Synthesis and anticancer cytotoxicity with structural context of an α-hydroxyphosphonate based compound library derived from substituted benzaldehydes

Article abstract of DOI:10.1039/c9nj02144b

We synthesized substituted benzaldehyde derived α-hydroxyphosphonates (αOHP), α-hydroxyphosphonic acids (αOHPA) and α-phosphinoyloxyphosphonates (αOPP) and characterized their cytotoxicity against a panel of cancer cell lines. A library containing 56 analogues was screened against Mes-Sa parental and Mes-Sa/Dx5 multidrug resistant uterine sarcoma cell lines, using a fluorescence-based cytotoxicity assay. The cytotoxicity screening revealed that dibenzyl-αOHPs and dimethyl-α-diphenyl-OPPs were the most active clusters, which encouraged us to synthesize further dibenzyl-α-diphenyl-OPP derivatives that elicited pronounced cell killing. Further structure-activity relationships showed the relevance of hydrophobicity and the position of substituents on the main benzene ring as determinants of toxicity. The most active analogs proved to be equally, or even more toxic to the multidrug resistant (MDR) cell line Mes-Sa/Dx5, suggesting these compounds may overcome P-glycoprotein mediated multidrug resistance by evading the drug transporter.

Full text of DOI:10.1039/c9nj02144b

View Article Online
View Journal
NJC
New Journal of Chemistry
Accepted Manuscript
A journal for new directions in chemistry
This article can be cited before page numbers have been issued, to do this please use: Z. Rádai, T. Windt,
V. Nagy, A. Füredi, N. Z. Z. Kiss, I. Ranelovi, J. Tóvári, G. Keglevich, G. Szakács and S. Tóth, New J. Chem.,
2019, DOI: 10.1039/C9NJ02144B.
Volume 42
This is an Accepted Manuscript, which has been through the
Number
6
21 March 2018
Pages 3963-4776
Royal Society of Chemistry peer review process and has been
accepted for publication.
NJC
New Journal of Chemistry
rsc.li/njc
A journal for new directions in chemistry
Accepted Manuscripts are published online shortly after acceptance,
before technical editing, formatting and proof reading. Using this free
service, authors can make their results available to the community, in
citable form, before we publish the edited article. We will replace this
Accepted Manuscript with the edited and formatted Advance Article as
soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes to the
text and/or graphics, which may alter content. The journal’s standard
Terms & Conditions and the Ethical guidelines still apply. In no event
shall the Royal Society of Chemistry be held responsible for any errors
or omissions in this Accepted Manuscript or any consequences arising
from the use of any information it contains.
ISSN 1144-2546
PAPER
Chechia Hu, Tzu-Jen Lin et al.
Yellowish and blue luminescent graphene oxide quantum dots
prepared via a microwave-assisted hydrothermal route using
H2O2 and KMnO4 as oxidizing agents
rsc.li/njc

Page 1 of 14
New Journal of Chemistry
1
2
3
4
5
View Article Online
DOI: 10.1039/C9NJ02144B
Synthesis and anticancer cytotoxicity with structural context of an α-hydroxyphosphonate
based compound library derived from substituted benzaldehydes
6
7
8
Zita Rádaia, Tímea Windtb, Veronika Nagyb, András Füredib,c, Nóra Zsuzsa Kissa, Ivan Ranđelovićd,e, József Tóvárid,e
,
György Keglevicha, Gergely Szakácsb,c, Szilárd Tóthb
9
a
Department of Organic Chemistry and Technology, Budapest University of Technology and
Economics, Budapest, Hungary.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
b Institute of Enzymology, Research Centre for Natural Sciences, Hungarian Academy of Sciences,
Budapest,
Hungary.
Correspondence
to:
toth.szilard.enzim@ttk.mta.hu
and
szakacs.gergely@ttk.mta.hu
c Institute of Cancer Research, Medical University Vienna, Vienna, Austria.
d National Institute of Oncology, Department of Experimental Pharmacology, Budapest, Hungary
e Kineto Lab ltd., Budapest, Hungary
ABSTRACT
We synthesized substituted benzaldehyde derived α-hydroxyphosphonates (αOHP), α-
hydroxyphosphonic acids (αOHPA) and α-phosphinoyloxyphosphonates (αOPP) and characterized
their cytotoxicity against a panel of cancer cell lines. A library containing 56 analogues was screened
against Mes-Sa parental and Mes-Sa/Dx5 multidrug resistant uterine sarcoma cell lines, using a
fluorescence-based cytotoxicity assay. The cytotoxicity screening revealed that dibenzyl-αOHPs and
dimethyl-α-diphenyl-OPPs were the most active clusters, which encouraged us to synthesize further
dibenzyl-α-diphenyl-OPP derivatives that elicited pronounced cell killing. Further structure-activity
relationships showed the relevance of hydrophobicity and the position of substituents on the main
benzene ring as determinants of toxicity. The most active analogs proved to be equally, or even more
toxic to the multidrug resistant (MDR) cell line Mes-Sa/Dx5, suggesting these compounds may
overcome P-glycoprotein mediated multidrug resistance by evading the drug transporter.
Footnotes
Electronic supplementary information (ESI) available: δP and MS value of the analogues; primary growth
inhibition values; reaction time and yield of products.
1. Introduction
Organophosphonate derivatives, such as α-hydroxyphosphonates (αOHP), α-hydroxyphosphonic acids
(αOHPA) and α-phosphinoyloxyphosphonates (αOPP) are a class of compounds of chemical and
biological relevance. In synthetic chemistry, αOHPs are used in the synthesis of α-aminophosphonates,
α-alkoxy-, or α-acyloxyphosphonates, ketophosphonates or α-halophosphonates.1 In addition, several
αOHPs, αOHPAs and αOPPs were found to be biologically active. For example, organic phosphonates
were reported to possess moderate viral cysteine protease inhibitory,2 antimicrobial or antifungal
activities.3-5 The αOPP SR-202 inhibits human peroxisome proliferator-activated receptor-γ (PPARγ),
influencing insulin sensitivity and glucose consumption of cells.6, 7 Studies have shown that organic
phosphonates may influence cancer progression through the inhibition of farnesyl protein transferase
(FPT)8, which has emerged as a novel target for anti-cancer agents due to its role in the
posttranslational modification of the Ras oncogene,9 or purine nucleoside phosphorylase (PNP), as PNP
inhibitors are potentially applicable in the management of certain hematologic malignancies.10, 11
1

New Journal of Chemistry
Page 2 of 14
1
2
3
4
5
View Article Online
DOI: 10.1039/C9NJ02144B
However, to date, the direct anticancer activity of αOHP, αOHPA and αOPP analogues have not been
systematically tested.12
In a recent study we identified cytotoxic αOHP analogues using the uterine sarcoma Mes-Sa cells.13
Here our aim was to synthesize and characterize the cytotoxic profile of a diverse compound set, and
to draw basic structure activity relationships against a panel of cell lines, including Mes-Sa/Dx5, the
multidrug resistant (MDR) derivative of the Mes-Sa cell line.
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2. Results and discussion
2.1. Synthesis and compilation of the chemical library
The derivatives were synthesized through the Pudovik reaction and subsequent modifications.
Considering green chemical aspects, the reactions were performed either without solvent or in a
minimal quantity of acetone, and the pure products were crystallized after reflux and adding a small
amount of n-pentane14 (Scheme 1, Scheme 2 and Scheme 3).
R1
H
R2
Me
Y1
-
Cl
-
-
Y2
-
-
Cl
-
-
Y3
-
-
-
Cl
Yield (%)
1a
1b
1c
1d
95
93
93
90
1e
NO2
-
-
95
-
1f
1g
NO2
-
95
92
-
-
Me
-
1h
1i
1j
Me
-
-
89
68
98
-
O-Me
-
-
O-Me
1l
2a
2b
-
-
-
F
-
-
96
78
80
Et
-
Cl
-
-
2c
2d
Cl
-
-
98
79
-
Cl
2e
2f
NO2
-
-
Cl
-
-
-
89
88
95
93
88
95
91
99
88
94
96
48
82
56
38
89
72
65
-
-
-
Cl
-
NO2
-
-
-
Cl
-
NO2
-
Me
-
3a
3b
3c
3d
3e
3f
3g
3h
3i
Bn
-
-
NO2
-
-
-
Me
-
-
-
-
-
-
Br
-
-
O-Me
-
-
F
-
-
4a Me Me
-
4f
4k
4l
4m
4n
NO2
-
F
-
-
Br
NO2
5f
Et
-
Scheme 1. General scheme of the synthesis and the synthesized series of αOHPs (1, 2, 3, 4 and 5). Dash
signs in the table under Y1 – Y3 refer to hydrogen; Bn: benzyl.
2

Page 3 of 14
New Journal of Chemistry
1
2
3
4
5
6
7
8
9
View Article Online
DOI: 10.1039/C9NJ02144B
Y1
Y2
-
-
Cl
-
Y3
-
-
-
Cl
Yield (%)
6a
6b
6c
6d
-
Cl
-
80
85
76
88
-
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
-
6g
Me
-
-
77
-
6h
Me
-
90
72
-
-
6i O-Me
6o
-
NH2
50
Scheme 2. General scheme of the synthesis and the synthesized αOHPAs (6). Dash signs in the table
under Y1 – Y3 refer to hydrogen; Bn: benzyl.
R2
Y1
-
-
Y2
Y3
Y4
Y5 Yield (%)
7a Me
-
-
-
-
Cl
57
61
7d
Ph
Ph
7e
NO2
-
-
79
-
7f
7h
8a
8d
NO2
Me
-
70
49
59
54
-
-
-
-
-
-
-
-
Cl
8f
8h
9a
9d
NO2
Me
-
72
46
59
51
-
-
-
-
-
-
-
-
Cl
9f
9h
NO2
Me
-
80
50
81
57
-
-
-
-
-
Cl
10a Bn
10c
Ph
Ph
-
Scheme 3. General scheme of the synthesis and the synthesized series of αOPPs (7, 8, 9 and 10). Dash
signs in the table under Y1 – Y3 refer to hydrogen; Bn: benzyl; Ph: phenyl.
2.2. Primary screening of the phosphonate library
A library containing 56 derivatives of αOHPs, αOHPAs and αOPPs was screened against the Mes-Sa
parental and Mes-Sa/Dx5 multidrug resistant uterine sarcoma cell lines, both engineered previously to
express the mCherry fluorescent protein (Mes-Sa mCh and Mes-Sa/Dx5 mCh).15 Primary cytotoxicity
was defined as at least 50% growth inhibition compared to the untreated cells, and was determined
3

New Journal of Chemistry
Page 4 of 14
1
2
3
4
5
View Article Online
DOI: 10.1039/C9NJ02144B
based on the fluorescence intensity of mCherry. At 20 μM, none of the analogues were effective, while
at 200 μM, 11 entities (3a-c, 3e-i, 7a, 7d-e, 7h) were toxic to both Mes-Sa mCh and Mes-Sa/Dx5 mCh,
and 3 additional compounds (3d, 4f, 7f) were selectively toxic to Mes-Sa/Dx5 mCh (Table S2).
6
7
8
2.3. Confirmation of dose-dependent cytotoxicity
9
Based on their primary cytotoxic effect, dibenzyl-αOHPs (3a-i) and dimethyl-α-diphenyl-OPPs (7a, 7d-
f, 7h) were chosen for further investigations. As these two classes of analogues can be chemically
‘combined’, we designed and synthesized dibenzyl-α-diphenyl-OPP compounds (10a, 10c) to see if we
can produce analogues with increased cytotoxic potency.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
We acquired dose-response curves and determined cytotoxicity as IC50 values against Mes-Sa mCh and
Mes-Sa/Dx5 mCh cells. Against Mes-Sa mCh, 3a-i exerted moderate toxicity (IC50 values ranging from
83 to 105 μM), except for the para-chloro (3d) and para-methyl (3h) substituted entities, which were
remarkably less toxic with IC50 values exceeding 250 μM. Interestingly, multidrug resistant Mes-Sa/Dx5
mCh cells were more sensitive to these compounds (3a-i), with IC50 values in the range of 34 - 78 μM,
and 126 μM for 3h. Compounds 7a, 7d-f and 7h were also only moderately toxic against the Mes-Sa
mCh cell line (IC50 values between 98 - 270 μM) and against Mes-Sa/Dx5 mCh (35 - 221 μM).
Remarkably, the newly synthesized analogues 10a and 10c showed increased toxicity with IC50 values
around 10 μM against both Mes-Sa mCh and Mes-Sa/Dx5 mCh cells (Fig. 1A).
Since most compounds exerted significantly greater toxicity against the Mes-Sa/Dx5 mCh cell line than
against Mes-Sa mCh, and since Mes-Sa/Dx5 mCh cells are multidrug resistant due to the
overexpression of P-glycoprotein (P-gp), we checked if the paradoxical selective toxicity of the
analogues were linked to the function of P-gp, as in the case of the so-called MDR-selective
compounds.15-18 However, the presence of the P-gp inhibitor tariquidar did not influence the selective
toxicity of the compounds (Fig. 1B), suggesting that the observed collateral sensitivity was linked to
other, cell line-specific factor(s) acquired by Mes-Sa/Dx5 cells during doxorubicin selection. On the
other hand, no compounds were found to be effluxed by the transporter, thus the tested compounds
can overcome this clinically important form of multidrug resistance. 19-21
To test the broader potency of the analogues, we selected 3 further cell lines originating from tumor
types known to show poor response in the clinics. We investigated the cytotoxicity against HT-29
rectosigmoid adenocarcinoma and HOP-62 lung adenocarcinoma cell lines, and against the MALME-
3M metastatic melanoma cell line transfected with eGFP to allow detection of cell survival in a
fluorescent protein-based assay. 22 In addition to 9 analogues of dibenzyl-α-OHPs (3a-i) we tested 7a,
10a and 10c. We found that 10a and 10c were much more toxic to all 3 cell lines, than the other
analogues (Table 1).
As previously was suggested, organic phosphonates may have an effect on the function of RAS proteins
by inhibiting FTP which farnesylates RAS.8 To test whether the compounds we synthesized are showing
mutant KRAS-specific toxicity, we tested 4 cytotoxic compounds (3c, 7a, 10a and 10c) on a parental
(H838) and on a mutant KRAS expressing (H838-G12D) lung adenocarcinoma cell line. However, no
specific hypersensitivity or resistance to the tested molecules was observed against the H838-G12D
cells (Table 1). These results could be explained by several ways: (1) the used αOHPs do not inhibit FTP,
(2) growth and survival of H838 cells do not rely on KRAS activity or (3) FTP inhibition is not sufficient,
because geranylgeranylation might activate KRAS and suppresses the effect of FTP inhibition.23
Nevertheless, the selected 4 compounds showed toxicity against both cell lines suggesting a more
general effectivity toward cancer cells.
4

Page 5 of 14
New Journal of Chemistry
1
2
3
4
5
6
7
View Article Online
DOI: 10.1039/C9NJ02144B
Table 1. IC50 values of compounds 3a-i, 7a, 10a and 10c against HT-29 and HOP-62 cell lines, measured
by PrestoBlue viability reagent, against Malme-3M eGFP cell line measured by the fluorescent protein-
based assay, and against H838 and H838-G12D cell lines measured by MTT assay. Numbers represent
IC50 values and standard deviations (sd) calculated from the individual pIC50 values.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
MALME-3M
HT-29
HOP-62
eGFP
H838
H838-G12D
IC50 ± sd
IC50 ± sd
IC50 ± sd
IC50 ± sd
IC50 ± sd
+
-
+
-
+
-
74.1
58.9
25.7
23.2
37.1
28.1
3a
3b
3c
288.6
110.4
110.5
>>500
116.1
129.7
117.6
>>500
154.7
337.1
27.6
236.8
112.8
89.5
115.3
82.0
79.4
151.5
76.5
89.5
84.4
215.4
88.3
214.2
30.7
13.4
+
-
+
-
+
-
8.8
8.1
3.7
3.6
11.6
10.1
+
-
+
-
+
-
+
-
+
-
7.3
6.9
1.9
1.8
3.7
3.5
3.9
3.7
5.2
5.0
116.6
109.5
+
-
+
-
55.4
42.0
50.4
37.8
3d
3e
3f
173.4
97.8
+
-
+
-
+
-
16.7
14.6
18.5
15.6
8.4
7.6
+
-
+
-
+
-
16.5
14.6
5.3
5.0
10.9
9.7
104.9
95.3
+
-
+
-
+
-
10.3
9.4
7.5
7.0
12.7
11.1
3g
3h
3i
+
-
79.8
58.2
>>500
110.9
301.8
26.5
+
-
+
-
+
-
9.6
9.0
11.8
10.6
3.7
3.5
+
-
+
-
+
-
+
-
+
-
40.4
45.8
27.7
30.4
26.4
30.1
0.5
0.5
1.3
1.3
7a
10a
10c
226.0
18.6
14.4
180.1
14.4
13.7
+
-
+
-
+
-
+
-
+
-
1.1
1.1
2.3
2.1
2.7
2.4
0.6
0.5
0.1
0.1
+
-
+
-
+
-
+
-
+
-
4.4
3.7
5.3
3.9
0.1
0.1
0.3
0.3
0.6
0.6
24.5
15.1
2.4. Toxicity against the human non-cancerous cell line HFF
Cancer specific toxicity of representatives of our compound library (3c, 7a, 10a and 10c) was probed
against the human foreskin fibroblast (HFF) cell line. Based on the IC50 values (Table 2), we observed
selectivity over tumor cells (note that HOP-62 and HT-29 cell lines were probed with the same viability
reagent as HFF). The more pronounced vulnerability of cancer cells was not obvious, since in a study
reporting the cytotoxicity of ß-formyl-α-hydroxyphosphonate derivatives, 2 out of 3 test compounds
were more toxic to HFF as to the ID8 ovarian cancer cells.24
5

New Journal of Chemistry
Page 6 of 14
1
2
3
4
View Article Online
DOI: 10.1039/C9NJ02144B
5
6
7
8
Table 2. IC50 values of compounds 3c, 7a, 10a and 10c against HFF, measured by PrestoBlue viability
reagent. Numbers represent IC50 values and standard deviations (sd) calculated from the individual
pIC50 values.
9
HFF
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
IC50 ± sd
+ 23.4
- 25.9
3c
7a
242.2
>>400
61.2
+
-
1.6
1.6
3.4
10a
10c
+
32.8
-
3.7
2.5. Investigation of cell death
As apoptotic cells can be quickly cleared by macrophages, while cell debris derived from necrosis can
cause inappropriate inflammation, apoptosis is the preferred cell death mechanism, when drug
candidates are tested.25, 26 To elucidate the mechanism of cytotoxicity of our analogues, we performed
Annexin binding assays27. As shown in Fig. 2, treatment with either 3c or 10c induced apoptosis, and
the proportion of late apoptotic/necrotic cells was increased remarkably after 48 h treatment.
2.6. Relationship of primary toxicity and lipophilicity
Growth inhibition at 200 μM (Table S2) shows correlation with the lipophilicity of the compounds (Fig.
3). With the exception of 4f and 5f (which had an approx. 50% inhibition against Mes-Sa/Dx5 mCh),
compounds below a logP (partition coefficient) of 3 were not toxic. With a logP between 4-6,
compounds were likely to be active. Furthermore, the most toxic analogues 10a and 10c have a logP
of 9.3 and 8.8, respectively. Based on drug-like filters such as Lipinksi’s rule of 5 and the Ghose filter,
logP of drugs are preferred to be under 5 or between -0.4 and +5.6, respectively.28, 29 One possibility
of further drug development of 10a and 10c is to increase their hydrophilicity by introducing certain
substituents, or/and design formulations that ensure their adequate bioavailability. As an example,
Navitoclax, an anticancer Bcl-2 inhibitor that has a logP of 8.06 can be administered orally in a lipid
solution.30 Similarly, formulations of venetoclax31 (logP: 6.76) and bexarothene32 (logP: 6.94), which
are also in use for cancer, play a key role.
2.7. Toxicity patterns
In general, αOHPAs (6), such as dimethyl- and diethyl-αOHPs (1, 2, 4, 5) were not toxic, although an
extra α-methyl-moiety improved toxicity (based on the activity of 4f and 5f against Mes-Sa/Dx5 mCh).
However, when dibenzyl-αOHPs were investigated (3), toxicity increased remarkably. This increase in
activity was observed also when dimethyl- and dibenzyl- αOPPs (7 and 10) were compared. We found
relevant features linked to the position of the substituents on the main benzene ring. When the
substituent was chlorine, the toxicity increased in the direction of para < ortho ≤ meta positions
6

Page 7 of 14
New Journal of Chemistry
1
2
3
4
5
6
7
8
9
View Article Online
DOI: 10.1039/C9NJ02144B
throughout the cell panel. In the case of the NO2 group, the ortho-position was also beneficial as
compared to the para-position, especially in the case of Mes-Sa/Dx5 mCh. Moreover, the meta-
position was much more beneficial than the para-position when methyl substitutions were present.
This pattern indicates that substituents in the meta-position were preferred among 3a-i (Table 3). Our
results provide the first detailed insights of the structure-anticancer activity relationship of the said
organophosphonates, as earlier studies on salicylaldehide derived αOHPs12 or on ß-formyl-αOHPs24
identified only a few compounds with cytotoxic potential, without relevant SAR observations.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 3. Relative toxicity of 3a-i. IC50 values were normalized to the IC50 of the unsubstituted benzene
ring containing compound (3a). Substituents on the benzene ring: o (ortho) refers to Y1; m (meta)
refers to Y2 and p (para) to Y3 (Scheme 1). Mean rel. IC50 is the average of the individual relative IC50
values.
Mes-
Sa/Dx5
mCh
substituent
(Y1-Y2-Y3)
Mes-Sa
mCh
Malme-3M
eGFP
HT-29
HOP-62
Mean rel.
IC50
3a
3b
3c
3d
H
1
1
1
1
1
1
o-Cl
m-Cl
p-Cl
0.88
0.89
2.86
0.78
0.54
1.02
0.69
0.66
1.31
0.37
0.37
1.69
0.47
0.38
0.76
0.64
0.57
1.53
3e
o-NO2
0.90
0.45
0.64
0.40
0.42
0.56
3f
3g
3h
3i
p-NO2
m-Me
p-Me
1.02
1.11
2.87
0.89
0.87
0.99
1.70
0.97
0.75
0.71
1.88
0.74
0.44
0.40
1.69
0.52
0.44
0.40
2.10
0.47
0.70
0.72
2.05
0.72
o-O-Me
3. Conclusions
By synthesizing and testing a library of αOHPs, αOHPAs and αOPPs, we identified potent anticancer
agents inducing apoptosis in several cell lines of different origin. The relation between growth
inhibition and hydrophobicity (logP) was revealed and quantified, which can be exploited as an in silico
pre-screening step in future studies. Based on the IC50 values, we observed the increased toxicity of
dibenzyl-αOHPs (3) and dibenzyl-αOPPs (10) compared to the other analogues, and found that the
most beneficial position for a substituent on the main benzene ring was the meta-position. These
results are so far the most detailed SAR observations pertaining to αOHPs, αOHPAs and αOPPs. We
also noticed the collateral sensitivity of the multidrug resistant Mes-Sa/Dx5 mCh cell line against most
of the tested analogues. Although this hypersensitivity was independent from the function of P-
glycoprotein, the results show that the tested analogues can overcome multidrug resistance by
evading the transporter.
4. Experimental
General procedure for the synthesis of α-hydroxyphosphonates 1, 2 and 3
A mixture of 11.0 mmol of aromatic aldehyde (benzaldehyde: 1.2 g, 2-chlorobenzaldehyde: 1.5 g, 3-
chlorobenzaldehyde: 1.5 g, 4-chlorobenzaldehyde: 1.5 g, 2-nitrobenzaldehyde: 1.7 g, 4-
nitrobenzaldehyde: 1.7 g, 3-methylbenzaldehyde: 1.3 g, 4-methylbenzaldehyde: 1.3 g, 2-
methoxybenzaldehyde: 1.5 g, 3-methoxybenzaldehyde: 1.5 g, 4-fluorobenzaldehyde: 1.4 g ), 11.0
mmol of dialkyl phosphite (dimethyl phosphite: 1.1 mL, diethyl phosphite: 1.4 mL and dibenzyl
phosphite: 2.4 mL) and 1.10 mmol (150 μl) of triethylamine was stirred in 1 mL of acetone at reflux.
7

New Journal of Chemistry
Page 8 of 14
1
2
3
4
5
View Article Online
DOI: 10.1039/C9NJ02144B
After 5–390 min, 6 mL of pentane was added to the reaction mixture. On cooling, the product
crystallized from the reaction mixture. Filtration afforded products 1, 2 and 3 in a purity of > 99%.
6
7
General procedure for the synthesis of α-hydroxyphosphonates 4 and 5
8
9
A mixture of 11.0 mmol of aromatic ketone (acetophenone: 1.3 g, 2-fluoroacetophenone: 1.5 g, 4-
fluoroacetophenone: 1.5 g, 3-bromoacetophenone: 2.2 g, 4-bromoacetophenone: 2.2 g, 4-
nitroacetophenone: 1.8 g), 11.0 mmol of dialkyl phosphite (dimethyl phosphite: 1.1 mL or diethyl
phosphite: 1.4 mL) and 11.0 mmol (1.5 mL) of triethylamine was stirred at 25 °C for 2–7 h. Completion
of the reaction was indicated by the crystallization of the product from the reaction mixture. The
reaction mixture was cooled to 5 °C. After the crystallization was complete, the white crystals were
filtered off and were washed with 2 mL of hexane to afford 4a, 4f, 4k, 4l, 4m, 4n and 5f in yields of 38–
89%.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
General procedure for the synthesis of α-hydroxyphosphonic acids 6
4.1 mmol of α-hydroxyphosphonate (3a: 1.5 g, 3b: 1.7 g, 3c: 1.7 g, 3d: 1.7 g, 3f: 1.7 g, 3g: 1.6 g, 3h: 1.6
g, 3i: 1.6 g) was hydrogenated in the presence of 10% Pd/C (Selcat Q) (0.08–0.09 g, the
catalyst/substrate ratio was 0.05 g/g) in 30 mL of methanol in a 80-mL stainless steel autoclave
equipped with a magnetic stirrer (stirring speed = 1100 rpm). The hydrogenations took place at 10 bar
and 25 °C in 5–150 minutes. Then, the catalyst was filtered off, and activated carbon (0.15–0.17 g) was
added to the solution. After 1 h of stirring, the absorbent was filtered off, and the organic solvent was
evaporated. 5 mL of CH2Cl2 was added to the crude product and stirred for 15 min at reflux. After
filtration, α-hydroxyphosphonic acids were obtained in yields of 50–90%.
General procedure for the synthesis of α-phosphinoyloxyphosphonates 7, 8, 9, and 10
A mixture of 1.0 mmol of -hydroxyphosphonate (1a: 0.22 g, 1d: 0.25 g, 1e: 0.26 g, 1f: 0.26 g , 1h: 0.23
g, 3a: 0.37 g, 3c: 0.40 g), 5 mL of toluene, 1.2 mmol (0.17 mL) of triethylamine and 1.1 mmol of
phosphinic chloride (diphenylphosphinic chloride: 0.21 mL, 1-chloro-3-methyl-3-phospholene-1-oxide:
0.17 g or 1-chloro-3,4-dimethyl-3-phospholene-1-oxide: 0.18 g) was stirred at 25 °C for 24-48 h under
N2 atmosphere. The precipitated triethylamine hydrochloride salt was filtered off, and the volatile
compounds were removed under vacuum. The purification of the crude product with column
chromatography on silica gel, using acetone:dichloromethane = 2:1 as the eluent afforded the
corresponding α-phospinoyloxyphosphonates (7, 8, 9 and 10).
The identity of the analogues that were known from the literature were validated by the δP and MS
values (Table S1).
7e. Yield: 79% 31P NMR (CDCl3) δ 17.0 (d, 3J = 26.9, P(O)(OCH3)2), 34.6 (d, 3J = 27.0, P(O)Ph2); 13C NMR
(CDCl3) δ 53.8 (d, 2J = 6.7, OCH3), 54.3 (d, 2J = 7.0, OCH3), 66.1 (dd, 1J = 170.5, 2J = 5.9, PCH), 124.9 (d, 4J
= 2.1, C3), 128.4 (d, 3J = 13.5, C3’), 128.6 (d, 3J = 13.4, C3’), 129.2 (d, 5J = 2.7, C4), 133.4 (d, 4J = 2.8, C5),
130.0 (d, 3J = 4.9, C6) overlapped by 130.0 (C1), 131.5 (d, 1J = 146.8, C1’), 131.6 (d, 2J = 10.5, C2’), 131.90
(d, 1J = 130.1, C1’), 131.93 (d, 2J = 10.5, C2’), 132.5 (d, 4J = 2.9, C4’), 132.8 (d, 4J = 3.1, C4’), 147.0 (d, 3J =
5.3, C2); 1H NMR (CDCl3) δ 3.62 (d, 3JP,H = 10.8, 3H, OCH3), 3.75 (d, 3JP,H = 10.8, 3H, OCH3), 6.80 (dd, 2,3JP,H
= 11.1, 15.5, 1H, PCH), 7.31-7.74 (m, 10H, Ar), 7.88-8.03 (m, 4H, Ar).
10a. Yield: 81% 31P NMR (CDCl3) δ 18.1 (d, 3J = 28.0, P(O)(OBn)2), 35.0 (d, 3J = 28.0, P(O)Ph2); 13C NMR
(CDCl3) δ 68.5 (d, 2J = 6.9, CH2), 68.6 (d, 2J = 7.2, CH2), 72.2 (dd, 1J = 173.1, 2J = 7.1, PCH), 127.2 (d, 3J =
6.0, C2), 127.8-128.6 (m, C4, C3’, C2”, C3”, C4”), 128.9 (d, 4J = 2.8, C3), 130.89 (d, 1J = 138.2, C1’), 130.97 (d,
1J = 134.0, C1’), 131.7 (d, 2J = 10.5, C2’), 131.8 (d, 2J = 10.6, C2’), 132.1 (d, 4J = 3.0, C4’), 132.4 (d, 4J = 2.9,
8

Page 9 of 14
New Journal of Chemistry
1
2
3
4
5
2
3
3
3
1
View Article Online
DOI: 10.1039/C9NJ02144B
C4’), 133.4 (t, J = 1.3, J = 1.3, C1), 135.8 (d, J = 6.0, C1”), 136.0 (d, J = 5.8, C1”); H NMR (CDCl3) δ 4.67-
5.10 (m, 4H, 2xCH2), 5.75 (dd, 2,3JP,H = 10.8, 13.1, 1H, PCH), 7.03-7.90 (m, 25H, Ar).
10c. Yield: 57% 31P NMR (CDCl3) δ 17.6 (d, 3J = 28.0, P(O)(OBn)2), 35.5 (d, 3J = 28.2, P(O)Ph2); 13C NMR
(CDCl3) δ 68.6 (d, 2J = 6,8, CH2), 68.8 (d, 2J = 7.3, CH2), 71.4 (dd, 1J = 173.2, 2J = 6.5, PCH), 126.5 (d, 3J =
5.9, C6), 127.9-128.7 (m, C3’, C2”, C3”, C4”) overlapped by 128.4 (d, 3J = 6.1, C2) 129.0 (d, 5J = 2.7, C4), 129.5
(d, 4J = 1.9, C5), 130.5 (d, 1J = 137.8, C1’), 130.7 (d, 1J = 134.5, C1’), 131.7 (d, 2J = 10.6, C2’), 131.8 (d, 2J =
10.6, C2’), 132.3 (d, 4J = 3.0, C4’), 132.5 (d, 4J = 2.9, C4’), 134.1 (d, 4J = 2.3, C3), 135.4 (t, 2J = 1.4, 3J = 1.4,
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
C1), 135.6 (d, 3J = 6.1, C1”), 135.8 (d, 3J = 5.5, C1”); 1H NMR (CDCl3) δ 4.76-5.05 (m, 4H, 2xCH2), 6.80 (dd,
2,3
J
P,H
= 10.3, 11.1, 1H, PCH), 7.07-7.89 (m, 24H, Ar).
4.1. Cell lines and culture conditions
To assess anticancer cytotoxicity, we used the HOP-62 lung adenocarcinoma, HT-29 rectosigmoid
adenocarcinoma, Malme-3M melanoma and Mes-Sa and the multidrug resistant Mes-Sa/Dx5 uterine
sarcoma cell lines, and in addition H838 and KRAS mutant H838-G12D lung adenocarcinoma cell lines.
The uterine sarcoma lines and the human foreskin fibroblast (HFF) cells were obtained from ATCC, and
were maintained in DMEM. H838 cells were from Horizon Discovery Group plc., while the other lines
were purchased from the NCI DTP, and were cultivated in RPMI. Media were supplemented with 10 %
FBS, 5 mmol/L glutamine, and 50 unit/mL penicillin and streptomycin (ThermoFisher), except for HFF,
where we used a 20 % FBS containing medium.
4.2. Fluorescent protein transfection.
Mes-Sa and Mes-Sa/Dx5 cell lines were transfected with the fluorescent protein mCherry (mCh), and
were already utilized in other studies.13, 15, 22 Establishment of the Malme-M3 eGFP was not described
earlier, but it was created with the same method. Briefly, the Malme-M3 cell line was transduced with
the fluorescent protein expressing lentiviral supernatants produced with pRRL-EF1-eGFP expression
plasmid. After the transduction, cell lines were sorted by flow cytometry based on fluorescent
intensity.
4.3. Cytotoxicity assays
Cells were seeded on 384 well plates 1 day prior to drug addition at a 2500 cells/well density in every
scenario but the H838 lines, where 5000 cells were seeded on 96 well plates. In the primary screening,
we assessed the cytotoxicity against Mes-Sa mCh and Mes-Sa/Dx5 mCh cell lines in 20 μM and in
200 μM concentrations. Drug treatment took 96 h, then the fluorescent intensity of mCherry (ex/em:
585/610 nm) was detected with an EnSpire multimode plate reader (Perkin Elmer). In the following
step, cells were treated with the serial dilution of the drugs showing activity in the primary screen, and
IC50 values were acquired after 144 h incubation time. Dose-response curves for 3a-i, 7a, 10a and 10c
were obtained against 3 additional cell lines to prove general anticancer toxicity of the selected
compounds. Cytotoxicity against Malme-M3 eGFP was assessed via the detection of eGFP fluorescence
(ex/em: 485/510 nm) after 144 h incubation, while the IC50 values against HOP-62 and HT-29 were
calculated from the detection of PrestoBlue viability reagent (Thermo Fisher) conversion at ex/em of
555/585 nm wavelengths, 72 h post drug addition. Against compounds 3c, 7a, 10a and 10c, the viability
of H838 lines were determined based on MTT assay, and based on PrestoBlue assay against HFF.
4.4. Cell death assessment
Annexin V/propidium iodide (PI) based apoptosis quantification was performed by using the Annexin
V, FITC Apoptosis Detection Kit (Dojindo Molecular Technologies) according to the protocol provided
by the manufacturer. Briefly, 1.5x105 HOP-62 cells were seeded on 12 well plates, treated with the
9

New Journal of Chemistry
Page 10 of 14
1
2
3
4
5
6
7
View Article Online
DOI: 10.1039/C9NJ02144B
given drug at IC50 concentrations for 24 and 48 hours. Cells in the supernatant were collected and
admixed with the cells that were detached with trypsin. Cells were stained for 15 minutes at room
temperature in dark with 5 μl Annexin V and 5μl PI solution in a 10-fold diluted Annexin V binding
buffer. Samples were analyzed by Attune NxT Flow Cytometer (Thermo Fisher Scientific).
8
9
Conflict of interest
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
There are no conflicts to declare.
Acknowledgements
The transfection of Malme-M3 cell line with the pRRL-EF1-eGFP expression plasmid was carried out by
Nóra Kucsma (Institute of Enzymology, RCNS, HAS, Budapest, Hungary).
This work was supported by a Momentum grant of the Hungarian Academy of Sciences. Support of
grant BME FIKP-BIO by Ministry of Human Capacities of Hungary (EMMI) is kindly acknowledged. The
project was sponsored by the National Research Development and Innovation Fund (K119202). Z.
Rádai was supported by the ÚNKP-18-3-IV-BME-265 New National Excellence Program of the Ministry
of Human Capacities, and N.Z. Kiss was supported by the János Bolyai Research Scholarship of the
Hungarian Academy of Sciences (BO/00130/19/7). J. Tóvári and I. Ranđelović were supported by the
National Research, Development and Innovation Office, NVKP_16-1-2016-0020.
References
1.
2.
Z. Rádai and G. Keglevich, Molecules, 2018, 23, 1493.
A. M. Prior, Y. Kim, S. Weerasekara, M. Moroze, K. R. Alliston, R. A. Uy, W. C. Groutas, K. O.
Chang and D. H. Hua, Bioorg Med Chem Lett, 2013, 23, 6317-6320.
A. H. Kategaonkar, R. U. Pokalwar, S. S. Sonar, V. U. Gawali, B. B. Shingate and M. S. Shingare,
Eur J Med Chem, 2010, 45, 1128-1132.
N. S. Patil, G. B. Deshmukh, S. V. Patil, A. D. Bholay and N. D. Gaikwad, Eur J Med Chem, 2014,
83, 490-497.
3.
4.
5.
6.
R. U. Pokalwar, R. V. Hangarge, P. V. Maske and M. S. Shingare, Arkivoc, 2006, 11, 196-204.
F. M. da Silva, J. C. dos Santos, J. L. Campos, A. C. Mafud, I. Polikarpov, A. C. Figueira and A. S.
Nascimento, Bioorg Med Chem Lett, 2013, 23, 5795-5802.
H. S. Wu, J. H. Yu, Y. Y. Li, Y. S. Yang, Q. J. He, Y. J. Lou and R. Y. Ji, Acta Pharmacol Sin, 2007,
28, 417-422.
7.
8.
D. L. Pompliano, E. Rands, M. D. Schaber, S. D. Mosser, N. J. Anthony and J. B. Gibbs,
Biochemistry, 1992, 31, 3800-3807.
9.
S. Bagchi, P. Rathee, V. Jayaprakash and S. Banerjee, Mini Rev Med Chem, 2018, 18, 1611-
1623.
10.
11.
12.
13.
A. Al-Kali, V. Gandhi, M. Ayoubi, M. Keating and F. Ravandi, Future Oncol, 2010, 6, 1211-
1217.
T. Yokomatsu, H. Abe, M. Sato, K. Suemune, T. Kihara, S. Soeda, H. Shimeno and S. Shibuya,
Bioorg Med Chem, 1998, 6, 2495-2505.
R. M. N. Kalla, H. R. Lee, J. Cao, J.-W. Yoo and I. Kim, New Journal of Chemistry, 2015, 39,
3916-3922.
Z. Rádai, P. Szeles, N. Z. Kiss, L. Hegedűs, T. Windt, V. Nagy and G. Keglevich, Heteroatom
Chemistry, 2018, 29, e21436.
14.
15.
G. Keglevich, Z. Rádai and N. Z. Kiss, Green Process Synth, 2017, 6, 197-201.
V. F. S. Pape, S. Tóth, A. Füredi, K. Szebényi, A. Lovrics, P. Szabó, M. Wiese and G. Szakács, Eur
J Med Chem, 2016, 117, 335-354.
10

Page 11 of 14
New Journal of Chemistry
1
2
3
4
5
6
7
8
9
View Article Online
DOI: 10.1039/C9NJ02144B
16.
G. Szakacs, J. P. Annereau, S. Lababidi, U. Shankavaram, A. Arciello, K. J. Bussey, W. Reinhold,
Y. Guo, G. D. Kruh, M. Reimers, J. N. Weinstein and M. M. Gottesman, Cancer Cell, 2004, 6,
129-137.
17.
18.
19.
20.
21.
22.
23.
24.
D. Turk, M. D. Hall, B. F. Chu, J. A. Ludwig, H. M. Fales, M. M. Gottesman and G. Szakacs,
Cancer Res, 2009, 69, 8293-8301.
A. Furedi, S. Toth, K. Szebenyi, V. F. Pape, D. Turk, N. Kucsma, L. Cervenak, J. Tovari and G.
Szakacs, Mol Cancer Ther, 2017, 16, 45-56.
A. Furedi, K. Szebenyi, S. Toth, M. Cserepes, L. Hamori, V. Nagy, E. Karai, P. Vajdovich, T. Imre,
P. Szabo, D. Szuts, J. Tovari and G. Szakacs, J Control Release, 2017, 261, 287-296.
G. Szakacs, J. K. Paterson, J. A. Ludwig, C. Booth-Genthe and M. M. Gottesman, Nat Rev Drug
Discov, 2006, 5, 219-234.
L. J. Goldstein, H. Galski, A. Fojo, M. Willingham, S. L. Lai, A. Gazdar, R. Pirker, A. Green, W.
Crist, G. M. Brodeur and et al., J Natl Cancer Inst, 1989, 81, 116-124.
T. Windt, S. Tóth, I. Patik, J. Sessler, N. Kucsma, Á. Szepesi, B. Zdrazil, C. Özvegy-Laczka and G.
Szakács, Archives of Toxicology, 2019, DOI: 10.1007/s00204-019-02417-6.
D. B. Whyte, P. Kirschmeier, T. N. Hockenberry, I. Nunez-Oliva, L. James, J. J. Catino, W. R.
Bishop and J. K. Pai, J Biol Chem, 1997, 272, 14459-14464.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
S. Perera, V. K. Naganaboina, L. Wang, B. Zhang, Q. Guo, L. Rout and C.-G. Zhao, 2011, 353,
1729-1734.
25.
26.
27.
S. Gordon and A. Pluddemann, Front Immunol, 2018, 9, 127.
S. Nagata, Annu Rev Immunol, 2018, 36, 489-517.
S. J. Martin, C. P. Reutelingsperger, A. J. McGahon, J. A. Rader, R. C. van Schie, D. M. LaFace
and D. R. Green, J Exp Med, 1995, 182, 1545-1556.
C. A. Lipinski, F. Lombardo, B. W. Dominy and P. J. Feeney, Adv Drug Deliv Rev, 2001, 46, 3-
28.
29.
30.
31.
26.
A. K. Ghose, V. N. Viswanadhan and J. J. Wendoloski, Journal of Combinatorial Chemistry,
1999, 1, 55-68.
J. Yang, H. Xiong, A. Krivoshik, K. Holen, R. Pradhan, Y.-L. Chiu, A. Nada, A. Tolcher, W. Awni
and R. Humerickhouse, 2011, 71, 1300-1300.
A. E. Place, K. Goldsmith, J. P. Bourquin, M. L. Loh, L. Gore, D. A. Morgenstern, Y. Sanzgiri, D.
Hoffman, Y. Zhou, J. A. Ross, B. Prine, M. Shebley, M. McNamee, T. Farazi, S. Y. Kim, M.
Verdugo, L. Lash-Fleming, C. M. Zwaan and J. Vormoor, Future Oncol, 2018, 14, 2115-2129.
H. M. Prince, C. McCormack, G. Ryan, C. Baker, H. Rotstein, J. Davison and R. Yocum,
Australas J Dermatol, 2001, 42, 91-97.
32.
11

New Journal of Chemistry
Page 12 of 14
1
2
3
4
5
6
7
8
9
View Article Online
DOI: 10.1039/C9NJ02144B
Mes-Sa mCh
Mes-Sa/Dx5 mCh
A)
350
300
250
200
150
100
50
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0
3a 3b 3c 3d 3e 3f 3g 3h 3i 7a 7d 7e 7f 7h 10a 10c
Mes-Sa mCh
Mes-Sa mCh (TQ)
Mes-Sa/Dx5 mCh
Mes-Sa/Dx5 mCh (TQ)
B)
350
300
250
200
150
100
50
0
3h
7d
7a
3b
3c
3e
Figure 1. (A) IC50 values of compounds 3, 7 and 10 against Mes-Sa mCh (red) and Mes-Sa/Dx5 mCh
(blue) cell lines. Results of compounds 3 against Mes-Sa mCh were taken from our previous study.13
(B) Cytotoxicity and selectivity ratio (SR = IC50 Mes-Sa mCh/IC50 Mes-Sa/Dx5 mCh) of compounds 3b-c,
3e, 3h, 7a and 7d. IC50 values, standard deviations and significance (*, P < 0.05; **, P < 0.01) were
calculated from the individual pIC50s. TQ refers to the P-glycoprotein inhibitor tariquidar (0.4 μM).
12

Page 13 of 14
New Journal of Chemistry
1
2
3
4
5
6
7
8
9
View Article Online
DOI: 10.1039/C9NJ02144B
Q1 Q2 Q3 Q4
100
80
60
40
20
0
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
control 3c
3c
10c
10c
24 h 48 h 24 h 48 h
Figure 2. Induction of apoptosis of HOP-62 cells by 3c or 10c. The four quadrants indicate viable cells
(Q1); only Annexin V positive cells (apoptosis, Q2); Annexin V and propidium-iodide positive cells (late
apoptosis/necrosis, Q3); and only propidium-iodide positive cells (Q4).
120
100
80
Mes-Sa mCh
60
40
20
0
-20
-40
-5
0
5
10
logP
120
100
80
Mes-Sa/Dx5 mCh
60
40
20
0
-20
-40
-5
0
5
10
logP
Figure 3. Growth inhibition of compounds at 200 μM after 96 h drug incubation as the function of logP.
Compound groups from Scheme 1, 2 and 3 are presented with different colors, 1: light grey, 2: yellow;
3: purple; 4: green; 5: black; 6: orange; 7: brown; 8: dark grey; 9: blue; 10: red. logP values were
calculated by the Instant JChem software (ChemAxon Ltd.)
13

New Journal of Chemistry
Page 14 of 14
View Article Online
DOI: 10.1039/C9NJ02144B
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
60x15mm (300 x 300 DPI)