Synthesis and photocytotoxic activity of [1,2,3]triazolo[4,5-h][1,6]naphthyridines and [1,3]oxazolo[5,4-h][1,6]naphthyridines

Article abstract of DOI:10.1016/j.ejmech.2018.10.071

[1,2,3]Triazolo[4,5-h][1,6]naphthyridines and [1,3]oxazolo[5,4-h][1,6]naphthyridines were synthesized with the aim to investigate their photocytotoxic activity. Upon irradiation, oxazolo-naphtapyridines induced light-dependent cell death at nanomolar/low

Full text of DOI:10.1016/j.ejmech.2018.10.071

Accepted Manuscript
Synthesis and photocytotoxic activity of [1,2,3]triazolo[4,5-h][1,6]naphthyridines and
[1,3]oxazolo[5,4-h][1,6]naphthyridines
Ilaria Frasson, Virginia Spanò, Simona Di Martino, Matteo Nadai, Filippo Doria,
Barbara Parrino, Anna Carbone, Stella Maria Cascioferro, Patrizia Diana, Girolamo
Cirrincione, Mauro Freccero, Paola Barraja, Sara N. Richter, Alessandra Montalbano
PII:
S0223-5234(18)30948-6
DOI:
https://doi.org/10.1016/j.ejmech.2018.10.071
EJMECH 10857
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 31 May 2018
Revised Date: 2 October 2018
Accepted Date: 31 October 2018
Please cite this article as: I. Frasson, V. Spanò, S. Di Martino, M. Nadai, F. Doria, B. Parrino, A.
Carbone, S.M. Cascioferro, P. Diana, G. Cirrincione, M. Freccero, P. Barraja, S.N. Richter, A.
Montalbano, Synthesis and photocytotoxic activity of [1,2,3]triazolo[4,5-h][1,6]naphthyridines and
[1,3]oxazolo[5,4-h][1,6]naphthyridines, European Journal of Medicinal Chemistry (2018), doi: https://
doi.org/10.1016/j.ejmech.2018.10.071.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Graphical abstract

ACCEPTED MANUSCRIPT
Synthesis and photocytotoxic activity of
[1,2,3]triazolo[4,5-h][1,6]naphthyridines and
[
1,3]oxazolo[5,4-h][1,6]naphthyridines
a,1
b,1
b
a
Ilaria Frasson , Virginia Spanò , Simona Di Martino , Matteo Nadai ,
c
b
b
b
Filippo Doria , Barbara Parrino , Anna Carbone , Stella Maria Cascioferro ,
b
b
c
b
Patrizia Diana , Girolamo Cirrincione , Mauro Freccero , Paola Barraja ,
a*
b*
Sara N. Richter , Alessandra Montalbano
a
Dipartimento di Medicina Molecolare, Università degli Studi di Padova, Via Gabelli 63, 35121 Padova, Italy; b Dipartimento di
Scienze e Tecnologie Biologiche Chimiche e Farmaceutiche (STEBICEF), Università degli Studi di Palermo, Via Archirafi 32,
9
1
0123 Palermo, Italy; c Dipartimento di Chimica, Università degli Studi di Pavia, Viale Taramelli 10, 27100 Pavia, Italy
.
The authors contributed equally to this work.
ABSTRACT
[1,2,3]Triazolo[4,5-h][1,6]naphthyridines
and
[1,3]oxazolo[5,4-
h][1,6]naphthyridines were synthesized with the aim to investigate their photocytotoxic activity.
Upon irradiation, oxazolo-naphtapyridines induced light-dependent cell death at nanomolar/low
micromolar concentrations (EC50 0.01–6.59 µM). The most photocytotoxic derivative showed
very high selectivity and photocytotoxicity indexes (SI=72-86, PTI>5000), along with a triplet
excited state with exceptionally long lifetime (18.0 µs) and high molar absorptivity (29781±180
-1
-1
M cm at λmax 315 nm). The light-induced production of ROS promptly induced an
unquenchable apoptotic process selectively in tumor cells, with mitochondrial and lysosomal
1

ACCEPTED MANUSCRIPT
involvement. Altogether, these results demonstrate that the most active compound acts as a
promising singlet oxygen sensitizer for biological applications.
KEYWORDS:
[1,2,3]Triazolo[4,5-h][1,6]naphthyridines,
[1,3]oxazolo[5,4-
h][1,6]naphthyridines, photochemiotherapy, photosensitizing agents, reactive oxygen species.
1
. Introduction
Naphthyridines (also known as pyridopyridines or diazanaphtalenes) are nitrogen heterocyclic
compounds widely used in agriculture, in analytical chemistry, in diagnostics and for the
treatment of different human diseases. [1]
Among the six possible isomers (1,5-, 1,6-, 1,7-, 1,8- 2,6- and 2,7-pyridopyridines),
[1,6]naphthyridines are endowed with interesting biological properties including antibacterial, [2,
3
] and anticonvulsant activity. [4] They have also been used as potent antiviral agents against
both the human cytomegalovirus (HCMV) [5-7] and human immunodeficiency virus-1 (HIV-1).
In fact, 8-hydroxy[1,6]naphthyridines 1 and 2 (Chart 1) proved to reduce HIV-1 integrase
activity and to inhibit 95% of the spread of viral infection in cell culture at 0.39 µM and 18.6 nM
respectively. [8,9] Further studies revealed that new derivatives of type 3, obtained from
modification of the decoration of the lead structure at the 8-position, were endowed with
significant cytotoxicity against a panel of cancer cell lines and inhibited selected oncogenic
kinases. [10] FGFR-1 and VEGFR-2 kinases inhibitory activities at low nanomolar level were
also exhibited by 7-acetamido[1,6]naphthyridines (4) synthetized by Thompson et al. [11,12]
2

ACCEPTED MANUSCRIPT
Chart 1. Structures of [1,6]naphthyridine derivatives 1-4.
Thus, considering the interesting properties showed by the [1,6]naphthyridine scaffold we
synthetized tricyclic systems of type 5-8 (Chart 2) in which the triazole or oxazole moieties are
annelated to the diazabicyclic portion. These new ring systems can be also regarded as hetero-
analogues of quinolines of type 9 and 10, potent photosensitizers previously reported by us,
containing a pyrrole and a pyrazole ring respectively. [13-16] In particular, one class of
pyrroloquinolones (9), upon light activation at a proper wavelength, showed potent growth
inhibitory activity at submicromolar level (0.5-7.2 µM). Cellular photocytotoxicity was related to
production of reactive oxygen species (ROS) with the involvement of both mitochondria and
lysosomes, alteration of the cell cycle profile and membrane lipid peroxidation, without DNA
photodamage. Condensation of the quinoline moiety to the pyrazole ring, led to pyrazolo[3,4-
h]quinolines (10); this structural modification led to the improvement of photocytotoxicity with
3

ACCEPTED MANUSCRIPT
growth inhibition of 50% of the cell population (GI50) reaching nanomolar level (0.04-14.52
µM). Derivatives 10 were able to generate high levels of superoxide anion and singlet oxygen,
suggesting involvement of both type I and type II oxygen-dependent photosensitization
mechanisms. [16]
More recently we have studied pyrrolo[3',2':6,7]cyclohepta[1,2-d]pyrimidines (11) and
pyrrolo[3’,2’:6,7]cyclohepta[1,2-b]pyridines (12) which showed very promising antitumor
properties. [17,18] Upon photoactivation with light of the proper wavelength, they showed
cytotoxic effect at 0.06–4.96 µM and 0.19–10.70 µM respectively, inducing apoptosis associated
with rapid and massive production of large amounts of ROS, mitochondrial morphology
modifications and activation of lysosomes. Unfortunately, although these two classes of
derivatives were endowed with potent photocytotoxic activity, they were not very selective for
tumor cell lines.
4

ACCEPTED MANUSCRIPT
.
Chart 2. Structures of [1,2,3]triazolo[4,5-h][1,6]naphthyridines (5, 7), [1,3]oxazolo[5,4-
h][1,6]naphthyridines (6, 8), pyrroloquinolines (9), pyrazoloquinolines (10),
pyrrolo[3',2':6,7]cyclohepta[1,2-d]pyrimidines (11) and pyrrolo[3’,2’:6,7]cyclohepta[1,2-
b]pyridines (12).
Thus, as part of our search for nitrogen heterocycles [19-37] and considering that the antitumor
effect of [1,6]naphthyridines has been widely investigated, but no examples of their
photochemotherapeutic activity have been reported so far, herein we report the synthesis of the
new ring systems [1,2,3]triazolo[4,5-h][1,6]naphthyridines 5a-l, 7a-c, and [1,3]oxazolo[5,4-
h][1,6]naphthyridines 6a-l, 8a-c with the aim of investigating their antiproliferative effect either
in the dark and under light irradiation.
5

ACCEPTED MANUSCRIPT
2
. Results and Discussion
2
.1. Chemistry
The key intermediates 7-amino-5-methyl-3-substituted-3H-[1,2,3]triazolo[4,5-b]pyridine-6-
carbaldehydes (23a-c) and 7-amino-5-methyl-2-substituted-[1,3]oxazolo[5,4-b]pyridine-6-
carbaldehydes (24a-c) were prepared starting from the suitable 5-amino-1-substituted-1H-1,2,3-
triazole-4-carbonitrile derivatives (17a-c) or 5-amino-2-substituted-[1,3]oxazole-4-carbonitrile
derivatives (18a-c) respectively (Scheme 1).
In particular, compounds 17a-c were obtained through a 1,3-dipolar cycloaddition of the proper
azido-derivatives (13a-c) with malononitrile (14) in the presence of sodium ethoxide. 5-Amino-
2
-substituted-[1,3]oxazole-4-carbonitrile derivatives 18a-c were prepared through the van
Leusen oxazole synthesis, according to the general method reported by Freeman and Kim. [38]
Thus a solution of commercially available aminomalononitrile tosylate (16) in 1-methyl-2-
pyrrolidone was reacted with the proper benzoyl chloride (15a-c). The desired compounds 17a-c
and 18a-c were subsequently refluxed with ethyl acetoacetate, in the presence of stannic chloride
as strong Lewis acid, allowing the isolation of 7-amino-5-methyl-3-substituted-3H-
[1,2,3]triazolo[4,5-b]pyridine-6-carboxylate derivatives (19a-c) and ethyl 7-amino-5-methyl-2-
substituted[1,3]oxazolo[5,4-b]pyridine-6-carboxylates (20a-c). These latter were subjected to
reduction of the ethoxycarbonyl group with lithium aluminium hydride in THF to give
compounds 21a-c and 22a-c, which were in turn submitted to oxidation reaction with Dess-
Martin periodinane in dichloromethane, leading to the corresponding key intermediate o-
aminoaldehydes 23a-c and 24a-c which were used without further purification.
6

ACCEPTED MANUSCRIPT
Scheme 1. Synthesis of key intermediates 23a-c, and 24a-c.
Reagents: (i) EtONa, ethanol, 0 °C-rt, 12 h, 60-70%; (ii) 1-methyl-2-pyrrolidinone, rt, 12 h, 51-
7
9%; (iii) ethyl acetoacetate, toluene, stannic chloride, reflux, 4 h, 45-62% (for derivatives 19a-
c), 45-65% (for derivatives 20a-c); (iv) LiAlH , THF, 0 °C-rt, 12 h, 58-92% (for derivatives 21a-
c), 55-80% (for derivatives 22a-c); (v) Dess-Martin periodinane, DCM, rt, 5 h.
4
7

ACCEPTED MANUSCRIPT
Scheme 2. Synthesis of naphthyridine derivatives 5a-l, 6a-l and naphthyridine-8-one derivatives
7
a-c, 8a-c.
Reagents: (i) KOH, ethanol and suitable ketone, or KOH and acetone (for derivatives 5d, 5h, 5l,
6
d, 6h, 6l) reflux, 12 h, 42-65% (for derivatives 5a-l), 40-58% (for derivatives 6a-l); (ii) K CO ,
2 3
ethanol, reflux, 12 h, 40-60% (for derivatives 7a-c), 45-62% (for derivatives 8a-c).
Cyclization to naphthyridine derivatives 5a-l and 6a-l was achieved through Friendlander
condensation of compounds 23a-c and 24a-c, respectively, with acetophenone or substituted
acetophenones, such as 2-methoxy-acetophenone, 4-methoxy-acetophenone and acetone, in basic
media (Table 1).
8

ACCEPTED MANUSCRIPT
Table 1. Naphthyridine derivatives 5a-l and 6a-l and spectrophotometric properties of selected
naphthyridine derivatives of type 5, 6.
1
Compd
R
Ph
R
λ
max (nm)
ε
max
ε
365 (nm)
6624
-
5a
Ph
337
-
9200
-
5b
Ph
2-OMe-Ph
4-OMe-Ph
Me
5c
Ph
365
310
-
8188
5200
-
8188
80
5d
Ph
5e
4-OMe-Ph
4-OMe-Ph
4-OMe-Ph
4-OMe-Ph
Ph
-
5
f
2-OMe-Ph
4-OMe-Ph
Me
354
351
302
334
352
-
5600
6400
3520
5200
5148
6352
2316
3560
5g
5h
5
i
3,4,5-(OMe)
3
3
3
3
-Ph
Ph
5
j
3,4,5-(OMe)
-Ph 2-OMe-Ph
-Ph 4-OMe-Ph
11600 11172
5
k
3,4,5-(OMe)
-
-
-
-
-
-
5
l
3,4,5-(OMe)
-Ph
Me
Ph
-
6a
Ph
Ph
Ph
Ph
-
6b
2-OMe-Ph
4-OMe-Ph
Me
360
-
11428 11224
6c
-
-
6d
338
10632
2396
9

ACCEPTED MANUSCRIPT
6
e
4-OMe-Ph
4-OMe-Ph
4-OMe-Ph
4-OMe-Ph
Ph
2-OMe-Ph
4-OMe-Ph
Me
325
-
21760 15384
6
f
-
-
-
-
6g
-
6h
350
330
327
-
10588
7588
6
i
3,4,5-(OMe)
3,4,5-(OMe)
3,4,5-(OMe)
3,4,5-(OMe)
3
3
3
3
-Ph
Ph
15780 11800
6
j
-Ph 2-OMe-Ph
-Ph 4-OMe-Ph
9976
-
9084
-
6
k
6
l
-Ph
Me
300
13940
5688
The absorption spectra were collected in 20 mM phosphate buffer (PB), pH 7.4 and ε were
calculated.
Alternatively, o-aminoaldehydes 23a-c and 24a-c were reacted with triethyl phosphonoacetate,
using potassium carbonate as base, allowing the isolation of naphthyridine-8-ones derivatives 7a-
c and 8a-c (Table 2).
Table 2. Naphthyridine-8-one derivatives 7a-c and 8a-c and spectrophotometric properties of
selected naphthyridine-8-one derivatives of type 7, 8.
10

ACCEPTED MANUSCRIPT
Compd
R
Ph
λmax (nm) ε max ε365 (nm)
7
a
325
325
-
8000
924
7
b
4-OMe-Ph
5600
3476
7
c
3,4,5-(OMe)
Ph
3
-Ph
-
-
176
-
8
a
300
-
21188
8
b
4-OMe-Ph
3,4,5-(OMe) -Ph
-
-
8
c
3
-
-
The absorption spectra were collected in 20 mM phosphate buffer (PB), pH 7.4
and ε were calculated.
2
.2 Biology
2
.2.1 Spectrophotometric properties
The essential requirement for a photosensitizer compound is the absorption of UV-Vis light. The
absorption spectra of compounds were collected in phosphate buffer (20 mM, pH 7.4) and ε
values were calculated in correspondence of both λmax and λ365 nm. All compounds absorbed UV-
Vis light with maximum peaks in the UV-A region (300-365 nm), a fundamental feature for a
photosensitizer (Tables 1, 2 and Figure S1).
2
.2.2 Antiproliferative assays
The antiproliferative activity of the tested compounds was evaluated on a panel of cultured
human tumor cell lines, representing the most common neoplasms: MCF7 (mammary gland
adenocarcinoma), HT-29 (colorectal adenocarcinoma) and A375 (malignant melanoma) cells.
Besides tumor cell lines, the compounds were tested also on the non-tumor cell line MCF10A
(mammary gland epithelial cells) to assess their selectivity towards malignancies. Moreover, the
11

ACCEPTED MANUSCRIPT
antiproliferative experiments were performed without irradiation by incubating cells with
compounds for 72 h to evaluate the UV light-independent cytotoxicity.
Prior to the administration of the compounds, all cell lines were tested for UV phototoxicity to
achieve the highest compounds’ activation in the absence of cellular cytotoxicity. We set up a
2
light dose of 1.5 J/cm (corresponding to 5 min of light exposure at 365 nm) at which the amount
of living cells in the UV-treated and UV-untreated samples was the same. Phototoxicity was
assessed by MTT test 72 h after UV irradiation in the presence of the compounds. [17,18]
Phototoxicity was evaluated only for compounds with adequate solubility in DMSO and reported
as the effective concentration able to kill 50% of the cell population (EC50, Table 3, Figure S2
for photocytotoxicity curves of most active compounds); toxicity in the absence of UV exposure
was indicated as the cytotoxic concentration able to kill 50% of the cell population (CC ).
50
12

ACCEPTED MANUSCRIPT
Table 3. Photocytotoxicity EC50 (µM) and cytotoxicity CC50 (µM) in human tumor and non-tumor cell lines measured 72 h post
administration of selected naphthyridine derivatives of type 5, 6 and naphthyridine-8-one derivatives of type 7, 8.
Compd
Human cell lines
A375
HT-29
MCF7
MCF10A
a
b
c
EC (µM)
CC (µM)
PTI
EC (µM)
CC (µM) PTI
EC (µM)
CC (µM) PTI
EC (µM)
CC (µM) PTI
5
0
50
50
50
50
50
50
50
5
a
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
n.d.
>5
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
n.d.
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
n.d.
>1.5
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
>50±0
0.099±0.009
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
> 50±0
> 50±0
> 50±0
> 50±0
>50±0
>50±0
>50±0
n.d.
>505
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
5
c
10.100±6.920
>50±0
0.099±0.017
>50±0
>508 33.958±12.600
5
d
n.d.
>46
>57
n.d.
n.d.
n.d.
n.d.
n.d.
>50±0
5
f
1.094±0.033
0.875±0.031
>50±0
>50±0
1.735±0.870
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
5
g
>29
n.d.
n.d.
n.d.
>50±0
5
h
>50±0
5
i
>50±0
>50±0
>50±0
5
j
>50±0
>50±0
>50±0
6
6
b
d
0.633±0.081
>50±0
>79
n.d.
0.197±0.067
6.587±0.158
>254
>8
1.323±0.371
>50±0
>38
n.d.
1.170±0.099
>50±0
>43
n.d.
>58
n.d.
6
e
0.011±0.009
>50±0
>4545 0.012±0.002
n.d. 4.283±0.339
>4167
>12
0.010±0.008
>50±0
>5000 0.864± 0.150
n.d. >50±0
6
h
13

ACCEPTED MANUSCRIPT
6
i
1.441±0.600
0.764±0.034
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>35
>65
n.d.
n.d.
n.d.
0.620±0.069
0.197±0.020
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>81
>254
n.d.
n.d.
n.d.
n.d.
5.944±0.301
3.529±0.226
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>8
4.380±0.099
1.860±0.086
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>50±0
>11
>27
n.d.
n.d.
n.d.
n.d.
6j
>14
n.d.
n.d.
n.d.
n.d.
6l
7
a
>50±0
>50±0
>50±0
>50±0
7
b
>50±0
>50±0
>50±0
>50±0
8
a
1.490±0.035
>34
>50±0
>50±0
>50±0
a
Data are expressed as mean ± SD of at least three different independent experiments. EC : Photocytotoxic concentration able
50
b
c
to kill 50% of the cell population. CC : Cytotoxic concentration able to kill 50% of the cell population. PTI: Phototoxic
5
0
index corresponds to CC /EC ratio; it indicates the effectiveness of the tested compounds upon photoactivation. n.d.: not
50
50
determined.
14

ACCEPTED MANUSCRIPT
The CC50/EC50 ratio was defined as the photoactivation toxicity index (PTI), i.e. the compounds’
ability to kill cells only upon light activation. Moreover, the selectivity index (SI), expressed as
EC of the photoactivated compound in the normal cell line/EC of the same compound in the
50
50
cancer cell line, was assessed (Table 4).
Notably, no cytotoxicity was observed in all the three cell lines by MTT analysis after 72 h of
incubation with up to 50 µM of the compounds in the absence of UV-light exposure (Table 3).
Although clear-cut structure-activity relationships (SAR) are not easily assessed, some
hypotheses can be formulated. Comparing the two classes of compounds, triazole derivatives 5
and 7 displayed lower cytotoxicity and tumor selectivity, compared to oxazolo-naphthyridines 6
and naphthyridine-8-ones 8. Triazolo-naphthyridines, of which only three derivatives (5c, 5f and
5
g) out of the eight could be screened, demonstrated photocytotoxicity with EC values at the
50
submicromolar/micromolar level (EC50 0.099–33.958 µM); in particular they showed high
cytotoxicity on the A375 cell line, with EC values from submicromolar (5g: EC 0.875 µM) to
5
0
50
micromolar (5c: EC 10.100 µM, Figure S2). The latter derivative 5c was the only one, among
5
0
the three, which proved to be photocytotoxic also against HT-29 (EC50 0.099 µM) and MCF7
cell line (EC50 33.958 µM) and displayed antiproliferative activity also against the non-tumor
cell line MCF10A. In contrast, triazolo-naphthyridine-8-ones 7a and 7b were not cytotoxic
against the three tested tumor cell lines, likely because of their low molar extinction coefficient
at 365 nm.
The best results were obtained with oxazolo-naphthyridines 6: upon irradiation, four out of seven
tested derivatives were able to induce UV dependent cell death at nanomolar/low micromolar
concentrations (EC50 0.010–6.587 µM). The most potent derivative was 6e (EC50 0.010–0.012
µM, Figure S2), which bears a 4-methoxy-phenyl group at position 2 and a phenyl ring at
15

ACCEPTED MANUSCRIPT
position 8. Replacement of the phenyl ring with a methyl group in position 8 caused a drastic
loss of activity (compare 6e with 6h). Moreover, removal of 4-methoxy-phenyl group and its
substitution with a phenyl (6b) or 3,4,5-trimethoxyphenyl group (6i, 6j) decreased phototoxicity
(
6b: EC 0.197–1.323 µM, 6i: EC 0.620–5.944 µM, 6j: EC 0.197–3.529 µM). In the 3,4,5-
50 50 50
trimethoxy series, the methyl group in position 8 caused again a complete loss of phototoxicity
6l).
(
Oxazolo-naphthyridine-8-one derivative 8a expressed photocytotoxicity at low micromolar level
only against the A375 cell line (EC50 1.490 µM).
Crucial requirements for photosensitizing agents are photocytotoxicity coupled to high ε and SI
values; the most photocytotoxic compound 6e fulfils these features. It induced UV-dependent
cell death at nanomolar concentrations, the same order of magnitude of
pyrrolo[3',2':6,7]cyclohepta[1,2-d]pyrimidines 11, ROS-inducer photosensitizers previously
reported by us (EC50 0.06–4.96 µM), [17] but in comparison with these, 6e is endowed with
much higher CC /EC ratio (PTI >5000 in MCF7 cell line) and selectivity for the tumor cell
5
0
50
lines (SI 72-86) (Table 4). Interestingly 6e is very active on the MCF7 cell line (breast cancer),
which is one of the most studied human breast cancer cell line and usually not very sensitive to
drugs. [39,40] When compared to other published PS, such as zinc Phthalocyanine, 6e
demonstrated no cytotoxicity in the absence of irradiation and much lower EC (nanomolar vs
5
0
micromolar range) than phthalocyanine on tumor cell lines. [41] Moreover, taking into
consideration compounds with similar absorption spectra and intracellular accumulation, i.e.
Chlorine e and derivatives, 6e was characterized by a higher cytotoxic effect (EC in the low
6
50
nanomolar range vs micromolar-high nanomolar range of Chlorine e
6
and derivatives) and a
16

ACCEPTED MANUSCRIPT
6
remarkably higher PTI (>5000 for 6e vs 466.6 of the most active Chlorine e derivatives). [42]
These are very promising results, which pave the way to further drug development.
Table 4. SI values of selected naphthyridine derivatives of type 5-6 and of naphthyridine-8-one
derivatives of type 7-8.
Compd
MCF10A/A375
MCF10A/HT-29
MCF10A/MCF7
5
a
c
n.d.
0.009
n.d.
46
n.d.
1.01
n.d.
>1
n.d.
0.003
n.d.
>1
5
5
d
5
f
5
g
57
29
>1
5
h
n.d.
n.d.
n.d.
2
n.d.
n.d.
n.d.
6
n.d.
n.d.
n.d.
0.884
n.d.
86
5
i
5
j
6
6
b
d
n.d.
79
8
6
e
72
6
h
n.d.
3
12
n.d.
0.737
0.527
n.d.
n.d.
n.d.
n.d.
6
i
7
6
j
2
9
6
l
n.d.
n.d.
n.d.
34
n.d.
n.d.
n.d.
n.d.
7
a
7
b
8
a
17

ACCEPTED MANUSCRIPT
SI: selectivity index corresponds to EC50 of photoactivated compound in the normal cell
line/EC50 of the same compound in the cancer cell line and indicates the degree of tumor
selectivity of the photoactivated compounds.
2
.2.3 Cell cytotoxicity localization: cytosolic ROS induction is required for
naphthyridines-induced cell death
To explore the cellular mechanisms underlying phototoxicity of the tested compounds, we chose
the most active and selective compound (6e) for further biological assays. To assess the cellular
compartments affected by the compound, we administrated 6e to MCF7 cells and monitored its
localization over 1 h time (Figure 1 and Figure S3), taking advantage of the residual fluorescence
(
Figure S4). In fact,6e exhibited a maximum in the fluorescent emission at 468 nm with a non-
negligible fluorescent quantum yield (ΦF = 0.2).
Figure 1. Subcellular localization of derivative 6e in MCF7 cells. After 10 min (A) the
compound (2 µM, blue signal) distributes within the cytoplasm, the same localization was
recorded after 30 min (B) and 1 h (C) post administration. Cellular nuclei were counterstained in
red.
Thanks to its fluorescent properties, 6e was detected using a confocal microscope (λ 405 nm,
ex
λem 415-460 nm) (Figure S4). As depicted in Figure 1, 6e rapidly entered MCF7 cells and widely
18

ACCEPTED MANUSCRIPT
distributed into the whole cytoplasm, with some punctate accumulation visible within 10 min of
exposure (Figure 1, panel A, 10 min, blue signal; Figure S3, panel 1). Thirty min - 1 h post
administration (Figure 1, panels B and C; Figure S3, panels 2 and 3), 6e subcellular localization
remained in the cytoplasm: areas of more intense signal were still visible but randomly
distributed, thus independent of intra-organelle accumulation, as reported for other
photosensitizer. [43] No nuclear signal was ever detected.
Taken together, these results led us to investigate a cytosolic DNA-unrelated mechanism of
cellular toxicity.
2
.2.4 Spectroscopic and kinetic characterization of the triplet excited state of 6e
To clarify the origin of the photocytotoxicity, we decided to provide a full characterization of the
newly synthesized naphthyridine derivative 6e as singlet oxygen sensitizer. Therefore, we
preliminarily detected its lowest triplet-excited state (T ) defining spectroscopic and kinetic
1
features. In fact, it is well known that the population of a triplet excited state and its lifetime are
both directly correlated to the efficiency of the generation of singlet oxygen. We set out to
compare these issues for 6e, the most photocytotoxic naphthyridine, with 5h, the least
photocytotoxic one. For solubility reasons, it was not possible to compare the two naphthyridines
in aqueous media. Consequently, 6e and 5h were irradiated in acetonitrile by nanosecond laser
flash photolysis (LFP), using a Nd:YAG laser, operating at 355 nm. Upon laser excitation at 355
nm, we recorded the transient absorption spectra in the range of 350-700 nm for both 6e and 5h
(
Figure 2a and 2b, respectively).
19

ACCEPTED MANUSCRIPT
Figure 2. Time-resolved difference absorption spectra (∆A) flashing argon-purged acetonitrile
-4
-4
solutions of 6e (1 x 10 M) (a) and 5h (5 x 10 M) (b), at 355 nm.
A maximum of absorbance was recorded for both naphthyridine 6e (λmax 650, Figure 2a) and 5h
(
λmax 530 nm, Figure 2b). The transient absorption profiles decayed monoexponentially in air-
equilibrated solutions with an identical lifetime of τ = 0.2 µs. In argon-saturated solutions, the
lifetime of the transient specie from 6e became bi-exponential (τ = 4.7 µs, 18%; τ = 18 µs,
2%, Figure 3). Such an oxygen-dependent lifetime indicates that at the LFP pulse end we are
very likely observing the triplet-state absorption of 6e. A similar behaviour was also recorded for
h. The intensity of the signal assigned to the triplet excited states (T ) of 6e monitored at 650
1
2
8
5
1
-2
nm (A = 4.3 X 10 ) was approximately 3 times more intense than that assigned to the T of 5h
1
-2
recorded at 530 nm (A = 1.4 X 10 ). Assuming that the molar extinction coefficients of T are
1
similar, all the data collected suggest that 5h does not efficiently generate a triplet-excited state.
Conversely, 6e exhibiting T
1
with an exceptionally long 18.0 µs lifetime and a high molar
-1
-1
absorptivity (29781±180 M cm at λ
315 nm) may act as an interesting singlet oxygen
max
sensitizer for biological applications.
20

ACCEPTED MANUSCRIPT
Figure 3. Triplet decay trace of 6e monitored at 630 nm in neat acetonitrile under argon
saturated condition.
1
2
.2.5 Generation and trapping of O by 9,10-anthracenedipropionic acid in aqueous
2
acetonitrile
To further evaluate the efficiency of 6e and 5h as O
1
2
photosensitizer, we investigated the photo-
1
oxidation of 9,10-anthracenedipropionic acid (ADPA), a water-soluble and efficient O
2
trap,
forming a colourless endoperoxide via a [4 + 2] Diels-Alder cycloaddition. The photo-oxidation
1
of ADPA by O was followed monitoring the absorption at 400 nm, which indicates the
2
consumption of the ADPA anthracene chromophore. [44] The experiments were carried out in
aqueous acetonitrile (acetonitrile: water = 3:2). We did not use higher water content to avoid
solubility issues for both 6e and 5h. The continuous photo-bleaching of the anthracene
absorption in the presence of both 6e and 5h, monitored for 1 h at 315 nm excitation wavelength,
1
indicated the photogeneration of O
2
(Figures S5 and S6). By-product formation was not detected
during irradiation both by UV absorption and HPLC analysis. We did not record any decrease in
the anthracene absorbance in the solutions without the sensitizer. We measured by ferrioxalate
actinometry the efficiency of the ADPA consumption (ΦR-315: quantum yield) irradiating at 315
21

ACCEPTED MANUSCRIPT
nm both naphthyridines. Surprisingly, the quantum yields for 6e and 5h were fairly low and very
1
similar with ΦR-315 = 0.020 ± 0.002. The quantum yield for ADPA consumption using 6e as O
2
sensitizer measured irradiating at 360 nm (ΦR- 360) was slightly higher (ΦR- 360 = 0.025 ± 0.003).
Buffering conditions in the same solvent did not substantially affect the quantum yield. We were
not able to reliably measure ΦR- 360 of 5h irradiated at 360 nm, as its absorbance was negligible at
that wavelength (Figure 4).
-4
-4
Figure 4. Absorbance comparison of 6e (10 M) and 5h (10 M) in acetonitrile: water = 3:2
solutions.
1
These data suggest that both 6e and 5h are capable to act as O photosensitizers with
2
comparable efficiency at 315 nm. The different photocytotoxicity described above, has to be
ascribed to their differential optical properties, as 6e exhibits much higher absorptivity at longer
wavelength (with a tail above 400 nm) than 5h (Figure 4).
22

ACCEPTED MANUSCRIPT
2
.2.5 Production of reactive oxygen species (ROS) upon UV irradiation
Generation of Reactive Oxygen Species (ROS) in cancer cells was next evaluated for compounds
e and 5h. We employed the cell-permeant dichlorofluorescein (DCF) probe, which is converted
to a highly fluorescent dye by ROS. Fluorescence in the presence/absence of UV irradiation
λ~365 nm) was quantitatively evaluated in cells treated with the selected compounds; non-
treated cells were used as negative controls. Cells treated with phorbol-12-myristate-13-acetate
6
(
(PMA) and light activated were used as a positive control of intracellular ROS generation.
[45,46] After 2 h from photoactivation, increasing concentrations of 6e (0.5-2 µM) showed a
steady increase in fluorescence, which reached 227% of the untreated control (100%) at the
highest concentration (Figure 5).
2
2
2
1
1
1
1
50
25
00
75
50
25
00
no scavenger
NAC 0.2 mM
NAC 0.5 mM
7
5
2
5
0
5
0
ND
6e 0.5 µM
6e 1 µM
6e 2 µM
5h 2 µM PMA 0.4 µM
Figure 5. ROS production upon photoactivation of 6e and 5h. The increase in intracellular ROS
levels was determined by a DCF-probe and was calculated as fluorescence increase compared to
untreated irradiated controls. Assays were performed on non-treated (control), 6e, 5h and PMA
treated cells, after 2 h from light exposure, in minimal medium (no scavenger) and in medium
23

ACCEPTED MANUSCRIPT
containing an antioxidant compound at various concentration (i.e. N-acetylcysteine, NAC 200
nM and 500 nM). Results from three independent experiments are shown.
Conversely, treatment with 5h or the administration of compounds without UV exposure did not
modify control levels of fluorescence (Figure S7). The positive control sample (PMA, 400 nM)
reached 120% of the untreated control 2 h after photoactivation (Figure 5). These data indicate
that generation of ROS in cells by 6e is massive and rises more rapidly than that mediated by the
ROS inductor PMA. Moreover, compared to other ROS-inducer photosensitizers previously
published by us, [17,18] 6e resulted more potent and active at much lower concentration. To
support the relevance of 6e-induced ROS, we pre-incubated cells with the ROS scavenger N-
acetylcysteine (NAC, 200-500 nM) and observed that NAC significantly reduced the generation
of reactive species, in a dose dependent manner (Figure 5). [47] These results indicate that 6e
promptly and massively produces ROS in cells. Statistical analysis (Two tailed Student’s t-test,
P-value < 0.01) indicated that ROS levels measured in 6e-treated cells were significant, while
5
h-treated cells were not significantly different from control samples. NAC-induced ROS
depletion was significant in all samples.
2
.2.6 Cytotoxicity mechanisms: cell death pathways
As production of copious amounts of ROS can trigger cell death by several different pathways,
we next examined which mechanism (apoptosis or necrosis) of cell death was induced by 6e.
The first feature that was taken into consideration was the distribution of MCF-7 cells in the cell
cycle phases 24 h after administration of compounds 6e and 5h. The exposure of MCF7 cells to
6
e upon UV-activation did not alter the distribution of the cells in the different phases of the cell
cycle (Figure S8), as assessed by flow cytometry analysis (PI staining, living cell population).
24

ACCEPTED MANUSCRIPT
Statistical analysis also confirmed that cell population distribution in treated cells was not
significantly altered in treated samples. This result suggested that cell death that could be related
to a rapid and harsh intracellular damage rather than block during the normal cell cycle. To shed
light on this aspect, a double staining (PI and Annexin V FITC) flow cytofluorimetry analysis
was performed to discriminate the early and late stages of apoptosis. The presence of necrosis
was also evaluated. MCF7 cells were treated with 6e and 5h at increasing concentrations (EC ,
2
5
EC50 and EC75 or EC50 respectively) and the whole cell population was analyzed. [48] At 4 h post
photoactivation, 44-60% (depending on the drug concentration) of cells treated with compound
6
e showed features of non-living cells and were classified as apoptotic (34-50%), or necrotic
(10%) cells (Figure 6).
25

ACCEPTED MANUSCRIPT
Figure 6. Cell population analysis. MCF7 cells were treated with 6e at various concentrations
(
EC , EC and EC ) and 5h at EC and double stained with 7AAD and PE Annexin V 4 h
25 50 75 50
after photoirradiation. Reported values represent the mean of three independent experiments.
In contrast, control cells and cells treated with 5h were characterized by a large fraction of living
cells (95% and 90%, respectively) and displayed only low amounts of apoptotic and necrotic
cells (4% and 7%, respectively). These results evidenced that the administration and subsequent
photoactivation of active compound 6e caused a massive and very rapid cell death in tumor cells.
Student T-test (two tailed, P<0.01) was performed and assessed that data obtained for 6e-treated
cells were statistically significant, while 5h-treated cells were comparable to control non-treated
cells. As compared to other photoactivated compounds, oxazolo-derivatives pushed the majority
of cell population into an irreversible apoptotic process in a shorter time lapse. [17,18] The
triggering of apoptosis is a favourable feature for a chemotherapic drug, since, unlike necrosis,
the apoptotic process does not cause an inflammatory reaction.
2
.2.7 Cytotoxicity mechanisms: mitochondrial and lysosomal-dependent ROS response
Different intracellular signaling pathways, including mitochondrial dysfunction and lysosome
activation, trigger and carry on the apoptotic process; we thus proceeded examining both
mitochondrial and lysosomal involvement in 6e-treated photoactivated cells. The induction of
mitochondrial dysfunction was assessed with a mitochondrial membrane fluorescent probe,
which allows to monitor morphology and mitochondria distribution within the cells. Untreated
and treated cells with/without UV irradiation, were evaluated by confocal imaging. Control cells
(untreated and non-irradiated) displayed the typical pattern of mitochondrial organization
26

ACCEPTED MANUSCRIPT
reported for the MCF7 cell line, characterized by a more fragmented network compared to non-
metastatic breast cancer cells (Figure 7, panel A). [49-51]
Figure 7. Mitochondrial morphology modifications induced by light-activated 6e and 5h.
Mitochondrial membranes were visualized with a membrane potential-specific red fluorescent
probe. A: non-treated and non-irradiated cells; B: non-treated and irradiated cells and C: PMA-
treated cells (400 nM). D: 6e, 2 µM, 20 min after UV exposure; E: 6e, 2 µM, 2 h after UV
exposure; F: 5h, 2 µM, 2 h after UV exposure. Well-defined mitochondrial network morphology
was stained in non-treated cells, irradiated and non-irradiated, as well as in 5h treated and
irradiated cells (A, B and C). Cells treated with 6e, following photoactivation, underwent
significant mitochondrial network impairment and organelles swelling, already visible after 20
min (D and E). PMA was used as ROS-inducer positive control: a condensed mitochondrial
network was visible 6 h post treatment (F).
27

ACCEPTED MANUSCRIPT
Cells irradiated in the absence of ROS-generating compounds as well as cells treated with the
compounds and non-irradiated were characterized by a similar mitochondrial network (Figure 7
and Figure S9, panel B; Figure S10, panels B-C). Notably, UV-activated 6e caused an early
remodeling of mitochondrial network, characterized by perinuclear clustering (Figure 7 and S9,
panel D), followed by organelles fusion and mild swelling (Figure 7 and S9, panel E). This effect
burst soon after UV exposure (20 min) and was even more intense 2 h after compound
photoactivation. Conversely, 5h in the same conditions did not induce any detectable
mitochondrial perturbation (Figure 7, panel F). Treatment with PMA induced moderate
mitochondria fusion after 6 h of treatment (Figure 7, panel C).
To further characterize the UV-activated 6e induced cell death, the lysosomal involvement in the
apoptotic process was also evaluated. A fluorescent probe able to stain acidic subcellular
compartments, such as lysosomes and endosomes, was used. Cells treated with 6e that also
underwent UV irradiation showed considerable lysosomes activation, visible as bright green dots
within the cytoplasm. Lysosomes activation was detectable 20 min after compound activation
(
green signal, Figure 8, panels D-E).
28