braunschweig, Germany), and breast cancer

Molecules 2021, 26, 737
26 of 30
(T47D, and MDA-MB-231, ATCC, Rockville, MD, USA). Cell cultures were maintained
using DMEM (Lonza, Basel, Switzerland) supplemented with 10% fetal bovine serum (FBS)
(
(
Biological Industries Cromwell, CT, USA) and penicillin (10,000 U/mL) –streptomycin
10 mg/mL) mix (Lonza, Basel, Switzerland).
3.5.2. Cell Viability Studies
Cultures were performed in 96-well plates (Nunc Thermo Fisher Scientific, Waltham,
4
◦
10 /well and incubated for 24 h (at 37 C, 5% CO ,
2
MA, USA). Cells were seeded at 5
×
constant humidity). Then, the medium was replaced with a fresh one with the addition
of test compounds in the concentration: 25; 12.5; 6.25; 3.12; 1.56; 0.78; 0.39; 0 g/mL
µ
and incubated again for a further 72 h. After this time, the WST-1 test (Roche Molecular
Biochemicals, Mannheim, Germany) was performed to assess the metabolic activity of
the cells. The principle of the WST-1 colorimetric test is based on the ability of live cells
to decompose bright red tetrazole salts of WST-1 to dark red formazan. The reaction is
mediated by mitochondrial dehydrogenases and correlates with the number of cells and
their metabolic activity. The absorbance was measured at
λ = 450 nm with a UVM340
microplate reader (Biogenet, Józefów, Poland). Results were expressed as a mean value of
at least three independent experiments performed in triplicate.
3.5.3. Analysis of the Expression of H3, TP53, BCL-2, and BAX by qRT-PCR
The influence of tested compounds on the transcriptional activity of cells of the
genes H3, TP53, BCL-2, and BAX was investigated. Assays were performed by real-time
®
qRT-PCR on an Opticon TM DNA Engine (MJ Research, USA) using the QuantTect
®
SYBR Green qRT-PCR Kit (Qiagen, Valencia, Spain). Cells were exposed to compound
7
a
and 7b for 24 h at a concentration equal to half of IC50 for appropriate cell line. For
the reference compound, cisplatin, concentrations were 0.8 g/mL (SNB-19 line) and
.6 g/mL (C-32 line). RNA extraction was performed on spin columns using the Quick
µ
0
µ
RNATM MiniPrep kit (Zymo Research, Irvine, CA, USA). RNA extracts were assessed
qualitatively and quantitatively. Total RNA integrity was checked electrophoretically on
a 1.2% agarose gel with ethidium bromide. The amount and purity of total RNA in the
extracts were determined spectrophotometrically with the HP8452A spectrophotometer
(
Hewlett Packard, Waldbronn, Germany).
3.5.4. Statistical Analysis
All experiments were repeated in triplicate. Control treatment refers to untreated cells.
As a positive control, used cells were treated with cisplatin. The data were expressed as
mean standard deviation (SD). One-way ANOVA with post-hoc Tukey’s test was used to
compare the data from different groups, p < 0.05 was considered statistically significant.
±
3.6. Physicochemical Descriptors (Lipophilicity, ADME Properties)
Chromatographic procedure by RP-TLC [62] is included in the Supplementary Materi-
als. The ADME parameters for all derivatives were calculated using ACD/iLab software
Advanced Chemistry Development, Inc., Toronto, ON, Canada) [42].
(
3.7. Molecular Docking
The three-dimensional (3D) structures of studied compounds were generated in their
low-energy conformation using Gaussian 16 (revision A.03) computer code [63] at the
density functional theory (DFT, B3LYP) and 6–311 + G(d,p) basis sets. Target macromolecule
for molecular docking studies was obtained from the Protein Data Bank (https://www.
rcsb.org/). We used 3D crystal structures of EGFR (PDB ID: 1M17). Protein Preparation
Wizard of UCSF Chimera (DockPrep) software [64] was used to process and prepare the
protein. Missing chains were added, the implicit hydrogen atoms also were added to the
atoms to satisfy their appropriate valences, and the ligand of no significance present in the
protein structure was deleted. Then the structure was optimized by assigning the bond

Molecules 2021, 26, 737
27 of 30
orders, bond angles, and topology. The formal atomic charges were fixed for the amino
acid residues, and energy minimization was carried out.
Genetic Optimization for Ligand Docking (GOLD) 2020.1 [50] was used for the dock-
ing analysis. The region of interest used for GOLD docking was defined as the EGFR
protein residues within the 10 Å of the reference erlotinib (X = 22.01, Y = 0.25, Z = 52.79 Å).
For remaining water molecules, the “Toggle” option was on. Default values of all other
parameters were used, and the complexes were submitted to 100 genetic algorithm runs
using the ChemScore fitness function. After calculations, only the ten highest scored pose
was returned as a docking result for ligand–cavity configuration. All obtained results were
ranked according to their score value and presented in GOLD arbitrary units (a.u.). Cal-
culation of protein–ligand binding free energy was performed using the KDEEP predictor
based on DCNNs (https://playmolecule.org/Kdeep) [52]. Molecular docking details were
visualized using the BIOVIA Discovery Studio virtual environment [65] and Chimera [63].
Molecular dynamics simulation was performed with Nanoscale Molecular Dynamics soft-
ware ver. 2.13 (NAMD, https://www.ks.uiuc.edu/Research/namd/) [57]. All input files
were prepared using QwikMD [66] computer program based on GOLD output complexes.
Protein–ligand systems have been solvated with 0.15 mol/L NaCl water box. Then system
was minimized, annealed, and equilibrated. After that, 10 ns production simulation was
performed. The MolAICal [56] is used to calculate the MM/GBSA between ligands and
EGFR protein based on molecular dynamical (MD) simulated results by NAMD.
4
. Conclusions
In the presented study, new 30-phosphate derivatives of betulin were synthesized,
and evaluated as anticancer agents. Taking into account all the tested neoplastic cell
lines, it can be seen that in the examined group of alkynyl derivatives, monoesters 7a 7e
IC50 in range 0.76–15.40 M) showed higher activity than diesters 6a 6e (IC50 in range
.65–31.46 M). Compounds 7a and 7b were characterized by potent in vitro activity
–
(
5
µ
–
µ
against the C-32 and SNB-16 cell lines. In addition, derivative 7a characterized antitumor
activity higher than that of the reference drug cisplatin against C-32, SND-19, and T47D
cells. Comparing the activity of 28-propynoylbetulin with its 30-phosphate analogue 7a in
relation to appropriate tumor cell lines shows that the introduction of diethylphosphate
substituent to the isopropenyl moiety resulted in an improvement in activity. With respect
to T47D cells, alkynyl esters with 30-phosphate substituent (7a
–7c) showed weaker activity
than their phosphonate analogs (11 13). In relation to the SNB-16 and C-32 lines, no
–
relationship between the activity and the type of the phosphorus substituent was observed.
The most active in vitro compounds, 7a and 7b, were tested on expression of H3, TP53,
BAX, and BCL-2 genes. The results have shown that 7a and 7b can probably activate the
apoptosis process in SNB-19 and C-32 cells. Further research is necessary to determine
the exact path of cell death of the studied lines. The antitumor activity of triterpenes
against the tested cell lines was more dependent on their chemical structure and less on
lipophilicity. It was observed that the correlation between lipophilicity and antitumor
activity was significant for the monoesters 7a–7e containing a free hydroxyl group at the
C3 position. The obtained results suggest that the relationship between ADME properties
and antitumor activity is a complex issue.
Taking into account the potential mechanisms of antitumor activity, molecular docking
of the most active in vitro derivatives, 7a and 7b, to the selected receptor, EGFR, was also
carried out. In the library of designed phosphate derivatives of betulin, compounds 7a
and 7b are reported to have good binding interactions as compared to standard erlotinib.
Compound 7b was found to have the highest binding affinity and form a perfect binding
pose with EGFR, justified by KDEEP score. Furthermore, the molecular dynamic study of
the most effective ligands 7a,b has proved the stability of ligand–protein complexes with
little deviation and fluctuation from 7 ns up to 10 ns. Compounds 7a and 7b demonstrated
comparable degree of fit to the reference ligand erlotinib. Such a result may indicate a
potential molecular target with which the antitumor activity of the test compounds is related.

Molecules 2021, 26, 737
28 of 30
Supplementary Materials: Figure S1a. 1H NMR, 3,28-diacetoxy-30-diethoxyphosphoryloxybetu-
, Figure S1b. 13C NMR, 3,28-diacetoxy-30-diethoxyphosphoryloxy-betulin
4, Figure S1c.
31
P
lin
4
NMR, 3,28-diacetoxy-30-diethoxyphosphoryloxybetulin
4
, Figure S1d. HR-MS, 3,28-diacetoxy-30-
1
diethoxyphosphoryloxybetulin
4
, Figure S2a. H NMR, 30-diethoxyphosphory-loxybetulin
5, Figure
S2b. 13C NMR, 30-diethoxyphosphoryloxybetulin
5, Figure S2c. P NMR, 30-diethoxyphosphoryloxy-
31
1
betulin
5, Figure S2d. HR-MS, 30-diethoxyphosphoryloxybetulin 5, Figure S3a. H NMR, 30-
1
3
diethoxyphosphoryloxy-28-propynoylbetulin 7a, Figure S3b. C NMR, 30-diethoxyphosphoryloxy-
3
1
2
7
2
8-propynoylbetulin 7a, Figure S3c. P NMR, 30-diethoxyphosphoryloxy-28-propynoylbetulin
a
1
, Figure S3d. HR-MS, 30-diethoxyphosphoryloxy-28-propynoylbetulin 7a, Figure S4a. H NMR,
8-(2-butynoyl)-30-diethoxyphosphoryloxybetulin 7b, Figure S4b. C NMR, 28-(2-butynoyl)-30-
diethoxyphosphoryloxybetulin 7b, Figure S4c. P NMR, 28-(2-butynoyl)-30-diethoxyphosphoryloxy-
betulin 7b, Figure S4d. HR-MS, 28-(2-butynoyl)-30-diethoxyphosphoryloxy-betulin 7b, Figure S5a.
H NMR, 28-(2-cyclopropylpropynoyl)-30-diethoxyphosphoryloxybetulin 7c, Figure S5b. C NMR,
1
3
3
1
1
13
3
1
28-(2-cyclopropylpropynoyl)-30-diethoxyphosphoryloxybetulin 7c, Figure S5c. P NMR, 28-(2-
cyclopropylpropynoyl)-30-diethoxyphosphoryloxybetulin 7c, Figure S5d. HR-MS, 28-(2-cyclopropyl-
propynoyl)-30-diethoxyphosphoryloxybetulin 7c, Table S1. Calculated Raman band frequencies and
assignments for compounds
data collection and refinement details for compound
Selected hydrogen bonds in the crystal structure of compound
RP-TLC, Table S5. The theoretical values of lipophilicity for tested compounds
Table S6. The correlation matrix for theoretically obtained lipophilicity parameters of compounds
–5, 6a–6e and 7a–7e.
5
and 10, X-ray diffraction experiment, Table S2 Crystal parameters,
, Table S3 Monocrystal characteristic, Table S4
, Chromatographic procedure by
6a 6e and 7a 7e
5
5
3–
5,
–
–
,
3
Author Contributions: E.C. and S.B. developed the concept of the work. E.C. performed experiments,
analyzed spectra, and wrote the paper. M.J. and R.W. made measurements and calculations of Raman
spectra. K.M. performed and described the molecular docking study and the molecular dynamics
simulations. M.L. conducted a study of the biological activity. E.B. and M.K.-T. contributed to the
synthesis and purification of all compounds. J.K.: X-ray diffraction measurement. All authors have
read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Acknowledgments: This work was supported by the Medical University of Silesia Grants nos. PCN-
1
-004/K/0/F and KNW-1-106/K/9/B. Energy-minimization calculations have been carried out using
resources provided by Wroclaw Centre for 818 Networking and Supercomputing (https://wcss.pl),
Grant no. 382. The authors are grateful to Czestochowa University of Technology for granting access
to HPC platform provided by the MICLAB project no. POIG.02.03.00.24-093/13.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
2.
3.
4.
5.
6.
7.
World Health Organization. WHO Report on Cancer, Setting Priorities, Investing Wisely and Providing Care for All. 2020.
Available online: https://apps.who.int/iris/handle/10665/330745 (accessed on 30 January 2021).
Rastogi, S.; Pandey, M.M.; Kumar, A.; Rawat, S. Medicinal plants of the genus Betula—traditional uses and a phytochemical–
pharmacological review. J. Ethnopharmacol. 2015, 159, 62–83. [CrossRef] [PubMed]
Zhang, L.; Ji, M.-Y.; Qiu, B.; Li, Q.-Y.; Zhang, K.-Y.; Liu, J.-C.; Dang, L.-S.; Li, M.-H. Phytochemicals and biological activities of
species from the genus Maytenus. Med. Chem. Res. 2020, 29, 575–606. [CrossRef]
Tolmacheva, I.A.; Shelepen’Kina, L.N.; Vikharev, Y.B.; Anikina, L.V.; Grishko, V.V.; Tolstikov, A.G. Synthesis and biological activity
of S-containing betulin derivatives. Chem. Nat. Compd. 2005, 41, 701–705. [CrossRef]
Kazakova, O.; Medvedeva, N.; Tolstikov, G.A.; Kukovinets, O.S.; Yamansarov, E.Y.; Spirikhin, L.V.; Gubaidullin, A.T. Synthesis of
terminal acetylenes using POCl3 in pyridine as applied to natural triterpenoids. Mendeleev Commun. 2010, 20, 234–236. [CrossRef]
Sidova, V.; Zoufaly, P.; Pokorny, J.; Dzubak, P.; Hajduch, M.; Popa, I.; Urban, M. Cytotoxic conjugates of betulinic acid and
substituted triazoles prepared by Huisgen Cycloaddition from 30-azidoderivatives. PLoS ONE 2017, 12, e0171621. [CrossRef]
Chrobak, E.; B e˛ benek, E.; Kadela-Tomanek, M.; Latocha, M.; Jelsch, C.; Wenger, E.; Boryczka, S. Betulin phosphonates; synthesis,
structure, and cytotoxic activity. Molecules 2016, 21, 1123. [CrossRef]

Molecules 2021, 26, 737
29 of 30
8
9
.
.
Spivak, A.Y.; Keiser, J.; Vargas, M.; Gubaidullin, R.R.; Nedopekina, D.A.; Shakurova, E.R.; Khalitova, R.R.; Odinokov, V.N.
Synthesis and activity of new triphenylphosphonium derivatives of betulin and betulinic acid against Schistosoma mansoni
in vitro and in vivo. Bioorganic Med. Chem. 2014, 22, 6297–6304. [CrossRef]
Chrobak, E.; Marciniec, K.; Da˛browska, A.; P e˛ cak, P.; B e˛ benek, E.; Kadela-Tomanek, M.; Bak, A.; Jastrz e˛ bska, M.; Boryczka, S.
New phosphorus analogs of bevirimat: Synthesis, evaluation of anti-HIV-1 activity and molecular docking study. Int. J. Mol. Sci.
2
019, 20, 5209. [CrossRef]
1
1
1
0. Talele, T.T. Acetylene group, friend or foe in medicinal chemistry. J. Med. Chem. 2020, 63, 5625–5663. [CrossRef]
1. DrugBank. Available online: https://go.drugbank.com/drugs/ (accessed on 15 September 2020).
2. Csuk, R.; Barthel, A.; Sczepek, R.; Siewert, B.; Schwarz, S. Synthesis, encapsulation and antitumor activity of new betulin
derivatives. Arch. Pharm. Chem. Life Sci. 2011, 1, 37–49. [CrossRef]
1
1
1
3. Bi, Y.; Xu, J.; Wu, X.; Ye, W.; Yuan, S.; Zhang, L. Synthesis and cytotoxic activity of 17-carboxylic acid modified 23-hydroxy
betulinic acid ester derivatives. Bioorganic Med. Chem. Lett. 2007, 17, 1475–1478. [CrossRef] [PubMed]
4. Yamansarov, E.Y.; Saltykova, I.V.; Kovalev, S.V.; Petrov, R.A.; Shkil’, D.O.; Seleznev, E.I.; Beloglazkina, E.K.; Majouga, A.G.
Synthesis and cytotoxicity of new alkyne derivatives of pentacyclic triterpenoids. Russ. Chem. Bull. 2019, 68, 855–861. [CrossRef]
5. Marciniec, K.; Pawełczak, B.; Latocha, M.; Skrzypek, L.; Macia˛ z˙ ek-Jurczyk, M.; Boryczka, S. Synthesis, anti-breast cancer activity,
and molecular docking study of a new group of acetylenic quinolinesulfonamide derivatives. Molecules 2017, 22, 300. [CrossRef]
[PubMed]
1
1
1
6. Marciniec, K.; Latocha, M.; Boryczka, S.; Kurczab, R. Synthesis, molecular docking study, and evaluation of the antiproliferative
action of a new group of propargylthio- and propargylselenoquinolines. Med. Chem. Res. 2014, 23, 3468–3477. [CrossRef]
7. Boryczka, S.; B e˛ benek, E.; Wietrzyk, J.; Kempi n´ ska, K.; Jastrz e˛ bska, M.; Kusz, J.; Nowak, M. Synthesis, structure and cytotoxic
activity of new acetylenic derivatives of betulin. Molecules 2013, 18, 4526–4543. [CrossRef] [PubMed]
8. Kadela-Tomanek, M.; Jastrz e˛ bska, M.; Pawełczak, B.; B e˛ benek, E.; Chrobak, E.; Latocha, M.; Ksia˛ z˙ ek, M.; Kusz, J.; Boryczka, S.
Alkynyloxy derivatives of 5,8-quinolinedione: Synthesis, in vitro cytotoxicity studies and computational molecular modeling
with NAD(P)H: Quinone oxidoreductase 1. Eur. J. Med. Chem. 2017, 126, 969–982. [CrossRef] [PubMed]
9. Chrobak, E.; Kadela-Tomanek, M.; B e˛ benek, E.; Marciniec, K.; Wietrzyk, J.; Trynda, J.; Pawełczak, B.; Kusz, J.; Kasperczyk, J.;
Chodurek, E.; et al. New phosphate derivatives of betulin as anticancer agents: Synthesis, crystal structure, and molecular
docking study. Bioorganic Chem. 2019, 87, 613–628. [CrossRef]
1
2
0. Evers, M.; Poujade, C.; Soler, F.; Ribeill, Y.; James, C.; Lelièvre, Y.; Gueguen, J.-C.; Reisdorf, D.; Morize, I.; Pauwels, R.; et al.
Betulinic acid derivatives: A new class of human immunodeficiency virus type 1 specific inhibitors with a new mode of action. J.
Med. Chem. 1996, 39, 1056–1068. [CrossRef]
2
2
2
2
1. Lan, P.; Wang, J.; Zhang, D.-M.; Shu, C.; Cao, H.-H.; Sunab, P.; Wu, X.-M.; Ye, W.-C.; Chen, W.-M. Synthesis and antiproliferative
evaluation of 23-hydroxybetulinic acid derivatives. Eur. J. Med. Chem. 2011, 46, 2490–2502. [CrossRef]
2. Wuts, P.G.M.; Greene, T.W. Greene’s Protective Groups in Organic Synthesis; John Wiley & Sons: New Jersey, NJ, USA, 2007;
pp. 230–232.
3. Neises, B.; Steglich, W. Esterification of carboxylic acids with dicyclohexylcarbodiimide/4-Dimethylaminopyridine:tert-butyl
ethyl fumarate. Org. Synth. 1985, 63, 183. [CrossRef]
4. Lin-Vien, D.; Fateley, W.G.; Colthup, N.B.; Grasselli, J.G. The Handbook of Infrared and Raman Characteristic Frequencies of Organic
Molecules; Academic Press: Cambridge, MA, USA, 1991.
2
2
5. Socrates, G. Infrared and Raman Characteristic Group Frequencies, 3rd ed.; John Wiley & Sons, LTD.: Hoboken, NJ, USA, 2001.
6. Csuk, R.; Barthel, A.; Schwarz, S.; Kommera, H.; Paschke, R. Synthesis and biological evaluation of antitumor-active γ-
butyrolactone substituted betulin derivatives. Bioorganic Med. Chem. 2010, 18, 2549–2558. [CrossRef] [PubMed]
2
2
7. B e˛ benek, E.; Jastrz e˛ bska, M.; Kadela-Tomanek, M.; Chrobak, E.; Orzechowska, B.; Zwoli n´ ska, K.; Latocha, M.; Mertas, A.; Czuba,
Z.; Boryczka, S. Novel triazole hybrids of betulin: Synthesis and biological activity profile. Molecules 2017, 22, 1876. [CrossRef]
[
PubMed]
8. O’Boyle, N.M.; Greene, L.M.; Keely, N.O.; Wang, S.; Cotter, T.S.; Zisterer, D.M.; Meegan, M.J. Synthesis and biochemical activities
of antiproliferative amino acid and phosphate derivatives of microtubule-disrupting -lactam combretastatins. Eur. J. Med. Chem.
013, 62, 705–721. [CrossRef] [PubMed]
β
2
2
9. Hua, H.; Zhang, H.; Qingbin Kong, Q.; Wang, J.; Jiang, Y. Complex roles of the old drug aspirin in cancer chemoprevention and
therapy. Med. Res. Rev. 2019, 39, 114–145. [CrossRef] [PubMed]
3
0. Pisha, E.; Chai, H.; Lee, I.-S.; Chagwedera, T.E.; Farnsworth, N.R.; Cordell, G.A.; Beecher, C.W.; Fong, H.H.; Kinghorn, A.D.;
Brown, D.M.; et al. Discovery of betulinic acid as a selective inhibitor of human melanoma that functions by induction of
apoptosis. Nat. Med. 1995, 1, 1046–1051. [CrossRef] [PubMed]
3
1. Fulda, S.; Jeremias, I.; Steiner, H.H.; Pietsch, T.; Debatin, K.-M. Betulinic acid: A new cytotoxic agent against malignant brain-tumor
cells. Int. J. Cancer 1999, 82, 435–441. [CrossRef]
3
2. Prokhorova, E.A.; Zamaraev, A.V.; Kopeina, G.S.; Zhivotovsky, B.; Lavrik, I.N. Role of the nucleus in apoptosis: Signaling and
execution. Cell. Mol. Life Sci. 2015, 72, 4593–4612. [CrossRef]
3
3. Hemann, M.T.; Lowe, S.W. The p53–Bcl-2 connection. Cell Death Differ. 2006, 13, 1256–1259. [CrossRef]
3
4. Polo, S.E.; Almouzni, G. Histone metabolic pathways and chromatin assembly factors as proliferation markers. Cancer Lett. 2005
20, 1–9. [CrossRef]
,
2

Molecules 2021, 26, 737
30 of 30
3
3
5. Orchel, J.; Slowinski, J.; Mazurek, U.; Wilczok, T. H3 mRNA level as a new proliferative marker in astrocytomas. Biochim Biophys
Acta 2004, 1689, 42–46. [CrossRef]
6. Roos, W.P.; Kaina, B. DNA damage-induced apoptosis: From specific DNA lesions to the DNA damage response and apoptosis.
Cancer Lett. 2012, 332, 237–248. [CrossRef] [PubMed]
3
3
3
7. Yip, K.W.; Reed, J.C. Bcl-2 family proteins and cancer. Oncogene 2008, 27, 6398–6406. [CrossRef] [PubMed]
8. Pfeffer, C.M.; Singh, A.T.K. Apoptosis: A target for anticancer therapy. Int. J. Mol. Sci. 2018, 19, 448. [CrossRef] [PubMed]
9. Inglot, T.; Gumieniczek, A.; Komsta, Ł.; Kasi n´ ska, A. LC and principal component analysis in the lipophilicity study of seven
angiotensin II-AT1 receptor antagonists (sartans). Chromatographia 2008, 68, 977–983. [CrossRef]
0. Tetko, I.V.; Gasteiger, J.; Todeschini, R.; Mauri, A.; Livingstone, D.; Ertl, P.; Palyulin, V.A.; Radchenko, E.V.; Zefirov, N.S.;
Makarenko, A.S.; et al. Virtual computational chemistry laboratory—design and description. J. Comput. Aided Mol. Des. 2005, 19,
4
4
53–463. [CrossRef]
4
4
4
4
1. Keller, T.H.; Pichota, A.; Yin, Z. A practical view of ‘druggability’. Curr. Opin. Chem. Biol. 2006, 10, 357–361. [CrossRef] [PubMed]
2. ACD Labs. Available online: https://www.acdlabs.com (accessed on 21 August 2019).
3. Schyman, P.; Liu, R.; Desai, V.; Wallqvist, A. vNN web server for ADMET predictions. Front. Pharmacol. 2017, 8, 889. [CrossRef]
4. Dincela, E.D.; Gürsoya, E.; Yilmaz-Ozdenb, T.; Ulusoy-Güzeldemirci, N. Antioxidant activity of novel imidazo[2,1-b]thiazole
derivatives: Design, synthesis, biological evaluation, molecular docking study and in silico ADME prediction. Bioorganic Chem.
2
020, 103, 104220. [CrossRef]
4
5. Desai, V.; Bhushan, A. Natural bioactive compounds: Alternative approach to the treatment of glioblastoma multiforme Hindawi.
BioMed Res. Int. 2017, 2017, 1–10. [CrossRef]
4
4
6. Hanahan, D.; Weinberg, R.A. Hallmarks of cancer: The next generation. Cell 2011, 144, 646–674. [CrossRef]
7. Louis, D.N.; Ohgaki, H.; Wiestler, O.D.; Cavenee, W.K.; Burger, P.C.; Jouvet, A.; Scheithauer, B.W.; Kleihues, P. The 2007 WHO
classification of tumours of the central nervous system. Acta Neuropathol. 2007, 114, 97–109. [CrossRef] [PubMed]
8. Hanahan, D.; Weinberg, R.A. The hallmarks of cancer. Cell 2000, 100, 57–70. [CrossRef]
4
4
9. Chowdhury, F.A.; Hossain, K.; Mostofa, A.; Akbor, M.M.; Bin Sayeed, M.S. Therapeutic potential of thymoquinone in glioblastoma
treatment: Targeting major gliomagenesis signaling pathways. BioMed Res. Int. 2018, 2018, 1–15. [CrossRef] [PubMed]
0. Cole, J.C.; Nissink, J.W.M.; Taylor, R. Protein-ligand docking and virtual screening with GOLD in virtual screening in drug
discovery. In Virtual Screening in Drug Discovery; Shoichet, B., Alvarez, J., Eds.; CRC Press: Boca Raton, FL, USA, 2005.
1. Jones, G.; Willett, P.; Glen, R.C.; Leach, A.R.; Taylor, R. Development and validation of a genetic algorithm for flexible docking. J.
Mol. Biol. 1997, 267, 727–748. [CrossRef]
5
5
5
5
2. Jiménez, J.; Škali cˇ , M.; Martínez-Rosell, G.; De Fabritiis, G. KDEEP: Protein–ligand absolute binding affinity prediction via
3
D-convolutional neural networks. J. Chem. Inf. Model. 2018, 58, 287–296. [CrossRef] [PubMed]
3. Abdelatef, S.A.; El-Saadi, M.T.; Amin, N.H.; Abdelazeem, A.H.; Omar, H.A.; Abdellatif, K.R. Design, synthesis and anticancer
evaluation of novel spirobenzo[h]chromene and spirochromane derivatives with dual EGFR and B-RAF inhibitory activities. Eur.
J. Med. Chem. 2018, 150, 567–578. [CrossRef]
5
5
4. Beveridge, D.L.; DiCapua, F.M. Free Energy Via Molecular Simulation: Applications to Chemical and Biomolecular Systems.
Annu. Rev. Biophys. Biophys. Chem. 1989, 18, 431–492. [CrossRef]
5. Hou, T.; Wang, J.; Li, Y.; Wang, W. Assessing the performance of the molecular mechanics/Poisson Boltzmann surface area and
molecular mechanics/generalized Born surface area methods. II. The accuracy of ranking poses generated from docking. J.
Comput. Chem. 2011, 32, 866–877. [CrossRef]
5
5
5
6. Bai, Q.; Tan, S.; Xu, T.; Liu, H.; Huang, J.; Yao, X. MolAICal: A soft tool for 3D drug design of protein targets by artificial
intelligence and classical algorithm. Briefings Bioinform. 2020. [CrossRef]
7. Phillips, J.C.; Hardy, D.J.; Maia, J.D.C.; Stone, J.E.; Ribeiro, J.V.; Bernardi, R.C.; Buch, R.; Fiorin, G.; Hénin, J.; Jiang, W.; et al.
Scalable molecular dynamics on CPU and GPU architectures with NAMD. J. Chem. Phys. 2020, 153, 044130. [CrossRef]
8. Chrobak, E.; B e˛ benek, E.; Marciniec, K.; Kadela-Tomanek, M.; Siudak, S.; Latocha, M.; Boryczka, S. New 30-substituted derivatives
of pentacyclic triterpenes: Preparation, biological activity, and molecular docking study. J. Mol. Struct. 2021, 1226, 129394. [CrossRef]
9. Sheldrick, G. A short history of SHELX. Acta Crystallogr. Sect. A Found. Crystallogr. 2008, 64, 112–122. [CrossRef] [PubMed]
0. CrysAlis PRO; Agilent Technologies Ltd: Yarnton, UK, 2014.
5
6
6
6
1. Sheldrick, G.M. Crystal structure refinement with SHELXL. Acta Cryst. 2015, 71, 3–8. [CrossRef]
2. B e˛ benek, E.; Bober-Majnusz, K.; Siudak, S.; Chrobak, E.; Kadela-Tomanek, M.; Wietrzyk, J.; Boryczka, S. Application of TLC to
Evaluate the Lipophilicity of Newly Synthesized Betulin Derivatives. J. Chromatogr. Sci. 2020, 58, 323–333. [CrossRef] [PubMed]
3. Frisch, M.J.; Trucks, G.W.; Schlegel, H.B.; Scuseria, G.E.; Robb, M.A.; Cheeseman, J.R.; Scalmani, G.; Barone, V.; Petersson, G.A.;
Nakatsuji, H.; et al. Gaussian 16; Revision A. 03; Gaussian Inc.: Wallingford, CT, USA, 2016.
6
6
4. Pettersen, E.F.; Goddard, T.D.; Huang, C.C.; Couch, G.S.; Greenblatt, D.M.; Meng, E.C.; Ferrin, T.E. UCSF Chimera—A visualiza-
tion system for exploratory research and analysis. J. Comput. Chem. 2004, 25, 1605–1612. [CrossRef] [PubMed]
5. Dassault Systèmes. BIOVIA Discovery Studio 2020; Dassault Systèmes: San Diego, CA, USA, 2019.
6. Ribeiro, J.V.; Bernardi, R.C.; Rudack, T.; Stone, J.E.; Phillips, J.C.; Freddolino, P.L.; Schulten, K. QwikMD—integrative molecular
dynamics toolkit for novices and experts. Sci. Rep. 2016, 6, 26536. [CrossRef] [PubMed]
6
6