Genotoxic activity of halogenated phenylglycine derivatives

Article abstract of DOI:10.1016/j.bmcl.2006.08.111

The discovery of genotoxic amino acids derived from phenylglycine, and posessing halogen substituents, is described. The utility of hypervalent iodine reagents in the synthesis of this class of compounds is highlighted. The mechanism of action of the (hal

Full text of DOI:10.1016/j.bmcl.2006.08.111

Bioorganic & Medicinal Chemistry Letters 16 (2006) 6073–6077
Genotoxic activity of halogenated phenylglycine derivatives
a
a,
b
Alicia Boto,a, Juan A. Gallardo, Rosendo Hernandez, Francisco Ledo,
*
*
´
b
c,ꢀ
c,ꢀ
´
´
´
Ana Munoz, Jose R. Murguıa, Mauricio Menacho-Marquez,
˜
Aurelio Orjalesb,* and Carlos J. Saavedraa
aInstituto de Productos Naturales y Agrobiologı´a del CSIC, Avda. Astrofı´sico Fco. Sa´nchez, 3, 38206-La Laguna, Tenerife, Spain
bFAES FARMA, S.A., Ma´ximo Aguirre, 14, 48940-Leioa, Vizcaya, Spain
cUnidad de Investigacio´n, Hospital Universitario de Canarias, Ofra s/n, 38320-La Laguna, Tenerife, Spain
Received 27 July 2006; revised 28 August 2006; accepted 28 August 2006
Available online 20 September 2006
Abstract—The discovery of genotoxic amino acids derived from phenylglycine, and posessing halogen substituents, is described. The
utility of hypervalent iodine reagents in the synthesis of this class of compounds is highlighted. The mechanism of action of the
(haloaryl)glycines was studied in Saccharomyces cerevisiae.
Ó 2006 Published by Elsevier Ltd.
Tumour cells have very active transport systems to cap-
ture the amino acids required for protein and nitrogen
base biosynthesis. Since these systems are overexpressed
with respect to most non-tumour cells, the development
of cytotoxic amino acid derivatives1 is a promising way
to achieve more selective anticancer treatments.2 For in-
stance, melphalan 13 (Fig. 1) is a clinically used alkylat-
ing agent whose uptake is performed by active amino
acid transport systems.4 This drug was developed in an
effort to reduce the side effects produced by other
mustards.
H2N COOH
COOH
NH2
OH
N
NO
2
N
L-alanosine
Cl
Cl
1
CONHR
melphalan
acyl-HN
N
NH2
COOH
Cl
OR
Once into the cell, the amino acid analogues disrupt dif-
ferent physiological processes. For instance, L-alanosine
2 interferes with aspartic acid metabolism.5 The potent
antibiotic and antitumoural acivicin 36 is a specific
inhibitor of c-glutamyl transpeptidase and transmem-
brane glutathione transport, inducing apoptosis in hu-
man lymphoblastoid cells.7
H
N O
4
azatyrosine
derivatives
3
acivicin
Figure 1. Some cytotoxic amino acid derivatives.
amino acids, such as phenylalanine, phenylglycine and
tyrosine analogues. Few examples of cytotoxicity have
been reported for this class of compounds. For instance,
the azatyrosine derivatives 4 were patented for the treat-
ment of pancreas, colon and thyroid cancer.8 Recently,
phenylglycine and other amino acids were reported to
block the ATB0+ amino acid transport, which is overex-
pressed in tumoural cells. Since they were deprived of vi-
tal nutrients, a strong growth inhibition was observed
for human colon and breast cancer cell lines.9
In spite of their potential selectivity, the use of amino
acids as anticancer agents has yet to be fully explored.1
In an effort to develop new amino acid-based antitu-
moural drugs, we turned our attention to aromatic
Keywords: Cytotoxic; Genotoxic; Amino acids; Phenylglycine; Cancer.
*
Corresponding authors. Tel.: +34 922 260112; fax: +34 922
250135; e-mail: alicia@ipna.csic.es
´
Present address: Instituto de Biologıa Molecular y Celular de Plantas
ꢀ
Now we report new cytotoxic agents derived from phe-
nylglycine. The lead compound 5 (Fig. 2) was discovered
´
‘Primo Yu´fera’, Universidad Politecnica de Valencia-CSIC, c/Camino
de Vera, s/n, 46022-Valencia, Spain.
0960-894X/$ - see front matter Ó 2006 Published by Elsevier Ltd.
doi:10.1016/j.bmcl.2006.08.111

6074
A. Boto et al. / Bioorg. Med. Chem. Lett. 16 (2006) 6073–6077
Compound 13 was then transformed into derivatives
BzHN COOMe
BzHN COOMe
14–20 (Scheme 2), where the benzamide group was re-
placed by oxygen functions (compounds 14 and 15), a
thioaryl group (product 16) and other nitrogen func-
tions (compounds 17–20). Some of these functions have
a volume similar to the NHBz group, but lack the ability
to form hydrogen bonds and present different polarity
(such as the PhS or the phthalimide groups). Others
are able to form hydrogen bonds (OH, hydrazide, sul-
fonamide, etc.) but differ in volume and polarity.
7 Inactive
X
OMe
I
OMe
5 X= I Cytotoxic
6 X= H Inactive
8 Inactive
Figure 2. Discovery of the lead compound 5.
In other group of analogues, the ester function in prod-
uct 5 was replaced by amide, hydroxymethyl, acid, or
ketone groups (Scheme 3), using conventional method-
ologies. The resulting products 21–26 present differences
in hydrosolubility, volume and metabolization, with re-
spect to the lead compound.
during a screening of the cytotoxic activity of different
amino acids in Saccharomyces cerevisiae. This yeast is
used as a model system to study the mechanism of ac-
tion of antitumour drugs, and to select, from a given
batch, the most promising cytotoxic compounds. It
was also observed that the deiodinated analogue 6 did
not show activity, nor did methyl N-benzoylglycine 7
or 2-iodoanisole 8.
In order to obtain more structural diversity, the acid 26
was transformed into the phosphonate 27, using a one-
pot fragmentation–phosphorylation reaction developed
by our group.10
In order to determine the structure–activity relation-
ships (SAR), different analogues of compound 5 were
prepared: (a) by replacement of the benzamide group
by oxygen, sulfur and other nitrogen functions; (b) by
replacing the ester group by amide, hydroxymethyl,
acid, ketone and phosphonate groups; (c) changing the
aromatic ring substitution pattern.
HO COOMe
RO COOH
i
I
I
OMe
OMe
The cytotoxic activity of these analogues was then stud-
ied with three tumour cell lines: MCF7 (breast), NCI-
H460 (lung) and SF-268 (glioma). Their mechanism of
action was characterised using mutant strains of S. cere-
visiae, as will be commented below.
13
14 R = H
15 R = Ac
ii
iii, v
iii, iv
The first analogues were prepared to determine the
influence of the nitrogen function on the cytotoxic
activity. Thus, the aromatic ring of compound 9
(Scheme 1) was iodinated with hypervalent iodine re-
agents and iodine, generating products 10–12. The
acetate group in product 12 was hydrolysed, yielding
the alcohol 13.
PhS COOMe
X
COOMe
I
I
OMe
OMe
16
17 X =
O
O
N
COOMe
O
18 X =
X
I
OMe
NH
N
O
10 X = H
11 X = I
COOMe
O
i
Ph
19 X =
20 X =
N
NH
N
RO COOMe
OMe
O
9
-NHSO2Ph
I
OMe
Scheme 2. Replacement of the benzamide group in the lead compound
5 by other functionalities. Reagents and conditions: (i) 2 N NaOH,
MeOH, 92%; (ii) Ac2O, Py, 84%; (iii) MsCl, CH2Cl2, Et3N; (iv)
Cs2CO3, DMF, reflux, PhSH, 58% for both steps; (v) Cs2CO3, DMF,
reflux, nucleophile (phthalimide or phthalohydrazide or phenylurazole
or phenylsulfonamide). Compounds 17 (74%); 18 (74%); 19 (64%); 20
(46%).
12 R = Ac
13 R = H
ii
Scheme 1. Replacement of the nitrogen function by hydrogen or
oxygen functions. Reagents and conditions: (i) DIB, I2, CH2Cl2, dark,
20 h 10 (55%), 11 (14%), 12 (30%); (ii) MeONa, MeOH, 92%.

A. Boto et al. / Bioorg. Med. Chem. Lett. 16 (2006) 6073–6077
6075
BzHN CONHR
BzHN COOMe
BzHN COOMe
i or ii
Hal
X
OH
Hal
I
i
OMe
X
BzHN COOMe
28
29−36
21 R = H or
22 R = Bu
I
Derivatives 29−36, and formation of 37, 38 from 36:
OMe
H
H
N
BzHN
H
N
5
OR
I
Bz N COOMe
COOMe
OMe
COOMe
I
Bz
Bz
MeO
I
ii
OMe
I
OMe
29
iv
23 R = H
24 R = Ac
iii
31
30
H
H
N
H
N
O
O
Bz N COOMe
COOMe
Hal
COOMe
I
OMe
BzHN
BzHN
Bz
Bz
P
R
OMe
MeO
vi
I
I
R = OH
I
OMe
X
OMe
OMe
32
33 Hal = Cl
34 Hal = Br
35 X = NHCO2Me
36 X = OTBDPS
25 R = OH
26 R = Me
27
v
iii
Scheme 3. Replacement of the ester group in the lead compound 5 by
other functionalities. Reagents and conditions: (i) RNH2, THF, reflux;
21 (31%), 22 (81%); (ii) DIBAL-H, CH2Cl2, 37%; (iii) Ac2O, Py, 57%;
(iv) NaOH, MeOH, 92%; (v) BuLi, THF, ꢀ78 °C; then MeMgI, THF,
ꢀ78 °C ! rt; 26 (20%), 25 (61%); (vi) DIB, I2, hm, CH2Cl2, then
BF3ÆOEt2, P(OMe)3, 84%.
37 X = OH
iv
38 X = OAc
Scheme 4. Changes in the aromatic ring substituents. (i) (a) DIB, I2,
hm, CH2Cl2, then MeOH; (b) BF3ÆOEt2, 3-iodoanisole, CH2Cl2, 29
(31%), 30 (35%) and 31 (7%); (ii) DIB, I2, hm, CH2Cl2, then BF3ÆOEt2,
Ar. With Ar = 4-iodoanisole: 32 (43%); with Ar = 2-chloroanisole: 33
(71%); with Ar = 2-bromoanisole: 34 (76%); with Ar = 2-iodoaniline
methyl carbamate: 35 (30%); with Ar = 2-iodo-O-tertbutyldiphenylsi-
lylphenol: 36 (50%); (iii) TBAF, THF, 61%; (iv) Ac2O, Py, 65%.
The influence of the aromatic substituents was studied
afterwards. Different halogen and X groups were intro-
duced, and their positions changed, as shown in Scheme
4. Thus, starting from serine derivative 28, a one-pot
fragmentation–arylation reaction11 was carried out,
yielding the arylglycines 29–36. The silyl ether 36 was
then cleavaged to give the phenol 37, and this was trans-
formed into the acetate 38.
compound 5. However, the bromo analogue 34 present-
ed similar activity.
Some changes in the aromatic X group were also per-
formed. When X = NHCO2Me or OH (products 35
and 37) the compounds were not active. On the con-
trary, compound 38 (X = OAc) retained some cytotoxic-
ity, and the analogue 36 (X = OTBDPS) was more
active than the lead compound 5. Clearly, both substitu-
ent volume and polarity are important for activity.
The cytotoxic activity of all derivatives was studied
(Fig. 3), and those compounds in which the nitrogen func-
tion was replaced by hydrogen, oxygen or sulfur functions
(products 10–16) proved to be inactive. The phthalimide
17 also was inactive, suggesting that hydrogen bonding
was needed for activity. However, the hydrazide 18 and
urazole 19 showed little activity, probably due to changes
in the position of the NH group or by polarity reasons,
since the heteroatom attached to the NH group exerts a
strong electron-withdrawing effect. The sulfonamide 20,
which was able to form hydrogen bonds, was cytotoxic.
The derivatives 21–27, where the ester group has been re-
placed by amide, hydroxymethyl, acid, ketone, or phos-
phonate groups, showed little activity.
With these results in hand, chiral derivatives12 of the most
active arylglycine, compound 36, were prepared (Scheme
5). The synthesis was performed from commercial L-(4-
hydroxyphenyl)glycine (S)-39, which was N-benzoylated
to compound (S)-40 and then esterified and O-silylated.
The resulting product (S)-41 was iodinated, affording
product (S)-36 in satisfactory yield. The synthesis of the
D-enantiomer (R)-36 was performed in a similar way.
The derivatives 29–38, where modifications of the aro-
matic ring were made, showed that changes in the iodine
position were deleterious for activity, and thus products
29–31 showed little cytotoxicity.
The L-enantiomer (S)-36 showed similar activity to the
D-epimer (R)-36, although this result could be due to
in vivo epimerization.
The changes in the halogen were also important, and the
chloro derivative 33 was less active than the lead
The mechanism of action for the arylglycines and their
simplified analogues was then studied in S. cerevisiae.

6076
A. Boto et al. / Bioorg. Med. Chem. Lett. 16 (2006) 6073–6077
H
H
N
200
180
160
140
120
100
80
N
COOH
COOMe
X
MCF7 media 1 uM
MCF7 media 10 uM
Z
Bz
ii, iii
OH
O
TBDPS
(S)-41 X = H
(S)-36 X = I
(S)-39 Z = H
(S)-40 Z = Bz
iv
i
60
40
20
0
Scheme 5. Synthesis of chiral arylglycines. Reagents and conditions: (i)
BzCl, NaHCO3 (aq satd), THF; (ii) MeOH, AcCl (80% for the two
steps); (iii) TBDPSCl, imidazole, CH2Cl2, 96%; (iv) DIB, I2, CH2Cl2,
dark, 45% (plus 23% recovered starting material).
cytotoxic compounds which are currently used in cancer
therapy.13
200
180
160
140
120
100
80
NCI-H460 media 1 uM
NCI-H460 media 10 uM
To further characterise the toxicity mechanism of these
compounds, we explored the effect of compounds 5
and 34 on cell growth and viability, using a set of isogen-
ic yeast strains defective in DNA repair (rad52, rad52-
ku80 and rad14 strains) and DNA damage checkpoint
(mec1-1 and rad53-11) pathways (Fig. 4).
60
40
20
0
As shown in Figure 4, treatment with both compounds
affected yeast cell growth (Fig. 4A) and viability
(Fig. 4B) in all strains, indicating that compounds 5
and 34 were cytotoxic in yeast. Interestingly, the growth
defect produced by exposure to arylglycines 5 and 34
was exacerbated in the rad52 and rad52,ku80 mutants,
respectively. These strains are hypersensitive to
alterations in a number or processes involved in genome
stability such as DNA replication, DNA damage signal-
ing, double strand break repair, chromatin structure and
200
180
160
140
120
100
80
SF268 media 1 uM
SF268 media 10 uM
A
rad53-11
60
rad52
40
20
0
rad52,ku
rad14
product 5
product 34
mec1-1
-4
-3
-2
-1
0
1
Figure 3. Table of cell viability with respect to control (%) of tumour
cell lines SF680, MCF7 and NCI-H460 in the presence of represen-
tative compounds 5, 13, 16, 17, 27, 29, 32, 34, 36 and 37.
0
B
8
1
1
u
1
-
-
k
,
3
2
1
4
2
5
5
c
1
5
d
T
e
d
d
d
a
a
r
a
m
W
r
a
r
r
Cancers accumulate a large number of genetic changes
during progression towards malignancy due to an intrin-
sic genetic instability. These genetic alterations frequent-
ly affect DNA repair and cell cycle checkpoint pathways
thereby increasing tumour cell sensitivity towards DNA
damaging agents. As these pathways are conserved
throughout evolution one can explore the therapeutic
potential of molecules by using a panel of isogenic yeast
strains with defined genetic alterations in DNA repair or
checkpoint functions. Indeed, this approach has proven
to be extremely useful in the analysis of well-known
control
product 5
Figure 4. (A) Inhibitory effects of compounds 5 and 34 on the growth
of selected DNA damage repair/checkpoint yeast mutants in liquid
media (log[% yeast mutant growth/% WT growth]). (B) Growth of
selected DNA damage repair/checkpoint yeast mutants in solid media
(control vs compound 5).

A. Boto et al. / Bioorg. Med. Chem. Lett. 16 (2006) 6073–6077
6077
2. These transporters can be used to introduce cytotoxic
compounds into the tumour cell, or inversely, can be
destroyed with antibodies or inhibited with certain
drugs, depriving the cell of vital nutrients. For an
example, see: Palmer, C. F.; Mc Evoy-Bowe, E.;
Meehan, G. V.; Piva, T.; Rigano, D.; Favot, P.; West,
M.; Mccabe, M.; Grenville, P.; Miller, D. J. Patent
WO9003399, 1990.
3. (a) Mittal, S.; Song, X.; Vig, B. S.; Landowski, C. P.; Kim,
I.; Hilfinger, J. M.; Amidon, G. L. Mol. Pharm. 2005, 2,
37; (b) Balcome, S.; Park, S.; Quirk-Dorr, D. R.; Hafner,
L.; Phillips, L.; Tretyakova, N. Chem. Res. Toxicol. 2004,
17, 950; (c) Bissery, M. C. Patent WO0168066, 2001; (d)
Haines, D. R.; Fuller, R. W.; Ahmad, S.; Vistica, D. T.;
Marquez, V. E. J. Med. Chem. 1987, 30, 542.
4. (a) Goldenberg, G.; Begleiter, A. Pharmacol. Ther. 1980,
8, 237; (b) Goldenberg, G.; Lam, H. Y. P.; Begleiter, A.
J. Biol. Chem. 1979, 254, 1057; (c) Vistica, D. T.; Toal,
J. N.; Rabinovitz, M. Biochem. Pharmacol. 1987, 27,
2865.
assembly, chromosome segregation, telomere mainte-
nance and metabolism of reactive oxygen species.14
Therefore, these data are consistent with compounds 5
and 34 behaving as genotoxins. Consequently, an attrac-
tive hypothesis would be that compounds 5 and 34
induce genomic instability through alteration of one/
some of the above-mentioned processes. This hypothesis
is currently being addressed.
In summary, new cytotoxic amino acids derived from
3-iodo- or 3-bromo-phenylglycine are described herein.
Many of these compounds were synthesized using one-
pot fragmentation–arylation or halogenation reactions
with hypervalent iodine reagents. The mechanism of ac-
tion of the (haloaryl)glycines was studied in S. cerevisi-
ae, showing that these compounds were genotoxic.
Acknowledgments
5. Strazzolini, P.; Dall’Arche, M. G.; Zossi, M.; Pavsler, A.
Eur. J. Org. Chem. 2004, 4710, and references cited
therein.
6. (a) Conti, P.; Roda, G.; Stabile, H.; Vanoni, M. A.;
Curti, B.; DeAmici, M. Il Farmaco 2003, 58, 683; (b)
Antczak, C.; Beauvois, B.; Monneret, C.; Florent, J.-C.
Bioorg. Med. Chem. 2001, 11, 2843; (c) Mzengeza, S.;
Yang, C. M.; Whitney, R. A. J. Am. Chem. Soc. 1987,
109, 276.
The work carried out in CSIC was supported by FAES
FARMA S.A. (Research Contract OTT-2004066).
J.A.G. and M.M.M. have enjoyed a contract with FAES
FARMA, S.A. C.J.S. thanks a CSIC predoctoral fellow-
ship, also financed by FAES FARMA S.A.
7. Graber, R.; Losa, G. A. Int. J. Cancer 1995, 62, 443, and
references cited therein.
Supplementary data
8. (a) Wang, H. P.; Lee, O.; Lee, S. J. Patent WO9904792,
1999; See also: (b) Adamczyk, M.; Reddy, R. E. Tetra-
hedron: Asymmetry 2001, 12, 1047.
9. Hatanaka, T.; Ohizumi, I.; Nezu, J.-I. Patent EP1561471,
2005.
Spectroscopic data for selected compounds, other
cytotoxic activities, materials and general procedures
for the determination of the cytotoxic activity and the
mechanism of action studies. Supplementary data
associated with this article can be found, in the online
version, at doi:10.1016/j.bmcl.2006.08.111.
´
10. Boto, A.; Gallardo, J. A.; Hernandez, R.; Saavedra, C. J.
Tetrahedron Lett. 2005, 46, 7807.
´
´
11. Boto, A.; Hernandez, R.; Montoya, A.; Suarez, E.
Tetrahedron Lett. 2002, 43, 8269.
12. For a review on chiral arylglycines, see: Williams, R. M.;
Hendrix, J. A. Chem. Rev. 1992, 92, 889.
References and notes
13. Simon, J. A.; Szankasi, P.; Nguyen, D. K.; Ludlow, C.;
Dunstan, H. M.; Roberts, C. J.; Jensen, E. L.; Hartwell, L.
H.; Friend, S. H. Cancer Res. 2000, 60, 328.
14. Pan, X.; Ye, P.; Yuan, D. S.; Wang, X.; Bader, J. S.;
Boeke, J. D. Cell 2006, 124, 1069.
1. For a review on the subject, see: Metabolism and action of
amino acid analog anti-cancer agents Ahluwalia, G. S.;
Grem, J. L.; Hao, Z.; Cooney, D. A. Pharmacol. Ther.
1990, 46, 243.