Pyrano[2,3-e]isoindol-2-ones, new angelicin heteroanalogues

Article abstract of DOI:10.1016/j.bmcl.2009.01.096

A convenient synthesis of the pyrano[2,3-e]isoindol-2-one ring system, an heteroanalogue of angelicin, is reported. Our synthetic approach consists of the annelation of the pyran ring on the isoindole moiety using 5-dialkylamino- or 5-hydroxymethylene int

Full text of DOI:10.1016/j.bmcl.2009.01.096

Bioorganic & Medicinal Chemistry Letters 19 (2009) 1711–1714
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Pyrano[2,3-e]isoindol-2-ones, new angelicin heteroanalogues
Paola Barraja a, Virginia Spanò a, Diana Patrizia a, Anna Carbone a, Girolamo Cirrincione a,
,
*
Daniela Vedaldi b, Alessia Salvador b, Giampietro Viola b, Francesco Dall’Acqua b
a Dipartimento Farmacochimico, Tossicologico e Biologico, Università degli Studi di Palermo, Via Archirafi 32, 90123 Palermo, Italy
b Dipartimento di Scienze Farmaceutiche, Università degli Studi, Via Marzolo 5, 35131 Padova, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
A convenient synthesis of the pyrano[2,3-e]isoindol-2-one ring system, an heteroanalogue of angelicin, is
reported. Our synthetic approach consists of the annelation of the pyran ring on the isoindole moiety
using 5-dialkylamino- or 5-hydroxymethylene intermediates as building blocks. The photoantiprolifera-
tive activity of the new derivatives was studied. Some of them bearing the benzyl group at the 8 position
were active with IC50 in the micromolar range. Cell cytotoxicity involves apoptosis, alteration of cell cycle
profile and membrane photodamage.
Received 5 November 2008
Revised 28 January 2009
Accepted 28 January 2009
Available online 1 February 2009
Keywords:
Ó 2009 Elsevier Ltd. All rights reserved.
Pyrano-isoindoles
Angelicin heteroanalogues
Photochemotherapeutic activity
Apoptosis
Furocoumarins constitute a group of natural or synthetic com-
pounds of great pharmacological interest. One of the most impor-
tant applications of these compounds is in the field of PUVA
photochemotherapy (psoralen plus UVA) for the treatment of a
variety of skin diseases characterized by hyperproliferative condi-
tions such as psoriasis, vitiligo and tumors, such as T-cell lym-
phoma, when used in conjunction with long-wave (320–400 nm)
ultraviolet light (UVA). Psoralen 1 is the parent compound of a rel-
atively large number of linearly fused furocoumarins. The biologi-
cal activity of these linear compounds is related to their ability to
link DNA via intercalation in the dark between the base pairs of
the nucleic acid followed by a [2 + 2] photocycloaddition with
the pyrimidine bases, particularly thymine, when irradiated with
UV-A light. The so formed monoadduct undergoes further [2 + 2]
photocycloaddition thus generating inter-strand cross-links.1
Undesirable side effects of this kind of photoreactions are mutage-
nicity and carcinogenicity. It explains why considerable efforts
have been made to develop monofunctional furocoumarins such
as angelicin 2, parent compound of a much smaller and less avail-
able class of furocoumarins, which on account of their geometry
cannot crosslink with DNA, resulting less dangerous than the linear
compounds.
roanalogues exhibit a strongly increased DNA photoactivation
property, thus appearing very promising for photochemotherapy.2
O
O
O
O
O
O
1 Psoralen
2 Angelicin
O
O
X
O
NH
S
NH
NR
X
X
X
N
N
X
X
R
R
3
4
5
X-X= CH2-CH2 or CH=CH
With the aim of preparing and studying new photoreactive
agents with increased antiproliferative activity and decreased se-
vere toxic side effects, we have recently reported the synthesis of
the new ring systems pyrrolo[2,3-h]quinolin-2-one 3 and thiopyr-
ano[2,3-e]indole 4, isosters of angelicin, whose derivatives showed
generally high phototoxicity in comparison with 8-MOP (GI50 0.4–
In consideration of above, we studied the synthesis of heteroan-
alogues of angelicin in which a nitrogen or a sulfur atom replaces
one or both oxygen atoms of the pyran and furan rings. These hete-
16.4 lM and 0.07–3.9 l
M, respectively).3
On continuing our studies on photochemotherapeutic drugs,
here we report the synthesis of pyrano[2,3-e]isoindol-2-one, an
heteroanalogue of angelicin, in which the isoindole ring replaces
the benzofuran one. We chose to synthesize this ring system in
* Corresponding author. Tel.: +39 091 6161606; fax: +39 091 6169999.
E-mail address: gcirrinc@unipa.it (G. Cirrincione).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.01.096

1712
P. Barraja et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1711–1714
order to verify how the condensation of the pyrrole ring on the
benzopyran moiety can affect the photobiological activity, consid-
ering that derivatives of the pyrrolo[3,4-h]quinoline system 5,
compared to their positional isomers of the series 3, showed an in-
and 10a were functionalized on the nitrogen atom in THF or
DMF using NaH as the base and an alkylating agent such as MeI,
BnCl, p-MeBnCl or p-OMeBnCl (60–98%). In the attempt to obtain
the N-phenyl substituted derivatives, tetrahydroisoindoles 7a, 9a
and 10a were subjected to the Ulman cross-coupling reaction with
CuBr, iodobenzene and K2CO3 in 1-methyl-2-pyrrolidinone (NMP).
However, it was impossible to obtain the corresponding N-phenyl
derivatives from 9a and 10a whereas 7a furnished 7d in 65% yield.
To achieve the direct introduction of the enamino functionality
which allows to obtain derivatives of type 13–15, the isoindoles
7a–f, 9a–c,e,f and 10a–c were reacted with an excess of tert-but-
oxybis(dimethylamino)methane (TBDMAM). However, this reac-
tion succeeded only with derivatives 9a–c,e,f bearing the
ethoxycarbonyl group in the 1 position which were converted into
15a–c,e,f.
In order to get the enaminoketons 13 and 14, isoindoles 7b–f
and 10b,c were subjected to a two steps sequence starting with a
formylation in benzene at room temperature with ethyl formate
as the formylating agent, leading to the hydroxymethylene deriva-
tives 11b–f and 12b,c and subsequent reaction with diethylamine
at room temperature in benzene to produce enaminoketones 13b–
f and 14b,c.
To afford the final tricyclic heteroanalogues of angelicin, enami-
noketones 13b–f were reacted with diethyl malonate, used as sol-
vent, keeping the temperature at 80–100 °C. The reactions
generally gave quite complex mixtures and was only possible to
isolate derivatives 16–18 in moderate yields (30–42%); from
14b,c only 19 (57%) was isolated, whereas all enaminoketones
15a–c,e,f provided the corresponding tricyclic systems 20–24
(30–46%). Compounds 15c,e,f were also reacted with dimethyl
malonate to give the corresponding pyrano-isoindoles 25–27
(40–42%) (Scheme 1).
Considering the remarkable activity of angelicin, we also fo-
cused our attention on the synthesis of 3-H substituted derivatives
of the ring system that could have been obtained simply by hydro-
lysis and decarboxylation of the ester derivatives. However, several
attempts for the hydrolysis of the carboxyester functionality on the
pyran ring of 16–18 and 19 failed. Only in the case of the reaction
of 18 in ethanol/potassium hydroxide very poor yield (8%) of the
corresponding acid 28 was obtained making difficult any further
reaction.
creased phototoxicity (0.1–14.5 l
M).4 The active compounds of the
series 5 were the dihydro derivatives, which did not intercalated
into DNA but after irradiation they induced photodamage to the
macromolecule. Moreover, upon irradiation, they caused remark-
able oxidation of the cellular membrane lipids.4
Our synthetic approach to pyrano-isoindoles consisted in the
annelation of the pyran ring on the isoindole moiety using dialky-
laminomethylene isoindoles 13–15 as building blocks.
Tetrahydroisoindole-4-ones 7, 9 and 10, as suitable substrates
for our purpose, were prepared by two synthetic routes with a
good substitution pattern on the pyrrole moiety.5,6 In particular,
the reaction of 2-ciclohexen-1-one 6, a good Michael acceptor,
with TOSMIC afforded the unsubstituted isoindole derivative 7a
(85%), whereas the Knorr-type reductive condensation of 2-acetyl-
cyclohexane-1,3-dione 8 with freshly prepared diethyl hydroxyi-
minomalonate led to 1,3-disubstituted isoindole 9a (65%). The
latter was subjected to basic hydrolysis followed by decarboxyl-
ation, with dilute hydrochloric acid in ethanol leading to 3-methyl
substituted isoindole 10a (70%). The tetrahydroisoindoles 7a, 9a
O
O
O
Me
N-R
CO2Et
COMe
H
ii
O
6
i
8
9
iii
iv
O
a R=H
O
Me
b R=Me
c R=Bn
e R=p-OMeBn
f R=p-MeBn
N-R
N-R
7
10
a R=H
iv
a R=H
iv
vi
b R=Me
c R=Bn
d R=Ph
v
b R=Me
c R=Bn
NEt
2 O
Me
e R=p-OMeBn
f R=p-MeBn
N-R
vii
O
vii
CO2Et
OH
OH
O
Me
15a-c,e,f
ix
Therefore, we considered an alternative route that was success-
fully used in the past for the synthesis of angelicin, and its deriva-
tives, such as oroselone and oroselol, reacting 5-hydroxymethylene
N-R
N-R
12b,c
11b-f
viii
NEt
2 O
O
R1
viii
O
Me
NEt2 O
Me
ii
O
N-R
N-R
N-R
CO2Et
O
OH
O
O
CO2Et
14b,c
ix
13b-f
ix
O
iii
i
20 R=H, R1=CO2Et
N-R
N-R
N-Me
21 R=Bn, R1=CO2Et
O
R=Me
EtO2C
22 R=Me, R1=CO2Et
R1
O
11b,c
b R=Me
c R=Bn
Me
23 R=p-MeBn, R1=CO2Et
24 R=p-OMeBn, R1=CO2Et
25 R=Bn, R1=CO2Me
29 R=Me
30 R=Bn
33
O
N-Bn
N-R
O
O
26 R=p-MeBn, R1=CO2Me
27 R=p-OMeBn, R1=CO2Me
EtO2C
16 R=Bn, R1=CO2Et
17 R=Ph, R1=CO2Et
18 R=Me, R1=CO2Et
28 R=Me, R1=CO2H
19
EtO2C
O
O
Me
Me
N-R
CO2Et
iv
x
N-R
CO2Et
20 R=H
Scheme 1. Synthesis of compounds 16–28. Reagents and conditions: (i) TOSMIC,
t-BuOK–THF; (ii) HON@C(CO2Et)2, Zn–acetic acid; (iii) KOH-EtOH then 6 M HCl–
EtOH; (iv) NaH, MeI or BnCl, or p-MeBnCl, or p-OMeBnCl/DMF or THF; (v) CuBr, PhI,
K2CO3/NMP; (vi) TBDMAM–toluene; (vii) HCOOEt, t-BuOK–benzene; (viii) HNEt2/
benzene; (ix) CH2(CO2Et)2 or CH2(CO2Me)2, piperidine, 80–100 °C; (x) KOH–EtOH,
reflux.
31 R=H
32 R=Me
22 R=Me
Scheme 2. Synthesis of compounds 29–33. Reagents and conditions: (i)
Ph3PCH2CO2Et/xylene; (ii) xylene or DMA or TEG; (iii) Ph3PCH2CO2Et/TEG or
DMA; (iv) DDQ/benzene.

P. Barraja et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1711–1714
1713
Table 1
Phototoxicity of pyrano[2,3-e]isoindol-2-ones 16–32 at different UVA doses
Compound
IC50 (l
M)a,b
NCTC-2544
2.5
LoVo
K562
Jurkat
0c
3.75
0
2.5
3.75
0
2.5
3.75
0
2.5
3.75
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
14.6 1.3
>20
>20
>20
14.0 1.2
>20
>20
>20
>20
>20
>20
>20
5.5 0.6
4.2 0.5
>20
>20
>20
>20
>20
10.9 0.9
>20
17.0 1.7
>20
10.0 0.8
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
10.5 1.0
>20
>20
>20
7.7 0.6
>20
>20
>20
>20
>20
>20
16.2 1.8
1.1 0.4
1.6 0.2
>20
>20
>20
>20
>20
9.6 0.8
11.1 1.1
15.0 1.7
>20
7.3 0.6
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
7.5 0.8
13.9 1.4
13.5 1.4
>20
6.1 0.5
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
8.2 0.9
15.4 1.7
16.1 1.8
>20
8.4 0.8
>20
>20
>20
>20
>20
>20
>20
>20
>20
7.2 0.7
15.2 1.6
12.8 1.4
13.5 1.8
3.7 0.4
17.8 2.1
11.8 1.1
>20
>20
>20
>20
>20
n.d.
n.d.
>20
7.3 0.6
13.1 1.2
9.3 0.8
14.0 1.7
2.7 0.4
16.6 1.7
12.6 1.9
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
n.d.
>20
>20
>20
1.2 0.3
0.9 0.2
>20
17.8 1.9
6.7 0.5
0.7 0.1
0.9 0.1
32
8MOP
Ang
>20
n.d.d.
n.d.
n.d.
n.d.
n.d.
a
Concentration of compound required to inhibit the cell growth by 50% after 72 h of exposure as determined by MTT assay.
b
c
Values are means SEM of three experiences.
UVA doses are expressed as J/cm2.
Not determined.
d
benzofuranones with (carbethoxymethylene) triphenylphospho-
such as N,N-dimethylaniline (DMA) or triethylene glycol (TEG) to
force the ring closure. However, this reaction did not give the desired
result and only the starting material from the reaction mixture was
recovered. On the contrary, straight heating of 5-hydroxymethylene
derivatives 11b,c with (carbethoxymethylene) triphenylphospho-
rane in the already mentioned high boiling solvents pyrano-isoin-
dole derivatives 29 and 30 were isolated in good yields (60–73%)
rane in refluxing xylene.7 However, under these conditions the
5-hydroxymethylene derivative 11b reacted only yielding the
uncyclized ester 33 as main product (70%). No evidence of the de-
sired tricyclic system was observed nor when the open chain inter-
mediate, once isolated, was heated in boiling xylene for 24 h. It was
thus reasonable to consider a solvent with an higher boiling point
Figure 1. DNA flow cytometric analysis. Jurkat cells were irradiated (2.5 J/cm2) with 5
l
M 21 and 25 (panels c and d, respectively). Panels a and b are non irradiated and 2.5 J/cm2
irradiated control, respectively. The data shown here are from a representative experiment repeated two times with similar results.

1714
P. Barraja et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1711–1714
(Scheme 2). Getting flat pyrano-isoindoles would allow to verify
whether such compounds, being aromatic could achieve intercala-
tion into DNA; the dihydro derivatives were therefore subjected to
oxidation with 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ)
in refluxing anhydrous benzene. From such a reaction only two
aromatic derivatives 31 and 32 in moderate yields (30–35%) were
isolated. The structure of all synthesized compounds was con-
firmed by spectroscopic data (IR, 1H and 13C NMR) and elemental
analysis (C, H, N) reported in the Supporting information.
All test compounds absorbed in UV–Vis range (230–500 nm)
and underwent photolysis after UVA irradiation.
TBA test as described by Morlière et al.9 and we found a clear in-
crease in lipid peroxidation in Jurkat cells irradiated in the pres-
ence of test compounds (data reported in the Supporting
information).
In conclusion, we found a convenient and versatile synthetic
pathway to achieve pyrano[2,3-e]isoindol-2-ones in moderate to
good yield. The new compounds tested as photoantiproliferative
agents showed no activity in the dark, whereas some of them
showed photoantiproliferative activity in the micromolar range
under UVA activation. Cell cytotoxicity involves apoptosis, alter-
ation of cell cycle profile and membrane photodamage.
The phototoxicity of pyrano[2,3-e]isoindol-2-ones was evalu-
ated on a panel of cultured human cell lines: NCTC-2544 (immor-
talized keratinocytes), LoVo (intestinal adenocarcinoma), K562
(chronic myeloid leukemia) and Jurkat (T-cell leukemia). No cyto-
toxicity was found in the dark by MTT8 test after 72 h from the
incubation with these compounds. Table 1 shows the extent of cell
survival expressed as IC50, which is the concentration that induces
50% inhibition of cell growth, after irradiation at different UVA
doses (2.5 and 3.75 J/cm2).
Some of the test compounds resulted phototoxic in the micro-
molar range: in particular, the lowest cellular survival was as-
sessed irradiating cells in the presence of 21 and 25. The
presence of a COOEt group in position 7 and of a benzyl one in
the isoindole nitrogen seemed to be important for phototoxicity.
It is worth remarking that the most active compounds – which
were active against all the four cell lines – are not fully unsaturated
such as angelicin but on the contrary are dihydroderivatives.
Whereas the aromatic pyrano-isoindoles 31 and 32 were active
against LoVo cell line only. Parallel behaviour was also observed
in the pyrrolo[2,3-h]quinolin-2-one 3 series, in which the unsatu-
rated derivative was devoid of phototoxicity,3b and in the pyrrol-
o[3,4-h]quinolin-2-one 5 series as already mentioned in this text.4
In order to have better insight into the mechanism of action of
compounds 16–32, we used flow cytometry to study cell cycle
variations upon irradiation. InFigure 1, the effect of the most active
compounds 21 and 25 was shown after 24 h from Jurkat cells irra-
diation (2.5 J/cm2). The irradiation in the presence of the test com-
pounds caused the appearance of a subG1 peak (apoptotic peak) in
cell cycle profile and the latter consisted of apoptotic cells with
DNA content lower than G1, as a consequence of endonuclease
activation.
Acknowledgment
This work was financially supported by Ministero dell’Istruzi-
one dell’Università e della Ricerca.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2009.01.096.
References and notes
1. See for example: (a) Dall’Acqua, F.; Viola, G.; Vedaldi, D. In CRC Handbook of
Organic Photochemistry and Photobiology; Hoorspool, W. M., Lenci, F., Eds.; CRC
Press: Boca Raton, FL, 2004. Chapter 142; (b) Dall’Acqua, F.; Viola, G. In
Photobiology for the 21th Century; Coohil, T. P., Valenzeno, D. P., Eds.; Valdenmar:
Overland Park, KS, 2001; p 325.
2. (a) Jakobs, A. E.; Piette, J. J. Photochem. Photobiol. B 1994, 22, 9; (b) Quanten, E.;
Adriaens, P.; De Schryver, F. C.; Roelandts, R.; Degreef, H. Photochem. Photobiol.
1986, 43, 485; (c) Rodighiero, P.; Chilin, A.; Pastorini, G.; Guiotto, A.
J. Heterocyclic Chem. 1987, 24, 1041; (d) Rodighiero, P.; Guiotto, A.; Chilin, A.;
Bordin, F.; Baccichetti, F.; Carlassare, F.; Vedaldi, D.; Caffieri, S.; Pozzan, A.;
Dall’Acqua, F. J. Med. Chem. 1996, 39, 1293.
3. (a) Barraja, P.; Sciabica, L.; Diana, P.; Lauria, A.; Montalbano, A.; Almerico, A. M.;
Dattolo, G.; Cirrincione, G.; Disarò, S.; Basso, G.; Viola, G.; Dall’Acqua, F. Biorg.
Chem. Lett. 2005, 15, 2291; (b) Barraja, P.; Diana, P.; Montalbano, A.; Dattolo, G.;
Cirrincione, G.; Viola, G.; Vedaldi, D.; Dall’Acqua, F. Bioorg. Med. Chem. 2006, 14,
8712.
4. Barraja, P.; Diana, P.; Montalbano, A.; Dattolo, G.; Cirrincione, G.; Viola, G;
Salvador, A.; Vedaldi, D.; Dall’Acqua, F. Abstract of Papers, EORTC Pharmacology
and Molecular Mechanisms Group 29th Winter Meeting, Palermo January 2008,
p. 85.
5. Spreitzer, H.; Mustafa, S. Chem. Ber. 1990, 123, 413.
6. Clezy, P. S.; Fookes, C. J. R. Aus. J. Chem. 1977, 30, 1337.
7. Lee, Y. R. Tetrahedron 1995, 51, 3087.
8. Mosmann, T. J. Immunol. Meth. 1983, 65, 55.
9. Morlière, P.; Moysan, A.; Santus, R.; Huppe, R.; Mazière, J. C.; Dubertret, L.
Biochim. Biophys. Acta 1991, 1084, 261.
As pyrano[2,3-e]isoindol-2-ones were highly hydrophobic, we
assessed the lipid peroxidation of Jurkat cell membranes using