DNA damage and biological effects induced by photosensitization with new N1-unsubstituted furo[2,3-h]quinolin-2(1H)-ones

Article abstract of DOI:10.1016/S0968-0896(02)00145-1

New furoquinolinones unsubstituted at the N₁ position were prepared and their photobiological activities were studied in comparison with 4,6,8,9-tetramethylfuro[2,3-h]quinolin-2(1H)-one (HFQ) and 8-MOP. The anti-proliferative activity of furoqu

Full text of DOI:10.1016/S0968-0896(02)00145-1

Bioorganic & Medicinal Chemistry10 (2002) 2835–2844
DNA Damage and Biological Effects Induced by
Photosensitization with New N1-Unsubstituted
Furo[2,3-h]quinolin-2(1H)-ones
Cristina Marzano, Adriana Chilin, Franco Bordin,
Francarosa Baccichetti and Adriano Guiotto*
Department of Pharmaceutical Sciences, University of Padua, Via Francesco Marzolo 5,
I-35131 Padova, Italy
Received 18 March 2002; accepted 7 May2002
Abstract—New furoquinolinones unsubstituted at the N1 position were prepared and their photobiological activities were studied in
comparison with 4,6,8,9-tetramethylfuro[2,3-h]quinolin-2(1H)-one (HFQ) and 8-MOP. The anti-proliferative activityof furo-
quinolinones 3a–f was tested upon UVA irradiation in mammalian cells, studying DNA synthesis and clonal growth capacity, and
in micro-organisms, evaluating T2 infectivity. Almost all compounds appeared to be more active than 8-MOP, and free of any
mutagenic activityand skin phototoxicity. Among them, compound 3b was the most effective one. Similarlyto HFQ, compound 3b
appeared to be veryactive also in DNA damaging, forming monoadducts and DPC L=0, but no ISC and DPCL>0, both responsible
for furocoumarin genotoxicityand phototoxicity. Moreover, Ehrlich ascites cells, photoinactivated bythe new furoquinolinone 3b
and injected into recipient mice, proved to be capable of inducing protection against a successive challenge performed with the same
tumor cells. For all these features, 3b seemed to be a new promising potential drug for PUVA therapyand photopheresis. # 2002
Elsevier Science Ltd. All rights reserved.
Introduction
With the aim to obtain more effective and less toxic
drugs for PUVA therapyand photopheresis, we have
prepared and studied various angular compounds,
unable to induce ISC for geometrical reasons, such as
Linear furocoumarins (psoralens) are active sensitizers
widelyused in PUVA (Psoralen plus UVA) therapyfor
treatment of several skin diseases,1 and in photo-
pheresis, for preventing rejection in organ transplanta-
tion, to treat T-cell lymphoma and other autoimmune
diseases.2 The compounds currentlyused are 8-methoxy-
psoralen (8-MOP), 5-methoxypsoralen (5-MOP) or
4,50,8-trimethylpsoralen (TMP) (Fig. 1). Nevertheless,
their therapeutic use shows various toxic effects, such as
severe skin erythemas,1 and genotoxicity, with induction
of point mutations3,4 and chromosomal aberrations.5 In
addition, PUVA therapyis associated to a certain risk
of skin cancer.6 Various authors attributed these dan-
gerous effects to the severe damage induced into DNA
angelicin derivatives,9 and recentlysome angelicin
10ꢀ12
bioisosters, namelyfuroquinolinones.
Among
them, 1,4,6,8-tetramethylfuro[2,3-h]quinolin-2(1H)-one
(FQ), and 4,6,8,9-tetramethylfuro[2,3-h]quinolin-2(1H)-
one (HFQ) are the more interesting (Fig. 1). Both com-
pounds are characterized bya verystrong photo-
sensitizing activity. However, FQ also shows an evident
11
skin phototoxicityand a marked clastogenic activity,
while these side effects are reduced or absent with
HFQ.13,14 Both derivatives induce large amounts of MA
and DPC, without forming ISC. Moreover, the forma-
tion and the structures of DPC seemed to be different
for the two compounds. In fact, DPCFQ are real
bifunctional adducts, in which the FQ molecule is the
covalent linkage between DNA and proteins, and so,
theyare defined DPC at length greater than zero
(DPCL>0).13 Similarlyto ISC, these DPC are formed by
a two-step reaction, first forming a furan-side MA
between FQ and DNA pyrimidine base, and then react-
ing with proteins to yield DPC. Applying the well-known
7
byfurocoumarin sensitization. In fact psoralens form
covalent monoadducts (MA) with pyrimidine bases, or
diadducts: covalent inter-strands cross-links (ISC) with
two pyrimidine bases on the opposite DNA strands,7
and covalent DNA-protein cross-links (DPC).8
*Corresponding author. Tel.: +39-049-827-5351; fax: +39-049-827-
5366; e-mail: adriano.guiotto@unipd.it
0968-0896/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(02)00145-1

2836
C. Marzano et al. / Bioorg. Med. Chem. 10 (2002) 2835–2844
Figure 1. Structure of 8-methoxypsoralen (8-MOP), 5-methoxypsoralen (5-MOP), 4,50,8-trimethylpsoralen (TMP), 1,4,6,8-tetramethylfuro[2,3-
h]quinolin-2(1H)-one (FQ), and 4,6,8,9-tetramethylfuro[2,3-h]quinolin-2(1H)-one (HFQ).
Results and Discussion
double irradiation technique to FQ, we observed that
DPCL>0 are responsible for its veryhigh clastogenic
activity.11
Chemistry
The title compounds were synthesized according to the
pathways summarized in Schemes 1–3, which differed
for the cyclization conditions used to build the furan
ring. Starting products were methyl-7-hydroxyquinolin-
2-ones 1a–b,15 which were condensed with chloro-
acetone or chlorobutanone or propargyl chloride to give
the corresponding 7-O-oxoalkyl or 7-O-alkynyl ethers
2a–d. Compounds 2a–c were submitted to cyclization in
In contrast, HFQ induces DPC without participating in
the reaction and MAHFQ proved to be unconvertible
into diadducts.13 These lesions, in which DNA and
proteins are directlylinked together, resemble DPC
induced byUVC ( DPC at zero length, DPCL=0). Since
HFQ induces MA and DPCL=0 to a comparable extent
with FQ, the different nature of DPC is probably
responsible for the different toxicities of these two furo-
quinolinones.
Since the absence of N1-substituent seems to make
MAHFQ unable to convert into diadducts, due to the
tautomerism pyridone–pyridine,14 we synthesized a ser-
ies of furoquinolinones unsubstituted at N1 position.
We also studied some aspects of their potential thera-
peutic and toxic activityon various biological substrates,
using HFQ and 8-MOP as reference compounds.
Scheme 2. Reagents and conditions: (i) CHꢁCCH2Cl, acetone,
K2CO3, reflux, 40%; (ii) N,N-diethylaniline, CsF, 210 ꢂC, 10%.
Scheme 1. Reagents and conditions: (i) CH3COCH(R0)Cl, acetone,
K2CO3, reflux, 45–68%; (ii) toluene, methanesulfonic acid, reflux,
42–53%.
Scheme 3. Reagents and conditions: (i) concd H2SO4, 50 ꢂC, 32% (5a)
and 36% (5b).

C. Marzano et al. / Bioorg. Med. Chem. 10 (2002) 2835–2844
Table 1. Biological activityof the new furoquinolinones
2837
Compound
Antiproliferative activityMutations in
Skin
phototoxicitye
E. coli TM9d
DNA synthesis in
Ehrlich cellsa
Clonal growth of
HeLa cellsb
T2 phage
inactivationc
8-MOP
HFQ (3e)
3.7ꢃ0.20
0.40ꢃ0.02
7.74ꢃ1.20
0.80ꢃ0.60
1.43ꢃ0.20
2.71ꢃ0.18
3.43ꢃ0.51
0.790ꢃ0.020
0.025ꢃ0.004
0.350ꢃ0.020
0.018ꢃ0.005
0.140ꢃ0.010
1.450ꢃ0.070
0.058ꢃ0.020
0.80ꢃ0.110
0.04ꢃ0.001
0.25ꢃ0.040
0.08ꢃ0.002
0.40ꢃ0.060
1.65ꢃ0.200
0.47ꢃ0.070
2.96ꢃ0.12
0.04ꢃ0.20
0.23ꢃ0.03
0.23ꢃ0.01
0.16ꢃ0.01
0.44ꢃ0.05
0.31ꢃ0.03
+++
ꢃ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
3a
3b
3c
3d
3f
a
ID50ꢃSD: UVA dose which reduces to 50% DNA synthesis in Ehrlich cells when delivered in the presence of 20 mM drug concentration.
b ID50ꢃSD: UVA dose which reduces to 50% clonal growth of HeLa cells when delivered in the presence of 5 mM drug concentration.
c ID37ꢃSD: UVA dose which reduces to 37% the surviving fraction when delivered in the presence of 2 mM drug concentration.
d Revertants per 106 survivors scored at 0.37 surviving fraction.
e Phototoxicityon guinea-pig skin at 0.5 mM cm ꢀ1 drug concentration and 20 kJ mꢀ2 UVA dose. Symbols: ꢀno erythema,ꢃlight erythema, +++
strong erythema.
toluene in the presence of methanesulfonic acid, yielding
the corresponding methylfuro[2,3-h]quinolin-2(1H)-ones
3a–c, while compound 2d was submitted to Claisen
rearrangement in the presence of CsF,16 affording 4,6,8-
trimethylfuro[2,3-h]quinolin-2(1H)-one (3d).
Table 1, all derivatives showed a significant activity,
higher than 8-MOP in most cases; compound 3b
appeared to be the best one, with data comparable to
those of HFQ.
The anti-proliferative activitywas also tested in micro-
organisms, byexamining T2 phage inactivation, a very
useful system to detect DNA damage.17 As shown in
Table 1, again compound 3b was the most effective.
Since methylfuroquinolin-2-one derivatives were already
been synthetized using concentrated sulfuric acid to
cyclize the furan ring,10 we tried to use the same experi-
mental conditions for the ethers 2a–c. In concentrated
sulfuric acid onlycompound 2c gave the desired fur-
oquinolin-2-one 3c, but with lower yield as compared
with the method above mentioned, while compounds 2a
and b gave no cyclization products. Compound 2e in
concentrated sulfuric acid at room temperature gave
The mutagenic abilitywas assayed in Escherichia coli
TM9, a bacterial strain defective in DNA repair and very
sensitive to mutagens (Table 1). All new furo-
quinolinones seemed one order of magnitude less effec-
tive than 8-MOP, classified as a mutagen for the high
numbers of revertants induced.
only4,6,8,9-tetramꢂetlhfuyro[2,3-
h]quinolin-2(1H)-one
(3e),10 while at 50 C gave also a comparable amount of
8-hydroxymethyl-4,6,9-trimethylfuro[2,3-h]quinolin-2(1H)-
one (3f), deriving from the oxidation of the methyl
group in the 8 position. The same result was also
obtained directlytreating compound 3e with con-
centrated sulfuric acid at 50 ꢂC, obtaining derivative 3f,
together with manyco-products, especiallyif the reac-
tion time was prolonged until the starting product dis-
appeared. Since there were two methyl groups in the
furan ring, 13C NMR measurements were performed to
demonstrate that the oxidation occurred onlyat the 8
position. 13C NMR spectra, obtained byretaining the
proton–carbon couplings through gated decoupling
experiments, clearlyshowed that the C-9 molteplicity
was the same for compounds 3e and 3f, while for com-
pound 3e the C-8 was coupled with three protons
(quartet at 150.03 ppm), and for compound 3f the C-8
was coupled with two protons (triplet at 149.73 ppm), so
the hydroxymethyl group was in the 8 position.
Finally, skin phototoxic properties was determined on
guinea pig (Table 1); all derivatives appeared to be
completelyunable to form erythemas.
Experiments carried out with compound 3b. Since the
most active and interesting derivative seemed to be
compound 3b, its photobiological properties were care-
fullyinvestigated. Also in these experiments HFQ and
8-MOP were chosen as references.
The abilityto induce DNA damage was tested in vitro
byexposing 3b to increasing UVA light doses and
recording the corresponding UV spectra (Fig. 2). As
UVA dose increased, a marked decrease of optical den-
sityat 250 nm was observed, but no modifications at
335 nm, and onlymoderate increases at 270–320 and
350–400 nm, without wavelength shifts of absorption
picks. These results suggested that 3b probablybroke
down forming manydegradation products, but not one
defined compound.
Biology
Experiments for a preliminary screening. The anti-
proliferative activityof furoquinolinones 3a–f was tes-
ted upon UVA irradiation in mammalian cells, studying
the effect on DNA synthesis in Ehrlich cells, that is a
short term effect, and the clonal growth capacityof
HeLa cells, that is a long-term effect. As summarized in
As compound 3b showed high emission intensities in the
characteristic coumarin region (300–380 nm) of the UV
spectrum, its abilityto photoreact with DNA was
assayed by fluorescence determinations (Table 2). No
significant modifications were detected in fluorescence
spectrum after exposure of 3b aqueous solution to UVA

2838
C. Marzano et al. / Bioorg. Med. Chem. 10 (2002) 2835–2844
light, as well as for aqueous solutions of calf thymus
DNA, which gave always a fluorescence peak at 260 nm.
After exposure of aqueous solution of DNA containing
3b to increasing UVA doses, a second fluorescence peak
at 366 nm appeared in the spectrum, with intensity
increasing with UVA doses. This was attributed to the
formation of fluorescent furan-side monoadducts into
DNA.
The formation of ISC and DPC was studied in HeLa
18
cells byalkaline elution.
As shown in Table 3, furo-
quinolinone 3b appeared to be completelyunable to
induce ISC, like HFQ, due to the angular molecular
structure, which prevents the formation of such
lesions.14 This finding was consistent with the low
mutagenic activityshown in E. coli TM9. On the con-
trary, 8-MOP formed high levels of ISC, as expected.
Moreover, compound 3b induced noticeable amounts of
DPC, larger than 8-MOP and HFQ. This evidence can
explain the strong antiproliferative activityof 3b.
The formation of double-strand breaks (DSB) was also
studied in HeLa cells using neutral elution18 (Table 3).
When cells were submitted to DSB detection just after
sensitization, always negative results were scored. DNA
fragmentation was observed onlywhen cells were incu-
bated for a suitable time in growth medium after sensi-
tization. Even in these experimental conditions,
compound 3b and HFQ induced onlymoderate DSB
amounts, while 8-MOP yielded a very large DNA frag-
mentation. This result meant that DSB were probably
due to an enzymatic repair of the lesions induced into
DNA and not onlyto a simple photochemical event.
Figure 2. UV spectrum of compound 3b exposed to increasing UVA
light doses: 0 (thick line), 0.8, 1.6, 3.2, 6.4, 12.8 and 25.6 kJ m_2
.
All these data suggested that the biological activityof 3b
could be correlated to the formation of MA and DPC.
Table 2. Formation of furan side MA with calf thymus DNA in vitro
UVA dose
(kJ mꢀ2
l
max emission
(nm)
% Fluorescence
)
To investigate the mechanism and the nature of DPC
induced by 3b, some experiments based on the double
irradiation protocol were carried out. Bymeans of this
technique, MA or diadducts were subsequentlyinduced,
as theyrequired the absorption of one or two photons,
respectively. Moreover, this method allowed to dis-
criminate between DPCL>0 (requiring a biphotonic
reaction) and DPCL=0 (formed bya monophotonic
one).13 This procedure thus represented a practical way
of studying the biological consequences of these differ-
ent DNA lesions. Compound 3b could onlyinduce
DPCL=0 into DNA, since it was unable to form ISC;
therefore, evaluating its antiproliferative activitybefore
and after the second irradiation step, we could get
information on its mechanism of DPC induction. The
results obtained studying the clonal growth capacity of
HeLa cells, using the double irradiation method, are
shown in Figure 3. Panel A reports the data related to
compound 3b and HFQ; when the unbound drug was
not removed after the first irradiation step, the inhibi-
tion curves decreased according to the increase of UVA
doses delivered during the second step. On the contrary,
when cells were washed after the first irradiation step,
the second one did not reduce significantlythe surviv-
ing fraction. These results suggested that both com-
pounds formed MA incapable of further reacting and
3b alonea
0
10
10
0
2.5
5
400
400
260
350–450
366
366
60ꢃ2.1
51ꢃ2.3
49ꢃ2.1
4ꢃ1.5
DNA alonea
3b plus DNAb
33ꢃ0.8
43ꢃ1.1
48ꢃ1.1
10
366
aSolutions were exposed to UVA light and then the fluorescence was
determined.
bSolutions were exposed to UVA light, DNA was precipitated and
dissolved in the initial volume of water and the fluorescence was then
determined.
Table 3. DNA damage in HeLa cells
Compound
ISCa
DPCa
DSBb
8-MOP
HFQ
3b
4.5ꢃ0.51
0.01ꢃ0.001
0.01ꢃ0.001
2.05ꢃ0.12
5.25ꢃ1.20
7.46ꢃ0.91
14.1ꢃ0.80
3.1ꢃ0.10
5.5ꢃ0.12
aNumber of the lesions (ꢄ10ꢀ3) per 106 nucleotides,ꢃstandard devia-
tion, detected byalkaline elution after UVA irradiation (1 kJ mꢀ2) in
.
the presence of 1 mM drug.
bPercent of DNA fragmentationꢃstandard deviation, detected by
neutral elution after UVA irradiation (0.67 kJ mꢀ2) in the presence of
2 mM drug followed by24 h incubation in growth medium.

C. Marzano et al. / Bioorg. Med. Chem. 10 (2002) 2835–2844
2839
Figure 3. Clonal growth of HeLa cells after sensitization with the double irradiation method. Cells were exposed to UVA light in the presence of 3b
(2 mM), HFQ (2 mM), and 8-MOP (5 mM). The irradiated samples were washed with PBS (dotted lines) or not washed (solid lines), and then sub-
mitted to the second irradiation step performed with increasing UVA doses. Clonal growth capacitywas determined. Symbols: 3b, ꢀ; HFQ, &; 8-
MOP, *. Panel A: first irradiation step: 0.03 kJ mꢀ2; Panel B: first irradiation step: 0.12 kJ mꢀ2
.
Table 4. Inhibition of the tumor transfer capacityof Ehrlich cells by
UVA irradiation in the presence of compound 3b
which were then observed scoring the tumor growth
(Table 4). Untreated cells could transmit the tumor very
efficiently, since all animals developed the disease and
died after few days. In contrast, sensitized cells com-
pletely lost this property: in fact, after 90 days all the
treated animals survived. This suggested that the sensi-
tized cells became completelyunable to replicate.
Drug
concentration
(mM)
UVA dose
(KJ mꢀ2
%
Mortality
Mean survival
timeꢃSD
(days)
)
0
0
15
15
2.5
0
2.66
0
2.66
2.66
100
100
100
0
10ꢃ3.0
10ꢃ2.4
10ꢃ1.6
>90
Then, it was investigated if the injection of so inacti-
vated cells could induce in mice a protective effect
against a successive challenge performed byviable Ehr-
lich cells (Table 5). When untreated mice received
0.3–1ꢄ106 viable Ehrlich cells, theydeveloped the
tumor and died within a short time; the same result was
obtained with mice injected with 5ꢄ106 photo-
inactivated cells 10 days before the challenge with
0.3–1ꢄ106 viable cells. However, when this challenge
was performed after 21 days from the injection of the
inactivated cells, all animals survived, so showing an
acquired resistance against tumor development.
0
>90
Ehrlich cells were sensitized and then injected ip into health NCL mice
(5ꢄ106 cells per animal). The animals (groups of 10) were observed for
90 days.
Table 5. Protection against the transplant of viable Ehrlich cells
Protecting
treatment
(cellsꢄ106)a
Challenge
%
Mortalityb
Mean survival
time (days)
Days
Viable cells
(ꢄ106)
These data could be explained supposing that 3b sensi-
tization could modifythe antigenic character of the
photoinactivated cells, so that the injected mice devel-
oped an immunological reaction against the same viable
tumor cells.
—
—
5
5
5
—
—
10
21
21
0.3
1.0
0.3
0.3
1.0
100
100
100
0
10.1ꢃ2.1
9.7ꢃ3
11ꢃ1.3
>90
0
>90
aProtection was always performed by injecting 5ꢄ106 cells previously
bThe recipient animals (groups of 20) were observed for 90 days.
submitted to 3b sensitization (2.5 mM; 2.66 kJ mꢀ2).
It was interesting to observe that compound 3b could be
considered as an isoster of 4,6,40-trimetilangelicin
(TMA), a well-known angelicin derivative, presumably
the most effective one; the onlydifference was that the
d-lactonic ring of TMA was replaced bya d-lactamic
one. The comparison between the photobiological
properties of these two derivatives was verystimulating.
Both compounds efficientlyphotoreacted with DNA:
however 3b did not form ISC at all, while TMA could
induce moderate but significant amounts of ISC.19
Compound 3b produced numbers of DPCL=0, while
TMA induced low levels of DPC of no established
type.8,20 Therefore, these two compounds induced dif-
ferent bifunctional lesions. We must remember that
TMA proved to be genotoxic in various bacterial strains
consequentlyforming onlyDPC L=0. Panel B showed
the data obtained with 8-MOP, which formed ISC and
DPCL>0 both responsible for antiproliferative
effects;13,14 in this case, after the first irradiation step,
the inhibition curves generated removing and not
removing the unbound sensitizer are quite similar.
Finally, experiments were carried out with mice in vivo
to investigate the tumor transmission behavior of 3b.
Ehrlich cells were submitted to 3b photosensitization
and then injected intraperitoneouslyinto healthymice,

2840
C. Marzano et al. / Bioorg. Med. Chem. 10 (2002) 2835–2844
and in mammalian cells.21 On the other side, from the
preliminarydata until now available, 3b genotoxicity
seemed to be veryreduced in comparison with TMA
and with 8-MOP too. So this was more evidence that
different kinds of DNA damage induce different
biological consequences.
7-O-Ethers. General procedure
7-(20-Oxopropyloxy)quinolin-2-ones and 7-[20-(30-oxo)-
butyloxy]quinolin-2-ones (2a–c). General procedure. A
mixture of 1 (20.0 mmol), chloroacetone or 3-chloro-
2-butanone or propargyl chloride (24.0 mmol) and
anhydrous K2CO3 (10.0 g) in acetone (500 mL) was
refluxed until 1 disappeared (40 h, TLC). After cooling,
the solid was filtered off and washed with fresh acetone.
The solvent was evaporated from the combined filtrate
and washings and the residue was crystallized from
MeOH to give 2.
Conclusion
A new series of furoquinolinones lacking the substituent
at the N1 position was prepared and studied, in order to
confirm the interesting properties of the parent HFQ,
and with the aim to obtain new agents for PUVA ther-
apyand photopheresis with improved features, that is
low levels of clastogenicityand skin phototoxicity.
4-Methyl-7-(20-oxopropyloxy)quinolin-2-one (2a). Yield
54%; mp 206 ꢂC; 1H NMR (CDCl3) d 7.61 (d, 1H,
J=8.9 Hz, 5-H), 6.87 (dd, 1H, J=8.9, 2.5 Hz, 6-H), 6.80
(d, 1H, J=2.5 Hz, 8-H), 6.46 (q, 1H, J=1.2 Hz, 3-H),
4.68 (s, 2H, 10-H), 2.48 (d, 3H, J=1.2 Hz, 4-Me), 2.34
(s, 3H, 30-H); anal. (C13H13NO3): C, H, N.
Among the furoquinolinones synthesized, compound 3b
appeared verypromising. It demonstrated highlyeffec-
tive in inducing DNA damage, without forming ISC
and DPCL>0, both responsible for the genotoxicityand
phototoxicityof furocoumarins, and so showing very
strong antiproliferative activity. In addition, compound
3b showed no mutagenic effects and absence of skin
erythema.
4,6-Dimethyl-7-(20-oxopropyloxy)quinolin-2-one (2b). Yield
68%; mp 248 ꢂC; 1H NMR (CDCl3) d 7.49 (s, 1H, 5-H),
6.61 (s, 1H, 8-H), 6.21 (q, 1H, J=1.1 Hz, 3-H), 4.83 (s,
2H, 10-H), 2.37 (d, 3H, J=1.1 Hz, 4-Me), 2.26 (s, 3H,
6-Me), 2.23 (s, 3H, 30-H); anal. (C14H15NO3): C, H, N.
Moreover, this new furoquinolinone proved to be able
to protect against tumor growth, so predicting its
potential use in photopheresis: in this therapylympho-
cytes taken from the patients and specific for the reac-
tion of organ rejection after transplant, are modified by
furocoumarin sensitization, and therefore are recognized
bythe patient immune system as a not-self, thus per-
forming a vaccination against the same rejection process.
4-Methyl-7-[20-(30-oxo)butyloxy]quinolin-2-one (2c). Yield
45%; mp 186 ꢂC; 1H NMR (CDCl3) d 7.59 (d, 1H,
J=8.9 Hz, 5-H), 6.79 (dd, 1H, J=8.9, 2.5 Hz, 6-H), 6.76
(d, 1H, J=2.5 Hz, 8-H), 6.44 (q, 1H, J=1.1 Hz, 3-H),
4.83 (q, 1H, J=6.9 Hz, 20-H), 2.46 (d, 3H, J=1.1 Hz,
4-Me), 2.22 (s, 3H, 40-H). 1.55 (d, 3H, J=6.9 Hz, 10-H);
anal. (C14H15NO3): C, H, N.
4,6-Dimethyl-7-propargyloxyquinolin-2-one (2d). Yield
40%; mp 255 ꢂC; 1H NMR (CDCl3) d 7.42 (s, 1H, 5-H),
6.72 (s, 1H, 8-H), 6.41 (q, 1H, J=1.0 Hz, 3-H), 4.81 (d,
2H, J=2.4 Hz, 10-H), 2.57 (t, 3H, J=2.4 Hz, 30-H), 2.45
(d, 3H, J=1.0 Hz, 4-Me), 2.30 (s, 3H, 6-Me); anal.
(C14H13NO2): C, H, N.
All these findings require further studies to improve the
knowledge of the photobiological properties of com-
pounds 3b, which seems to be a verypromising agent
for PUVA therapyand photopheresis.
Methylfuro[2,3-h]quinolin-2(1H)-ones (3a–c). General
procedure. A mixture of 2 (10.0 mmol) and methane-
sulfonic acid (40.0 mmol) in toluene (500 mL) was
refluxed until 2 disappeared (40–60 h, TLC). After
cooling, the reaction mixture was washed with water
(2 ꢄ 200 mL) and the organic phase was evaporated
under reduced pressure. The residue was crystallized
from MeOH to give 3.
Experimental
Chemistry
Melting points were determined on a Gallenkamp
MFB-595–010M melting point apparatus and are
uncorrected. Analytical TLC was performed on pre-
coated 60 F254 silica gel plates (0.25 mm; Merck) devel-
oping with a CHCl3/MeOH mixture (9:1) unless other-
wise indicated. Preparative column chromatography
was performed using silica gel 60 (0.063–0.100 mm;
Merck), eluting with a CHCl3/MeOH mixture (98:2). 1H
NMR spectra were recorded on a Varian Gemini-200
spectrometer and 13C NMR were recorded on a Bruker
AMX300 spectrometer, with TMS as internal standard.
Mass spectra were recorded on a Varian MAT112
spectrometer. Elemental analyses were obtained on all
intermediates and were within ꢃ0.4% of theoretical
values. Starting 4-methyl-7-hydroxyquinolin-2-one 1a
and 4,6-dimethyl-7-hydroxyquinolin-2-one 1b were pre-
pared according to literature methods,15 as well as
4,6-dimethyl-7-(20-oxopropyloxy)quinolin-2-one 2e.10
4,9-Dimethylfuro[2,3-h]quinolin-2(1H)-one (3a). Yield
53%; mp 292 ꢂC; 1H NMR (CDCl3) d 7.57 (d, 1H,
J=8.9 Hz, 5-H), 7.46 (q, 1H, J=1.1 Hz, 8-H), 7.34 (d,
1H, J=8.9 Hz, 6-H), 6.50 (q, 1H, J=1.2 Hz, 3-H), 2.58
(d, 3H, J=1.2 Hz, 4-Me), 2.44 (d, 3H, J=1.1 Hz, 9-Me);
MS (EI) m/z 213 (M+); anal. (C13H11NO2): C, H, N.
4,6,9-Trimethylfuro[2,3-h]quinolin-2(1H)-one (3b). Yield
42%; mp 224 ꢂC; 1H NMR (CDCl3) d 7.46 (q, 1H,
J=1.1 Hz, 8-H), 7.35 (q, 1H, J=0.9 Hz, 5-H), 6.49 (q,
1H, J=1.2 Hz, 3-H), 2.57 (d, 3H, J=1.2 Hz, 4-Me), 2.54
(d, 3H, J=0.9 Hz, 6-Me), 2.53 (d, 3H, J=1.1 Hz, 9-Me);
MS (EI) m/z 227 (M+); anal. (C14H13NO2): C, H, N.

C. Marzano et al. / Bioorg. Med. Chem. 10 (2002) 2835–2844
2841
4,8,9-Trimethylfuro[2,3-h]quinolin-2(1H)-one (3c). Yield
47%; mp 233 ꢂC; 1H NMR (CDCl3) d 7.41 (d, 1H,
J=8.9 Hz, 5-H), 7.16 (d, 1H, J=8.9 Hz, 6-H), 6.32 (q,
1H, J=1.1 Hz, 3-H), 2.44 (d, 3H, J=1.1 Hz, 4-Me), 2.35
(q, 3H, J=0.8 Hz, 8-Me or 9-Me), 2.33 (q, 3H,
J=0.8 Hz, 8-Me or 9-Me); MS (EI) m/z 227 (M+); anal.
(C14H13NO2): C, H, N.
J=128.7 Hz, 9-Me); MS (EI) m/z 257 (M+); anal.
(C15H15NO3): C, H, N.
8-Hydroxymethyl-4,6,9-trimethylfuro[2,3-h]quinolin-2(1H)-
one (3f) from 3e. Compound 3e (0.5 g, 2.5 mmol) was
dissolved in concd. H2SO4 (10 mL) and the solution was
heated at 50 ꢂC until 3e disappeared (3 h). The mixture
was poured into cold water (100 mL), neutralized with
5% NaHCO3 and extracted with AcOEt (3 ꢄ 100 mL).
The organic phase was dried (Na2SO4), evaporated
under reduced pressure and the residue purified bycol-
umn chromatographyto give 3f (0.05 g, 10%), with mp,
NMR and MS data as above reported.
4,6,8-Trimethylfuro[2,3-h]quinolin-2(1H)-one (3d).
A
mixture of 2d (6.4 g, 28.0 mmol) and CsF (3.2 g,
20.4 mmol) in N,N-diethylaniline (150 mL) was heated
at 210 ꢂC until starting product disappeared (2 h, 1H
NMR). After cooling, the mixture was diluted with
AcOEt (200 mL), washed with HCl 1 N (5 ꢄ 100 mL)
and water (3 ꢄ 100 mL) and the solvent was evaporated
under reduced pressure. The residue was purified by
column chromatographyto give 3d (0.64 g, 10%): mp
Biology
8-MOP was obtained from Chinoin (Milan, Italy) and
calf thymus DNA from Sigma-Chemie (Deisenhofen,
Germany). The test compounds were dissolved in di-
methyl sulfoxide (DMSO; 4.5 mM) and the solutions
kept at ꢀ20 ꢂC in the dark. Just before the experiment, a
calculated amount of compound solution was added in
the dark to phosphate-buffered saline (PBS) or to the
growth medium containing cells, to a final solvent con-
centration never exceeding 0.5%. The same amount of
DMSO was added to untreated controls. Everyexperi-
ment was carried out at least three times.
>300 ꢂC; H NMR (CDCl3) d 7.29 (q, 1H, J=0.8 Hz,
1
5-H), 7.02 (q, 1H, J=1.0 Hz, 9-H), 6.55 (q, 1H,
J=1.0 Hz, 3-H), 2.56 (d, 6H, J=1.0 Hz, 4-Me and
8-Me), 2.55 (d, 3H, J=0.9 Hz, 6-Me); MS (EI) m/z 227
(M+); anal. (C14H13NO2): C, H, N.
4,6,8,9-Tetramethylfuro[2,3-h]quinolin-2(1H)-one (3e) and
8-hydroxymethyl-4,6,9-trimethylfuro[2,3-h]quinolin-2(1H)-
one (3f). Compound 2e10 (2.0 g, 7.7 mmol) was dissolved
in concd. H2SO4 (50 mL) and the solution was heated at
50 ꢂC for 1 h. The mixture was poured into cold water
(500mL) and the obtained precipitate was collected,
washed with water and purified bycolumn chromato-
graphyto give 3e (0.6 g, 32%), followed by 3f (0.7 g, 36%).
Animals and tumor line
All procedures involving animals and their care were
conducted in conformitywith the institutional guide-
lines, that were in compliance with National and Euro-
pean Economic CommunityCouncil Directives. Ehrlich
cells, Lettre strains from Heidelberg, were routinely
transferred byintraperitoneous injection (2 ꢄ106 cells
per mouse) into Swiss mice (LD50 after 30 days
5ꢄ102 cells). For the experiments mice at 8–10 weeks of
age and at a mean bodyweight of 23 ꢃ2 g were used.
Tumor cells were withdrawn from donor mice as a pure,
highlyconcentrated, single cell suspension. For the
experiments, theywere collected bycentrifugation and
suspended in PBS.
Compound 3e:10. Mp 231 ꢂC (MeOH); 1H NMR
(CD3OD) d 7.19 (q, 1H, J=0.9 Hz, 5-H), 6.32 (q, 1H,
J=1.0 Hz, 3-H), 2.44 (d, 1H, J=1.0 Hz, 4-Me), 2.41 (d,
1H, J=0.9 Hz, 6-Me), 2.37 (q, 1H, J=0.9 Hz, 8-Me or
9-Me), 2.34 (q, 1H, J=0.9 Hz, 8-Me or 9-Me); 13C
NMR (DMSO-d6) d 162.05 (m, C-2), 154.32 (m, C-6a),
151.01 (broad q, J=6.0 Hz, C-4), 150.03 (q, J=7.2 Hz,
C-8), 132.02 (d, J=9.0 Hz, C-9b), 121.26 (dq, J=160.0,
5.2 Hz, C-5), 119.01 (dq, J=164.0, 5.0 Hz, C-3), 116.33
(q, J=6.2 Hz, C-6), 116.02 (m, C-4a o C-9a), 115.62 (m,
C-4a o C-9a), 110.50 (q, J=4.8 Hz, C-9), 20.42 (qd,
J=127.8, 5.7 Hz, 4-Me), 15.49 (qd, J=123.0, 4.8 Hz,
6-Me), 12.20 (q, J=128.7 Hz, 8-Me), 10.46 (q,
J=129.7 Hz, 9-Me); MS (EI) m/z 241 (M+); anal.
(C15H15NO2): C, H, N.
UVA irradiation
Cell suspensions containing the test compound were
incubated at room temperature for 15 min in the dark,
put into Petri dishes (5 cm in diameter, 3 mL) and
exposed to UVA light. UVA exposure were performed
with a Philips HPW 125 lamp, provided with a built-in
Philips filter. The emission spectrum was in the range
320–400 nm, with a maximum, over 90% of the total, at
365 nm. Irradiation intensitywas determined on a UV
radiometer (model 97507, Cole-Parmer Instrument Co.,
Compound 3f. Mp 298 ꢂC (MeOH); 1H NMR
(DMSO-d6) d 11.49 (broad s, 1H, –NH), 7.44 (s, 1H,
5-H), 6.67 (t, 1H, J=4.4 Hz, 8-CH2OH), 6.36 (q, 1H,
J=1.2 Hz, 3-H), 4.77 (d, 2H, J=4.4 Hz, 8-CH2OH),
2.48 (m, 9H, 4-Me, 8-Me and 9-Me); 13C NMR
(DMSO-d6) d 162.10 (m, C-2), 154.56 (m, C-6a), 150.74
(qd, J=6.7, 4.0 Hz, C-4), 149.73 (t, J=5.2 Hz, C-8),
131.62 (d, J=8.6 Hz, C-9b), 121.65 (dq, J=159.7,
5.2 Hz, C-5), 119.18 (dq, J=164.0, 4.3 Hz, C-3), 116.14
(q, J=5.7 Hz, C-6), 116.05 (m, C-4a o C-9a), 115.60 (m,
C-4a o C-9a), 115.52 (q, J=3.8 Hz, C-9), 55.26 (td,
J=143.1, 6.6 Hz, 8-CH2), 20.30 (qd, J=127.8, 5.7 Hz,
4-Me), 15.49 (qd, J=127.8, 4.8 Hz, 6-Me), 12.76 (q,
Niles, IL) at 5.5ꢄ10ꢀ3 KJ sꢀ1 mꢀ2
.
DNA synthesis in Ehrlich cells
DNA synthesis was assayed in Ehrlich ascites tumor
11
cells (Lettre strain) as alreadydescribed.
Just after
UVA irradiation (cell density: 2ꢄ107 cells mL-1 in PBS),
the samples (106 cells in 0.5 mL PBS) were incubated for

2842
C. Marzano et al. / Bioorg. Med. Chem. 10 (2002) 2835–2844
30 min at 37 ꢂ in the presence of 40 kBq mL-1 of
[3H]thymidine (4.77 TBq mM-1; Amersham Inter-
national Inc., UK). The acid-insoluble fraction was
precipitated byadding ice-cold 5% trichloroacetic acid
and collected on Whatman GF/C filters (2.5 cm in dia-
meter). After several washings with cold 1% tri-
chloroacetic acid, the filters were dried and counted.
The results were calculated as the percentage of radio-
activityincorporated into the DNA with respect to
untreated control cells (approximately3–6 kBq). Filtra-
tions were carried out with a Sample Manifold appara-
tus (Millipore Corp., Bedford, USA). The filters were
dried and counted. Data are expressed as ID50, that is
the UVA dose (KJ mꢀ2) inducing 50% inhibition of the
macromolecular synthesis in the presence of 20 mM drug
concentration. Data were computed byprobit analysis.
reverted byUV light and bymost base pair substitution
mutagens23 and therefore byformation of C 4-cyclo-
adducts of furocoumarins.4 Bacteria were grown over-
night in a minimal Davis–Mingioli salt glucose medium
supplemented with tryptophan (20 mg L-1). E. coli cells
were washed and then suspended in PBS (pH 7.0) con-
taining the test compound (20 mM) at a densityof
108 cells mL-1. Bacteria were irradiated with UVA. For
the mutagenesis test, 0.1 mL aliquots of the irradiated
suspensions were added to 2 mL of molten 0.6% top
agar and poured onto plates containing 20 mL of SEM
agar (MMA fortified with 0.1 mg mL-1 Difco nutrient
broth). To determine the surviving fraction, the irra-
diated cells (0.1 mL) were diluted with phosphate buffer,
added to 2 mL of molten 0.6% agar, and plateled on
Davis–Mingioli minimal medium supplemented with
tryptophan. The plates were incubated for 24 h at 37 ꢂC
in the dark, and the colonies were then counted. The
Clonal growth of HeLa cells
6
mutation frequencywas expressed as mutants per 10
HeLa cells were grown in nutrient mixture F-12Ham
medium (Sigma Chemical Co, St. Louis, MO, USA),
containing 10% fetal calf serum, and supplemented with
antibiotics. Cell growth was accomplished at 37 ꢂC in a
5% carbon dioxide atmosphere.
survivors, computed bydividing the number of rever-
tants observed per plate bythe number of surviving
bacteria at the same treatment and subtracting from
result the number of revertant colonies per million sur-
vivors observed in controls (0.03ꢃ0.01).
HeLa cells (1.5–2ꢄ105) were seeded in Petri dishes in
growth medium (4 mL). After 24 h, the medium was
replaced with a fresh one containing the test compound.
The cells were incubated for 15 min at 37 ꢂC in the dark
and then exposed to UVA light. Aliquots of 100 cells
were seeded in the same medium, incubated for 7 days
and then the colonies were stained and counted, dis-
carding colonies with less than 50 cells. The efficiencyof
the clonal growth, that is the ratio between the number
of the formed colonies and the number of the cells see-
ded, was then calculated. The plating efficiencywas
about 90%. Data are expressed as ID50, that is the UVA
dose (KJ mꢀ2) inducing 50% inhibition of the clonal
growth in the presence of 5 mM drug concentration.
Calculations were accomplished byprobit analysis.
Skin phototoxicity
Skin phototoxicitywas tested applied on depilated
albino guinea-pigs (outbred Dunkin–Hartleystrain) as
24
alreadydescribed. Compounds were applied topically
.
as 0.5% methanolic solutions up to 5ꢄ10ꢀ2 mg cmꢀ2
The animals were kept in the dark for 15 min and then
exposed to UVA light (20 kJ mꢀ2). The animals were
observed for 7 days.
Spectrophotometric determinations
A 3b solution (8.2ꢄ10ꢀ5 M; EtOH/water, 1:1) was
exposed to increasing UVA light doses (0, 0.8, 1.6, 3.2,
6.4, 12.8 and 25.6 kJ mꢀ2) and the UV spectrum was
then recorded on a Kontron UVIKON-930 UV-Visible
spectrophotometer.
Experiments with T2 phage
The host bacteria (E. coli B48) were grown in cultures
prepared in brain-heart infusion (Difco Laboratories,
Radiochemical determinations
Detroit, MI, USA) at 37 ꢂC, collected in log phase, sus-
Filters from DNA synthesis determinations were coun-
ted using a toluene based scintillator (PPO 5 g, di-
methyl-POPOP 0.25 g, toluene up to 1 L of solution).
9
pended in MgSO4 (2 mM) at a densityof 10 cells mLꢀ1
,
and then infected with T2 phage at a multiplicityof 1.
The culture was then incubated at 37ꢂC for 3 h. Phage
titers were determined using the standard bilayer method22
and the same E. coli strain. Phage suspensions were diluted
to 1010 viral particles/mL with MgSO4 (2 mL) containing
the test compound. Aliquots (5 mL) of these virus suspen-
sions were poured into Petri dishes (5cm in diameter),
incubated at 37ꢂC for 25min in the dark and then
exposed to UVA light. After irradiation, the viral suspen-
sions were further diluted with the same medium and the
numbers of plaque-forming units/mL were scored.
Fractions from alkaline and neutral elution were coun-
ted using Ultima Gold XR (Packard Instruments, Mer-
iden, CT, USA) and a Packard Tri-Carb 1900TR
spectrometer. Counting was accomplished auto-
maticallyon the basis of quenching curves obtained
using [14C] and [3H] radioactivitystandards.
Photoreaction with DNA in vitro
Samples (3 mL) of an aqueous solution of DNA
(2.3 mM) containing 2 mM NaCl, 1 mM EDTA and
compound 3b (4 mM), were exposed to UVA light (10 kJ
Mutagenesis tests
The strain used was E. coli TM9 (WP2, uvrA, R46)
carrying a nonsense mutation in the trpE gene which is
m
ꢀ2), made up to 1 M in NaCl and treated with two
volumes of ethanol. The precipitated DNA, washed

C. Marzano et al. / Bioorg. Med. Chem. 10 (2002) 2835–2844
2843
with 80% ethanol, was dissolved in 3 mL of water. The
fluorescence was then determined on a Perkin Elmer
model LS-5 spectrophotofluorimeter.
SDS; pH 9.6). In these experiments no internal controls
were used. The data obtained byneutral elution assay
were computed as DNA fragmentation (F) on the basis
of its radioactivity, as shown by the equation:
Detection of DNA damage
Reluted
F ¼
DNA damage was detected byalkaline or neutral elu-
tion according to Kohn;17 each experiment was carried
out using an internal standard, that is untreated cells
labeled with [3H]thymidine and submitted only to a well
defined dose of gamma rays, while treated cells were
labeled with [14C]thymidine. HeLa cells in exponential
growth were labeled byovernight incubation in the
presence of 7.4 KBq mLꢀ1 [3H]thymidine or 3.7 KBq
mLꢀ1 [14C]thymidine. The radioactive medium was
removed and replaced bya fresh one containing the test
compound for [14C]cells, and bya fresh one containing
Reluted þ Rfilter
where Reluted is the total eluted radioactivityand Rfilter
is that retained on the filter.
The results are expressed as the percentage of DNA
fragmentation in comparison with the controls accord-
ing to the formula:
F ꢀ F0
Percent of DNA fragmentation ¼
ꢁ 100
3
only0.5% DMSO for [ H]cells. The cells were exposed
to UVA, washed, and submitted to alkaline elution.
1 ꢀ F0
Detection of inter-strand cross-links (ISC). About
0.5–1.0ꢄ106 of treated [14C]cells were mixed with equal
amounts of standard [3H]cells, the mixture was cooled
in ice, and exposed to 6 Gyof gamma rays. The mixture
was deposited on a polycarbonate filter (pores 2 mm in
diameter, Nucleopore Corp., Pleasanton, CA, USA) in
a Swinnex-25 filter holder (Millipore Corp., Bedford,
MA, USA) and immediatelylysed with 2% SDS, 0.1 M
glycine, 0.025 M Na2EDTA, pH 10, (5 mL). The solu-
tion was then allowed to flow out bygravity. The same
solution (2 mL) containing 0.5 mg L-1 of proteinase K
(Sigma Chemical Co., St. Louis, MO, USA) was gently
poured on to the filter, followed by40 mL of the eluting
solution (1 M tetrapropylammonium hydroxide, 0.02 M
EDTA, 0.1% SDS; pH 12.1). Elution was carried out
with a Gilson Minipuls peristaltic pump, at a flow rate
of 0.03–0.04 mL minꢀ1. The radioactivityof both iso-
topes was determined in the fractions collected with a
Gilson fraction collector (approximately, 3.5 mL per
fraction).
where F is the DNA fragmentation observed in the
treated sample and F0 is that scored in the controls.25
Experiments with the double irradiation protocol
The double irradiation procedure was carried out by
exposing HeLa cells to a small UVA dose in the pre-
sence of the compound to be tested. The drug-contain-
ing medium was then carefullydischarged and the cells
washed twice with aliquots (15 mL) of medium free of
drug and serum. A third portion of fresh medium was
then added and the cells were further exposed to UVA
light.14
Experiments in mice
Ehrlich cells were irradiated in the presence of the test
compound and then injected intraperitoneously
(106 cells per mouse) into healthymice. The animals
were then observed for 90 days, scoring mortality due to
tumor growth.
Detection of DNA-protein cross-links (DPC). The pro-
cedure was the same reported for ISC, but the gamma
raydose delivered to the mixed cell suspensions was
increased to 30 Gy; polyvinyl chloride filters (pores 5 mm
in diameter; Nucleopore Corp., Bedford, MA, USA)
were employed and treatment with proteinase K was
In the experiments concerning protection against the
tumor, Ehrlich cells were irradiated in the presence of
test compound in experimental conditions sufficient to
inhibit their abilityto transmit the tumor; and theywere
then injected into healthymice. The challenge was per-
formed byinjecting an established number of viable
Ehrlich cells on the selected day, according to the
experimental protocol. Animals were observed for 90
days.
omitted. All calculations were performed according to
14
Kohn17 and as alreadydescribed.
Gamma rayexpo-
sures were always performed on ice using a 60Co source
working at the Reparto Applicazioni, Legnaro, Padova,
Istituto di Fotochimica e Radiazioni d’Alta Energia
(FRAE), C.N.R., with a dose rate of 2.5 Gymin ꢀ1, as
determined byFricke solution.
References and Notes
1. Parrish, J. A.; Stern, R. S.; Pathak, M. A.; Fitzpatrick,
T. B., Photochemotherapyof Skin Diseases. In The Science of
Photomedicine; Regan, J. D., Parrish, J. A., Eds.; Plenum:
New York, 1982; p 595.
2. Gasparro, F. P. Extracorporeal Photochemoterapy-Clinical
Aspects and the Molecular Basis for Efficacy; Landes: George-
town, TX, 1994.
Detection of DNA double-strand breaks (DSB). DSB
were detected byneutral elution. HeLa cells were
labeled as before by7.4 KBq mL
then submitted to sensitization with the test compound
as described. The cells were washed with PBS and then
submitted to the elution procedure as described for ISC
detection, eluting with a different solution (1 M tetra-
propylammonium hydroxide, 0.02 M EDTA, 0.1%
-1
[3H]thymidine and
3. Kirkland, D. J.; Creed, K. L.; Mannisto, P. Mutat. Res.
1983, 116, 73.

2844
C. Marzano et al. / Bioorg. Med. Chem. 10 (2002) 2835–2844
4. Venturini, S.; Tamaro, M.; Monti-Bragadin, C.; Carlassare,
F. Mutat. Res. 1981, 88, 17.
15. Chilin, A.; Rodighiero, P.; Pastorini, G.; Guiotto, A.
J. Org. Chem. 1991, 56, 980.
5. Hook, G. J.; Heddle, J. A.; Marshall, R. R. Cytogenet. Cell
Genet. 1983, 35, 100.
6. Stern, R. S.; Lange, R. J. Invest. Dermatol. 1988, 91,
120.
16. Ishii, H.; Hishikawa, T.; Takeda, S.; Ueki, S.; Suzuki, M.;
Harayama, T. Chem. Pharm. Bull. 1990, 38, 1775.
17. Baccichetti, F.; Bordin, F.; Marzano, C.; Carlassare, F.;
Guiotto, A. Z. Naturforsch 1970, 34c, 811.
7. Ben-Hur, E.; Song, P. S. Adv. Rad. Biol. 1984, 11, 131.
8. Bordin, F.; Carlassare, F.; Busulini, L.; Baccichetti, F.
Photochem. Photobiol. 1993, 58, 133.
9. Bordin, F.; Dall’Acqua, F.; Guiotto, A. Pharm. Ther. 1991,
52, 331.
10. Rodighiero, P.; Guiotto, A.; Chilin, A.; Bordin, F.; Bac-
cichetti, F.; Carlassare, F.; Vedaldi, D.; Caffieri, S.; Pozzan,
A.; Dall’Acqua, F. J. Med. Chem. 1996, 39, 1293.
11. Bordin, F.; Marzano, C.; Carlassare, F.; Rodighiero, P.;
Guiotto, A.; Caffieri, S.; Baccichetti, F. J. Photochem. Photo-
biol., B: Biol. 1996, 34, 159.
18. Khon, K. W. Pharm. Ther. 1991, 49, 55.
19. Bordin, F.; Marzano, C.; Gatto, C.; Carlassare, F.;
Rodighiero, P.; Baccichetti, F. J. Photochem. Photobiol., B:
Biol. 1994, 26, 197.
20. Bordin, F.; Marzano, C.; Baccichetti, F.; Carlassare, F.;
Vedaldi, D.; Falcomer, S.; Lora, S.; Rodighiero, P. Photo-
chem. Photobiol. 1998, 68, 157.
21. Stivala, L. A.; Pizzala, R.; Rossi, R.; Melli, R.; Verri,
M. G.; Bianchi, L. Mutat. Res. 1995, 327, 227.
22. Adams, M.H. In Bacteriophages; Interscience/John Wiley
and Sons: New York, 1959.
12. Fossa, P.; Mosti, L.; Menozzi, G.; Marzano, C.; Bacci-
chetti, F.; Bordin, F. Bioorg. Med. Chem. 2002, 10, 743.
13. Bordin, F.; Baccichetti, F.; Marzano, C.; Carlassare, F.;
Miolo, G.; Chilin, A.; Guiotto, A. Photochem. Photobiol.
2000, 71, 254.
23. Bridges, B. A.; Mottershead, R. P.; Rothwell, M. A.;
Green, M. H. L. Chem.-Biol. Interact. 1972, 5, 77.
24. Carlassare, F.; Baccichetti, F.; Guiotto, A.; Rodighiero,
P.; Gia, O.; Capozzi, A.; Pastorini, G.; Bordin, F. J. Photo-
chem. Photobiol., B: Biol. 1990, 5, 25.
14. Marzano, C.; Baccichetti, F.; Carlassare, F.; Chilin, A.;
Lora, S.; Bordin, F. Photochem. Photobiol. 2000, 71, 263.
25. Bertrand, R.; Kohn, K. W.; Solary, E.; Pommier, Y. Drug
Dev. Res. 1995, 34, 128.