Prodrugs of butyric acid. Novel derivatives possessing increased aqueous solubility and potential for treating cancer and blood diseases

Article abstract of DOI:10.1016/S0223-5234(00)01199-5

The synthesis and biological activities of acidic, basic and neutral types of butyric acid (BA) prodrugs possessing increased aqueous solubility are described. The compounds are butyroyloxyalkyl derivatives of carboxylic acids, which possess functionalities suitable for aqueous solubilization. The anticancer activity of the prodrugs in vitro was evaluated by examining their effect on the growth of human colon, breast and pancreatic carcinoma cell lines, and their solubility in aqueous media was determined. The most promising compounds, with respect to activity and solubility, were found to be the butyroyloxymethyl esters of glutaric 2a and nicotinic acids 4a and phosphoric acid as its diethyl ester 10a, which displayed IC₅₀ values of 100 μM or lower. These prodrugs are expected to release formaldehyde upon metabolic hydrolysis. The corresponding butyroyloxyethyl esters (2b, 4b and 10b) that release acetaldehyde upon metabolism were significantly less potent. A similar correlation was observed for growth inhibition of the human prostate carcinoma cell lines PC-3 and LnCap and for induction of differentiation and apoptosis in the human myeloid leukemia cell line HL-60. The higher biological activity of the formaldehyde-releasing prodrugs 2a and 10a was further confirmed when induction of hemoglobin (Hb) synthesis in the human erythroleukemic cell line K562 was measured. Moreover, a therapeutic index (IC₅₀/ED₅₀) of ca. 5 was observed. The acute i.p. toxicity LD₅₀ in mice for 2a, 2b, 10a and 10b was similar and in the range of 400-600 mg kg^-1. The results obtained support the potential use of the butyric acid prodrugs for the treatment of neoplastic diseases and β-globin disorders.

Full text of DOI:10.1016/S0223-5234(00)01199-5

Eur. J. Med. Chem. 36 (2001) 63–74
´
© 2001 Editions scientifiques et me´dicales Elsevier SAS. All rights reserved
PII: S0223-5234(00)01199-5/FLA
Prodrugs of butyric acid. Novel derivatives possessing increased aqueous
solubility and potential for treating cancer and blood diseases
Abraham Nudelmana*, Elisheva Gnizia, Yifat Katza, Revital Azulaia, Mirit Cohen-Ohanaa,
Regina Zhuka, Sanford R. Sampsonb, Leah Langzamb, Eitan Fibachc, Eugenia Prusc,
Victoria Pugachd, Ada Rephaelid
aChemistry Department, Faculty of Life Sciences, Bar Ilan University, Ramat Gan 52900, Israel
bGonda-Goldschmied Center, Faculty of Life Sciences, Bar Ilan University, Ramat Gan 52900, Israel
cDepartment of Hematology, Hadassah University Hospital, Ein-Kerem, Jerusalem, Israel
dFelsenstein Medical Research Center Petach Tikva-Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel
Received 30 August 2000; revised 30 October 2000; accepted 30 October 2000
Abstract – The synthesis and biological activities of acidic, basic and neutral types of butyric acid (BA) prodrugs possessing
increased aqueous solubility are described. The compounds are butyroyloxyalkyl derivatives of carboxylic acids, which possess
functionalities suitable for aqueous solubilization. The anticancer activity of the prodrugs in vitro was evaluated by examining their
effect on the growth of human colon, breast and pancreatic carcinoma cell lines, and their solubility in aqueous media was
determined. The most promising compounds, with respect to activity and solubility, were found to be the butyroyloxymethyl esters
of glutaric 2a and nicotinic acids 4a and phosphoric acid as its diethyl ester 10a, which displayed IC50 values of 100 mM or lower.
These prodrugs are expected to release formaldehyde upon metabolic hydrolysis. The corresponding butyroyloxyethyl esters (2b, 4b
and 10b) that release acetaldehyde upon metabolism were significantly less potent. A similar correlation was observed for growth
inhibition of the human prostate carcinoma cell lines PC-3 and LnCap and for induction of differentiation and apoptosis in the
human myeloid leukemia cell line HL-60. The higher biological activity of the formaldehyde-releasing prodrugs 2a and 10a was
further confirmed when induction of hemoglobin (Hb) synthesis in the human erythroleukemic cell line K562 was measured.
Moreover, a therapeutic index (IC50/ED50) of ca. 5 was observed. The acute i.p. toxicity LD50 in mice for 2a, 2b, 10a and 10b was
similar and in the range of 400–600 mg kg–1. The results obtained support the potential use of the butyric acid prodrugs for the
´
treatment of neoplastic diseases and b-globin disorders. © 2001 Editions scientifiques et me´dicales Elsevier SAS
prodrugs / butyric acid / anticancer / water solubility
1. Introduction
does BA [1, 2]. Moreover, AN-9 penetrates 100-fold
faster than BA into cancer cells in vitro [6].
Previous publications from our laboratories have
described the synthesis and anticancer activity of bu-
tyric acid (BA) prodrugs, that have been shown to
modulate gene expression, induce histone hyperacety-
lation, differentiation, and apoptosis of cancer cells
[1–5]. Pivaloyloxymethyl butyrate (AN-9), the best-
studied prodrug, affects cancer cells about 100-fold
faster and at about 10-fold lower concentration than
Esterification of BA improves its permeability
across cell membranes and enables efficient delivery
of BA to a subcellular target. AN-9 was shown to
inhibit the proliferation of a variety of cancer cell
lines [1, 2] as well as primary human tumors, includ-
ing colorectal, breast, lung, ovarian, renal cell and
urinary bladder [7]. The response to AN-9 was dose-
dependent and was significantly greater than that of
BA. AN-9 displayed low toxicity in mice and was
effective in prolonging survival of mice bearing
melanoma, lung carcinoma and monocytic leukemia
* Correspondence and reprints
E-mail addresses: nudelman@mail.biu.ac.il (A. Nudelman);
adarep@post.tau.ac.il (A. Rephaeli).

64
A. Nudelman et al. / European Journal of Medicinal Chemistry 36 (2001) 63–74
[1, 2, 8]. It induces transient hyperacetylation of his-
tones leading to relaxation of the chromatin structure
and allowing access to transcription factors [3, 4, 9].
This activity is likely to be an important mechanism
by which the prodrugs exert their effect on gene
modulation. AN-9 modulates the expression of the
early regulatory genes, c-myc and c-jun, and the tu-
mor suppressor gene RB as well as the anti-apoptotic
gene bcl-2 in WEHI and HL-60 cells [3, 5, 10]. A
synergistic effect of AN-9 and daunorubicin was ob-
served in murine monocytic leukemia cells resulting in
markedly prolonged survival of mice with monocytic
leukemia [8]. AN-9, formulated in lipid emulsion,
labeled – PIVANEX®, in a phase I clinical study
displayed low toxicity and was reported to have an
estimated MTD of 2.69 g m–2 day–1 [11]. It is
presently in Phase II clinical studies with non-small-
cell lung carcinoma.
Another important application of BA prodrugs is
in the treatment of b-globin disorders [12–16]. BA-
prodrugs have been shown to increase hemoglobin
(Hb) expression in several model systems. Among
them, butylidene dibutyrate (AN-10) was found to be
the most effective [17]. In humans, Hb production is
characterized by two major ‘switches’ that involve the
production of different b-type chains at different
stages of development. The production of embryonic
Hb switches into fetal hemoglobin (HbF) after the
first two months of gestation and then again into
adult hemoglobin before birth (HbA). In adult life,
the major Hb is HbA, while HbF comprises <1% of
the total Hb content. Inherited blood disorders,
known as b-hemoglobinopathies, are characterized by
reduced, or lack of, production of adult-type b-chains
in b-thalassemia or in the formation of structurally
abnormal b-globin chain in sickle cell anemia (SCA).
In both cases, an increase in HbF production amelio-
rates the clinical symptoms of the underlying disease.
Although much is known about the biochemical and
molecular basis of these inherited blood diseases in
man, similar progress in the development of suitable
therapies has not been made [13].
The design of the AN-9 and AN-10, as well as other
early analogs [1], resulted in highly lipophilic deriva-
tives better able to cross cell membranes than BA [6].
However, they are virtually devoid of water solubility,
requiring non-aqueous media for their clinical formu-
lation. This report describes the synthesis and biolog-
ical activity of novel types of BA prodrugs that may
be more readily formulated in aqueous media, retain
high anticancer activity and augment production of
HbF.
2. Chemistry
In view of our earlier successful experience with
acyloxyalkyl ester prodrugs of BA [1–6], analogous
derivatives possessing water solubilizing functional
groups were designed. These acetal diester prodrugs
(V) are known to release equimolar amounts of the
acidic drug (I), an aldehyde (R1CHO, VI) and a
second acid (R2COOH, IV) upon metabolic break-
down [19]. The latter is the acidic component of the
‘carrier’. The selection of ‘carrier’ moieties that upon
metabolic cleavage would release fragments possess-
ing minimal toxicity was of major importance in the
design of the new prodrugs.
For this purpose, the acidic ‘carriers’ chosen were
natural compounds found in the human body such as
p-amino-benzoic acid (PABA) (LD50 po in rats >6 g
kg–1) [20a], nicotinic acid (LD50 sc in rats 5 g kg–1)
[20b], and phosphoric acid as its diethyl ester. Deriva-
tives of other acids expected to possess low toxicity,
such as glutaric acid (found in green sugar beets [20c],
LD50 po in mice 6 g kg–1, [21]), 2-(2-methoxy-
ethoxy)acetic and [2-(2-methoxyethoxy)ethoxy]acetic
acids were also prepared. Although no prohibitive
toxicity has been reported with the use of prodrugs
that release formaldehyde such as Pivampicillin [20d]
or PIVANEX®, for comparative purposes acy-
loxymethyl derivatives that release formaldehyde
(R1 =H) and the corresponding 1-acyloxyethyl
derivatives that release acetaldehyde (R1 =Me) were
prepared (figure 1).
In clinical trials conducted for the treatment of
b-thalassemia and SCA, BA salts were reported to
elevate expression of HbF [14, 16], thereby favorably
modifying the disease symptoms. BA prodrugs, which
possess low toxicity, may provide significant advan-
tages for treatment of b-thalassemia and SCA, as
compared to the cytotoxic agent hydroxyurea cur-
rently approved for the chronic treatment of SCA
patients [18].
The novel BA prodrugs prepared are of acidic,
basic and neutral types.
2.1. Acidic derivatives
Chloromethyl butyrate, 1a, was prepared from bu-
tyryl chloride/paraformaldehyde/ZnCl2 at 80 °C, as de-

A. Nudelman et al. / European Journal of Medicinal Chemistry 36 (2001) 63–74
65
scribed for chloromethyl pivalate [22], whereas 1-
chloroethyl butyrate, 1b, was prepared from butyryl
chloride/acetaldehyde/ZnCl2 at 0 °C [23]. Reaction of 1a
and 1b with glutaric acid gave 2a and 2b accompanied
by the respective neutral side products 3a and 3b stem-
ming from esterification at both ends of the glutaric acid
(figure 2).
The water solubilities of the prodrugs were deter-
mined as described in the legend of table I and were
compared to those of AN-9 and AN-10. The latter
exhibited poor solubility (<1 mg mL–1), whereas the
acidic prodrugs 2a and 2b were found to be slightly
soluble per se, their alkali metal or lysine salts were
highly soluble. Basic prodrugs 4a and 5 were poorly
soluble, but their acid addition salts were highly soluble.
Neutral prodrugs 8 and 9 derived from methoxy-alkoxy-
acetic acids, and 10a and 10b derived from diethyl
phosphate, had significantly improved solubilities at-
tributed to their high oxygen content. As expected,
methylidene derivatives 2a and 10a were more soluble
than the corresponding ethylidene ones 2b and 10b
(table I).
2.2. Basic derivatives
Reaction of nicotinic acid with 1a or 1b, led to the
respective nicotinic acid esters 4a and 4b (figure 2).
Reaction of PABA with 1a gave 5. The basic prodrugs
4a, 4b and 5 were converted into their respective acid
addition salts with inorganic acids.
2.3. Neutral derivatives
3. Biology
Since polyalkoxyacetic acid derivatives have been re-
ported to provide water solubility to their esters [24],
prodrugs 8 and 9 were prepared from chloromethyl
butyrate, 1a, and 2-(2-methoxyethoxy)acetic and [2-(2-
methoxyethoxy)ethoxy]acetic acids, respectively (figure
2). Prodrugs 10a and 10b, which upon hydrolysis are
expected to give BA, the corresponding aldehydes, etha-
nol and phosphoric acid as final metabolites, were pre-
pared by esterification of 1a and 1b with diethyl
phosphate [25]. Other neutral, water insoluble deriva-
tives, 3a, 3b, 7a and 7b, were also included in the
comparative activity studies. The latter two, derived
from PABA, were prepared from amide 6.
3.1. Growth inhibition of human colon, breast and
pancreatic cancer cell lines
Growth inhibition activity on the colon HT-29, breast
MCF-7 and pancreatic MIAPaCa2 carcinoma cell lines
was determined by the sulforhodamine blue (SRB) assay
(table II). Similar tests have been used by the NCI for in
vitro screening of potential anticancer drugs [27].
Among the group of acidic prodrugs the most active one
was mono-butyroyloxymethyl glutarate 2a. In the group
of basic prodrugs the nicotinic acid derivative 4a dis-
played similar activity to 2a. The PABA derivative 5
was less active than 4a. In the group of the neutral
prodrugs, compounds 3a, 8, 9 and 10a had antiprolifer-
ative activity comparable to that of 2a.
Although on a molar basis, 2a releases 1 equiv. of BA
upon hydrolysis, whereas 3a releases 2 equiv. of the
acid, both compounds were found to display similar
activity. This indicates that the total number of equiva-
lents of BA present in the molecule is not the determin-
ing factor in the potency of prodrug. The lack of
correlation between the number of BA equivalents re-
leased and the activity of the compounds has been
observed previously, e.g. AN-9 vs. AN-10 [1]. The lower
activity of 7a as compared to 5 could be attributed to
slow intracellular amide hydrolysis. For practical pur-
poses, 2a and 5 have the advantage of lower molecular
weight and higher aqueous solubility than 3a and 7a,
respectively.
It should be noted that exposure to the nicotinic acid
derivatives 4a and 4b caused a potent, albeit transient
Figure 1. General scheme for synthesis and hydrolysis of
butyric acid (BA) prodrugs.

66
A. Nudelman et al. / European Journal of Medicinal Chemistry 36 (2001) 63–74
Figure 2. Acidic, basic and neutral prodrugs of butyric acid (BA).
peripheral skin vasodilation of the persons handling
these compounds and therefore they were excluded from
further evaluation. This effect is similar to that reported
for other nicotinic acid esters and amides, such as
Nicametate [20c].
that intracellularly released formaldehyde affects cancer
cells specifically, without significantly contributing to
the overall toxicity. Therefore, the formaldehyde releas-
ing BA-prodrugs, as anticancer agents, have an advan-
tage over the acetaldehyde-releasing ones.
All the BA prodrugs that release formaldehyde upon
hydrolysis were found to be considerably more active
than the corresponding analogs that release acetalde-
hyde. Since formaldehyde is a more reactive species, it
may contribute to the higher antiproliferative activity
observed in vitro for the former compounds (table II). It
is reasonable to assume that the released formaldehyde
may also contribute to the in vivo toxicity in mice. To
clear up this assumption, compounds 2a, 2b, 10a, and
10b were evaluated for in vivo toxicity in mice. How-
ever, the acute toxicity values of a single i.p. dose (LD50
in Balb-c mice), of all four compounds were similar
ranging between 400–600 mg kg–1. It can be concluded
Although 2a and 2b displayed satisfactory aqueous
solubility, they were converted to their lysine salts to
further increase their solubility profile. The salts were
indeed more soluble and had lower in vivo toxicity than
the corresponding free acids, (LD50>1000 mg kg–1), but
were found to be less potent inhibitors of human
prostate cancer cells than the respective free acids (table
III).
In view of the water solubility and activity profiles of
the acidic 2a and neutral 10a prodrugs (tables I and II),
subsequent biological evaluations were expanded for
these compounds in parallel with their analogs 2b and
10b, as described below.

A. Nudelman et al. / European Journal of Medicinal Chemistry 36 (2001) 63–74
67
Table II. Proliferation inhibition of human colon-HT-29,
breast-MCF-7 and pancreatic-MIAPaCa2 cancer cell lines by
the acidic, basic and neutral BA prodrugs.a
3.2. Inhibition of proliferation of human prostate cell
lines
The inhibition of proliferation of the hormone-depen-
dent (LnCap) and hormone-independent (PC-3) human
prostate carcinoma cell lines, was determined using an
XTT assay [28]. The results obtained confirm the higher
antiproliferative activity (at least 7-fold) of the methyli-
dene derivatives 2a and 10a as compared to their respec-
tive ethylidene analogs 2b and 10b (table III).
Prodrug
IC50 mMb
HT-29
MCF-7
MIAPaCa2
2a
2b
3a
3b
4a
4b
5
7a
7b
8
98
3060
95
110
1900
94
\3000
126
2900
390
570
1540
95
160
95
2900
32
1900
84
\3000
58
1500
220
450
1500
185
210
91
1270
\3000c
100
3100
190
360
1450
164
210
3.3. Induction of differentiation and apoptosis, and
inhibition of proliferation of human leukemia cells
The differentiation activity of prodrugs 2a, 2b, 10a
and 10b on the myeloid leukemia cell line HL-60 was
determined by nitro blue tetrazolium (NBT) reduction
activity, and inhibition of proliferation was evaluated by
3[H]-thymidine incorporation [1, 2]. Apoptosis induction
was evaluated by FITC-conjugated annexin and ana-
lyzed by flow cytometry [29]. The results demonstrate
that in addition to inhibiting proliferation, the formalde-
hyde-releasing prodrugs 2a and 10a, also displayed
higher potency (at least 3-fold), than 2b and 10b in
promoting differentiation and apoptosis (table IV).
9
10a
10b
100
2420
a
Cells at a density of 5×104 cells mL–1 were plated in 96-well
plates and after 24 h the test drugs were added. The drugs,
dissolved in DMSO, were diluted with growth media to give a
final DMSO concentration in the well of 0.1%. Ten prodrug
concentrations ranging from 4 to 0.008 mM were used in
triplicate. The negative controls were untreated cells and posi-
tive controls were cells treated with 5 mM doxorubicin. After
seven days of incubation the samples were stained with 0.4%
SRB and read for absorbance at 570 nm.
b
The correlation factors obtained for the above dose–re-
sponse studies were \0.9.
c
At concentration above 3 mM, 3b could not be solubilized in
Table I. Aqueous solubility of the prodrugs.a
the media used.
Prodrug
Solubility mg mL–1
3.4. Induction of Hb synthesis
AN-9
AN-10
2a
B1
B1
22
Induction of Hb synthesis is considered as a marker
of differentiation activity in the erythroleukemic cell
line, K562 [30]. Since salts of BA have been reported to
elevate expression of Hb [14–16], the effect of the
prodrugs on the induction of cellular Hb (ED50) was
evaluated by the HPLC method described in [31]. Paral-
lel inhibition of proliferation of these cells (IC50) was
2b
6
3a
B1
B1
3.5
3b
4a
4b
B1
B1
B1
B1
\40
\40
18.5
3
5
7a
3
measured by [H]-thymidine incorporation into cellular
7b
DNA, which took place at much higher doses (table V).
While BA increased the production of Hb in the cells
with an ED50 of 179 63 mM, prodrugs 2a and 10a
displayed ED50 values of ca. 30 mM and IC50 above 100
mM, giving an overall IC50/ED50 ratio of about 5. Com-
pound 2b had lower activity. It is important for treat-
ment of chronic b-hemoglobinopathies, that the drugs
should induce Hb synthesis with high potency and dis-
play low cytotoxicity, in particular toward Hb prod-
8
9
10a
10b
a
Solubility in 25 mM phosphate buffer, pH 7, at 25 °C was
measured by the shake flask method [26], suspending 40 mg of
the prodrug in 1 mL of buffer and increasing the volume of
buffer, up to 40 mL, until complete dissolution occurred.
Compounds that did not dissolve at this concentration are
indicated as possessing solubility of B1 mg mL–1.

68
A. Nudelman et al. / European Journal of Medicinal Chemistry 36 (2001) 63–74
Table III. Growth inhibition of human prostate cancer cell
Table V. The effect of BA prodrugs on Hb synthesis and
inhibition of proliferation in the human erythroleukemic cell
line K562.a
lines.
Prodrug
IC50 mMa
Compound
ED50 mM
IC50 mM
PC-3
LnCap
8995
650917
12297
9819123
569928
1081982
2a
2297
89915
2690.3
ndb
164
601
119
553
2a
2b
10a
10b
2a lysine salt
2b lysine salt
150922
13959420
128933
32499421
ndb
2b
10a
10b
a
Hb synthesis was determined by HPLC and proliferation
ndb
inhibition was measured by [H]-thymidine incorporation into
3
K562 cells, as described in the experimental section. The
linearity of the IC50 values obtained was high as is evident by
the high correlation coefficient ]0.90.
a
IC50 values were determined by the XTT assay after 48 h of
exposure to the prodrugs. The IC50 values represent the aver-
age and standard deviation of at least three independent
experiments.
b
not determined.
b
not determined.
series of derivatives ranging from highly lipophilic to
hydrophilic were obtained. The compounds possess-
ing moderate water solubility in the range of 3–22 mg
mL–1 were found most advantageous for drug formu-
lation. In addition, other structural features of the
compounds turned out to be important for their bio-
logical activities. A comparison of inhibition of hu-
man carcinoma cell lines by the prodrugs that release
formaldehyde (compounds 2a, 4a, 7a and 10a) with
their acetaldehyde-releasing counterparts (compounds
2b, 4b, 7b and 10b) showed that the former were
significantly more active. A similar correlation was
found for induction of differentiation and apoptosis
in the human myeloid leukemia cell line HL-60 as
well as for induction of Hb synthesis in the ery-
throleukemic cell line K562. We postulate that this
difference is related to the higher reactivity of the
released formaldehyde. Although formaldehyde-re-
leasing prodrugs display higher in vitro activity com-
pared to acetaldehyde releasing ones, both series of
compounds displayed similar acute toxicity LD50 val-
ues in mice. The mechanism by which the aldehyde
moiety of the prodrugs contributes to the anticancer
activity in now under investigation.
ucing cells. Prodrugs 2a and 10a meet these criteria. It is
worth noting that to date all the agents that induced Hb
accumulation in K562 cells were also effective in induc-
ing HbF in vivo (both in experimental animals and
patients) [15, 16]. Therefore, the results obtained in this
experimental model suggest that 2a and 10a may have a
therapeutic efficacy in patients with b-globin disorders
(e.g. sickle cell anemia and b-thalassemia).
4. Conclusions
The chemical design of BA prodrugs led to modifi-
cation of their physico–chemical properties so that a
Table IV. Induction of apoptosis and differentiation, and
inhibition of proliferation of HL-60 cells.
Prodrug Apoptosis Proliferation
Differentiation
ED50, mMa IC50, mMb
ED50, mMc
2a
77
220
70
72
205
50
81928
150942
5899.6
218954
2b
10a
10b
230
192
5. Experimental protocols
a
The values were determined for cells stained with annexin
and propidium iodide and subjected to FACS analysis.
b
Proliferation was determined by 3[H]-thymidine incorpora-
5.1. Chemistry
tion, as described in the experimental section. The data pre-
sented for apoptosis and proliferation were based on a
representative single set of data. The linearity of the data is
reflected by the correlation factor \0.95.
1H-NMR and 13C-NMR spectra were obtained on
Bruker AC-200, DPX-300 and DMX-600 spectrometers.
For CDCl3 and acetone-d6 solutions, chemical shifts are
expressed in ppm downfield from Me4Si used as internal
standard; for D2O solutions the HOD peak was taken as
c
The ED50 values for differentiation, measured by % NBT
positive cells, were derived from linear regression of three
dose–response experiments.

A. Nudelman et al. / European Journal of Medicinal Chemistry 36 (2001) 63–74
69
l 4.80 (1H-spectra), or the peak of a small amount of
5.1.3. Diethyl phosphate [25]
added MeOH taken as l 49.50 (13C). For 31P spectra an
external standard of 85% H3PO4 was used. Multiplicities
in the 13C-NMR spectra were determined by off-reso-
nance decoupling. Mass spectra were obtained on a
Finnigan 4021 spectrometer operating in CI (chemical
ionization), DCI (desorption chemical ionization), EI
(electron impact) or HRMS (high-resolution) modes.
Progress of the reactions was monitored by TLC on
silica gel or alumina. Melting points were determined on
a Fisher–Johns apparatus. Analyses indicated by the
symbols of the elements were within 0.4% of the
theoretical values
NaOH (1N, 100 mL) was added dropwise to an
ice-cold solution of diethyl chlorophosphate (17.3 g, 0.1
mol) in THF. After stirring the mixture for 0.5 h, the
solvent was evaporated and the product obtained was a
clear oily residue (95% yield), which was used without
1
further purification. H-NMR (CDCl3) l 9.92 (brs; 1H,
OH); 4.12 (q; J=7.0 Hz, 2H, CH2Me); 4.09 (q; J=7.0
Hz, 2H, CH2Me); 1.34 (td; J=7.0, 1.0 Hz, 6H, two
Me). 13C-NMR (CDCl3) l 63.63 (d; JCP=5.5 Hz, CH2);
15.96 (d; JCP=7.0 Hz, Me).
5.1.4. 4-Butyrylamino-benzoic acid 6
Butyric anhydride (18 mL, 17.3 g, 1 equiv.) was added
dropwise to an ice-cold solution of 4-amino-benzoic acid
(15.12 g, 109 mmol) in dry pyridine (100 mL) under N2.
The ice bath was removed and the reaction was stirred
overnight, whereby TLC (EtOAc:hexane 2:1, developed
with ninhydrin) showed that most of the acid was
consumed. The pyridine was evaporated, and the white
solid residue was washed with ether, filtered and dried.
The product obtained was a white solid (18 g, 80%
yield) and was used without further purification, m.p.:
236–238 °C. 1H-NMR (acetone-d6) l 9.40 (brs; 1H,
OH); 7.97 (d; J=9.0 Hz, 2H, Ar-H2); 7.79 (d; J=9.0
Hz, 2H, ArꢀH3); 2.38 (t; J=7.5 Hz, 2H, CH2CO); 1.70
(sextet; J=7.5 Hz, 2H, MeCH2); 0.96 (t; J=7.5 Hz,
3H, Me). 13C-NMR (acetone-d6) l 172.39 (s; CONH);
167.28 (s; COOH); 144.60 (s; ArꢀC4); 131.44 (d; ArꢀC2
and ArꢀC6); 125.68 (s; COOH); 119.12 and 119.04 (d;
ArꢀC3 and ArꢀC5); 39.59 (t; CH2CO); 19.40 (t;
MeCH2); 13.92 (q; Me). HRMS (DCI-CH4) 208.0890
calc. 208.0974 (MH+). Anal. C11H13NO3 (C, H).
5.1.1. Butyric acid chloromethyl ester 1a
The ester was prepared by a modification of the
reported procedure [22]. A mixture of butyryl chloride
(944.2 g, 8.86 mol), paraformaldehyde (265.9 g, 8.86
mol) and ZnCl2 (cat.) were stirred at r.t. An exothermic
reaction occurred after 10 min, whereby the temperature
reached 75–80 °C. After 20–25 min, the temperature
dropped and the reaction was heated at 75 °C for 3 h.
The mixture was distilled at 40 Torr, whereby four
fractions having the following boiling points (b.p.) were
collected: a) 47–87 °C, b) 87–91 °C, c) 91–92 °C, d)
>92 °C. The combined fractions c and d were redistilled
to give the product as a colorless oil (769.4 g, 64%
1
yield). H-NMR (CDCl3) l 5.69 (s; 2H, OCH2Cl); 2.36
(t; J=7.5 Hz, 2H, CH2CO); 1.68 (sextet; J=7.5 Hz,
2H, CH2Me); 0.97 (t; J=7.5 Hz, 3H, Me). 13C-NMR
(CDCl3) l 171.39 (s; CO); 68.52 (t; OCH2Cl); 35.78 (t;
CH2CO); 18.04 (t; MeCH2); 13.44 (q; Me).
5.1.2. Butyric acid 1-chloroethyl ester 1b [23]
Cooled acetaldehyde (190 mL, 3.3 mol) was added
dropwise to a cooled (−20 °C, ice-salt bath), stirred
mixture of butyryl chloride (320 g, 3 mol) and ZnCl2
(ca. 0.3 g) under N2 in the course of 0.5 h. At the end of
the addition, the mixture was cooled for an additional 1
h. A 1H-NMR spectrum taken at room temperature
after 1 h showed almost pure product. The crude mate-
rial was dissolved in petroleum ether, and was eluted
through a silica column (silica gel 60 H, 6×8 cm) under
suction. The eluted solution was filtered, evaporated,
and the residue was vacuum distilled (water pump) at
89–105 °C, to give the product (330 g, 74% yield).
1H-NMR (CDCl3) l 6.55 (q; J=6.0 Hz, OCHO, 1H);
2.34 (t; J=7.5 Hz, 2H, CH2CO); 1.78 (d; J=6.0 Hz,
1H, CHMe); 1.68 (sextet; J=7.5 Hz, 2H, MeCH2); 0.96
(t; J=7.5 Hz, 3H, Me).
5.1.5. Pentanedioic acid monobutyryloxymethyl ester 2a
and pentanedioic acid dibutyryloxymethyl ester 3a
Et3N (165 mL, 120 g, 1.2 mol) was added to a
solution of chloromethyl butyrate (133.8 g, 0.98 mol)
and glutaric acid (129 g, 0.98 mol) in acetone (500 mL).
The mixture was refluxed for 12 h, whereupon a large
amount of precipitate formed. The precipitate was
filtered and washed with acetone. The filtrate was evap-
orated and the residue was partitioned with aqueous 1
M K2CO3 (500 mL) at a pH of about 8 and EtOAc (500
mL). The aqueous phase was extracted several times
with EtOAc whilst maintaining the basic pH by continu-
ous addition of solid K2CO3. The organic phase was
washed (3×100 mL) with water at pH 8, and the
washings were added to the aqueous phase. The com-
bined aqueous phase was acidified with 2 M HCl and

70
A. Nudelman et al. / European Journal of Medicinal Chemistry 36 (2001) 63–74
was extracted with EtOAc (4×200 mL). The organic
phase was collected, treated with charcoal, dried
(MgSO4) and evaporated to give the product 2a as an
oil, 110 g, 0.47 mol (48% yield) which was used without
(t; J=7.5 Hz, 6H, two Me). 13C-NMR (CDCl3) l
171.54 (s; MeCH2CH2CO); 170.79 (s; CH2CO); 88.38
(d; OCHO); 35.90 (t; MeCH2CH2); 32.82 (t;
CH2CH2CH2); 19.50 (superimposed
q
and t;
1
further purification. H-NMR (CDCl3) l 5.76 (s; 2H,
CH2CH2CH2, O2CHMe); 18.11 (t; MeCH2CH2); 13.46
OCH2O); 2.47 and 2.44 (two t; J=7.5 Hz, 4H,
CH2CH2CH2); 2.35 (t; J=7.5 Hz, 2H, MeCH2CH2);
1.98 (quintet; J=7.5 Hz, 2H, CH2CH2CH2); 1.67 (sex-
tet; J=7.5 Hz, 2H, MeCH2CH2); 0.95 (t; J=7.5 Hz,
3H, Me). 13C-NMR (CDCl3) l 178.72 (s; COOH);
172.35 (s; MeCH2CH2CO); 171.61 (s; CH2CO); 79.24 (t;
OCH2O); 35.83 (t; MeCH2CH2); 32.83 and 32.79 (two t;
CH2CH2CH2); 19,52 (t; CH2CH2CH2); 18.14 (t;
MeCH2CH2); 13.51 (q, Me). MS (CIꢀNH3) 250 (MNH4+
), 233 (MH+). Anal. C10H16O6 (C, H). The EtOAc
extract from the basic aqueous solution was dried
(MgSO4) and evaporated to give an oily residue (76 g)
containing traces of chloromethyl butyrate (as detected
by NMR). The residue was kugelrohr distilled (190 °C/
0.5Torr) to give 70 g, 0.20 mole (20% yield) of pure
bis-ester 3a. 1H-NMR (CDCl3) l 5.74 (s; 4H, two
OCH2O); 2.43 (t; J=7.5 Hz, 4H, CH2CH2CH2); 2.33 (t;
J=7.5 Hz, 4H, two MeCH2CH2); 1.96 (quintet; J=7.5
Hz, 2H, CH2CH2CH2); 1.65 (sextet; J=7.5 Hz, 4H, two
MeCH2CH2); 0.94 (t, 6H, two Me).
(q; Me).
5.1.7. Butyric acid 2-(2-methoxy-ethoxy)-acetoxymethyl
ester 8
Chloromethyl butyrate (5 mL, 5.4 g 30.5 mmol) was
added to a solution of 2-(2-methoxyethoxy)acetic acid
(4.5 mL, 5.45 g 40.6 mmol) in acetone (30 mL), followed
by the dropwise addition of Et3N (6.6 mL, 1.2 equiv.).
The mixture was heated at 40 °C for 20 h. The precipi-
tate was filtered and washed with acetone. The filtrate
was evaporated, and the residue (11 g) was washed
thrice with water, thrice with 5% NaHCO3 and thrice
with brine, to give the crude product as a colorless oil,
4.6 g. The crude product was distilled at 140 °C/1 Torr,
1
to give the product as an oil, 3.8 g (44% yield). H-
NMR (CDCl3) l 5.83 (s; 2H, OCH2O); 4.30 (s; 2H,
COCH2O); 3.69 (m; 4H, O(CH2)2O); 3.43 (s; 3H, OMe);
2.42 (t; J=7.3 Hz, 2H, MeCH2CH2); 1.62 (sextet; J=
7.3 Hz, 2H, MeCH2); 0.90 (t; J=7.3 Hz, 3H, Me).
13C-NMR (D2O) l 175.72 (s; OCH2CO); 171.62 (s;
CH2CH2CO);
80.53
(t;
OCH2O);
71.53
(t;
5.1.6. Pentanedioic acid mono-(1-butyryloxy-ethyl) ester
2b and pentanedioic acid bis-(1-butyryloxy-ethyl) ester
3b
MeOCH2CH2); 70.83 (t; MeOCH2CH2); 68.15 (t;
COCH2O); 58.65 (q; MeO); 35.95 (t; CH2CH2CO);
18.26 (t; MeCH2); 13.26 (q; Me). MS (CIꢀNH3) 252
(MNH+4). Anal. C10H18O6 (C, H).
Obtained in 36 and 25% yield, respectively, as de-
scribed for 2a and 3a. 2b, b.p. 165 °C/0.3 Torr (kugel-
rohr). 1H-NMR (CDCl3) l 6.68 (q; J=5.5 Hz, 1H,
OCHO); 2.43 and 2.41 (two t; J=7.5 Hz, 4H,
CH2CH2CH2); 2.29 (td; J=7.5, 1.0 Hz, 2H,
MeCH2CH2); 1.95 (quintet; J=7.5 Hz, 2H,
CH2CH2CH2); 1.65 (sextet; J=7.5 Hz, 2H,
MeCH2CH2); 1.47 (d; J=5.5 Hz, 3H, CHMe); 0.94 (t;
J=7.5 Hz, 3H, Me). 13C-NMR (CDCl3) l 177.80 (s;
COOH); 170.69 (s; MeCH2CH2CO); 169.86 (s;
CH2CO); 87.48 (d; OCHO); 34.92 (t; MeCH2CH2CO);
31.94 and 31.76 (two t; CH2CH2CH2); 18.52 (superim-
posed q and t; CH2CH2CH2, OCH(Me)O), 17.13 (t;
MeCH2CH2); 12.47 (q; Me). HRMS (CIꢀCH4)
247.111000 (MH+), calc. 247.154549, C11H18O6. 3b, b.p.
5.1.8. Butyric acid 2-[2-(2-methoxy-ethoxy)-ethoxy]-
acetoxymethyl ester 9
1
Obtained in 53% yield as described for 8. H-NMR
(CDCl3) l 5.83 (s; 2H, OCH2O); 4.30 (s; 2H, COCH2O);
3.52–3.80 (m; 8H, [O(CH2)2O]2); 3.38 (s; 3H, OMe);
2.43 (t; J=7.3 Hz, 2H, CH2CH2CO); 1.62 (sextet; J=
7.3 Hz, 2H, MeCH2); 0.97 (t; J=7.3 Hz, 3H, Me).
13C-NMR (D2O) l 175.61 (s; OCH2CO); 171.57 (s;
CH2CH2CO);
80.49
(t;
OCH2O);
71.50
(t;
MeOCH2CH2); 70.94 (t; COCH2OCH2CH2); 71.11 (t;
MeOCH2CH2); 70.01 (t; COCH2OCH2CH2); 68.13 (t;
COCH2O); 58.65 (q; MeO); 35.94 (t; MeCH2CH2);
18.27 (t; MeCH2CH2); 13.28 (q; Me). MS (CIꢀNH3) 296
(MNH+4).
1
165–170 °C/0.3 Torr. H-NMR (CDCl3) l 6.85 (q; J=
5.5 Hz, 1H, OCHO); 2.38 (t; J=7.5 Hz, 4H,
CH2CH2CH2); 2.28 (td; J=7.5, 1.0 Hz, 4H, two
MeCH2CH2); 1.94 (quintet; J=7.5 Hz, 2H,
CH2CH2CH2); 1.64 (sextet; J=7.5 Hz, 4H, two
MeCH2CH2); 1.46 (d; J=5.5 Hz, 6H, two CHMe); 0.94
5.1.9. Butyric acid diethoxy-phosphoryloxymethyl ester
10a
Et3N (22.2 mL, 1.2 equiv.) was added dropwise to a
solution of diethyl phosphate (20.55 g, 0.133 mol) and

A. Nudelman et al. / European Journal of Medicinal Chemistry 36 (2001) 63–74
71
chloromethyl butyrate (18.62 g, 1 equiv.) in dry DMF
(30 mL) under N2. The reaction mixture was heated at
70 °C for 4 h, during which time a large amount of
precipitate formed and TLC (EtOAc:hexane:i-PrOH
5:5:1, detection-vanillin) showed that most of the acid
had reacted. The precipitate was filtered and washed
with EtOAc. The filtrate was partitioned between water
and EtOAc. The aqueous phase was washed with a
small amount of EtOAc, and the combined organic
phase was washed thrice with water, twice with 5%
solution of NaHCO3, twice with brine, dried (MgSO4)
and evaporated to give an oily crude product (25.5 g)
which was distilled at 125 °C/0.1–0.2 Torr to give 10a as
small amount of EtOAc (3×10 mL) and the combined
organic phase was washed twice with brine, dried
(MgSO4) and evaporated. The product was obtained as
a brownish oil (31.7 g), that was distilled at 109–110 °C
(0.05 Torr) to give a colorless liquid (25 g, 59% yield)
1
which was used without further purification. H-NMR
(CDCl3) l 9.26 (dd; J=2.0, 1.0 Hz, 1H, ArꢀH2); 8.82
(dd; J=5.0, 2.0 Hz, 1H, ArꢀH6); 8.34 (dt; J=8.0, 2.0
Hz, 1H, ArꢀH4); 7.43 (ddd; J=8.0, 5.0, 1.0 Hz, 1H,
ArꢀH5); 6.04 (s; 2H, OCH2O); 2.39 (t; J=7.5 Hz, 2H,
CH2CO); 1.70 (sextet; J=7.5 Hz, 2H, MeCH2); 0.97 (t;
J=7.5 Hz, 3H, Me). 13C-NMR (CDCl3) l 171.83 (s;
CH2CO); 163.68 (s; PyCO); 153.76 (d; PyꢀC2); 150.88
(d; PyꢀC6); 137.04 (d; PyꢀC4); 124.74 (s; PyꢀC3);
123.10 (d; PyꢀC5); 79.32 (t; OCH2O); 35.35 (t; CH2CO);
17.77 (t; MeCH2); 13.17 (q; Me). HRMS (CIꢀCH4)
224.0900 (MH+), calc. 224.0923, C11H14NO4.
1
a colorless oil (17.2 g, 50%). H-NMR (CDCl3) l 5.63
(d; J=13 Hz, 2H, OCH2O); 4.13 (dquintets; J=7.0, 0.8
Hz, 4H, two CH2OP); 2.36 (t; J=7.5 Hz, 2H,
CH2COO); 1.69 (sextet; J=7.5 Hz, 2H, CH2CH2CO);
1.34 (td; J=7.0, 1.1 Hz, 6H, two MeCH2OP); 0.96 (t;
J=7.5 Hz, 3H, Me). 13C-NMR (CDCl3) l 171.91 (s;
CO), 82.40 (td; OCH2O); 64.20 (td; MeCH2OP); 35.78
(t; CH2COO); 17.99 (t; CH2CH2CO); 15.50 (td;
MeCH2OP); 13.44 (q; Me). 31P-NMR CDCl3 l −4.51.
Anal. C9H19O6P.0.5H2O (C, H).
5.1.12. Nicotinic acid 1-butyryloxy-ethyl ester 4b
Obtained in 48% yield as described for 4a, b.p. 135–
1
140 °C/0.5 Torr. H-NMR (CDCl3) l 9.20 (dd; J=2.0,
1.0 Hz, 1H, PyꢀH2); 8.78 (dd; J=5.0, 2.0 Hz, 1H,
PyꢀH6); 8.28 (dt; J=8.0, 2,0 Hz, 1H, PyꢀH4); 7.39
(ddd; J=8.0, 5.0, 1.0 Hz, 1H, PyꢀH5); 7.12 (q; J=5.5
Hz, 1H, OCHO); 2.32 (td; J=7.5, 1.0 Hz, CH2COO,
2H); 1.65 (sextet; J=7.5 Hz, 2H, MeCH2); 1.61 (d;
J=5.5 Hz, 3H, CHMe); 0.94 (t; J=7.5 Hz, 3H, Me).
13C-NMR (CDCl3) l 171.59 (s; COꢀBu); 163.31 (s;
PyCO); 153.83 (d; PyꢀC2); 151.09 (d; PyꢀC6); 137.33 (d;
PyꢀC4); 125.41 (s; PyꢀC3); 123.37 (d; PyꢀC5); 89.13 (d;
OCHO); 35.92 (t; COCH2), 19.65 (q; CHMe); 18.17 (t;
MeCH2); 13.52 (q; Me). HRMS (CIꢀCH4) 238.1070
(MH+), calc. 238.1079, C12H16NO4.
5.1.10. Butyric acid 1-(diethoxy-phosphoryloxy)-ethyl
ester 10b
Obtained in 62% yield as described for 10a, b.p.
95–100 °C/0.3Torr. 1H-NMR CDCl3 l 6.47 (dq; J=
7.0, 5 Hz, 1H, OCHO); 4.08 (ddquintets; J=7.0, 5.0,
0.8 Hz, 4H, two CH2OP); 2.28 (t; J=7.5 Hz, 2H,
CH2COO); 1.62 (sextet; J=7.5 Hz, 2H, CH2CH2CO);
1.49 (d; J=5 Hz, 3H, CHMe); 1.29 (tdd; J=7.0, 1.5,
0.8 Hz, 6H, two MeCH2OP); 0.91 (t; J=7.5 Hz, 3H,
Me). 13C-NMR CDCl3 l 171.43 (s; CO); 91.07 (d;
OCHO); 64.11 (td; CH2OP); 35.99 (t; CH2COO); 21.40
(qd; CHMe); 18.09 (t; CH2CH2CO); 16.06 (qd;
MeCH2OP); 13.55 (q; Me). HRMS (CIꢀCH4), 267.0993
(MꢀH+). Anal. C10H20O6P (C, H).
5.1.13. 4-Amino-benzoic acid butyryloxymethyl ester 5
4-Amino-benzoic acid (21.05 g, 153 mmol) was added
to a solution of chloromethyl butyrate (21.5 g, 157
mmol) in DMF (60 mL) under N2, followed by the
dropwise addition of Et3N (25.5 mL, 18.5 g, 1.2 equiv.)
until the acid dissolved. The reaction mixture was
heated at 60 °C overnight, during which time a large
amount of precipitate formed and TLC (EtOAc:hexane
2:1) showed that most of the acid has reacted. The
precipitate was filtered and washed with EtOAc. The
solution was partitioned between water and EtOAc. The
aqueous phase was extracted with a small amount of
EtOAc, and the combined organic phase was washed
thrice with a 5% solution of NaHCO3, twice with water,
twice with brine and dried (MgSO4). Evaporation of the
solvent gave a pasty residue that was triturated with
5.1.11. Nicotinic acid butyroyloxymethyl ester 4a
Nicotinic acid (23.4 g, 190 mmol) was added to a
solution of 1a (26 g, 190 mmol) in DMF (90 mL) under
N2, followed by the dropwise addition of Et3N (32 mL,
23.23 g, 1.2 equiv.) until the acid dissolved. The mixture
was heated at 60 °C overnight, during which time a
large amount of precipitate formed and TLC
(EtOAc:hexane 2:1) showed that most of the acid was
consumed. The precipitate was filtered and washed with
EtOAc (300 mL). The solution was washed four times
with water, the aqueous phase was extracted with a

72
A. Nudelman et al. / European Journal of Medicinal Chemistry 36 (2001) 63–74
hexane to give 18 g of a white solid that was recrystal-
118.70 (d; ArꢀC3 and ArꢀC5); 88.79 (d; OCHO); 39.62
(t; NCOCH2); 35.95 (t; OCOCH2); 19.58 (t; OCHO);
18.82, 18.15 (two t; two CH2CH2CO); 13.67, 13.46 (two
q; two Me). HRMS (DCIꢀCH4) 322.1640, calc. 322.1654
(MH+) C17H24NO5.
lized from i-PrOH to give 15 g (41% yield) of product,
1
m.p. 130–132 °C. H-NMR (CDCl3) l 7.88 (d; J=9.0
Hz, 2H, ArꢀH2); 6.63 (d, J=9.0 Hz, 2H, ArꢀH3); 5.95
(s; 2H, OCH2O); 4.75 (s; 2H, NH2); 2.35 (t; J=7.5 Hz,
2H, COCH2); 1.67 (sextet; J=7.5 Hz, 2H, MeCH2);
0.94 (t; J=7.5 Hz, 3H, Me). HRMS (CIꢀCH4):
238.1080, calc. 238.1079 (MH+) C12H16NO4. In order to
obtain the HCl salt, the product (pure or crude) was
treated with 1 M HCl, in EtOAc (4 equiv.), and precip-
itated with dry ether. The precipitate was filtered,
washed with ether and hexane. The crude salt was
crystallized from CH2Cl2:ether to give the pure product,
5.2. Biology
5.2.1. Materials
RPMI-1640 medium, fetal calf serum (FCS) and all
other tissue culture materials were obtained from Bio-
logical Industries Beth-Haemek, Israel; 3[H]-thymidine
from ICN, England; and other reagents from Sigma.
1
m.p. 133–135 °C. H-NMR (CDCl3) l 7.56 (d; J=9.0
Hz, 2H, ArꢀH2); 6.63 (d; J=9.0 Hz, 2H, ArꢀH3); 5.95
(s; 2H, OCH2O); 2.37 (t; J=7.5 Hz, 2H, CH2CO); 1.67
(sextet; J=7.5 Hz, 2H, MeCH2); 0.94 (t; J=7.5 Hz,
3H, Me).
5.2.2. Cells
HT-29, MCF-7, MIAPaCa2, PC-3, LnCap, HL-60
and K-562 cells were acquired from ATCC. They were
grown in RPMI-1640 medium supplemented with 10%
FCS, penicillin (250 U mL–1), and streptomycin (125 mg
mL–1). The cells were transferred into fresh medium
twice weekly and were incubated at 37 °C in a humi-
dified 5% CO2 incubator.
5.1.14. 4-Butyramino-benzoic acid butyryloxymethyl
ester 7a
Obtained in 77% yield as described for 4a, m.p.
1
89–90 °C. H-NMR (CDCl3) l 8.03 (d; J=8.0 Hz, 2H,
Ar-H2); 7.62 (d; J=8.0 Hz, 2H, ArꢀH3); 5.98 (s; 2H,
OCH2O); 2.37, 2.36 (two t; J=7.5 Hz, 4H, OCOCH2);
1.76 (sextet; J=7.5 Hz, 2H, CH2 CH2CON); 1.67 (sex-
tet; J=7.5 Hz, 2H, MeCH2CH2COO); 1.01 and 0.94
(two t; J=7.5 Hz, 6H, two Me). 13C-NMR (CDCl3) l
172.47 (s; CONH); 171.52 (s; CO-acetal); 164.70 (s;
ArꢀCO); 142.85 (s; ArꢀC4); 131.35 (d; ArꢀC2 and
ArꢀC6); 124.12 (s; ArꢀC1); 118.73 (d; ArꢀC3 and
ArꢀC5); 79.42 (t; OCH2O); 39.71 (t; NCOCH2); 35.83
(t; OCOCH2); 18.85, 18.11 (two t; two MeCH2CH2CO);
13.70 and 13.49 (two q; two Me). HRMS (DCIꢀCH4)
308.1480, calc. 308.1498 (MH+) C16H22NO5. Anal.
C16H21NO5 (C, H, N).
5.2.3. Treatment with prodrugs
The lipophilic prodrugs were solubilized in DMSO.
The concentrated DMSO solutions were diluted with
medium so that the highest concentration of DMSO in
the test medium was 0.1%. Solutions of the prodrugs
were handled using Hamilton syringes and Teflon or
glass vials. In parallel to the test compounds, appropri-
ate controls were run using 0.1% DMSO in medium.
5.2.4. Proliferation assays
Cells, 100–200 mL, at a density of 2–5×104 cells
mL–1 were seeded in tissue culture 96-well plates (in
triplicate) for 24 h. They were then exposed to different
concentrations of the prodrugs for the specified times.
Three methods were used for proliferation measure-
ments: a) SRB assay. A modified procedure of Papazisis
et al. [27], was employed for HT-29, MCF-7 and
MIAPaCa2 cell lines. Briefly, after 7 days of cell incuba-
tion with the test compounds the samples were fixed by
the addition of 50% trichloroacetic acid at 4 °C and
stained with 0.4% SRB in 1% acetic acid. The samples
were then rinsed with 1% acetic acid, air-dried and the
SRB dye was solubilized in 10 mM TRIS, pH 10, and
absorbance was read at 570 nm. b) XTT proliferation
assay. This assay was used with LnCap and PC-3 hu-
man prostate cancer cells. The cells were exposed to the
test compound for 48 h and analyzed as described [28].
5.1.15. 4-Butyramino-benzoic acid 1-butyryloxy-ethyl
ester 7b
Obtained in 14% yield as described for 4a, m.p.
1
65–66 °C. H-NMR (CDCl3) l 7.98 (d; J=8.0 Hz, 2H,
ArꢀH2); 7.61 (d; J=8.0 Hz, 2H, ArꢀH3); 7.45 (brs; 1H,
NH); 7.10 (q; J=5.5 Hz, 1H, OCHO); 2.37 (t; J=7.5
Hz, 2H, NHCOCH2); 2.32, 2.31 (two t; J=7.5 Hz, 2H,
OCOCH2); 1.76 (sextet; J=7.5 Hz, 2H, CH2CH2CON);
1.66 (sextet; J=7.5 Hz, 2H, CH2CH2COO); 1.59 (d;
J=5.5 Hz, 3H, CHMe); 1.00, 0.94 (two t; J=7.5 Hz,
6H, two Me). 13C-NMR (CDCl3) l 171.73 (s; CONH,
CO-acetal); 164.01 (s; ArꢀCO); 142.72 (s; ArꢀCNH);
131.08 (d; ArꢀC2 and ArꢀC6); 124.40 (s; ArꢀCO);

A. Nudelman et al. / European Journal of Medicinal Chemistry 36 (2001) 63–74
73
c) Thymidine uptake. HL-60 and K562 cells at 1×105
5.2.8. Toxicity studies in mice
mL–1 were treated with the prodrugs for 3 days.
Compounds 2a, 2b or 10a were dissolved in a solution
of 20 mM acetate buffer, pH 6, and 10% DMSO to give
a final concentration of 20 mg mL–1. The 2a lysine salt
was dissolved in 20 mM acetate buffer, pH 6, to give a
final concentration of 30 mg mL–1. Male Balb-c mice,
8–10 weeks old, were obtained from Bar-Ilan University
animal colony. The formulated compounds were in-
jected i.p. into the mice at a rate of 0.4 mL min–1. Each
animal was weighed and adjustments of dose and vol-
ume were made. All the experiments with mice were
conducted following the NIH laboratory animal care
guidelines and in accordance with the guidelines of
Beilinson Campus Helsinki committee for animal care.
Aliquots of 150 mL were transferred into 96-well plates,
3
and 20 mL of [H]-thymidine, 6.7 Ci mmole–1 specific
activity (final radioactivity 5 mCi mL–1) was added.
After 16 h, the cells were washed and harvested (PHD,
Cambridge Technology, Inc., Cambridge, MA). The
filters were dried and 2 mL scintillation fluid added.
Radioactivity was measured for 5 sec/sample in a 1217
RackBeta (LKB, Sweden) counter. Background counts
were deducted from the radioactivity of the cell samples.
The mean value obtained from 4 wells was computed by
the program.
5.2.5. Induction of differentiation
Cancer cell differentiation was evaluated with the
human myeloid leukemic cell line HL-60 by NBT reduc-
tion activity [1, 2]. Cell cultures, after 3 days of exposure
to the test compounds, were stimulated with 0.4 mM
12-O-tetradecanoyl-phorbol 13-acetate (PMA) in the
presence of 0.1% NBT. After 30 min of incubation at
37 °C they were examined microscopically by scoring at
least 200 cells. The % of NBT positive cells was calcu-
lated from the ratio the stained cells to the total cells
scored and the % of positive cells in untreated culture
was subtracted. NBT reduction activity correlated with
morphological changes observed with May–Grunwald
Giemsa stained cytospin smears of the cells.
Acknowledgements
Generous support for this work by Beacon Laborato-
ries and the ‘Marcus Center for Pharmaceutical and
Medicinal Chemistry’ and the ‘Bronia and Samuel
Hacker Fund for Scientific Instrumentation’ at Bar Ilan
University, are gratefully acknowledged.
References
[1] a) Nudelman A., Shaklai M., Aviram A., Rabizadeh E., Zimra
Y., Ruse M., Rephaeli A., J. Med. Chem. 35 (1992) 687–694.
b) Nudelman A., Rephaeli A., J. Med. Chem. 43 (2000) 2962–
2966.
[2] Rephaeli A., Shaklai M., Ruse M., Nudelman A., Int. J.
Cancer. 49 (1991) 66–72.
5.2.6. Apoptosis
The cells were exposed to the drugs for 3 days,
harvested, washed in PBS and resuspended in binding
buffer (20 mM HEPES, pH 7.4, 150 mM NaCl, 2.5 mM
CaCl2). FITC-conjugated annexin (1:100 dilution of 50
mg mL–1) and 1:100 dilution of propidium iodide (0.5
mg mL–1 PBS) were added. After 10 min on ice, the
cells were analyzed by flow cytometry. Labeled cells
were run in a FACStarplus flow cytometer (Becton-
Dickinson, Immunofluorometry systems, Mountain
View, CA). Cells, negative for propidium iodide and
positive for annexin, were regarded as apoptotic cells
[29].
[3] Rabizadeh E., Shaklai M., Nudelman A., Eisenbach L.,
Rephaeli A., FEBS Lett. 328 (1993) 225–229.
[4] Aviram A., Zimra Y., Shaklai M., Nudelman A., Rephaeli A.,
Int. J. Cancer 56 (1994) 906–909.
[5] Zimra Y., Wasserman L., Maron L., Shaklai M., Nudelman
Rephaeli A., J. Cancer Res. Clin. Oncol. 123 (1997) 152–160.
[6] Zimra Y., Nudelman A., Zhuk R., Rabizadeh E., Shaklai M.,
Aviram A., Rephaeli A., J. Cancer Res. Clin. Oncol. 126 (2000)
693–698.
[7] Siu L.L., Von Hoff D.D., Rephaeli A., Izbicka E., Cerna C.,
Gomez L., Rowinsky E.K., Eckhardt S.G., Invest. New Drugs
16 (1998) 113–119.
[8] Kasukabe T., Rephael A., Honma Y., Br. J. Cancer 75 (1997)
850–854.
[9] Rabizadeh E., Shaklai M., Eisenbach L., Nudelman A.,
Rephaeli A., Israel J. Med. Sci. 32 (1996) 1186–1191.
5.2.7. Quantitative measurement of Hb
K562 cells were cultured at 2×105 cells mL–1, 3
mL/well, harvested on day 3, counted, lyzed in water
and Hb was determined by ion-exchange HPLC as
described [31].
[10] Aviram A., Rephaeli A., Shaklai M., Nudelman A., Ben Dror
I., Maro L., Rabizadeh E., J. Cancer Res. Clin. Oncol. 123
(1997) 267–271.
[11] Eckhart S.D., Villaona-Calero M.A., Hommond L., Proc.
AACR (1998) 3449.

74
A. Nudelman et al. / European Journal of Medicinal Chemistry 36 (2001) 63–74
[12] Faller D.V., Perrine S.P., Current Opinion Hematol. 2 (1995)
109–117.
[21] Material Safety Data Sheets (MSDS), Aldrich Chemical Co.
Inc.
[13] Bunn H.F., Blood 93 (1999) 1787–1789.
[22] Rasmussen M., Leonard N.J., J. Am. Chem. Soc. 33 (1967)
5439–5445.
[14] Perrine S.P., Ginder G.D., Faller D.V., Dover G.H., Ikuta T.,
Witkowska H.E., Cai S.P., Vichinsky E., Olivieri N.F., New
Eng. J. Med. 328 (1993) 81–86.
[23] Yoshimura Y., Hamaguchi N., Yashiki T., J. Antibiotics 39
(1986) 1329–1342.
[15] Aweh G.F., Sutton M., Nassif I., Boosalis V., Dover G.J.,
Wallenstei S., Wright E., McMahon L., Stamatoyannopoulos
G., Faller D.V., Perrine S.P., Blood 93 (1999) 1790–1797.
[24] Harada N., Ozaki K., Oda K., Nakanishi N., Ohashi M.,
Hashiyama T., Tsujihara K., Chem. Pharm. Bull. 145 (1997)
1156–1162.
[16] Dover G.J., Brusilow S., Samid D., New Engl. J. Med. 327
(1992) 569–570.
[25] Drushel W.A., Selty A.R., Am. J. Sci. 43 (1917) 57–66.
[26] EPA Product properties test guidelines. OPPTS 830.7840. Wa-
ter solubility: Column elution method.
[17] Rephaeli A., Zimra Y., Nudelman A., Ben-Dror I., Fibach E.,
Shaklai M., in: Y. Beuzard, B. Lubin, J. Ross (Eds.), Sickle Cell
Disease and Thalassaemias: New Trends in Therapy, Colloque
INSERM//John Libbey Eurotext Ltd. 234 (1995) 215–216.
[27] Papazisis K.T., Geromichalos G.D., Dimitriadis K.A., Kort-
saris A.H., J. Immunol. Methods 208 (1997) 151–158.
[28] Sutherland M.W., Learmonth B.A., Free Radic. Res. 27 (1997)
283–289.
[18] Charache S., Terrin M.L., Moore R.D., Dover G.J., Barton
F.B., Eckert S.V., McMahon R.P., Bonds D.R., N. Engl. J.
Med. 332 (1995) 1317–1322.
[29] Choi B.-Y., Kim H.-Y., Lee K.-H., Cho Y.-H., Kong G.,
Cancer Lett. 147 (1999) 85–93.
[19] Bundgaard H.R., Nielsen N.M., J. Med. Chem. 30 (1987)
451–454.
[20] The Merck Index 12th Edition, Merck & Co., Inc. 1996. a)
entry 443; b) entry 6612; c) entry 4841; d) entry 7669; e) entry
6577.
[30] Kawasaki N., Morimoto K., Tanimoto T., Hayakawa T., Arch.
Biochem. Biophys. 328 (1996) 289–294.
[31] Fibach E., Kollia P., Schechter A.N., Noguchi C.T., Rodgers
G.P., Blood 85 (1995) 2967–2974.
.