Cytoprotective activities of kinetin purine isosteres

Article abstract of DOI:10.1016/j.bmc.2021.115993

Kinetin (N⁶-furfuryladenine), a plant growth substance of the cytokinin family, has been shown to modulate aging and various age-related conditions in animal models. Here we report the synthesis of kinetin isosteres with the purine ring replace

Full text of DOI:10.1016/j.bmc.2021.115993

Bioorg. Med. Chem. 33 (2021) 115993
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Cytoprotective activities of kinetin purine isosteres
a
a
b
a
,
g
b
Barbara Makov a´ , V a´ clav Mik , Barbora Li ˇs kov a´ , Gabriel Gonzalez , Dominik Vítek ,
b
d
c
a
Martina Medvedíkov a´ , Beata Monfort , Veronika Ru ˇc ilov a´ , Alena Kadlecov ´a ,
e
a
f
h
Prashant Khirsariya , Zoila G a´ ndara Barreiro , Libor Havlí ˇc ek , Marek Zatloukal ,
c
e
d
b
a
Miroslav Soural , Kamil Paruch , Benoit D’Autr ´e aux , Mari a´ n Hajdúch , Miroslav Strnad ,
Ji ˇr í Vollerb
,i,*
a
ˇ
Department of Experimental Biology, Faculty of Science, Palacký University, Slechtitelů 27, Olomouc CZ-78371, Czech Republic
Institute of Molecular and Translational Medicine, Faculty of Medicine and Dentistry, Palacký University, Hn ˇe votínsk a´ 3, Olomouc CZ-77515, Czech Republic
Department of Organic Chemistry, Faculty of Science, Palacký University, 17. listopadu 1192/12, Olomouc CZ-783-71, Czech Republic
Universit ´e Paris-Saclay, CEA, CNRS, Institute for Integrative Biology of the Cell (I2BC), 91198 Gif-sur-Yvette, France
b
c
d
e
f
Department of Chemistry, CZ Openscreen, Faculty of Science, Masaryk University, Brno, Czech Republic
Isotope Laboratory, The Czech Academy of Science, Institute of Experimental Botany, Víde nˇ sk a´ 1083, Praha 4 CZ-14220, Czech Republic
g
h
Department of Neurology, Palacký University Olomouc, Faculty of Medicine and Dentistry and University Hospital, Olomouc, Czech Republic
ˇ
Department of Chemical Biolology and Genetics, Centre of the Region Hana for Biotechnological and Agricultural Research, Palacký University, Slechtitelů 27, Olomouc
CZ-78371, Czech Republic
i
ˇ
Laboratory of Growth Regulators, Palacký University & Institute of Experimental Botany AS CR, Slechtitelů 27, Olomouc CZ-78371, Czech Republic
A R T I C L E I N F O
A B S T R A C T
6
Keywords:
Cytokinin
Kinetin
Kinetin (N -furfuryladenine), a plant growth substance of the cytokinin family, has been shown to modulate
aging and various age-related conditions in animal models. Here we report the synthesis of kinetin isosteres with
the purine ring replaced by other bicyclic heterocycles, and the biological evaluation of their activity in several in
́
vitro models related to neurodegenerative diseases. Our findings indicate that kinetin isosteres protect Friedreichs
bioisostery – Friedreichs ataxia
́
Mitoprotection – familial dysautonomia
ataxia patient-derived fibroblasts against glutathione depletion, protect neuron-like SH-SY5Y cells from
glutamate-induced oxidative damage, and correct aberrant splicing of the ELP1 gene in fibroblasts derived from a
familial dysautonomia patient. Although the mechanism of action of kinetin derivatives remains unclear, our
data suggest that the cytoprotective activity of some purine isosteres is mediated by their ability to reduce
oxidative stress. Further, the studies of permeation across artificial membrane and model gut and blood-brain
barriers indicate that the compounds are orally available and can reach central nervous system. Overall, our
data demonstrate that isosteric replacement of the kinetin purine scaffold is a fruitful strategy for improving
known biological activities of kinetin and discovering novel therapeutic opportunities.
Neuroprotection
1
. Introduction
in other organisms, including humans.1 In plants, they occur as free
cytokinin bases (the active forms) as well as ribosides and glucosides
6
Cytokinins, N -substituted adenines in chemical terms, are major
regulators of plant growth and development but they also have activities
(transport and storage forms). The first studies of their pharmacological
activity in humans focused on the anti-cancer activity of certain
Abbreviations: ADME, Absorption, distribution, metabolism and excretion; ATCC, American Type Culture Collection,; ATRA, All-trans retinoic acid; BSA, Bovine
serum albumin; BSO, L-buthionine sulfoximine; DCM, Dichloromethane; DFO, Deferoxamine; DHE, Dihydroethidium; DIC, N,N’-Diisopropylcarbodiimide; DMF, N,N-
Dimethylformamide; DMSO, Dimethyl sulfoxide; DMEM, Dulbecco’s modified Eagle’s medium; ECACC, European Collection of Authenticated Cell Culture; EDIPA, N-
Ethyldiisopropylamine; GSH, Reduced glutathione; FA, Friedreich’s ataxia; FBS, Fetal bovine serum; FXN, Frataxin; ISC, Iron-sulfur cluster; HBBS, Hank’s balanced
buffer solution; HOBt, 1-hydroxybenzotriazole; MW, Microwave; PAMPA, Parallel Artificial Membrane Permeability Assay; PBS, Phosphate buffer saline; PI, Pro-
pidium iodide; RED, Rapid equilibrium dialysis; R-LA, R-lipoic acid; TEA, Triethylamine; TFA, Trifluoroacetic acid,.
ˇ
*
Corresponding author at: LGR proper and Department of Experimental Biology, Palacký University & Institute of Experimental Botany AS CR, Slechtitelů 27,
Olomouc CZ-78371, Czech Republic.
E-mail address: jiri.voller@upol.cz (J. Voller).
https://doi.org/10.1016/j.bmc.2021.115993
Received 2 December 2020; Accepted 31 December 2020
Available online 6 January 2021
0
968-0896/© 2021 Published by Elsevier Ltd.

B. Makov a´ et al.
Bioorganic & Medicinal Chemistry 33 (2021) 115993
cytokinin ribosides. Later research focused on cytokinin bases, for which
diverse activities (including neuroprotective, antipsoriatic, and anti-
aggregatory activities) have been reported, as reviewed by Kadlecova´
Table 1
Overview of the tested compounds.
No.
C2
C6 substituent
Isostere core
2
et al. in 2019. In contrast to some cytokinin ribosides, cytokinin bases
Subst.
show no or limited toxicity toward human cells. The most intensively
1
2
3
-H
-Cl
-H
furfurylamino-
furfurylamino-
furfurylamino-
purine
6
studied cytokinin base is kinetin (N -furfuryladenine), in large degree
purine
3
because it delays the aging of human cells in vitro and may naturally
3H-imidazo[4,5-b]pyridine, 1-
deazapurine
occur in humans.4 An important impetus for kinetin research was the
4
5
6
7
8
9
-Cl
-H
-Cl
-H
-Cl
-H
-Cl
-H
furfurylamino-
furfurylamino-
furfurylamino-
furfurylamino-
furfurylamino-
furfurylamino-
furfurylamino-
furfurylamino-
3H-imidazo[4,5-b]pyridine, 1-
deazapurine
discovery of its ability to correct the aberrant ELP1 and NF1 pre-mRNA
splicing responsible for the neurodegenerative disorders familial dys-
autonomia and neurofibromatosis type I, respectively.5 Follow-up
studies demonstrated that kinetin can increase the abundance of ELP1
wild-type transcripts in transgenic mouse familial dysautonomia
models. The first human studies demonstrated that kinetin is orally
available.7 Recently, promising in vitro results have been reported for
combination of kinetin and phoshatidylserine.8 Other investigations
1H-imidazo[4,5-c]pyridine, 3-
deazapurine
,6
1H-imidazo[4,5-c]pyridine, 3-
deazapurine
7H-pyrrolo[2,3-d]pyrimidine, 7-
deazapurine
7H-pyrrolo[2,3-d]pyrimidine 7-
deazapurine
9
5H-pyrrolo[3,2-d]pyrimidine, 9-
deazapurine
showed that kinetin has protective activity in models of Parkinson’s
1
0
and Huntington’s diseases.
1
0
5H-pyrrolo[3,2-d]pyrimidine, 9-
deazapurine
In plants, cytokinin signaling is well characterized, from signal
recognition by plasma membrane receptors (e.g., AHK2, AHK3 and
11
12
3H-[1,2,3]triazolo[4,5-d]
pyrimidine, 8-azapurine
pyrazolo[1,5-a]pyrimidine
pyrazolo[4,3-d]pyrimidine
1-methyl-1H-pyrazolo[3,4-d]
pyrimidine
AHK4 in the model plant Arabidopsis thaliana) to activation of the
response genes.11 In contrast, mechanisms of action of cytokinins,
-H
-H
-H
furfurylamino-
furfurylamino-
furfurylamino-
1
1
3
4
including kinetin, in human cells remain largely unknown. Studies of
cytokinin bases often focus on their ability to protect biological mac-
romolecules, cells, and tissues against stress by direct antioxidant ac-
tivity or induction of antioxidant defenses, as reviewed by Kadlecova´
15
-H
-H
5-methylfurfurylamino-
(furfuryl)(methyl)amino-
1H-pyrazolo[3,4-d]pyrimidine
1-methyl-1H-pyrazolo[3,4-d]
pyrimidine
1
6
2
et al. in 2019. Recent exciting studies have shown that kinetin riboside-
1
7
-H
-H
(1-(5-methylfuran-2-yl)
ethyl)amino-
1-methyl-1H-pyrazolo[3,4-d]
pyrimidine
′
5
-triphosphate, a metabolite of kinetin, acts as a neosubstrate of the
9
mitoprotective kinase PINK1 and CK2 kinase, which phosphorylates
huntingtin aggregates.10 Perturbance of these kinases’ activities plays
major roles in Parkinson’s and Huntington’s diseases, respectively, so
restoring or modulating their activities using kinetin could provide
potent novel therapeutic options.
18
(5-methylfurfuryl)(methyl)
amino-
1-methyl-1H-pyrazolo[3,4-d]
pyrimidine
1
2
2
9
0
1
-H
-H
-H
-H
furfurylamino-
furfurylamino-
furfurylamino-
furfurylamino-
thieno[2,3-d]pyrimidine
thieno[3,2-d]pyrimidine
furo[2,3-d]pyrimidine
5,6-dimethylfuro[2,3-d]
pyrimidine
22
In contrast to cytokinin ribosides, for which several series of de-
rivatives have been reported1
2–19
23
-H
furfurylamino-
quinazoline
, studies of pharmacological activities
of cytokinin bases have been largely limited to naturally occurring
bases. However, a correction of aberrant splicing of ELP1 by 2-chlor-
okinetin20 and some 6,8-disubstituted purines was reported and it was
also demonstrated that 9-substituted derivatives of kinetin protect skin
cells against UV light.22 In those studies, effects of substitution of the
purine ring or replacement of the furan moiety was explored. Pharma-
cological activity of kinetin isosteres is subject of two recent US patents.
and the following settings: stirring – high, power – 200 W, ramp time – 2
21
◦
min, temperature 120 C, hold time – 10 min cycles, powermax - on.
Solid-phase synthesis was carried out in plastic reaction vessels (sy-
ringes, each equipped with a porous disk) using a manually operated
synthesizer. All reactions were carried out at ambient temperature (25
◦
C) unless stated otherwise. The volume of wash solvent was 10 mL per
2
3
Whereas 10676475 deals with effects on alternative pre-RNA splicing,
016728624 deals with PINK1 activation in the context of mitopro-
tection and neuroprotection. Previously, kinetin isosteres were studied
1
g of resin. For washing, resin slurry was shaken with the fresh solvent
1
for at least 1 min before changing the solvent. Resin-bound in-
termediates were dried by a stream of nitrogen for prolonged storage
and/or quantitative analysis.
as plant hormones.2
5–27
Here, we report an evaluation of effects of replacing the purine
For LC/MS analysis a sample of resin (~5 mg) was treated by TFA in
DCM, the cleavage cocktail was evaporated by a stream of nitrogen, and
cleaved compounds were extracted into 1 mL of MeCN/H2O (1:1). Prior
to biological testing all the residual solvents were removed from the
moiety of kinetin (1) and 2-chlorokinetin (2) by other fused heterocycles
(
Table 1) on biological activities in several assays related to neurode-
generative diseases.
◦
samples by careful lyophilization at ꢀ 110 C using a ScanVac Coolsafe
2
. Material and methods
1
10-4 freeze-dryer. Thin layer chromatography (TLC) was performed on
Silica gel 60 F254 plates (Merck) and migrating compounds were
detected by 254 nm UV light and/or visualized by vanillin staining.
Crude products were purified by column chromatography using 40–63
2
.1. Synthesis
The test compounds include both commercial compounds, from
μ
m Silikagel (VWR). Compounds cleaved from the resin were purified
chromatographically using a C18 reverse phase column (YMC Pack ODS-
m particles), and mobile phase consisting of a
Olchemim (kinetin ꢀ 1) and Molport (14–23), and prepared com-
pounds. The synthesis of the latter is described in detail below. An
exception is 2-chlorokinetin (2) that was provided by Dr. Mik (Palacký
University) and its analytical data correspond to the previously pub-
lished data.28
A, 20 × 100 mm, 5
μ
gradient formed from 10 mM aqueous ammonium acetate and MeCN,
with a flow rate of 15 mL/min.
The LC-MS analyses were performed following published protocols29
All chemicals and solvents for synthesis were purchased in analytical
or HPLC grade quality from available suppliers and used without further
purification. For microwave-assisted synthesis steps, reagents were
placed in 10 mL closed vessels and subjected to treatment with a
Discover SP microwave reactor (CEM Corporation) using dynamic mode
and compounds synthesized on solid phase were analysed using a
UHPLC-MS system consisting of an Acquity UHPLC chromatograph,
coupled to a single quadrupole mass spectrometer (Waters), and
equipped with a photodiode array detector and X-Select C18 column
2

B. Makov a´ et al.
Bioorganic & Medicinal Chemistry 33 (2021) 115993
◦
maintained at 30 C. The mobile liquid chromatography phase consisted
of a linear gradient of 20–80% MeCN balanced with 0.01 M ammonium
acetate in H2O, over 2.5 min, followed by a hold at 80% MeCN for 1.5
washed 5× with DMSO, shaken with triethyl orthoformate/DMSO (5
◦
mL, 1:1) for 16 h at 80 C and washed 5× with DMSO and 5× with DCM.
The resins VII(H) or VII(Cl) were shaken with TFA/DCM (5 mL, 1:1) for
90 min at rt, then washed 5× with TFA/DCM (1:1). TFA/DCM fractions
were collected and evaporated with a stream of nitrogen. The crude
products 3 or 4 were dissolved in MeCN (10 mL) and purified with
semipreparative HPLC.
min and re-equilibration with 20% B for 1 min (flow rate: 600 L/min).
μ
The ESI I source of the MS was operated at discharge current of 5
μ
A,
◦
◦
vaporizer temperature of 350 C, and capillary temperature of 200 C.
Melting points of synthesized compounds reported here were
determined using a Kofler, Stuart SMP30 or Stuart SMP 40 melting point
apparatus and are uncorrected. The elemental analysis of prepared
compounds was performed on CHN-O Analyzer EA Flash 1112 Series
White solid, overall yield 20% (calculated from the loading of resin
1
II). H NMR (500 MHz, DMSO‑d
6
) δ (ppm): 4.62 (d, J = 6.1 Hz, 2H), 6.31
(d, J = 2.9 Hz, 1H), 6.37 (dd, J = 3.2, 1.8 Hz, 1H), 6.41 (d, J = 5.6 Hz,
1H), 7.08 (t, J = 5.9 Hz, 1H), 7.50–7.61 (m, 1H), 7.88 (d, J = 5.6 Hz,
(
Thermo Finnigan). Analyses indicated by the symbols of the elements
were within ±0.4% of the theoretical values
1H), 8.06 (s, 1H). 13C NMR (125 MHz, DMSO‑d
6
) δ (ppm): 99.0, 107.1,
NMR spectra were recorded on Bruker Avance (300 or 500 MHz),
Jeol ECA-500 MHz, and Jeol ECX500 spectrometers operating at 300
110.4, 139.0, 142.1, 144.4, 145.2, 148.5, 152.8. HRMS (ESI-TOF): m/z
calcd. for C11H10N4O [M+H]+ 215.0927, found 215.0930.
1
13
1
13
MHz ( H), 75 MHz ( C) or 500 MHz ( H), and 125 MHz ( C), fre-
quencies, respectively. Samples were dissolved in DMSO‑d or CDCl3
and the chemical shifts (δ) reported in ppm were calibrated against a
residual solvent peak (2.49 ppm for proton - DMSO‑d , 7.26 ppm for
proton – CHCl3) and DMSO‑d (39.5 ppm for carbon) or CDCl3 (77.0
ppm for carbon). The NMR spectra are available as Appendix A. Pre-
pared compounds were subjected to HRMS analysis using Agilent 6224
Accurate-Mass TOF LC-MS with dual electrospray/chemical ionization
mode with mass accuracy greater than 2 ppm or Dionex Ultimate 3000
chromatograph and Exactive Plus Orbitrap Elite high-resolution mass
spectrometer (ThermoFisher, MA, USA) operating in positive full scan
mode (120 000 FWMH) in the range of 100–1000 m/z The oven tem-
perature and source voltage for electrospray ionization were set at 150
6
2.1.1.2. 5-chloro-N-(furan-2-ylmethyl)-3H-imidazo[4,5-b]pyridin-7-
amine (4). White solid, overall yield 25% (calculated from the loading
1
5
of resin II). H NMR (500 MHz, DMSO‑d ) δ (ppm): 4.64 (bs, 2H), 6.32
6
6
(d, J = 2.9 Hz, 1H), 6.39 (dd, J = 3.2, 1.8 Hz, 1H), 6.42 (s, 1H),
7.27–7.47 (m, 1H), 7.58 (dd, J = 1.8, 0.8 Hz, 1H), 8.09 (s, 1H), 12.71
(bs, 1H). 13C NMR (125 MHz, DMSO‑d ) δ (ppm): 39.1, 97.6, 106.9,
6
110.2, 139.1, 141.9, 145.5, 146.3, 152.1. HRMS (ESI-TOF): m/z calcd.
for C11H9ClN4O [M+H]+ 249.0538, found 249.0540.
2.1.2. Synthesis of 1H-imidazo[4,5-c]pyridine (3-deazapurine) derivatives
(VIII-X, 5, 6)
Resin II (500 mg) was washed 5× with DMSO and shaken with 2-
chloro-3-nitropyridin-4-amine or 2,6-dichloro-3-nitropyridin-4-amine
◦
C and 3.6 kV, respectively. The acquired data were internally calibrated
with phthalate as a contaminant in MeOH (m/z 297.15909). Samples
(2.5 mmol) with EDIPA (436 L) in DMSO (5 mL) for 16 h at rt, then
μ
were diluted to a final concentration of 0.1 mg/mL in H2O and MeOH
washed 3× with DMSO and 3× with DCM. Na2S2O4 (1050 mg), K2CO3
(960 mg) and TBAHS (170 mg) were dissolved in DCM/H2O (10 mL,
1:1) and added to resins VIII(H) or VIII(Cl). The resins were shaken 16 h
at rt, then washed 10× with MeOH/H2O (1:1), 3× with MeOH/DCM
(1:1) and 3× with DCM. The resulting resins IX(H) or IX(Cl) were
washed 5× with DMSO, shaken with triethyl orthoformate/DMSO (5
(
50:50, v/v). Before HPLC separation (using a Phenomenex Gemini, 50
×
2.00 mm, 3 µm particles, C18 column), the samples were injected by
direct infusion into the LC-MS system using an autosampler. The mobile
phase was isocratic MeCN/iPrOH/0.01 M ammonium acetate (40:5:55)
with a flow rate of 0.3 mL/min.
◦
mL, 1:1) for 16 h at 100 C and washed 5× with DMSO and 5× with
2
.1.1. Synthesis of 3H-imidazo[4,5-b]pyridine (1-deazapurine) derivatives
DCM. The resins X(H) or X(Cl) were shaken with TFA/DCM (5 mL, 1:1)
90 min at rt, then washed 5× with TFA/DCM (1:1). TFA/DCM fractions
were collected and evaporated with a stream of nitrogen. The crude
products 5 or 6 were dissolved in MeCN (10 mL) and purified with
semipreparative HPLC.
(
I-VII, 3, 4)
2
.1.1.1. N-(furan-2-ylmethyl)-3H-imidazo[4,5-b]pyridin-7-amine (3).
Aminomethyl resin (1 g, 0.098 mmol/g) was washed 3× with DCM and
3
1
× with DMF then 10% TEA in DMF (10 mL) was added. After shaking
0 min at rt, the resin was washed 5× with DMF and 4-(4-formyl-3,5-
2.1.2.1. N-(furan-2-ylmethyl)-1H-imidazo[4,5-c]pyridin-4-amine
(5).
dimethoxyphenoxy)butanoic acid (524 mg), HOBt (229 mg) and DIC
White solid, overall yield 21% (calculated from the loading of resin IV).
1
(
300 mL) in DMF/DCM (1:1, 10 mL) were added. The resin was shaken
for 16 h at rt, then washed 3× with DMF, 3× with DCM, 3× with
anhydrous THF and 3× with anhydrous DMF. Furfurylamine (442 L)
H NMR (500 MHz, DMSO‑d ) δ (ppm): 4.69 (d, J = 5.9 Hz, 2H),
6
6.30–6.39 (m, 1H), 6.21 (s, 1H), 6.79 (d, J = 4.9 Hz, 1H), 6.98 (bs, 1H),
1
3
μ
7.51 (s, 1H), 7.70 (d, J = 5.8 Hz, 1H), 8.08 (s, 1H), 12.50 (bs, 1H).
C
was dissolved in 10% AcOH/anhydrous DMF (10 mL), the mixture was
added to the resin and it was shaken for 16 h at rt. NaBH(OAc)3 (600
mg) was dissolved in 5% AcOH/anhydrous DMF (5 mL), and three equal
portions of the solution were added at 1 h intervals to the reaction
mixture. After the last addition, the resin was shaken for 16 h at rt,
washed 5× with DMF, then shaken with 20% piperidine in DMF (10 mL)
for 10 min at rt and washed 5× with DMF and 5× with DCM to obtain
the resin I. The resin I (500 mg) was washed 3× with DMSO and shaken
with a solution of 4,6-dichloro-3-nitropyridin-2-amine or 2,4-dichloro-
NMR (125 MHz, DMSO‑d ) δ (ppm): 37.1, 99.2, 106.0, 110.1, 138.9,
6
139.6, 141.3, 147.6, 153.7. HRMS (ESI-TOF): m/z calcd. for
+
C11H10N4O [M+H] 215.0927, found 215.0929.
2
.1.2.2. 6-chloro-N-(furan-2-ylmethyl)-1H-imidazo[4,5-c]pyridin-4-
amine (6). White solid, overall yield 28% (calculated from the loading
1
of resin IV). H NMR (500 MHz, DMSO‑d
6
) δ (ppm): 4.64 (bs, 2H), 6.25
(
d, J = 2.9 Hz, 1H), 6.37 (dd, J = 3.0, 1.9 Hz, 1H), 6.79 (s, 1H), 7.41 (bs,
1
3
1
H), 7.55 (d, J = 0.9 Hz, 1H), 8.15 (s, 1H). C NMR (125 MHz,
3
1
-nitropyridine (1.54 mmol) and EDIPA (436 L) in DMSO (5 mL) for
μ
DMSO‑d
6
) δ (ppm): 37.1, 96.9, 106.7, 110.4, 124.2, 140.4, 140.7, 140.8,
6 h at rt. The resulting resins were washed 3× with DMF and 3× with
1
41.7, 148.9, 153.3. HRMS (ESI-TOF): m/z calcd. for C11H9ClN4O
[
M+H]+ 249.0538, found 249.0539.
DCM to obtain the intermediates II or IV. Resin IV was shaken with a
◦
solution of 25% aq⋅NH3 in DMSO for 16 h at 100 C (in a sealed vial),
washed 3× with DMF and 3× with DCM to obtain the resin V. Na2S2O4
2
.1.3. Synthesis of 7H-pyrrolo[2,3-d]pyrimidine (7-deazapurine)
(
1050 mg), K2CO3 (960 mg) and TBAHS (170 mg) were dissolved in
derivatives (7, 8)
DCM/H2O (10 mL, 1:1), added to resins III or V, the resins were shaken
for 16 h at rt, then washed 10× with MeOH/H2O (1:1), 3× with MeOH/
DCM (1:1) and 3× with DCM. The resulting resins VI(H) or VI(Cl) were
2
.1.3.1. N-(furan-2-ylmethyl)-7H-pyrrolo[2,3-d]pyrimidin-4-amine (7).
A mixture of 4-chloro-7H-pyrrolo[2,3-d]pyrimidine (100 mg, 0.65
3

B. Makov a´ et al.
Bioorganic & Medicinal Chemistry 33 (2021) 115993
mmol), furfurylamine (90 µL, 0.98 mmol) and TEA (225
μ
L, 1.62 mmol)
mixture was stirred in the dark at room temperature for 9 h. The reaction
was quenched by adding ice cold water and the product was extracted by
Et2O. Combined organic layers were carefully concentrated under
reduced pressure to give an oily liquid that was used further without
in dry MeOH (2.6 mL) was heated in a Discover SP microwave reactor in
abovementioned conditions for four cycles. Solvent was evaporated
under reduced pressure, the residue was treated with water (10 mL) and
product was extracted by EtOAc (4 × 10 mL). Combined organic layers
were washed with brine (10 mL), dried (Na2SO4) and concentrated in
vacuo. Crude material was purified by silica gel column chromatography
using CHCl3/MeOH as a mobile phase with a MeOH gradient. Yellow
1
purification. Oily liquid; yield 79%. H NMR (300 MHz, CDCl3) δ (ppm):
1
3
7.24–7.48 (m, 5H), 4.35 (s, 2H). C NMR (125 MHz, CDCl3) δ (ppm):
54.9, 128.4 (2 × C), 128.5, 129.0 (2 × C), 135.5.
+
+
1
solid; yield 60%; ESI -MS m/z: 215.4 [M+H] . H NMR (500 MHz,
DMSO‑d
) δ (ppm): 4.68 (d, J = 5.8 Hz, 2H), 6.26 (dd, J = 3.2, 0.8 Hz,
H), 6.37 (dd, J = 3.1, 1.8 Hz, 1H), 6.57 (dd, J = 3.4, 1.8 Hz, 1H), 7.06
2.1.5.2. 5-amino-1-benzyl-1H-1,2,3-triazole-4-carboxamide
(XII). 2-
6
Cyanoacetamide (11.83 g, 0.14 mol) was added to a solution of sodium
ethoxide (Na ꢀ 3.23 g, 0.14 mol, EtOH – 190 mL) to give a white sus-
pension. Benzyl azide (XI) (18.2 g, 0.13 mol) was added and the reaction
1
(
dd, J = 3.2, 2.3 Hz, 1H), 7.55 (dd, J = 1.7, 0.8 Hz, 1H), 7.84 (t, J = 5.7
1
3
◦
Hz, 1H), 8.11 (s, 1H), 11.51 (bs, 1H). C NMR (125 MHz, DMSO‑d
6
) δ
mixture was heated at 90 C overnight. The resulting solid was filtered,
(
ppm): 36.5, 98.6, 102.6, 106.8, 110.4, 121.0, 141.9, 150.2, 151.2,
washed with EtOH, then water and dried. A second portion of product
was obtained after concentration of the reaction mixture under reduced
pressure and treatment with water. Off-white solid, yield 69%, m.p.
+
1
2
53.1, 155.6. HRMS (ESI-TOF): m/z calcd. for C11H10N4O [M+H]
15.0927, found 215.0926.
2
36–237 ◦C, ESI -MS m/z: 217.8 [M+H] (100). 1H NMR (500 MHz,
+
+
2
.1.3.2. 2-chloro-N-(furan-2-ylmethyl)-7H-pyrrolo[2,3-d]pyrimidin-4-
DMSO‑d
6
) δ (ppm): 5.38 (s, 2H), 7.17. (d, J = 7.3 Hz, 2H), 7.26 (t, J =
13
amine (8). A mixture of 2,4-dichloro-7H-pyrrolo[2,3-d]pyrimidine
L,
.33 mmol) was refluxed in nPrOH (2.13 mL) for 4 h. After cooling at 4
7.3 Hz, 1H), 7.32 (t, J = 7.4 Hz, 2H). C NMR (125 MHz, DMSO‑d
6
) δ
(
100 mg, 0.53 mmol), furfurylamine (74 µL, 0.8 mmol) and TEA (185
μ
(ppm) 48.8, 122.2, 127.9 (2 × C), 128.2, 129.1 (2 × C), 136.5, 145.4,
1
164.8.
◦
C for 2 h the resulting solid material was filtered, washed with cold
nPrOH (3 × 1 mL), cold H2O (5 × 1 mL) and dried. Crude material was
purified by silica gel column chromatography using CHCl3/MeOH as a
2.1.5.3. 3-Benzyl-3,6-dihydro-7H-[1,2,3]triazolo[4,5-d]pyrimidin-7-one
(XIII). Ethyl formate (2.22 mL, 27.6 mmol) was added to a mixture of
XII (2 g, 9.21 mmol) in sodium ethoxide solution (Na – 1.06 g, 46 mmol,
+
mobile phase with a MeOH gradient. Yellow solid; yield 62%; ESI -MS
m/z (rel %): 249.4 [35Cl-M+H] (100), 251.4 [ Cl-M + H] (35). 1
+
37
+
H
◦
EtOH – 50 mL). The mixture was heated in a pressure tube at 95 C for
NMR (500 MHz, DMSO‑d
6
) δ (ppm): 4.63 (d, J = 5.5 Hz, 2H), 6.30 (dd, J
24 h. After cooling the pH of the reaction mixture was adjusted to 6,
=
3.2, 0.8 Hz, 1H), 6.40 (dd, J = 3.4, 1.8 Hz, 1H), 6.58 (dd, J = 3.4, 1.8
Hz, 1H), 7.08 (dd, J = 3.5, 2.3 Hz, 1H), 7.59 (dd, J = 1.8, 0.9 Hz, 1H),
resulting in formation of solid material. The product was filtered off and
◦
recrystallized from ethanol. White solid; yield 84%; m.p. 194–195 C.
1
3
+
+
1
8
.32 (t, J = 5.7 Hz, 1H), 11.70 (bs, 1H). C NMR (125 MHz, DMSO‑d
6
) δ
ESI -MS m/z: 228.8 [M+H] . H NMR (500 MHz, DMSO‑d ) δ (ppm):
6
13
(
ppm): 36.6, 99.1, 101.2, 107.3, 110.5, 121.6, 142.1, 150.8, 152.2,
5.70 (s, 2H), 7.24–7.31 (m, 5H), 8.22 (s, 1H), 12.70 (s, 1H). C NMR
+
1
2
52.3, 156.4. HRMS (ESI-TOF): m/z calcd. for C11H9ClN4O [M+H]
(125 MHz, DMSO‑d ) δ (ppm): 50.3, 128.4 (2 × C), 128.7, 129.3 (2 × C),
6
49.0538, found 249.0537.
130.2, 135.9, 149.1, 150.5, 155.9.
2
.1.4. Synthesis of 5H-pyrrolo[3,2-d]pyrimidine (9-deazapurine)
2
.1.5.4. 3-Benzyl-7-chloro-3H-[1,2,3]triazolo[4,5-d]pyrimidine (XIV). A
derivatives (9, 10)
mixture of SOCl2 (10.13 mL, 139.6 mmol) and DMF (1.82 mL) was
added to a solution of XIII (2.54 g, 11.17 mmol) in CHCl3 (100 mL). The
reaction mixture was refluxed for 2 h then evaporated in vacuo. The
residue was mixed with ice-cold water (30 mL) and the resulting mixture
was extracted by Et2O (100 mL). After evaporation of Et2O, crude
product was purified by silica column chromatography using petroleum
ether/ethyl acetate (9:2, v/v) as a mobile phase. Off-white solid; yield
2
.1.4.1. N-(furan-2-ylmethyl)-5H-pyrrolo[3,2-d]pyrimidin-4-amine (9).
Compound 9 was prepared similarly to the 7-deazapurine analogue (see
Section 2.1.3.1), except that five reaction cycles were needed. Yellow
+
+
1
solid; yield 25%; ESI -MS m/z: 216.4 [M+H] . H NMR (500 MHz,
DMSO‑d
6
) δ (ppm): 4.71 (d, J = 5.5 Hz, 2H), 6.36 (d, J = 3.1 Hz, 2H),
◦
+
+
1
6
0
.41 (dd, J = 3.4, 1.8 Hz, 1H), 7.50 (d, J = 2.8 Hz, 2H), 7.61 (dd, J = 1.8,
50%; m.p. 223–239 C. ESI -MS m/z (rel. %): 246.0 [M+H] (100). H
1
3
6
.9 Hz, 1H), 8.20 (s, 1H), 10.97 (bs, 1H). C NMR (125 MHz, DMSO‑d )
NMR (500 MHz, CDCl3) δ (ppm): 5.87 (s, 2H), 7.29–7.35 (m, 3H), 7.45
1
3
δ (ppm): 36.5, 101.2, 107.4, 110.6, 113.7, 127.9, 142.4, 146.1, 148.9,
(dd, J = 7.9, 1.5 Hz, 2H), 8.91 (s, 1H). C NMR (125 MHz, CDCl3) δ
(ppm): 51.5, 128.6 (2 × C), 129.0, 129.2 (2 × C), 133.9, 134.2, 149.8,
154.3, 155.7.
+
1
2
49.8, 152.4. HRMS (ESI-TOF): m/z calcd. for C11H10N4O [M+H]
15.0927, found 215.0926
2
.1.4.2. 2-chloro-N-(furan-2-ylmethyl)-5H-pyrrolo[3,2-d]pyrimidin-4-
2.1.5.5. 3-Benzyl-N-(furan-2-ylmethyl)-3H-[1,2,3]triazolo[4,5-d]pyr-
amine (10). Compound 10 was prepared similarly to the 7-deazapurine
imidin-7-amine (XV). A mixture of XIV (0.2 g, 0.81 mmol), furfuryl-
+
analogue (see Section 2.1.3.2). Pale yellow solid; yield 57%; ESI -MS m/
amine (68
μ
L, 0.77 mmol) and EDIPA (0.7 mL, 40.7 mmol) in iPrOH (8
z (rel. %): 249.1 4 [35Cl-M + H] (100), 251.0 [ Cl-M + H]+ (31). H
+
37
1
◦
mL) was heated at 100 C for 4.5 h. After cooling the product started to
NMR (500 MHz, DMSO‑d
6
) δ (ppm): 4.66 (d, J = 4.6 Hz, 2H), 6.32 (d, J
precipitate. The solid was filtered, washed with iPrOH (10 mL), then
◦
=
3.1 Hz, 1H), 6.39 (dd, J = 3.2, 0.8 Hz, 1H), 6.43 (dd, J = 3.1, 1.8 Hz,
H), 7.53 (d, J = 3.1 Hz, 1H), 7.63 (dd, J = 1.8, 0.9 Hz, 1H), 7.85 (bs,
H), 11.03 (bs, 1H). 13C NMR (125 MHz, DMSO‑d
water (10 mL) and dried. Pale yellow solid; yield 84%; m.p. 169–170 C.
+
+
1
1
1
1
ESI -MS m/z (rel. %): 306.8 [M+H] (100). H NMR (500 MHz, CDCl3)
δ (ppm): 4.88 (d, J = 5.8 Hz, 2H), 5.75 (s, 2H), 6.32 (s, 2H), 6.80 (s, 1H),
6
) δ (ppm): 36.6,
1
3
01.2, 107.8, 110.6, 112.4, 129.3, 142.6, 148.3, 149.9, 150.8, 151.6.
7.27–7.41 (m, 6H), 8.53 (s, 1H). C NMR (125 MHz, CDCl3) δ (ppm):
37.6, 50.6, 108.2, 110.6, 124.9, 128.4 (2 × C), 128.6, 129.0 (2 × C),
134.9, 142.7, 148.7, 150.6, 154.4, 157.1.
+
HRMS (ESI-TOF): m/z calcd. for C11H9ClN4O [M+H] 249.0538, found
2
49.0537.
2
.1.5. Synthesis of N-(furan-2-ylmethyl)-3H-[1,2,3]triazolo[4,5-d]
2
.1.5.6. N-(furan-2-ylmethyl)-3H-[1,2,3]triazolo[4,5-d]pyrimidin-7-
pyrimidin-7-amine (XI-XV, 11)
◦
amine (11). Sodium (37 mg, 1.63 mmol) was added at ꢀ 78 C to liquid
ammonia (25 mL) and stirred until complete dissolution. XVII (0.1 g,
◦
2
.1.5.1. Benzyl azide (XI). Benzyl bromide (17.84 mL, 0.15 mol) was
0.33 mmol) was added and the reaction mixture was stirred at ꢀ 78 C
added to a solution of NaN3 (11 g, 0.17 mol) in DMSO (300 mL) and the
for 1 h. After evaporation, the residue was treated with water and the pH
4

B. Makov a´ et al.
Bioorganic & Medicinal Chemistry 33 (2021) 115993
◦
of the mixture adjusted to 5–6, resulting in precipitation of product. The
serum (FBS) (Sigma) in standard culture conditions (5.5% CO2, 37 C,
solid was filtered off, washed with water and dried. Off-white solid;
100% relative humidity). The cells were subcultured twice a week.
◦
+
+
yield 62%; m.p. 244–245 C. ESI -MS m/z (rel. %): 217.1 [M+H]
1
(
100). H NMR (500 MHz, DMSO‑d
6
) δ (ppm): 4.74 (d, J = 5.8 Hz, 2H),
.29 (d, J = 3.0 Hz, 1H), 6.37 (d, J = 1.8 Hz, 1H), 7.56 (s, 1H), 8.38 (s,
H), 9.28 (s, 1H). 13C NMR (125 MHz, DMSO‑d
2.3. Fibroblast culture stress assays
6
1
1
6
) δ (ppm): 37.0, 107.7,
GM04078 fibroblasts were trypsinized, diluted in DMEM medium
with 10% FBS and seeded into 384-well plates (1500 cells in 30 µL
medium/well). After 24 h incubation, the cells were pre-treated by the
test compounds dissolved in DMSO using an Echo 555 system (Labcyte).
DMSO vehicle served as a negative control. Stressors were added in 15
11.0, 123.9, 142.6, 143.0, 152.4, 154.4, 156.7. HRMS (ESI-TOF): m/z
+
calcd. for C9H8N6O [M+H] 217.0838, found 249.0845.
2
.1.6. Synthesis of N-(furan-2-ylmethyl)pyrazolo[1,5-a]pyrimidin-7-
amine (12)
Furfurylamine (180
were added to a solution of 7-chloropyrazolo[1,5-a]pyrimidine (200 mg,
µL of the medium used for seeding after 8 h. After another 48 h, 15
μ
L of
μ
L, 1.95 mmol) and K2CO3 (360 mg, 2.60 mmol)
a 4-fold concentrated solution of resazurin (Sigma) in the culture me-
dium was added to the cells to a final concentration of 0.0125 mg/mL.
Fluorescence was measured after 3 h incubation using an M2 reader
(Biotek) with excitation and emission wavelengths of 570 and 610 nm,
respectively.
1
.30 mmol) in anhydrous acetonitrile (8 mL) and DMF (2 mL) and the
◦
reaction mixture was stirred at 90 C for 3 h. The solvents were evap-
orated under reduced pressure, the residue was mixed with H2O (30
mL), and the mixture was extracted with EtOAc (2 × 50 mL). Combined
organic extracts were washed with brine (50 mL), dried over MgSO4 and
concentrated under reduced pressure. The residue was purified by silica
gel flash column chromatography using DCM/EtOAc (3:2) as a mobile
phase. White solid; yield 34%; m.p. 120–122 ◦C. H NMR (500 MHz,
CDCl3) δ (ppm): 4.59 (d, J = 5.8 Hz, 2H), 6.06 (d, J = 5.2 Hz, 1H),
2.4. Reduced glutathione measurements
GM04078 fibroblasts were trypsinized, diluted in DMEM medium
with 10% FBS and seeded into 384-well plates (1500 cells in 30 µL
medium/well). After 24 h incubation, the cells were pre-treated by
adding compound 6 dissolved in DMSO (final concentration 10 and 50
µM) using the Echo 555 system. DMSO vehicle served as a negative
control. BSO was added (to serial concentrations of 10, 100 and 1000
1
6
1
.32–6.38 (m, 2H), 6.53 (d, J = 2.3 Hz, 1H), 6.84 (s, 1H), 7.35–7.45 (m,
13
H), 8.00 (d, J = 2.3 Hz, 1H), 8.25 (d, J = 5.2 Hz, 1H). C NMR (125
MHz, CDCl3) δ (ppm): 39.4, 85.3, 95.9, 108.6, 110.7, 143.0, 143.9,
+
1
2
46.9, 149.6, 149.7. HRMS (APCI): m/z calc. for C11H11N4O [M+H]
µM) in 15
μ
L portions of medium after 8 h. The effect was evaluated after
15.0927, found 215.0925.
24 h, when the medium was removed and, after a washing with 100 µL
DMEM medium, replaced by DMEM medium containing 40 µM mono-
bromobimane. After 40 min incubation, images were recorded, at 4×
magnification using the blue channel of a Cell Voyager confocal imaging
system. The intensity of the signal was evaluated by an image analysis
tool in the Cell Profiler program. Control measurement using spectro-
photometer (ex 395 nm, em 490 nm) was also performed.
2
.1.7. Synthesis of 1H-pyrazolo[4,3-d]pyrimidine derivative (XVI, XVII,
1
3)
2
.1.7.1. 1H-pyrazolo[4,3-d]pyrimidine-7-thiol (XVI). 1H-pyrazolo[4,3-
d]pyrimidine-7-thiol was prepared from 1H-pyrazolo[4,3-d]pyrimidin-
7
-ol according to the published procedure.30
2
.5. Frataxin assay
2
.1.7.2. 7-(methylthio)-1H-pyrazolo[4,3-d]pyrimidine (XVII). A solution
◦
To test the ability of compounds to replace frataxin (FXN) func-
of MeI (144 mg, 1 mmol) in DMF (2.8 mL) was added dropwise at 0 C to
a mixture of XVI (153 mg, 1 mmol) and K2CO3 (152 mg, 1.1 mmol) in
DMF (7.5 mL). The reaction mixture was allowed to warm to rt then
stirred for an additional 10 min. After evaporation in vacuo the residue
was partitioned between H2O and DCM. The product was finally ob-
tained by crystallization from DCM. White solid; yield 84%; m.p.
tionally, enzymatic assays were performed using mouse ISC machinery
reconstituted in vitro with purified proteins 31. The assays were per-
formed under anaerobic conditions in Tris buffer (50 mM Tris, 150 mM
NaCl, pH 8) containing 50 µM of apo-ISCU loaded with one equivalent of
2
+
Fe , 5 µM of the NFS1-ISD11-ACP complex, 5 µM of FDX2, 2 µM of
FDXR and 100 µM of NADPH. Positive controls were prepared by adding
◦
-
ꢀ
1
95–200 C; UV (nm): max. 308, sh 319. ESI -MS (m/z): 165.1 [Mꢀ H] .
5
µM of FXN to the reaction mixture. Test compounds diluted in DMSO
were added to the reaction mixture at concentrations of 30, 50, 100, 200
and 500 M and supplementary DMSO was added to the reaction mix-
2
.1.7.3. N-(furan-2-ylmethyl)-1H-pyrazolo[4,3-d]pyrimidin-7-amine
μ
(
13). A mixture of XVII (93 mg, 0.56 mmol) in furfurylamine (3 mL)
◦
tures to keep its concentration the same in each reaction. Reference
reaction mixtures were prepared with the same concentration of DMSO
as the mixtures with test compounds. The reactions were initiated by
adding 50 µM of L-cysteine and the formation of Fe-S clusters was
was heated at 125 C for 8 h. After evaporation under reduced pressure
the crude material was purified by silica gel column chromatography
using stepwise increments of MeOH (3, 4, and 5%) in CHCl3. Product
was finally obtained by crystallization from CHCl3. White solid; yield
◦
◦
+
monitored by UV–visible spectroscopy at 456 nm and 30 C. Rates of the
6
6%, m.p. 182–185 C; UV λ (nm): min. 253, max. 292, sh 305. ESI -MS
+
-
+
1
reactions were determined by linear regression of the slopes of the
resulting curves in their linear domain between 2 and 10 min, using
GraphPad Prism. The compounds’ activities were calculated by dividing
slopes of their reaction curves by those of the corresponding reference
reactions, which yielded activation factors. Activation factors higher
and lower than 1 indicate active and inactive compounds, respectively.
(
m/z): 216.1 [M+H] , ESI -MS (m/z): 214.1 [M+H] . H NMR (500
MHz, DMSO‑d
6
) δ (ppm): 4.75 (bs, 2H), 6.38 (s, 1H), 6.41 (s, 1H), 7.61
s, 1H), 7.93 (bs, 1H), 8.12 (s, 1H), 8.26 (s, 1H), 12.72 (bs, 1H). 13C NMR
(
(
125 MHz, DMSO‑d
6
) δ (ppm): 36.4, 107.6, 110.5, 121.0, 133.1, 141.4,
1
42.4, 148.7, 151.7. Anal. (C10H9N5O) C,H,N.
2
.2. Fibroblast cell culture
2.6. SH-SY5Y cultivation and differentiation
Skin fibroblasts GM04078 homozygous for the GAA expansion in the
The SH-SY5Y human neuroblastoma cell line obtained from ATCC
(American Type Culture Collection, Manassas, VA, USA) was cultivated
in DMEM medium and Ham’s F12 Nutrient Mixture (DMEM:F-12, 1:1),
supplemented with 10% FBS and 1% of both penicillin and streptomycin
frataxin gene and skin fibroblast GM04663, which are homozygous for
the 2507 + 6 T > C mutation in the ELP1 gene, were obtained from the
Coriell repository. The cells were maintained in Dulbecco’s modified
Eagle’s medium (DMEM) containing 5 g/L glucose, 2 mM L-glutamine,
◦
at 37 C in a humidified atmosphere of 5% CO2. Only the cells passaged
1
00 U/mL penicillin, 100
μg/mL streptomycin and 10% fetal bovine
less than 20 times were used in the experiments. The cells were seeded in
5

B. Makov a´ et al.
Bioorganic & Medicinal Chemistry 33 (2021) 115993
′
35
9
6-multiwell plates, in 100
μ
L total volume of medium per well. Next
CATTTCCAAGAAACACCTTAGGG-3 primers and JumpStart Taq DNA
day, 10 µM of ATRA in DMEM/F12 medium with 1% FBS was added to
each well to a final concentration 10 µM to induce neuronal differenti-
ation. The cells were used for experiments 48 h later. The same protocol
was used for the SH-SY5Y obtained from ECACC (European Collection of
Authenticated Cell Culture).
Polymerase with MgCl2 (Sigma-Aldrich). The temperature program
◦
involved: initial denaturation at 94 C for 5 min followed by 34 cycles of
◦
◦
denaturation at 94 C for 40 s, annealing at 63.8 C for 30 s and elon-
◦
◦
gation at 72 C for 90 s, with a final elongation step at 72 C for 5 min.
The PCR products corresponding to wild-type ELP1 transcript (422 bp)
and transcript lacking exon 20 (348 bp) were separated by electropho-
resis in 1.5% agarose gel and visualized by GelRed staining. For final
comparison of the compounds identified as active during the initial
screening, microfluidic electrophoresis (with an Agilent 2100 Bio-
analyzer and 1000 LabChip kit) was used. A comparison with DNA
standards of known size and concentration allowed precise quantifica-
tion of the wild-type and aberrant transcripts’ concentrations. The area
under the peak calculations were carried out in the 2100 Expert
Software.
2
.7. Evaluation of compounds’ cytoprotective activity toward
differentiated SH-SY5Y cells
Portions of culture containing 20,000 SH-SY5Y cells were seeded
into 96-well plates and differentiated by ATRA as outlined above. The
differentiation medium was removed and replaced by fresh DMEM/F12
medium with 1% FBS containing test compounds at 0.1, 1 or 10 M
μ
concentration. DMSO vehicle was used as a negative control. The
differentiated cells were immediately exposed to glutamate (160 mM),
and the frequency of dead cells was assessed using propidium iodide (PI)
as a probe after another 24 h, following the published protocol32 with
modification. Briefly, PI (Sigma Aldrich) was diluted to 1 mg/mL in
DMSO. This solution was further diluted in PBS then added to the cells’
medium to a final concentration of 1 µg/mL, then the cells were incu-
bated at rt for 15 min. Proportions of PI-stained cells were quantified
using an Infinite M200 Pro reader (Tecan) with 535 excitation and
emission wavelengths of 535 and 617 nm, respectively.
2.11. Western blotting and immunodetection of frataxin
For this purpose, GM4078 cells in DMEM medium with 10% FBS
were seeded in 10 cm Petri dishes (about 1.4 million cells per dish) and
24 h later treated with 10 and 50 M of compound 6. The cells were
μ
harvested after 48 h. Medium was removed and the cell monolayer was
washed twice with ice-cold PBS. The cells were scraped into 2 mL ice-
cold PBS and centrifuged. The supernatant was removed and each pel-
let was flash-frozen by dipping the tube into liquid nitrogen. The sam-
◦
2
.8. Measurement of caspase 3/7 activity
ples were stored at ꢀ 80 C.
Total protein was extracted by direct cell lysis in Laemmli buffer (50
mM Tris-HCl, pH 6.8, 2% SDS, 10% glycerol, 0.04% bromophenol blue,
and 250 mM mercaptoethanol). The protein concentrations in the
samples were determined by Bradford’s method and standardized. The
To assess the effects of test compounds on glutamate-induced
apoptosis, SH-SY5Y cells were subjected to the differentiation and
treatment steps described in the previous section, and then to the one-
step caspase 3/7 assay described by Carrasco et al. in 200333 with
modification of the caspase-3,7 substrate concentration, which was
decreased to 75 µM. Caspase-3,7 activity was quantified using an Infinite
M200 Pro (Tecan, Austria) microplate reader with excitation and
emission wavelengths of 346 and 438 nm, respectively after 3-hour in-
cubation. The average signal of the samples treated with glutamate
corresponds to 100% caspase 3/7 activity.
◦
samples were boiled for 10 min and stored at ꢀ 20 C.
The protein lysates (25–40 µg) were separated by SDS-PAGE (12.5%
gel). Proteins were transferred to polyvinylidene difluoride membrane
(Bio-Rad) pre-treated by incubation in a 5% solution of bovine serum
albumin (BSA) with 1x PBS-Tween (PBS-T) for 2 h. The membrane was
then incubated with anti-frataxin mouse antibody (18A5DB1 Abcam) in
BSA for 2 h in PBS-T, followed by three washes with PBS-T. For loading
standardization, a portion of the membrane was incubated with anti-
actin antibody. The membrane was then incubated with a secondary
2
.9. Measurement of oxidative stress
(
horseradish peroxidase-conjugated anti-mouse) antibody for 60 min
To assess the effects of test compounds on glutamate-induced
followed by three washes with PBS-T. The immunoblots were then
developed using Amersham ECL reagents.
oxidative stress, SH-SY5Y cells were subjected to the differentiation
and treatment steps described in the Section 3.5, and then the dihy-
droethidium assay described by Kim et al. in 201734 with the following
modifications. After 4 h of exposure to glutamate, cells were centrifuged
at 500 g for 330 s with subsequent replacement of media by PBS con-
2.12. In vitro permeability assays
Parallel artificial membrane permeability, chemical stability, sta-
bility in human plasma, microsomal stability and protein plasma bind-
ing assays were performed using methods described by Borkov a´ et al. In
2019.36
taining 10 M dihydroethidium (DHE). Finally, plates with cells were
μ
kept at rt for 30 min, then superoxide radical formation (DHE) signal
was quantified using an Infinite M200 Pro reader (Tecan) with excita-
tion and emission wavelengths of 500 and 580 nm, respectively.
2
.12.1. Quantification of test compounds in the permeability assays
2
.10. Evaluation of compounds’ effects on alternative splicing of ELP1
Preparation of lyophilized samples for this analysis is described in
transcripts
the sections dedicated to the in vitro ADME assays. Test compounds in
the samples were quantified using a RapidFire RF300 system (Agilent
Technologies) interfaced with a QTRAP 5500 mass spectrometer fitted
with an electrospray ionization source (AB Sciex, Concord, Canada)
running in multiple-reaction monitoring mode.
GM04663 cells in DMEM medium with 10% FBS were seeded in 10
cm Petri dishes (about 500,000 cells per dish) and grown overnight
before treatment with the test compounds at 50 M concentration to test
μ
their effects on alternative splicing of ELP1 transcripts. Kinetin (com-
pound 1) and 2-chlorokinetin (compound 2) were used as positive
controls. DMSO vehicle served as a negative control. After 24 h, total
RNA was extracted from the cells using Trizol reagent. Isolated RNA was
transcribed into cDNA using a High Capacity cDNA Reverse Transcrip-
tion Kit (Applied Biosystems). The ELP1 cDNA segment between nu-
cleotides 2194 and 2593 (numbering corresponding to the
NM_003640.5 transcript) was amplified by the polymerase chain reac-
Lyophilized samples were dissolved in the mobile phase (95% water,
5% acetonitrile, 0.1% formic acid) with respective internal standards.
The dissolved samples were aspired directly from 96-well plates into a
10 L sample loop and passed through a C4 cartridge (Agilent Tech-
μ
nologies) with solvent A (95% water, 0.01% formic acid, 5% acetoni-
trile) at a flow rate of 1.5 mL/minute for 3 s. After the desalting step, the
analyte retained on the cartridge was eluted with solvent B (95%
acetonitrile, 5% 0.01% formic acid) into the mass spectrometer at a flow
rate of 0.4 mL/minute for 7 s, where it was subjected to electrospray
′
′
′
tion
using
5 -CAGGTGTCGCTTTTTCATCA-3
and
5 -
6

B. Makov a´ et al.
Bioorganic & Medicinal Chemistry 33 (2021) 115993
ionization in the positive ion mode and daughter ion was identified.
P-gp substrates.
2
.12.2. Studies of transport across Caco-2 and MDCK-MDR1 cell
monolayers
To generate cell monolayers for transport studies,37,38 the cells were
2
.13. Data analysis
Statistical significance of the compound effect was evaluated by non-
trypsinized and seeded on tissue culture polyester membrane filters with
pore size 1 µm in 96-well Transwell® plates (Corning, NY, USA). Culture
medium was added to both the donor and the acceptor compartments
and the cells were allowed to differentiate and form monolayers. The
culture medium was changed every other day.
parametric Kruskal-Wallis test followed by post-hoc Man-Whitney test
and sequential Bonferroni correction using the GrahpPad and PAST (ver.
3
9
1
.97) software packages. Differences with the corrected p-values <
0
.05 were considered statistically significant. The error bars in graphs
are standard errors, unless stated otherwise. Standard errors for drug
MDR1-MCDK differentiated monolayers were used only if they were
intact, as confirmed by the Lucifer Yellow Rejection Assay. Before each
experiment, the cells were washed twice with Hank’s balanced buffer
solution (HBBS) (Gibco, Waltham, USA) and pre-equilibrated with HBSS
buffered at pH 7.4 for 1 h. After removing the medium, MDCK-MDR1
4
0
efflux ratios were calculated by the Delta method described in
implemented in R.
3
. Results and discussion
A series of kinetin (1) and 2-chlorokinetin (2) purine bioisosteres
cells were treated with test compounds at 10
μM in HBSS (pH 7.4) for
1
h, then samples were removed from both donor and acceptor com-
were prepared (Figs. 1, 2) and screened for their biological activity.
Although the majority of the compounds is known in either peer
reviewed literature or patents, synthesis and characterization was re-
ported only for compounds 3, 5, 7 and 11. Moreover, with exception of
compound 7, we used different synthetic routes as described below. We
also evaluated biological effects of commercially available derivatives.
Those included 1H-pyrazolo[3,4-d]pyrimidine analogues methylated on
the N1 nitrogen atom and/or furfurylamino moiety (14–18) as well as
thieno[2,3-d]pyrimidine (19), thieno[3,2-d]pyrimidine (20), furo[2,3-
d]pyrimidine (21, 22), and quinazoline (23) (Fig. 3).
partments and lyophilized. All experiments were done in duplicates.
Apparent permeability coefficients were calculated as Papp = (dQ/
dt)/(C0 × A), where dQ/dt is the rate of permeation of the test com-
pound across the cell monolayer, C0 is the donor compartment con-
centration at time t = 0 and A is the area of the cell monolayer. The
efflux ratio R was defined as the ratio PBA/PAB where PBA and /PAB are
the apparent permeability coefficients of the test compound from the
basal to apical and apical to basal sides of the cell monolayer, respec-
tively. Compounds with an efflux ratio of ≥ 2 were considered potential
Fig. 1. Synthesis of 2-H and 2-Cl imidazo[4,5-b]pyridines (1-deazapurine, scheme A) and imidazo[4,5-c]pyridines (3-deazapurine, scheme B) kinetin isosteres.
Reagents and conditions: (i) (a) furfurylamine, 10% AcOH in anh. DMF, 16 h, rt; (b) NaBH(OAc)
3
, 5% AcOH in anh. DMF, 16 h, rt; (ii) 4,6-dichloro-3-nitropyridin-2-
◦
3
amine or 2,4-dichloro-3-nitropyridine, EDIPA, DMSO, 16 h, rt; (iii) DMSO, 25% aq⋅NH , 16 h, 100 C; (iv) Na
2
S
2
O
4
, K
2
CO
3
, TBAHS, DCM/H
2
O, 16 h, rt; (v) triethyl
◦
orthoformate/DMSO (1:1), 16 h, 80 C; (vi) TFA/DCM (1:1), 90 min, rt; (vii) 2-chloro-3-nitropyridin-4-amine or 2,6-dichloro-3-nitropyridin-4-amine, EDIPA, DMSO,
1
6 h, rt.
7

B. Makov a´ et al.
Bioorganic & Medicinal Chemistry 33 (2021) 115993
Fig 2. Synthesis of 2-H and 2-Cl pyrrolo[2,3-d]pyrimidines (7-dezapurines, scheme A), pyrrolo[3,2-d]pyrimidines (9-deazapurines, scheme B), triazolo[4,5-d]py-
rimidine (8-azapurine, scheme C), pyrazolo[1,5-a]pyrimidine (scheme D), and pyrazolo[4,3-d]pyrimidine (scheme E) kinetin isosteres. Reagents and conditions: (i)
furfurylamine, TEA, 120–170 ◦C, MW irradiation, 40–50 min; (ii) furfurylamine, TEA, nPrOH, reflux, 4 h; (iii) NaN
, DMSO, rt (XI); (iv) 2-cyanoacetamide, EtONa,
3
◦
◦
◦
, ꢀ 78 ◦C, 1 h;
EtOH, 90 C; (v) ethyl formate, EtONa, EtOH, 95 C; (vi) SOCl
2
, DMF/CHCl
3
, reflux; vii) furfurylamine, EDIPA, iPrOH, 100 C, 4.5 h (XV); (viii) Na, NH
3
◦
◦
◦
(
ix) furfurylamine, K
2
CO
3
, MeCN/DMF, 90 C, 3 h; (x) P
2
S
5
, Py, reflux 3 h; (xi) MeI, K
2
CO
3
, DMF, 0 C to rt, 10 min; (xii) furfurylamine, 125 C, 8 h.
Fig. 3. Structures of commercially available kinetin isosteres.
solid-phase synthesis (Fig. 1). Aminomethyl resin with acid-labile
3
.1. Chemistry
benzaldehyde linker I was reductively aminated with furfurylamine to
give the resin II. To synthesize the chlorinated imidazo[4,5-b]pyridine
VII(Cl), the resin II was arylated with 4,6-dichloro-3-nitropyridin-2-
Imidazo[4,5-b]pyridines (3, 4) and imidazo[4,5-c]pyridines (5, 6)
were synthesized similarly to previously reported procedures41,42 using
8

B. Makov a´ et al.
Bioorganic & Medicinal Chemistry 33 (2021) 115993
amine and intermediate III was obtained. After reduction with sodium
dithionite, the resulting resin-bound intermediate VI(Cl) was cyclized
with triethyl orthoformate, which afforded the polymer-supported imi-
dazo[4,5-b]pyridine VII(Cl). Final cleavage using trifluoroacetic acid
yielded the desired product 4 (Fig. 1A).
buthionine sulfoximine (BSO), an inhibitor of glutathione de novo syn-
thesis. The effect on cell viability was evaluated using the resazurin
reduction assay.47 Protective activity was observed not only for com-
pound 6 but also for compound 5, an analogue with no 2-chloro sub-
stituent. Neither kinetin nor 2-chlorokinetin was active (data not
shown). Follow-up experiments showed that both compounds 5 and 6
have dose-dependent effects (Fig. 4A), but the effect of compound 6 was
much stronger. Compound 6 showed some protective activity at 10 or
The corresponding unsubstituted imidazo[4,5-b]pyridine 3 was
prepared analogically using 2,4-dichloro-3-nitropyridine in the aryla-
tion step. The resin IV was subjected to amination with a mixture of
aqueous ammonia and dimethylsulfoxide (DMSO) to obtain the resin V
followed by the reduction and cyclization of the resin-bound interme-
diate VI(H) with triethyl orthoformate. Cleavage from the resin yielded
crude imidazo[4,5-b]pyridine 3. Cleaved compounds 3 and 4 were pu-
rified using semipreparative reverse-phase chromatography. Imidazo
50
μ
M against an even higher dose of BSO (Fig. 4B).
The protective effect of the compounds was not mediated by
restoring levels of GSH, which were measured by quantifying fluores-
cent adducts with monobromobimane (not shown). We also addressed
the possibility that the effect of compound 6 could be explained by
restoration of FXN expression, but treating FA fibroblasts with the
[
4,5-c]pyridines 5 and 6 were prepared in a similar way using either 2-
chloro-3-nitropyridin-4-amine or 2,6-dichloro-3-nitropyridin-4-amine
to arylate the resin II. The rest of the reaction sequence was identical
to preparation of imidazo[4,5-b]pyridines 3 and 4 (Fig. 1B). An alter-
native synthesis of compounds 3 and 5 using the substitution of the
complete purine isostere ring with furfurylamine was reported.27
Unsubstituted pyrrolo[2,3-d]pyrimidine 7 and pyrrolo[3,2-d]py-
rimidine 9 isosteres were prepared by microwave-assisted synthesis. A
mixture of 4-chloro-7H-pyrrolo[2,3-d]pyrimidine or 4-chloro-5H-pyr-
rolo[3,2-d]pyrimidine with furfurylamine and triethylamine in meth-
anol was subjected to focused microwave irradiation (MW) using a
Discover SP reactor (Fig. 2A, B). Similar synthesis of compound 7 was
reported by Davoll in 196025 but our protocol using MW provided much
higher yields (60% vs. 26.2%)0.2-Cl Analogues (8, 10) were prepared by
conventionally heating starting material with furfurylamine, and trie-
thylamine in refluxing n-propanol (Fig. 2A and B, respectively).
Synthesis of 8-azakinetin (11) was started by azidation of benzyl
bromide with sodium azide in anhydrous DMSO to give XI, followed by
cyclization of the triazole ring of XII with 2-cyanoacetamide. Pyrimidine
ring closure was accomplished with ethyl formate and intermediate XIII
was subsequently chlorinated by thionyl chloride to obtain compound
XIV. Substitution of the chlorine atom with furfurylamine yielded
compound XV. Final debenzylation was achieved by treating XV with
sodium in liquid ammonia (Fig 2C). Pyrazolo[1,5-a]pyrimidine kinetin
analogue (12) was prepared in just one step by reacting 7-chloropyr-
azolo[1,5-a]pyrimidine with furfurylamine (Fig. 2D), whereas synthesis
of compound 13 required a three-step procedure. First, the hydroxy
group of 1H-pyrazolo[4,3-d]pyrimidin-7-ol was converted to thiol by
heating the starting compound with phosphorous pentasulfide accord-
ing to a published protocol,30 followed by methylation of XVI with
methyl iodide. Finally, substitution of the methylsulfanyl group of XVII
with furfurylamine furnished compound 13 (Fig. 2E). An alternative
synthesis of compound 11 from 4-amino-6-chloro-5-nitropyrimidin was
reported previously.26
compound at 10 and 50 M had no clear effects as assessed by western
μ
blotting and immunodetection (not shown). We also tested whether
compounds 5 and 6 could directly replace the primary function of FXN
in the stimulation of iron-sulfur cluster (ISC) biosynthesis.4 To this end,
we used functionally relevant reconstituted ISC machinery without
frataxin to test effects of these compounds on the rate of Fe-S cluster
8
synthesis.31 At 200
μM, compound 5 slightly activated Fe-S cluster
biosynthesis by a factor of 1.1, which is only about 0.5% of the activa-
tion afforded by FXN (see Appendix B – Fig. B.1). Such a small effect is
not consistent with the compound’s marked protective effect on BSO-
treated cells. Moreover, compound 6, which also protects against BSO
toxicity in FA fibroblasts, slightly inhibits rather than stimulates Fe-S
cluster synthesis. Therefore, compound 5 or 6 may influence second-
ary effects of the lack of FXN in FA fibroblasts rather than Fe-S cluster
biosynthesis. Yet another mechanism that could potentially explain
protective activity in FA models is iron chelation; however compound 6
had no effect in calcein quenching assay49 (data not shown).
3.3. The effect on oxidative stress injury in SH-SY5Y cells
Observations of the protective activities of compounds 5 and 6
inspired us to test activities of kinetin isosteres in other disease models
where glutathione depletion plays a role. The SH-SY5Y neuroblastoma
cell line is often used in neuroprotection research, as reviewed by Xicoy
5
0
et al. in 2017. One of the frequently used stressors in SH-SY5Y studies
is glutamate.51 At high concentrations, it causes massive oxidative stress
by inhibiting the cystine/glutamate Xc-antiporter which resulted in GSH
5
1
depletion and negative regulation of superoxide dismutase (SOD).
Resulting phenotypes are complex; features of necroptosis, apoptosis,
and ferroptosis may be present.52 The changes are probably not trig-
gered by glutamate excitotoxicity, as SH-SY5Y cells probably express
only one of the NMDA receptor subunits.53 However, some other
mechanisms have been suggested including, for example, Rac-NADPH-
oxidase-driven superoxide radical formation.51
3
.2. Protection of fibroblasts derived from Friedreich ataxia patients
We exposed SH-SY5Y cells from American Tissue Type Collection
(ATTC) to 160 mM glutamate for 24 h – following enhancement of their
against oxidative stress
differentiation and neuronal phenotype by 48 h treatment with 10 M
μ
Friedreich ataxia (FA) is a mitochondrial disease affecting the ner-
vous system and heart caused by a homozygotic mutation in the gene
encoding frataxin (FXN), a crucial protein for Fe-S cluster assembly. As a
consequence, several essential proteins for mitochondrial energy pro-
duction are affected, as reviewed by Bürk in 2017.43 The current therapy
only eases the symptoms without addressing the pathological changes of
mitochondria and cells. Established screening protocols for candidate
all-trans retinoic acid (ATRA) - and assessed the effect of kinetin iso-
steres on the proportion of dead cells by propidium iodide staining. The
iron chelator deferoxamine was used as a positive control. Pilot exper-
iments (not shown) allowed us to focus on several candidates for which
protective activity was suspected. Significant activity was confirmed for
compounds 6, 10 and 22 (Fig 5A). At 10 M concentration, they reduced
μ
death-indicating propidium iodide signals in cells by more than 10%.
However, their effect was lower than that of equimolar deferoxamine.
We also assessed the possibility that those compounds may reduce
oxidative stress, by quantifying superoxide radical production by dihy-
droethidium staining, and found that all the candidates had a dose-
dependent effect (Fig. 5B).
drugs exploit the increased sensitivity of FA fibroblasts to stress.4
4–46
In
a recent high-throughput screening campaign aimed at finding potential
new treatments for FA based on evaluation of protection against gluta-
thione depletion (Voller et al., in preparation), we identified 2-chloro-3-
deazakinetin as a highly active derivative (compound 6 herein). We
decided to further search the chemical space of kinetin derivatives for
other active compounds using the same protocol. GM04078 FA fibro-
blasts were pretreated with the test compounds and then exposed to L-
Protective activity of compound 6 against glutamate-induced stress
was further validated by testing in a SH-SY5Y cell line from European
Collection of Authenticated Cell Culture (ECCAC) with a different
9

B. Makov a´ et al.
Bioorganic & Medicinal Chemistry 33 (2021) 115993
Fig. 4. The protective effect of compounds 5 and 6 against L-buthionine sulfoximine (BSO) in FA-fibroblasts GM04078. Panel A: Pretreatment with serial con-
centrations of the compounds followed by exposure to 15 µM BSO. Panel B: Pretreatment with compound 6 at 10 or 50 µM followed by exposure to serial con-
centrations of BSO.
morphology and sensitivity to various toxins.54 Compound 6 at 10
μ
M
derivatives may have beneficial effects in models of neurodegenerative
decreased oxidative stress and had a protective activity, albeit the effects
were lower than those observed in the cells from ATTC. We also show
that the protective effect was partially mediated by inhibition of caspase
activation (see Appendix B Fig. B.2). A detailed study of kinetin’s
behaviour in neuron-like SH-SY5Y will be published separately
diseases in which kinetin (1) or 2-chlorokinetin (2) are active. One such
disease is familial dysautonomia, a genetic disorder due (as already
mentioned) to aberrant splicing of ELP1 transcripts.2
0,55
It is known that
patient-derived skin fibroblasts can be used for identification of com-
pounds that increase exon 20 inclusion, as in wild-type splicing.
Here, we evaluated the effect of kinetin isosteres at 50 µM on the
ratio of wild-type to mutant transcripts in the patient-derived skin
fibroblast cell line GM04663. mRNA was reverse-transcribed and ELP1
DNA was amplified by PCR. During the screening phase of the study, the
effect of the compounds was evaluated using standard agarose electro-
phoresis and GelRed staining (not shown). The activity of the hits was
then validated by capillary electrophoresis (Fig. 6, Appendix B Fig. B.3).
The results demonstrate that replacement in either the pyrimidine or
imidazole rings of the purine moiety can yield active compounds.
However, some purine ring modifications led to loss of activity. 1-deaza
derivatives of both kinetin and 2-chlorokinetin were active, but not their
(
Gonz a´ lez et al., submitted to Molecules journal). Compounds 1-deaza
and 3-deazakinetin showed a protective activity in undifferentiated
SH-SY5Y against proteotoxic stress induced by MG-132 treatment ac-
cording to the US patent 20170190704.24 The effect is supposed to be
mediated by PINK1 activation, however no supporting data are shown.
As PINK1 activation by kinetin requires its conversion into the corre-
sponding nucleoside monophosphate through the action of adenine
phosphoribosyl transferase followed by further enzymatic phosphory-
9
lation, the combination of purine ring modification with the presence of
2
-chloro substitent suggests that compound 6 and 10 act by a different
mechanism.
3
-deaza counterparts, suggesting that the N3 nitrogen is crucial for the
interaction with an unknown molecular target. The activity was retained
if either N7 (compounds 7 and 8) or N9-nitrogen (compounds 9 and 10)
were replaced by a carbon atom. The activity of 1-deaza, 7-deaza and 9-
3
.4. Effect of compounds on aberrant splicing of ELP1
We were also interested in the possibility that compound 6 and other
1
0

B. Makov a´ et al.
Bioorganic & Medicinal Chemistry 33 (2021) 115993
Fig. 5. Protective effects of selected kinetin isosters on glutamate-induced death (panel A) and oxidative stress (panel B) of SH-S5Y5 from ATCC collection. DFO –
deferoxamine.
Fig. 6. Effects of the selected isosteres on alternative splicing of mutant ELP1 transcript: wild-type (wt)/wild-type + mutant transcript ratios.
deaza derivatives is also reported in US patent 10676475,23 where a
dual color luminiscence-based reporter system was used for the effect
quantification. Activity can also be retained if the imidazole ring is
replaced by other 5-membered heterocycles like furan (compound 21)
and thiophene (compound 19 and 20). On the other hand, 8-aza
properties.56,57 The compounds selected for these assays excluded
methylated derivatives, except 22, where the substitution is on the pu-
rine ring.
The parallel artificial membrane permeability assay (PAMPA),58
which measures compounds’ permeation across hexadecane mem-
branes, was used to estimate the selected compounds’ absorption by
passive diffusion. For this, we assessed their rates of transport across
Caco-2 (human colon carcinoma cells with transport properties similar
to enterocytes) and MDCK-MDR1 (dog kidney cells expressing a human
efflux transporter) monolayers, which are reliable predictors of com-
(
compound 11) and 9-deaza-8-aza (compound 13) derivatives were
inactive. 7-deaza-8-azapurine derivatives substituted by a methyl group
6
at position 9, the N -nitrogen or furan ring (compounds 15–18) were
also inactive. The binding site of the unknown target can accommodate
even larger substituents, as demonstrated by the activity of the de-
rivatives where the imidazole is replaced by either furan ring substituted
by two methyl groups (compound 22) or benzene (compound 23). It has
already been reported that the 2-chloro derivative of kinetin is an even
more potent regulator of ELP1 splicing.20 Similarly, 2-chloro- de-
rivatives (where available) in our series of compounds retained the ac-
tivity of their unsubstituted parent compounds.
3
8
37
pounds’ ability to cross intestinal and blood-brain barriers (BBB)
,
respectively. Both apical-to-basal and basal-to-apical movements
through the barriers were studied, in efforts to identify compounds that
are subject to active efflux. Metabolization by human plasma and liver
microsomal fractions was used to evaluate the selected compounds’
stability. Finally, binding of the compounds to human plasma proteins
was studied. The results of all the assays are summarized in Fig. 7.
The compounds showed good plasma stability (83–100% remaining
after 2 h incubation) and most of them also had good microsomal sta-
bility (median 78% remaining after 1 h incubation). Degradation
exceeded 50% only in two cases: 20 and 22. Compound 22 was
particularly prone to degradation, possibly because of the dimethyl
3
.5. In vitro ADME assays
Kinetin (1), 2-chlorokinetin (2) and their isosteres 3–13 and 19–23
were subjected to biochemical assays that can be used to predict ab-
sorption,
distribution,
metabolism
and
excretion
(ADME)
1
1

B. Makov a´ et al.
Bioorganic & Medicinal Chemistry 33 (2021) 115993
Fig. 7. Overview of the results of ADME in vitro assays. Error bars are ranges of values (triplicates in biochemical assays, duplicates in cell-based permeability assays).
4
0
Standard errors calculated by the Delta method ) are only given for results of efflux ratio experiments.
1
2

B. Makov a´ et al.
Bioorganic & Medicinal Chemistry 33 (2021) 115993
substitution on the 5-membered ring of the purine moiety. Plasma
protein binding was highly variable (12.9–96.6%, median 76.7%). Re-
sults of the PAMPA assays (log Papp ꢀ 6.3 to ꢀ 4.8 cm/s, 1st quartile
that isosteric replacement of kinetin’s purine moiety is a fruitful strategy
for not only improving its known activities but also for exploring ther-
apeutic applications.
ꢀ
5.8 cm/s, median ꢀ 5.5 cm/s, 3rd quartile ꢀ 5.3 cm/s) indicate that
majority of compounds can readily cross biological barriers by passive
diffusion. Compounds with log Papp > ꢀ 6 cm/s are classified as having
high permeability. Results of the Caco-2 and MDCK-MDR1 assays (Papp
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
ꢀ
6
=
7.6–108.8 and 16.3–207.3 in the units of 10 cm/s, respectively) are
6
ꢀ
substantially higher than thresholds for high oral absorbance (5 × 10
ꢀ 6
cm/s) and BBB penetration (10 × 10 cm/s). Remarkably high values
were obtained for kinetin, suggesting possible facilitated transport
across the cell membrane. Comparably high absorption was observed for
Acknowledgement
5
, 7, 19 and 22 in Caco-2, but not MDCK-MDR1 cells. The reason for this
Funding: This work was supported by the Ministry of Education,
Youth and Sports of the Czech Republic (INTER-COST project [grant
number LTC18078]), European Structural and Investment Funds,
Operational Programme Research, Development and Education – proj-
ect “Preclinical Progression of New Organic Compounds with Targeted
Biological Activity” (Preclinprogress) [grant number CZ.02.1.01/0.0/
difference is unclear, although differences in the abundance and/or
types of membrane transporters present could be responsible. Differ-
ences in compounds’ efflux rates could also contribute to this effect, as
three of these compounds are among those with the highest efflux rates
from MDCK-MDR1 cells. However, compound fluxes are influenced by
complex interactions of multiple factors, which would require substan-
tial further elucidation.
0
.0/16_025/0007381] and the European Regional Development Fund –
project ENOCH [grant number CZ.02.1.01/0.0/0.0/16_019/0000868].
The collaboration of the Czech and French teams was realized within the
frame of COST Action CA15133 (FeSBioNet). The authors are grateful to
Kate ˇr ina Faksov a´ , Dita Jordov a´ and Marianna R o´ zsa for their excellent
technical assistance.
3
.6. Conclusion
Kinetin has been proposed as a potential therapeutic for diverse
conditions and promotion of healthy aging (reviewed by Kadlecov a´ et al.
2
59
in 2019 ). It even reached clinical trials for treating skin photoaging
Appendix A. Supplementary material
7
and familial dysautonomia . However, few previous peer reviewed
studies have assessed pharmacological activities of kinetin deriva-
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.bmc.2021.115993.
2
0–22
tives
, and none of them explored effects of modifications of the
purine ring system. Some purine isosters are revealed in patent litera-
ture, however.2
3,24
References
Here, we report the pharmacological activity of kinetin and 2-chlor-
okinetin purine isosteres, and show that some of them have activities not
reported for the parent compounds. For example, compound 6 was
originally identified as a compound that can protect FA fibroblasts
against oxidative stress in our previously mentioned, unpublished high-
throughput campaign. The same compound, together with several
others, also protected neuron-like cells against oxidative stress damage
after high dose glutamate treatment. It would be interesting to evaluate
those compounds in other models of mitochondrial and neurodegener-
ative diseases.
1
2
Voller J, Makov a´ B, Kadlecov ´a A, Gonzalez G, Strnad M. Plant hormone cytokinins
for modulating human aging and age-related diseases. In: Rattan S, Sharma R, eds.
Hormones in Ageing and Longevity. Healthy Ageing and Longevity6.
Kadlecov ´a A, Makov a´ B, Artal-Sanz M, Strnad M, Voller J. The plant hormone kinetin
in disease therapy and healthy aging. Age Res Rev. 2019;55:100958. https://doi.org/
10.1016/j.arr.2019.100958.
3
4
Rattan SI, Clark BF. Kinetin delays the onset of ageing characteristics in human
fibroblasts. Biochem Biophys Res Commun. 1994;201:665–672.
Barciszewski J, Siboska GE, Pedersen BO, Clark BFC, Rattan SIS. A mechanism for the
in vivo formation of N6-furfuryladenine, kinetin, as a secondary oxidative damage
product of DNA. FEBS Lett. 1997. https://doi.org/10.1016/S0014-5793(97)01037-5.
Slaugenhaupt SA, Mull J, Leyne M, et al. Rescue of a human mRNA splicing defect by
the plant cytokinin kinetin. Hum Mol Genet. 2004;13:429–436. https://doi.org/
10.1093/hmg/ddh046.
5
6
To our disappointment, compound 6 did not correct aberrant splicing
of ELP1 transcripts. It would be advantageous to have a compound with
such combined activity for treating familial dysautonomia, where stress
contributes to the neurodegeneration progression. Overall, activity in
the splicing assay was most tolerant of isosteric replacement in the pu-
rine core. This tolerance was remarkable in the case of the 5-membered
ring, suggesting that it faces toward the cytosolic part of the unknown
target’s binding site.
We anticipate that isosteric purine replacement could be beneficial
not only for improving kinetin’s activity toward some targets by
increasing interaction with the binding sites but also for increasing the
safety of the treatment by limiting affinity for off-targets. This may be
particularly important for a compound such as kinetin that has adenine
moiety that is present in many important regulatory and signaling
molecules. A lower propensity to interact with the purinome generally is
thus clearly a desirable property in most cases. Although molecular
targets of kinetin remain mostly elusive, it seems highly unlikely that the
wide range of reported activities is mediated by a single target or
pathway. Kinetin isosteres could therefore also be used for studies of the
independence of its various activities. Isosteric replacement also mod-
ulates pharmacokinetic properties. Although our data suggest that ma-
jority of kinetin isosteres can reach CNS after oral administration, they
also show that replacement of the purine ring system of kinetin may
result in reduced bioavailability, making development of derivatives
with the better activity in the target tissues more difficult. We conclude
Pros E, Fern ´a ndez-Rodríguez J, Benito L, et al. Modulation of aberrant NF1 pre-
mRNA splicing by kinetin treatment. Eur J Hum Genet. 2010;18:614–617. https://doi.
org/10.1038/ejhg.2009.212.
Axelrod FB, Liebes L, Gold-von Simson G, et al. Kinetin improves IKBKAP mRNA
splicing in patients with familial dysautonomia. Pediatr Res. 2011;70:480–483.
Yannai S, Zonszain J, Donyo M, Ast G. Combinatorial treatment increases IKAP levels
in human cells generated from Familial Dysautonomia patients. PLoS One. 2019;14.
https://doi.org/10.1371/journal.pone.0211602.
7
8
9
Hertz NT, Berthet A, Sos ML, et al. A neo-substrate that amplifies catalytic activity of
parkinson’s-disease- related kinase PINK1. Cell. 2013;154:737–747. https://doi.org/
1
0.1016/j.cell.2013.07.030.
1
0
1
Bowie LE, Maiuri T, Alpaugh M, et al. N6-Furfuryladenine is protective in
Huntington’s disease models by signaling huntingtin phosphorylation. PNAS. 2018;
1
15:E7081–E7090.
1
Werner T, Schmülling T. Cytokinin action in plant development. Curr Opin Plant Biol.
2
009;12:527–538.
12 Dole ˇz al K, Popa I, Hauserov ´a E, et al. Preparation, biological activity and endogenous
occurrence of N6-benzyladenosines. Bioorg Med Chem. 2007;15:3737–3747.
1
3
Colombo F, Falvella FS, De Cecco L, et al. Pharmacogenomics and analogues of the
antitumour agent N 6 -isopentenyladenosine. Int J Cancer. 2009;124:2179–2185.
Ottria R, Casati S, Maier JAM, Mariotti M, Ciuffreda P. Novel isopentenyladenosine
analogues: synthesis, characterization, and evaluation of antiproliferative activity on
bladder carcinoma cells. Nucleosides Nucleotides Nucleic Acids. 2009;28:736–751.
https://doi.org/10.1080/15257770903155550.
1
4
1
1
5
6
Ottria R, Casati S, Baldoli E, Maier JAM, Ciuffreda P. N6-Alkyladenosines: synthesis
and evaluation of in vitro anticancer activity. Bioorg Med Chem. 2010;18:8396–8402.
https://doi.org/10.1016/j.bmc.2010.09.030.
Amiable C, Pochet S, Padilla A, Labesse G, Kaminski PA. N6-substituted AMPs inhibit
mammalian deoxynucleotide N-hydrolase DNPH1. Fraidenraich D, ed. PLoS One.
2013;8. https://doi.org/10.1371/journal.pone.0080755.
1
3

B. Makov a´ et al.
Bioorganic & Medicinal Chemistry 33 (2021) 115993
1
7
Castiglioni S, Ciuffreda P, Casati S, Ottria R, A.M. Maier J. N6-Isopentenyladenosine
and its analogue n6-benzyladenosine induce cell cycle arrest and apoptosis in
bladder carcinoma T24 cells. Anticancer Agents Med Chem. 2013;13:672–678. https://
doi.org/10.2174/1871520611313040016.
39 Hammer O, Harper D, Ryan P. PAST: paleontological statistics software package for
education and data analysis. Palaeontol Electron. 2001;4:1–9.
40 Wu J. Statistical inference for tumor growth inhibition T/C ratio. J Biopharm Stat.
2010;20:954–964.
1
1
8
9
Oslovsky VE, Drenichev MS, Alexeev CS, Solyev PN, Esipov RS, Mikhailov SN.
Synthesis of Cytokinins via Enzymatic Arsenolysis of purine nucleosides. Curr Protoc
Nucleic Acid Chem. 2018;75. https://doi.org/10.1002/cpnc.61.
41 Lemrov a´ B, Smyslov ´a P, Popa I, et al. Directed solid-phase synthesis of trisubstituted
imidazo[4,5- c ]pyridines and imidazo[4,5- b ]pyridines. ACS Comb. Sci. 2014;16:
558–565.
′
Matu ˇs kov ´a V, Zatloukal M, Voller J, et al. New aromatic 6-substituted 2 -deoxy-9-
42 Vanda D, Soural M, Canale V, et al. Novel non-sulfonamide 5-HT 6 receptor partial
inverse agonist in a group of imidazo[4,5- b ]pyridines with cognition enhancing
properties. Eur J Med Chem. 2018;144:716–729.
(
β)-d-ribofuranosylpurine derivatives as potential plant growth regulators. Bioorg
Med Chem. 2020;28:115230. https://doi.org/10.1016/j.bmc.2019.115230.
Yoshida M, Kataoka N, Miyauchi K, et al. Rectifier of aberrant mRNA splicing
recovers tRNA modification in familial dysautonomia. PNAS. 2015;112:2764–2769.
Salani M, Urbina F, Brenner A, et al. Development of a screening platform to identify
small molecules that modify ELP1 Pre-mRNA splicing in familial dysautonomia.
SLAS Disc: Adv Sci Drug Disc. 2019;24:57–67.
2
2
0
1
43 Bürk K. Friedreich Ataxia: current status and future prospects. Cerebellum Ataxias.
2017;4:1–9. https://doi.org/10.1186/s40673-017-0062-x.
44 Richardson TE, Yang S-H, Wen Y, Simpkins JW. Estrogen protection in Friedreich’s
Ataxia skin fibroblasts. Endocrinology. 2011;152:2742–2749. https://doi.org/
10.1210/en.2011-0184.
2
2
H o¨ nig M, Plíhalov a´ L, Spíchal L, et al. New cytokinin derivatives possess UVA and
UVB photoprotective effect on human skin cells and prevent oxidative stress. Eur J
Med Chem. 2018;150:946–957. https://doi.org/10.1016/j.ejmech.2018.03.043.
Slaugenhaupt SA, Johnson G, Paquette WD, Zhang W, Marugan J. United States
Patent: 10676475 – Compounds for improving mRNA splicing. 2020.
45 Richardson TE, Simpkins JW. R- and S-Equol have equivalent cytoprotective effects
in Friedreich’s Ataxia. BMC Pharmacol Toxicol. 2012;13:12. https://doi.org/
10.1186/2050-6511-13-12.
46 Ferriero R, Manco G, Lamantea E, et al. Phenylbutyrate therapy for pyruvate
dehydrogenase complex deficiency and lactic acidosis. Sci Transl Med. 2013;5.
https://doi.org/10.1126/scitranslmed.3004986.
2
2
2
2
3
4
5
6
de Roulet, DeVita. United States Patent: 10167286 – Compositions and methods using
the same for treatment of neurodegenerative and mitochondrial disease. 2019.
Davoll J. 26. Pyrrolo[2,3-d]pyrimidines. J Chem Soc. 1960:131–138. https://doi.org/
47 Bonnier F, Keating ME, Wr o´ bel TP, et al. Cell viability assessment using the Alamar
blue assay: a comparison of 2D and 3D cell culture models. Toxicol In Vitro. 2015;29:
124–131. https://doi.org/10.1016/j.tiv.2014.09.014.
1
0.1039/JR9600000131.
Weiss R, Robins RK, Noell CW. Potential purine antagonists. XXIII. Synthesis of some
48 Pastore A, Puccio H. Frataxin: a protein in search for a function. J Neurochem. 2013;
126:43–52.
7
-substituted amino-v-triazolo(d)pyrimidines1. J Org Chem. 1960:765–770. https://
doi.org/10.1021/jo01075a023.
49 Ali A, Zhang Q, Dai J, Huang X. Calcein as a fluorescent iron chemosensor for the
determination of low molecular weight iron in biological fluids. Biometals. 2003;16:
285–293. https://doi.org/10.1023/A:1020642808437.
2
2
7
8
Rogozi n´ ska JH, Kroon C, Salemink CA. Influence of alterations in the purine ring on
biological activity of cytokinins. Phytochemistry. 1973;12:2087–2092.
Savelieva EM, Oslovsky VE, Karlov DS, et al. Cytokinin activity of N6-benzyladenine
derivatives assayed by interaction with the receptors in planta, in vitro, and in silico.
Phytochemistry. 2018;149:161–177.
50 Xicoy H, Wieringa B, Martens GJM. The SH-SY5Y cell line in Parkinson’s disease
research: a systematic review. Mol Neurodegener. 2017;12:1–11. https://doi.org/
10.1186/s13024-017-0149-0.
2
3
9
0
Kubiasov a´ K, Mik V, Nisler J, et al. Design, synthesis and perception of fluorescently
labeled isoprenoid cytokinins. Phytochemistry. 2018;150:1–11.
51 Kritis AA, Stamoula EG, Paniskaki KA, Vavilis TD. Researching glutamate – induced
cytotoxicity in different cell lines: a comparative/collective analysis/study. Front Cell
Neurosci. 2015;9:91. https://doi.org/10.3389/fncel.2015.00091.
52 Ito K, Eguchi Y, Imagawa Y, Akai S, Mochizuki H, Tsujimoto Y. MPP+ induces
necrostatin-1-and ferrostatin-1-sensitive necrotic death of neuronal SH-SY5Y cells.
Cell Death Discov. 2017;3:1–10. https://doi.org/10.1038/cddiscovery.2017.13.
53 Kulikov AV, Rzhaninova AA, Goldshtein DV, Boldyrev AA. Expression of NMDA
receptors in multipotent stromal cells of human adipose tissue under conditions of
retinoic acid-induced differentiation. Bull Exp Biol Med. 2007;144:626–629. https://
doi.org/10.1007/s10517-007-0390-6.
54 Nishida Y, Adati N, Ozawa R, et al. Identification and classification of genes
regulated by phosphatidylinositol 3-kinase- and TRKB-mediated signalling pathways
during neuronal differentiation in two subtypes of the human neuroblastoma cell line
SH-SY5Y. BMC Res Notes. 2008;1:95. https://doi.org/10.1186/1756-0500-1-95.
55 Slaugenhaupt SA, Blumenfeld A, Gill SP, et al. Tissue-specific expression of a splicing
mutation in the IKBKAP gene causes familial dysautonomia. Am J Hum Genet. 2001;
68:598–605. https://doi.org/10.1086/318810.
Robins RK, Furcht FW, Grauer AD, Jones JW. Potential purine antagonists. II.
Synthesis of some 7- and 5,7-substituted pyrazolo [4,3-d]pyrimidines. J Am Chem
Soc. 1956;78:2418–2422. https://doi.org/10.1021/ja01592a021.
Gervason S, Larkem D, Mansour AB, et al. Physiologically relevant reconstitution of
iron-sulfur cluster biosynthesis uncovers persulfide-processing functions of
ferredoxin-2 and frataxin. Nat Commun. 2019;10. https://doi.org/10.1038/s41467-
3
1
0
19-11470-9.
3
3
3
2
3
4
Stone JS, Shang J-L, Tomarev S. Expression of Prox1 defines regions of the avian
otocyst that give rise to sensory or neural cells. J Comp Neurol. 2003;460:487–502.
Carrasco RA, Stamm NB, Patel BKR. One-step cellular caspase-3/7 assay.
Biotechniques. 2003;34:1064–1067.
Kim MS, Seo JY, Oh J. Neuroprotective effect of halophyte Salicornia herbacea L. is
mediated by activation of heme oxygenase-1 in mouse hippocampal HT22 cells.
J Med Food. 2017;20:140–151.
3
3
5
6
Anderson SL, Qiu J, Rubin BY. EGCG corrects aberrant splicing of IKAP mRNA in
cells from patients with familial dysautonomia. Biochem Biophys Res Commun. 2003;
56 Di L, Kerns EH. Profiling drug-like properties in discovery research. Curr Opin Chem
Biol. 2003;7:402–408. https://doi.org/10.1016/S1367-5931(03)00055-3.
57 Zhang D, Luo G, Ding X, Lu C. Preclinical experimental models of drug metabolism
and disposition in drug discovery and development. Acta Pharmaceutica Sinica B.
2012;2:549–561. https://doi.org/10.1016/j.apsb.2012.10.004.
3
10:627–633.
Borkov a´ L, Frydrych I, Jakubcov ´a N, et al. Synthesis and biological evaluation of
triterpenoid thiazoles derived from betulonic acid, dihydrobetulonic acid, and
ursonic acid. Eur J Med Chem. 2020;185:111806. https://doi.org/10.1016/j.
ejmech.2019.111806.
58 Yu H, Wang Q, Sun Y, et al. A new PAMPA model proposed on the basis of a synthetic
phospholipid membrane. Hofmann A, ed. PLoS One. 2015;10. https://doi.org/
10.1371/journal.pone.0116502.
59 Wanitphakdeedecha R, Meeprathom W, Manuskiatti W. Efficacy and safety of 0.1%
kinetin cream in the treatment of photoaging skin. Indian J Dermatol Venereol Leprol.
2015;81:547. https://doi.org/10.4103/0378-6323.157446.
3
3
7
8
Irvine JD, Takahashi L, Lockhart K, et al. MDCK (Madin-Darby Canine Kidney) cells:
a tool for membrane permeability screening. J Pharm Sci. 1999;88:28–33.
Artursson P, Palm K, Luthman K. Caco-2 monolayers in experimental and theoretical
predictions of drug transport. Adv Drug Deliv Rev. 2001;46:27–43. https://doi.org/
1
0.1016/S0169-409X(00)00128-9.
1
4