Design and synthesis of a cephalosporin-retinoic acid prodrug activated by a monoclonal antibody-β-lactamase conjugate

Article abstract of DOI:10.1016/S0968-0896(01)00127-4

Two novel series of all-trans-β-retinoic acid derivatives were synthesized and found to possess anticancer activity. The first series, cephalosporin 3′-retinoic esters 6 and 7 were, respectively, obtained by the condensation of all-trans-β-retinoic acid (2) with cephalosporins 4 and 5. The second series, 7-(retinamido)cephalosporins 11 and 12, were synthesized, respectively, by the condensation of 2 with cephalosporins 9 and 10. These four heretofore undescribed compounds 6, 7, 11, and 12 showed inhibitory activity against murine leukemias (L1210 and P388), sarcoma 180, breast carcinoma (MCF7), and human T-lymphocytes (Molt4/C8 and CEM/0). They also inhibited squamous metaplasia and keratinization in tracheal organ cultures derived from vitamin-A-deficient hamsters. Moreover, cephalosporin 3′-retinoic ester 7 exhibited enhanced activity against keratinization with ED₅₀ = 3.91 × 10^-11 M in the presence of a β-lactamase from Staphylococcus aureus 95. A tumor targeting fusion protein (dsFv3-β-lactamase) was also used in conjunction with cephem-based retinoid 7 and the potency of 7 toward L1210, P388, and MCF7 was found to approach that of the free retinoic acid (2). In the presence of dsFv3-β-lactamase, tumor cells were found to be much more susceptible to retinoid 7 than normal human embryonic lung cells. These notions provide a new approach to the use of β-retinoic acid for antitumor therapy.

Full text of DOI:10.1016/S0968-0896(01)00127-4

Bioorganic & Medicinal Chemistry 9 (2001) 2139–2147
Design and Synthesis of a Cephalosporin-Retinoic Acid Prodrug
Activated by a Monoclonal Antibody–ꢀ-Lactamase Conjugate
Gholam H. Hakimelahi,a,* Tai Wei Ly,a Sheng-Fa Yu,a Maryam Zakerinia,b
Ali Khalafi-Nezhad,b Mohammad N. Soltani,b Mohsen N. Gorgani,c
Azra R. Chadeganic and Ali A. Moosavi-Movahedic
aInstitute of Chemistry, Academia Sinica, Taipei, Taiwan 115, Republic of China
bDepartments of Chemistry and Medicine, Shiraz University, Shiraz, Iran
cInstitute of Biochemistry-Biophysics, Tehran University, Tehran, Iran
Received 14 February 2001; accepted 9 April 2001
Abstract—Two novel series of all-trans-b-retinoic acid derivatives were synthesized and found to possess anticancer activity. The
first series, cephalosporin 30-retinoic esters 6 and 7 were, respectively, obtained by the condensation of all-trans-b-retinoic acid (2)
with cephalosporins 4 and 5. The second series, 7-(retinamido)cephalosporins 11 and 12, were synthesized, respectively, by the
condensation of 2 with cephalosporins 9 and 10. These four heretofore undescribed compounds 6, 7, 11, and 12 showed inhibitory
activity against murine leukemias (L1210 and P388), sarcoma 180, breast carcinoma (MCF7), and human T-lymphocytes (Molt4/
C8 and CEM/0). They also inhibited squamous metaplasia and keratinization in tracheal organ cultures derived from vitamin-A-
deficient hamsters. Moreover, cephalosporin 30-retinoic ester 7 exhibited enhanced activity against keratinization with
ED50=3.91Â10À11 M in the presence of a b-lactamase from Staphylococcus aureus 95. A tumor targeting fusion protein (dsFv3–b-
lactamase) was also used in conjunction with cephem-based retinoid 7 and the potency of 7 toward L1210, P388, and MCF7 was
found to approach that of the free retinoic acid (2). In the presence of dsFv3–b-lactamase, tumor cells were found to be much more
susceptible to retinoid 7 than normal human embryonic lung cells. These notions provide a new approach to the use of b-retinoic
acid for antitumor therapy. # 2001 Elsevier Science Ltd. All rights reserved.
Introduction
Chemical modification of the polar carboxylic
group,20À23 the cyclic end group,24À26 or the polyene
chain6,27 in all-trans-b-retinoic acid could lead to new
retinoids with high potency and low toxicity. Results
from a systematic study on the structure–activity rela-
tionship reveal that retinoic acid derivatives of azetidi-
none exhibit special affinity towards rapidly growing
cells and possess anti-leukemic activity in laboratory
animals.28À30
Retinoids are of importance to epithelial tissue for the
maintenance of its normal growth.1 They also prevent
cancer in experimental animals,2 particularly effective
for carcinoma of bladder,3 breast,4 and skin.5 Retinoids
can induce differentiation of human promyelocytic leu-
kemia cell line HL-60 from mature granulocytes.6À8 The
mechanism of retinoids in cell differentiation and pro-
liferation is one of the most important topics under
investigation in biology.9,10 All-trans-b-retinoic acid
often binds to specific nuclear receptors in order to reg-
ulate special gene expressions.11À17 Effective doses of
retinoids (e.g., all-trans-b-retinoic acid) in most pro-
phylactic studies, however, produce undesirable side
effects.18,19
The mechanism by which most b-lactamases (bLs)
inactivate b-lactam antibiotics is through the acylation
of a serine residue at the active site of the enzyme.31À40
In the case of a cephalosporin, a potential leaving group
at the C-30 position is eliminated during the enzymatic
reaction,33,35,40 which most probably proceeds in two
steps.34,39 The leaving group could be an anticancer
agent. Thus, the enzymatic acylation of the cephalos-
porin moiety by a serine residue in endogenous proteins
releases the anticancer agent.36,37
*Corresponding author. Tel.:+886-2-2789-8682; fax: +886-2-2783-
1237; e-mail: hosein@chem.sinica.edu.tw
0968-0896/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(01)00127-4

2140
G. H. Hakimelahi et al. / Bioorg. Med. Chem. 9 (2001) 2139–2147
The advent of monoclonal antibodies has revolutionized
many aspects of drug development in tumor immu-
notherapy, where monoclonal antibodies have been
used as directed messengers through conjugation, deliv-
ery, and release of drugs at the tumor site.41À48 Most of
the cytotoxic agents exert their activities once inside the
cell, requiring that the monoclonal antibody carrier
facilitate the delivery of the drug to its precise site of
activity within the cell. An exciting development in this
area of research has involved the conjugation of specific
enzymes to monoclonal antibodies, and the enzymatic
release of a drug at the tumor site.48 In principle, one
can deliver many drugs to the tumor site with each
immunoconjugate molecule and limitations in cancer
chemotherapy could perhaps be overcome.
Results
Synthesis of cephalosporin 30-retinoic esters 6 and 7
(Scheme 2). For the preparation of cephem–retinoic
acid conjugate 6, we silylated (+)-30-iodocephalosporin
340,49,50 with trimethylsilyl bromide in CH2Cl2 (Scheme
2). Without isolation, the resultant silylated ester 4 was
treated with the sodium salt of all-trans-b-retinoic
acid (2) in DMF to give the desired ester 6 (48%
overall yield) as a single isomer as evidenced by HPLC
analysis.
For the synthesis of oxocephem–retinoic acid conjugate
7, (+)-30-iodocephalosporin 3 was similarly silylated
and then treated, in situ, with 3-chloroperoxybenzoic
acid to afford 1-oxocephalosporin 5 (Scheme 2). We
then esterified the intermediate 5 with the sodium salt of
all-trans-b-retinoic acid (2) to produce the target ester 7,
as a single isomer, in 40% overall yield.
We planned to attach all-trans-b-retinoic acid (2) onto a b-
lactam antibiotic at the C-30 position as shown in Scheme
1. Cephalosporins are highly versatile substrates in the
construction of enzyme-activatable prodrugs, including
nitrogen mustard drugs,43,44 a carboplatin analogue,45
and doxorubicin.46À48 bLs are an attractive choice of
enzymes for prodrug activation because of their high cat-
alytic efficiency and substrate specificity (see Scheme 1).42
Synthesis of 7-(retinamido)cephalosporins 11 and 12
(Scheme 3). We considered an alternative way to attach
the b-retinoic acid (2) onto b-lactams by using their
amino group at the N-7 position. Accordingly, we sily-
lated (+)-7-aminocephalosporin 840 with trimethylsilyl
bromide to give silyl ester intermediate 9 (Scheme 3). In
the meantime, all-trans-b-retinoic acid (2) was treated
with ethyl chloroformate in pyridine and CH2Cl2 to give
the corresponding anhydride. We added this anhydride
to the silylated 7-aminocephalosporin 9 to afford a sin-
gle isomer of retinamido-b-lactam conjugate 11 in 60%
overall yield.
As a general program aimed at the development of
novel and effective retinoids, we have been investigating
the prospects of applying the monoclonal antibody–bL
conjugates in releasing the retinoic acid in vivo and the
obvious potential in tumor directed chemotherapy.
Herein, we report the chemical syntheses and anticancer
properties of two classes of new retinoic acid–cephem
conjugates. They include all-trans-b-retinoic acid
attached to cephalosporins at the C-30 position (i.e., 6
and 7) and at the N-7 position (i.e., 11 and 12). Enzyme-
activatable prodrug 7 in conjunction with antibody-
enzyme fusion protein (dsFv3–bL)42 exhibited compar-
able activity to that of all-trans-b-retinoic acid (2)
against certain malignant tumor cell lines. Its toxicity,
however, was found to be much less toward normal
human embryonic lung cells (HEL).
To obtain the corresponding S-oxide 12, we treated the
silylated intermediate 9 with 3-chloroperoxybenzoic
acid in situ to afford sulfoxide 10 (Scheme 3). By the
same way described above, all-trans-b-retinoic acid (2)
was converted to the corresponding anhydride, which
was added to sulfoxide 10. In this way, the desired reti-
namido–b-lactam conjugate 12 was obtained as a single
isomer in 42% overall yield.
Scheme 1. Liberation of b-retinoic acid (2) from b-lactam-containing prodrug 1 upon enzyme activation.

G. H. Hakimelahi et al. / Bioorg. Med. Chem. 9 (2001) 2139–2147
2141
Solubility and stability of retinoic acid–cephem conju-
Biological activity
gates 6, 7, 11, 12, and retinoic acid (2) in water. We
found that the solubility of cephalosporin 30-retinoic
esters 6 and 7 in water was, respectively, 0.16 and
0.32 mg/mL. The compounds were also stable under
subdued light at physiological pH for 7 days. The ste-
reochemistry of the side chain remain intact as judged
Enzymatic hydrolysis study of cephalosporin–retinoic
1
acid conjugate 7 by H NMR. Because of the poor solu-
bility of retinoid 7 in neutral phosphate solution, a
mixture of DMSO-d6 and phosphate buffer solution
(pD 7.2, 1:1 mL/mL) was used for the 1H NMR study of
bL catalyzed hydrolysis.45 In the presence of sufficient
bL from S. aureus 95, the 1H NMR spectrum of 7
changed rapidly to that of the eliminated compounds 14
1
by HPLC and H NMR studies. However, at pH=12,
the b-lactam ring in 6 and 7 decomposed within 5 min.
After neutralization of the basic solution, all-trans-b-
retinoic acid 2 was isolated in 87% yield and character-
ized by 1H NMR. On the other hand, the solubility of 7-
(retinamido)cephalosporins 11 and 12 in water was
respectively 0.07 and 0.14 mg/mL. They were found to
be stable in a phosphate buffer (0.10 M, pH 7.2) in the
dark for 25 days. No change in the stereochemistry of
the side chain was observed as evidenced by HPLC
analysis. However, at pH=12, decomposition was evi-
dent after 35 min, at which time a mixture of uni-
dentified compounds was obtained. It should be noted
that the solubility of the reference compound, all-trans-
b-retinoic acid (2), in water, however, was determined to
be only 0.008 mg/mL.
1
and 2 (Scheme 4) as judged by H NMR and HPLC
analyses. In a control experiment, in the absence of the
bL, 7 was stable to hydrolysis for 7 days in the dark.
Kinetic parameters for purified ꢀL and dsFv3-ꢀL with
cephalosporin–retinoic acid conjugate 7. bL and dsFv3-
bL were secreted from Escherichia coli 27C7 and pur-
ified according to an established procedure.42b Kinetic
parameters for bL and dsFv3-bL with retinoid 7 were
determined by HPLC method as described.42c The con-
centrations of pro-drug 7 and the resultant all-trans-b-
retinoic acid (2) were monitored by absorbance at 352
nm. Retinoid 7 was found to be an excellent substrate
Scheme 2. Synthesis of cephalosporin 30-retinoic esters 6 and 7. Reagents: (a) (1) Me3SiBr, CH2Cl2, 25 ꢀC, 10 min; (2) NaHCO3, DMF, 25 ꢀC, 1 h,
3!4!6 (48%); (b) (1) Me3SiBr, CH2Cl2, 25 ꢀC, 10 min; (2) 3-ClC6H4CO3H, 1 h; (3) NaHCO3, DMF, 25 ꢀC, 1 h, 3!4!5!7 (40%).

2142
G. H. Hakimelahi et al. / Bioorg. Med. Chem. 9 (2001) 2139–2147
for bL (kcat 2460 sÀ1, KM 98 mM; kcat/KM 25 sÀ1 mMÀ1
)
was observed for 2, 6, 11, and 12 against squamous
metaplasia and keratinization. In contrast, cephalos-
porin–retinoic acid conjugate 7 was activated by the bL
and exhibited remarkable anti-keratosis activity with
ED50=3.91Â10À11 M.
and dsFv3-bL (kcat 1950 sÀ1, KM 74 mM; kcat/KM 26 sÀ1
mMÀ1).
Inhibition of squamous metaplasia and keratinization.
We studied the anti-hyperkeratosis activity of all-trans-
b-retinoic acid–cephem conjugates 6, 7, 11, and 12. The
experiments were performed according to their ability to
inhibit squamous metaplasia and keratinization in
organ cultures of trachea derived from vitamin-A-defi-
cient hamsters. b-Retinoic acid 2 with an all trans con-
figuration was used as the reference compound.51
Anticancer activity. The anticancer screening experi-
ments for compounds 6, 7, 11, 12, 7+bL from E. coli,
7+dsFv3–bL as well as the reference compounds 9-(b-
d-arabinofuranosyl)cytosine (ara-C) and all-trans-b–
retinoic acid (2) were carried out in vitro against murine
leukemias (L1210 and P388), sarcoma 180 (S-180),
breast carcinoma (MCF7), and human T-lymphocytes
(Molt4/C8 and CEM/0).54 The activity is expressed as
the concentration (mM) required to inhibit tumor cell
proliferation by 50% (IC50). Their toxicity toward nor-
mal human embryonic cell line (HEL) were also deter-
mined. Results are summarized in Table 1.
The percentage of the cultures exhibiting keratinization
is illustrated in Figure 1; untreated control exhibited
95% keratinization. We determined mean effective
doses (ED50) for the reversal of keratinization of 50%
of the explants by probit analysis.52,53 The ED50 values
were 1.46Â10À10, 3.68Â10À10, 2.15Â10À10, 4.16Â10À10
,
and 3.46Â10À10 M for 2, 6, 7, 11, and 12, respectively.
We also obtained the carcinostatic properties of reti-
noids in the presence of 0.85 nM of a bL from S. aureus
95 in vitro. No significant enhancement of the activity
Determination of LD50 for all-trans-ꢀ-retinoic acid (2)
and cephem–retinoic acid conjugate 7 in rats. All-trans-
b-retinoic acid (2) and cephem–retinoic acid conjugate 7
were administered subcutaneously (sc). All-trans-b-reti-
Scheme 3. Synthesis of 7-(retinamdo)cephalosporins 11 and 12. Reagents: (a) (1) Me3SiBr, CH2Cl2, 25 ꢀC, 10 min; (2) ClCO2 Et, CH2Cl2, pyridine,
À5 ꢀC, 1.5 h, 8!9!11 (60%); (b) (1) Me3SiBr, CH2Cl2, 25 ꢀC, 10 min; (2) 3-ClC6 H4CO3 H, 1 h; (3) ClCO2 Et, CH2Cl2, pyridine, À5 ꢀC, 1.5 h,
8!9!10!12 (42%).

G. H. Hakimelahi et al. / Bioorg. Med. Chem. 9 (2001) 2139–2147
2143
Scheme 4. The acyl–enzyme adduct 13 and the corresponding hydrolyzed product 14 resulting from the ring opening of cephalosporin 30-retinoic
ester 7.
noic acid (2) and the conjugate 7 did not show any
toxicity at concentration levels as high as 70 and
105 mg/kg, respectively. All rats were found to be in
good conditions after administration for 3 months.
Nevertheless, an LD50 (sc) value of 95 and 170 mg/kg
were obtained, respectively, for all-trans-b-retinoic acid
(2) and oxocephem–retinoic acid conjugate 7.
Discussion
Retinoids have been used in the prevention and therapy
of tumors.55 Cephalosporins can react with transpepti-
dases31,35 and/or b-lactamases42 (see Scheme 1). By
applying the dual targeting approach,56 we first com-
bined cephalosporins with all-trans-b-retinoic acid (2) to
produce novel anticancer agents 6, 7, 11, and 12. These
newly synthesized retinoids exhibited anti-hyperker-
atosis activity comparable with that of all-trans-b-reti-
Figure 1. Dose–response relationships of retinoids on the percentage
of tracheal organ cultures with keratin.
Table 1. Inhibitory concentrationsa (IC50, mM) of retinoids 6, 7, 11, 12, 7+bL, 7+dsFv3-bL as well as all-trans-b-retinoic acid 2 and ara-C on the
growth of malignant cell lines and normal human embryonic lung cells in vitro
Murine leukemias
Sarcoma 180
S-180
Breast carcinoma
MCF7
Human T-lymphocytes
Human lung cell
HEL
Compound
L1210
P388
Molt4/C8
CEM/0
ara-C
2
6
7
11
12
7+bL
0.16Æ0.02
0.06Æ0.01
0.26Æ0.07
0.17Æ0.02
1.01Æ0.21
0.96Æ0.04
0.04Æ0.00
0.05Æ0.02
0.14Æ0.03
0.04Æ0.00
0.70Æ0.13
0.24Æ0.04
2.30Æ0.08
0.87Æ0.10
0.03Æ0.01
0.04Æ0.00
0.38Æ0.05
0.47Æ0.21
1.99Æ0.15
1.18Æ0.03
3.60Æ0.21
1.67Æ0.06
0.41Æ0.24
1.27Æ0.29
0.78Æ0.13
0.02Æ0.00
2.02Æ0.12
0.86Æ0.02
2.16Æ0.41
0.97Æ0.06
0.01Æ0.00
0.01Æ0.00
0.70Æ0.12
0.63Æ0.05
1.77Æ0.03
0.98Æ0.04
1.89Æ0.18
1.21Æ0.15
0.56Æ0.07
1.01Æ0.60
0.77Æ0.09
0.43Æ0.01
1.84Æ0.10
1.36Æ0.09
2.99Æ0.05
2.15Æ0.00
0.29Æ0.03
1.35Æ0.32
1.98Â10À3
12.16 Æ 1.09
29.37 Æ 1.36
24.61 Æ 1.02
22.13 Æ 2.03
18.97 Æ 0.58
13.13 Æ 1.65
26.40Æ2.03
7+dsFv3–bL
aThe IC50 values were estimated from dose–response curves compiled from at least two independent experiments and represent the compound
concentration (mM) required to inhibit cell proliferation by 50%.

2144
G. H. Hakimelahi et al. / Bioorg. Med. Chem. 9 (2001) 2139–2147
noic acid (2) (ED50=1.46Â10À10 M). At the second
stage, we applied a bL from S. aureus 95 to these new
compounds individually. Only oxocephem–retinoic acid
Conclusions
For the development of new compounds as anticancer
agents, all-trans-b-retinoic acid (2) was attached to the
cephalosporin at either the C-30 or the N-7 position by
chemical methods. The first series involves the forma-
tion of an ester linkage by condensation of (+)-30-
iodocephalosporin 3 at the C-30 position with the
sodium salt of b-retinoic acid (2). In this way, the target
cephalosporin 30-retinoic esters 6 and 7 were synthesized
in 40–48% overall yields. The second series involves the
formation of an amide linkage by condensation of a
silylated 7-aminocephalosporin at the N-7 position with
an anhydride of all-trans-b-retinoic acid (2). The desired
retinamido-b-lactams 11 and 12 were thus obtained in
42–60% overall yields.
conjugate
7
showed greater anticancer property
(ED50=3.91Â10À11 M) than that of 2. These results
indicate that the all-trans-b-retinoic acid component can
be liberated effectively from 7 at the C-30 position
through a bL induced 1,4-elimination,40 as shown in
Scheme 4. In fact, release of all-trans-b-retinoic acid (2)
from cephalosporin 30-retinoic ester 7, by the bL, results
in anti-hyperkeratosis activity better than administra-
tion of retinoic acid (2) alone. This could be attributed
to the 40-fold increment in water solubility of 7 com-
pare to 2.
While bL enhanced only the activity of compound 7, all
compounds (i.e., 6, 7, 11, and 12) showed retinoid
activity. Considering the stability of all-trans-b-retinoic
acid–cephalosporin linkage, we believe that the retinoic
acid receptor b (RARb) induction57 by newly synthe-
sized retinoids 6, 7, 11, and 12 may play an important
role in mediating growth inhibitory effects of these reti-
noids in cancer cells. In fact, the induction of RARb by
retinoids is not surprising, because the RARb gene
promoter contains a retinoic acid responsive element
(RARE).
Among these four newly synthesized compounds, oxo-
cephem–retinoic acid conjugate 7 was found to be acti-
vated by a bL or the targeting fusion protein, dsFv3–
bL, and exhibited remarkable anti-hyperkeratosis
(ED50=3.91Â10À11 M) as well as inhibitory activity
against murine leukemias (L1210 and P388), sarcoma
180 (S-180), breast carcinoma (MCF7), and human T-
lymphocytes (Molt4/C8 and CEM/0). These results
indicate that the all-trans-b-retinoic acid component
could be effectively released from 7 through a 1,4-elim-
ination process. Furthermore, the LD50 (sc) value of 7
was found to be 170 mg/kg in rats.
Retinoids 6, 7, 11, and 12 were also found to exhibit
interesting activity against L1210, P388, S-180, MCF7,
Molt4/C8 and CEM/0. In comparison to reference
compounds ara-C and all-trans-b-retinoic acid (2), reti-
noids 6, 7, 11, and 12 displayed less inhibition on the
examined tumor cell lines. They were, however, less
toxic against HEL (see Table 1). In contrast, in the pre-
sence of free bL from E. coli, the potency of pro-all-
trans-b-retinoic acid 7 approaches that of all-trans-b-
retinoic acid (2). Prodrugs which are rapidly hydrolyzed
by bL, and which exhibit high affinity for the enzyme
(high kcat/KM) are expected to offer a therapeutic
advantage in an antibody-directed catalysis delivery
system. In the presence of the tumor-targeting fusion
protein, dsFv3–bL, prodrug 7 was active (kcat/KM 26 sÀ1
mMÀ1) against L1210, P388, and MCF7 and offer a
similar activity to that of all-trans-b-retinoic acid (2); yet
its activity was not enhanced against S-180, Molt4/C8,
and CEM/0 (see Table 1). Similarly, toxicity of prodrug
7 toward HEL was found to be similar to that of all-
trans-b-retinoic acid (2) in the presence of free bL, but
remained about the same in the presence of dsFv3–bL.
This could be due to the capability of monoclonal anti-
body–bL conjugate (i.e., dsFv3–bL) to bind to cell sur-
face antigenes. Consequently, binding activity of
monoclonal antibody may be correlated with the ability
of the conjugate to activate prodrugs.
Experimental
General methods
For anhydrous reactions, glassware was dried overnight
in an oven at 120 ꢀC and cooled in a desiccator over
powdered anhydrous calcium sulfate (i.e., CaSO4) or
silica gel. Reagents were purchased from Fluka Chemi-
cal Co. Solvents (Switzerland), including chloroform,
dichloromethane, dimethylformamide, ethyl acetate,
hexanes, and pyridine were distilled over CaH2 under
nitrogen. Ethanol was purchased from Merck (Ger-
many) and used as received. Reactions were performed
under a nitrogen atmosphere with shielding from light;
the apparatus was evacuated and filled with dry nitro-
gen at least three times.
Melting points were obtained with a Buchi 510 melting
point apparatus. Ultraviolet (UV) spectra were recorded
on a Cary 118 Spectrophotometer and lmax are reported
in nm (e) units. Infrared (IR) spectra were recorded on a
Beckman IR-8 Spectrophotometer. The wavenumbers
reported are referenced to the 1601 cmÀ1 absorption of
polystyrene. Proton NMR spectra were obtained on a
Varian XL-300 (300 MHz) Spectrometer. Chloroform-d
and D2O were used as solvent; Me4Si (d 0.00 ppm) was
used as an internal standard. All NMR chemical shifts
are reported as d values in parts per million (ppm) and
coupling constants (J) are given in hertz (Hz). The
splitting pattern abbreviations are as follows: s, singlet;
d, doublet; br, broad; and m, unresolved multiplet due
to the field strength of the instrument. Mass spectra
As such, the binding ability of monoclonal antibody
to cell surface antigen as well as the capability of its
protein conjugate in recognition of prodrugs in vitro
may pave the way to further studies involving the
effectiveness of monoclonal antibody–b-lactamase
conjugates in releasing the retinoic acid in vivo. This
strategy can be utilized in selective tumor directed
chemotherapy.

G. H. Hakimelahi et al. / Bioorg. Med. Chem. 9 (2001) 2139–2147
2145
were carried out on a VG 70-250 S mass spectrometer.
Microanalysis were performed on a Perkin-Elmer 240-B
microanalyzer. Purification on silica gel refers to gravity
column chromatography on Merck silica gel 60 (particle
size 230–400 mesh). Analytical TLC was performed on
precoated plates purchased from Merck (silica gel 60
F254); compounds were visualized by use of UV light.
Reverse phase HPLC (Hewlett-Packard, C18 column)
was used for identification of retinoids with 20–90%
1H NMR (CDCl3/D2O) d 1.16 (s, 6H, (CH3)2C), 1.32–
2.43 (m, 15H, 3ÂCH3+3ÂCH2), 3.80 (d, 2J=17 Hz,
1H, HCSO), 3.97 (d, 2J=17 Hz, 1H, HCSO), 4.61 (br s,
2
2H, OCH2CO), 4.86 (d, J=14 Hz, 1H, CHO), 5.02 (d,
3
2J=14 Hz, 1H, CHO), 5.29 (d, J=5.0 Hz, 1H, HC(6)),
5.79 (d, 3J=5.0 Hz, 1H, HC(7)), 5.60–6.72 (m, 6H,
6Â=CH), 7.01–7.42 (br s, 5H, C6H5). Anal. calcd for
C36H42N2O8S: C, 65.24; H, 6.39; N, 4.23; S, 4.84.
Found: C, 64.98; H, 6.41; N, 4.36; S, 4.97.
CH3CNin H O as the mobile phase. Peaks were iden-
2
tified as the recovered starting compounds or all-trans-
b-retinoic acid (2) on the basis of mass spectrometry.
3-(Retinamido)cephalosporanic acid (11). To a solution
of (+)-8 (0.144 g, 0.438 mmol) in dry CH2Cl2 (12 mL)
was added trimethylsilyl bromide (0.260 g, 1.65 mmol).
The reaction mixture was stirred at 25 ꢀC under N2 for
10 min. In a separate flask, all-trans-b-retinoic acid (2,
0.150 g, 0.499 mmol) in dry CH2Cl2 (10 mL) and pyri-
dine (0.201 g, 2.52 mmol) was treated with ethyl chlor-
oformate (0.062 g, 0.571 mmol) under N2 at À5 ꢀC for
10 min. The resultant solution was transferred by a syr-
inge into the above reaction mixture. After the solution
was stirred for 1.5 h, it was washed with H2O
(3Â40 mL), dried over Na2SO4 (s), filtered, and con-
centrated under reduced pressure. Crystallization from
a mixture of Et2O hexane (2:1) gave single isomer 11
(0.144 g, 0.263 mmol) as crystals in 60% yield: mp 80–
82 ꢀC; UV lmax (EtOH) 355 nm (E 43,000); IR (CH2Cl2)
3400–3300 (NH, COOH), 1785 (b-lactam), 1730 (ester),
3-[(Retinoyloxy)methyl]-7-(phenoxyacetamido)-3-cephem-
4-carboxylic acid (6). To a solution of (+)-3 (0.265 g,
0.499 mmol) in dry CH2Cl2 (12 mL) was added tri-
methylsilyl bromide (0.130 g, 0.824 mmol). The reaction
mixture was stirred at 25 ꢀC under N2 for 10 min. The
mixture was then injected into a solution of all-trans-b-
retinoic acid (2, 0.150 g, 0.499 mmol) and NaHCO3
(0.30 g, 3.6 mmol) in DMF (20 mL) and stirred under N2
for 1.0 h. The solution was diluted with EtOAc (60 mL)
and H2O (80 mL). The organic layer was separated and
washed with aqueous HCl solution (1%, 40 mL) and
H2O (50 mL). The organic layer was dried over MgSO4
(s), filtered, and concentrated under reduced pressure.
Purification by use of silica gel column chromatography
with CHCl3 as eluant afforded single isomer 6 (0.155 g,
0.240 mmol) as a foam in 48% yield: UV lmax (EtOH)
352 nm (E 41,500); IR (CH2Cl2) 3410–3300 (NH,
COOH), 1782 (b-lactam), 1730 (ester), 1720 (C¼O),
1710 (C¼O), 1668 (amide) cmÀ1
;
1H NMR (CDCl3/
D2O) d 1.14 (s, 6H, (CH3)2C), 1.30–2.41 (m, 15H,
3ÂCH3+3ÂCH2), 2.21 (s, 3H, CH3CO), 3.67
(d, 2J=19 Hz, 1H, CHS), 3.79 (d, 2J=19 Hz, 1H,
CHS), 4.81 (d, 2J=13 Hz, 1H, CHO), 5.19 (d,
1
1670(amide) cmÀ1; H NMR (CDCl3/D2O) d 1.15 (s,
3
6H, (CH3)2C), 1.30–2.41 (m, 15H, 3ÂCH3+3ÂCH2),
2J=13 Hz, 1H, CHO), 5.20 (d, J=5.0 Hz, 1H, HC(6)),
2
2
3.68 (d, J=19 Hz, 1H, CHS), 3.82 (d, J=19 Hz, 1H,
5.80 (d, 3J=5.0 Hz, 1H, HC(7)), 5.60–6.74 (m, 6H,
6Â=CH). Anal. calcd for C30H38N2O6S: C, 64.96; H,
6.91; N, 5.05; S, 5.78. Found: C, 64.90; H, 6.83; N, 5.16;
S, 5.61.
2
CHS), 4.61 (br s, 2H, OCH2CO), 4.88 (d, J=13 Hz,
1H, CHO), 5.11 (d, 2J=13 Hz, 1H, CHO), 5.18 (d,
3J=4.9 Hz, 1H, HC(6)), 5.78 (d, 3J=4.9 Hz, 1H,
HC(7)), 5.61–6.72 (m, 6H, 6Â=CH), 6.99–7.40 (br s,
5H, C6H5). Anal. calcd for C36H42N2O7S: C, 66.85; H,
6.55; N, 4.33; S, 4.96. Found: C, 66.79; H, 6.36; N, 4.41;
S, 4.99.
3-(Retinamido)-1-oxocephalosporanic acid (12). To a
solution of (+)-8 (0.140 g, 0.426 mmol) in dry CH2Cl2
(15 mL) was added trimethylsilyl bromide (0.253 g,
1.60 mmol). The reaction mixture was stirred at 25 ꢀC
under N2 for 10 min. Then 3-ClC6H4CO3H (0.097 g,
0.56 mmol) was added into the reaction mixture, which
was stirred for 1 h. In a separate reaction flask, all-trans-
b-retinoic acid (2, 0.146 g, 0.484 mmol) in dry CH2Cl2
(12 mL) and pyridine (0.195 g, 2.44 mmol) was treated
with ethyl chloroformate (0.0600 g, 0.554 mmol) under
N2 at À5 ꢀC for 15 min. The resultant solution was
transferred by a syringe into the above mixture. After
the solution was stirred for 1.5 h, it was washed with
H2O (3Â40 mL), dried over Na2SO4 (s), filtered, and
concentrated under reduced pressure. Crystallization
from Et2O gave 12 (0.102 g, 0.179 mmol) as a single
isomer in 42% yield: mp 89–92 ꢀC (dec.); UV lmax
(EtOH) 355 nm (E 43,100); IR (CH2Cl2) 3420–3310
(NH, COOH), 1791 (b-lactam), 1730 (ester), 1715
3-[(Retinoyloxy)methyl]-7-(phenoxyacetamido)-(1-oxo)-
3-cephem-4-carboxylic acid (7). To a solution of (+)-3
(0.260 g, 0.490 mmol) in dry CH2Cl2 (15 mL) was added
trimethylsilyl bromide (0.127 g, 0.808 mmol). The reac-
tion mixture was stirred at 25 ꢀC under N2 for 10 min.
Subsequently, 3-ClC6H4CO3H (0.086 g, 0.50 mmol) was
added into the reaction mixture, which was stirred for
1 h. Then, this reaction mixture was added to a solution
of all-trans-b-retinoic acid (2, 0.147 g, 0.490 mmol) and
NaHCO3 (0.30 g, 3.6 mmol) in DMF (25 mL). After the
solution was stirred for 1 h, it was diluted with EtOAc
(60 mL) and H2O (80 mL). The organic layer was sepa-
rated, washed with aqueous HCl solution (1%, 40 mL)
and H2O (50 mL), dried over MgSO4 (s), filtered, and
concentrated under reduced pressure. Purification by
use of silica gel column chromatography with CHCl3 as
eluant followed by crystallization from Et2O gave single
isomer 7 (0.130 g, 0.196 mmol) as yellow crystals in 40%
yield: mp 70–72 ꢀC (dec.); UV lmax (EtOH) 352 nm (E
42,000); IR (CH2Cl2) 3420–3300 (NH, COOH), 1789 (b-
1
(C¼O), 1669 (amide) cmÀ1; H NMR (CDCl3/D2O) d
1.15 (s, 6H, (CH3)2C), 1.35–2.42 (m, 15H, 3ÂCH3+
2
3ÂCH2), 2.22 (s, 3H, CH3CO), 3.78 (d, J=16 Hz, 1H,
HCSO), 3.99 (d, 2J=16 Hz, 1H, HCSO), 4.84 (d,
2
2J=13 Hz, 1H, CHO), 5.08 (d, J=13 Hz, 1H, CHO),
lactam), 1730 (ester), 1715 (C¼O), 1675 (amide) cmÀ1
;
5.30 (d, 3J=5.0 Hz, 1H, HC(6)), 5.80 (d, 3J=5.0 Hz,

2146
G. H. Hakimelahi et al. / Bioorg. Med. Chem. 9 (2001) 2139–2147
1H, HC(7)), 5.62–6.73 (m, 6H, 6Â=CH). Anal. calcd
for C30H38N2O7S: C, 63.14; H, 6.71; N, 4.91; S, 5.62.
Found: C, 63.34; H, 6.60; N, 5.03; S, 5.41.
Reversal of keratinization in tracheal organ culture by
retinoids
After incubation for 24 h, cultures were divided into six
groups (10 cultures/group) and treated in the following
manner. Group 1: untreated control; group 2: all-trans-
Determination of solubility of all-trans-ꢀ-retinoic acid–
cephem conjugates 6, 7, 11, 12, and all-trans-ꢀ-retinoic
acid (2) in water
b-retinoic acid (2) with 1.0Â10À8, 1.0Â10À9, 1.0Â10À10
,
and 1.0Â10À11 M; group 3: retinoids 6, 7, 11, and 12
with 1.0Â10À8 M; group 4: retinoids 6, 7, 11, and 12
with 1.0Â10À9 M; group 5: retinoids 6, 7, 11, and 12
with 1.0Â10À10 M; and group 6: retinoids 6, 7, 11, and
12 with 1.0Â10À11 M. In the meantime, retinoid stock
solutions were prepared in DMSO and stored at
À110 ꢀC. Aliquots of the retinoid solutions were added
to the culture medium to give the concentrations of
1.0Â10À8–10À11 M. Control cultures received equivalent
amounts of the solvent.
A solution of each compound (80 mg) in 1-octanol
(7.0 mL) was shaken in a 30-mL volumetric flask with
phosphate buffer (0.10 M, 10.0 mL) for 24 h. 1-Octanol
was evaporated and the aqueous suspension was agi-
tated for 10 h. This solution was filtered from undis-
solved solid through a sintered glass funnel (4.0–5.5
mesh ASTM), and the concentration of the solution was
determined by UV absorbance.
Enzymatic hydrolysis of all-trans-ꢀ–retinoic acid–
cephem conjugate 7 in a mixture of DMSO-d6 and
phosphate buffer (1:1 mL/mL, pD 7.2)—(1H NMR
study)
Under subdued light, fresh retinoids were added to the
culture medium every 4 h. After 96 h, all cultures were
fixed in 10% buffered formalin and processed for histo-
logical examination. Five micron sections were stained
in hematoxylin and eosin, and scored with a microscope
for the presence of keratin and keratohyaline granules.
Retinoid 7 (0.0650 g, 0.122 mmol) was dissolved in a
mixture of DMSO-d6 (3.0 mL) and 0.10 M deutrated
phosphate buffer (3.0 mL, pD 7.2) at 25 ꢀC. The 1H
NMR spectrum was taken at this temperature and then
0.60 mL (9.9 units) of bL (from S. aureus 95) buffer
A test result was scored as ‘inactive’ if both keratin and
keratohyaline granules were observed. On the other
hand, it was scored as ‘active’ if neither keratin nor
keratohyaline granules were observed, or if keratohya-
line granules alone were absent.51 The above experi-
ments were also performed in the presence of 0.85 nM of
a bL from S. aureus.
solution was added. The H NMR spectrum at 25 ꢀC
1
was taken immediately; the spectrum of 7 changed
rapidly to that of the eliminated compounds 14 and 2.
The mixture was extracted with CDCl3 (2Â3.0 mL) to
remove all-trans-b-retinoic acid (2), which was found to
be identical with an authentic sample. The aqueous
solution was frozen and lyophilized to afford a residue.
Purification by silica gel column chromatography using
EtOAc/MeOH 9:1 as eluant gave 14 (0.0420 g,
Anticancer test procedure in vitro
Murine leukemias (L1210 and P388), sarcoma 180 (S-
180), breast carcinoma (MCF7), human T-lymphoblasts
(Molt4/C8 and CEM/0), and normal human embryonic
lung (HEL) cell lines were cultured in DMEM supple-
mented with 10% FBS, 2.0 mM glutamine, 100 U/mL
penicillin, and 100 mg/mL streptomycin in a humidified
atmosphere with 5% CO2 at 37 ꢀC and pH 7–7.3.54
Under this condition, the generation time for L1210,
P388, S-180, MCF7, Molt4/C8, CEM/0, and HEL cells
was about 13, 12, 17, 17, 18, 21, and 29 h, respectively.
Compounds 6, 7, 11, 12, 7+bL (120 nM), 7+dsFv3-bL
(120 nM) as well as the reference compounds ara-C and
all-trans-b-retinoic acid (2), at various concentrations,
were added to L1210, P388, S-180, MCF7, Molt4/C8,
CEM/0, and HEL cells (250 cells/mL) in their exponen-
tial phase of growth. The cell numbers of the control
cultures, as well as that of the cultures supplemented
with the test compounds, were determined after 24, 48,
and 72 h of growth. The IC50 values were estimated from
dose–response curves compiled from two independent
experiments and represent the compound concentration
(mM) required to inhibit proliferation of the respective
cell lines by 50% after 72 h incubation (Table 1).
0.105 mmol) in 86% yield: mp 198–199 ꢀC; H N MR
1
2
(DMSO-d6/D2O) d 3.48 (d, J=13.5 Hz, 1H, CHSO),
3.47 (d, 2J=13.5 Hz, 1H, CHSO), 4.71 (br s, 2H,
OCH2CO), 4.82 (d, 3J=4.2 Hz, 1H, HC(6)), 5.36 (d,
3J=4.2 Hz, 1H, HC(7)), 5.58 (s, 1H, ¼CH), 5.81 (s, 1H,
¼CH), 7.38 (br s, 5H, C6H5). CIMS: 381 (M++1), 380
(M+).
Kinetic procedure
Determination of the kinetic parameters for bL and
dsFv3–bL (secreted from E. coli)42b with the substrate 7
was done according to an established procedure.42c
Briefly, three sets of vials containing 1.5-mL solutions of
varying concentrations of substrate 7 at 37 ꢀC in PBS/
DMSO (9.5:0.5, pH 7.2) were treated with 0.15 nM of
dsFv3–bL. Individual sets of samples were quenched
after 60, 90, and 120 s, respectively, by adding 0.50 mL
of the reaction solution to 0.50 mL of 40% CH3CNin
180 mM KH2PO4 (pH 4.2). Samples of the quenched
reaction mixtures were injected onto a 0.46Â15-cm C18
reversed-phase HPLC column (50% CH3CNas eluant)
at 1.0 mL/min. Prodrug 7 and product 2 concentrations
were monitored by absorbance at 352 nm. Linear rate
plots were used to obtain reaction velocities. KM and
kcat were determined from the slope and intercept of
Lineweaver–Burk plots.
Acknowledgements
For financial support, we thank the National Science
Council of Republic of China, Academia Sinica, the

G. H. Hakimelahi et al. / Bioorg. Med. Chem. 9 (2001) 2139–2147
2147
Ministry of Medical Sciences of Iran, Pasteur Institute,
Tehran and Shiraz Universities. The effort of Dr. Shwu-
Chen Tsay for drawing the schemes is greatly appreciated.
29. Hakimelahi, G. H.; Zakerinia, M. Iran J. Med. Sci. 1986,
13, 25.
30. Hakimelahi, G. H.; Shiao, M.-J.; Hwu, J. R. Helv. Chim.
Acta 1992, 75, 1840.
31. Waxman, D. J.; Strominger, J. L. J. Biol. Chem. 1980,
255, 3964.
References and Notes
32. Hamilton-Miller, J. M. T.; Newton, G. G. F.; Abraham,
E. P. Biochem. J. 1970, 116, 371.
1. Moore, T. The Vitamins, 2nd ed.; Academic: New York,
1967; Vol. 1, p 286.
2. Moon, R. C.; Grubbs, C. J.; Sporn, M. B.; Goodman, D. G.
Nature (London) 1977, 267, 620.
3. Sporn, M. B.; Squire, R. A.; Brown, C. C.; Smith, J. M.;
Wenk, M. L.; Springer, S. Science 1977, 195, 487.
4. Moon, R. C.; Thompson, H. J.; Becci, P. J.; Grubbs, C. J.;
Gander, R. J.; Newton, D. L.; Smith, J. M.; Phillips, S. L.;
Henderson, W. R.; Mullen, M. T.; Brown, C. C.; Sporn, M. B.
Cancer Res. 1979, 39, 1339.
33. O’Callaghan, C. H.; Kirby, S. M.; Morris, A.; Waller,
R. E.; Duncombe, R. E. J. Bacteriol. 1972, 110, 988.
34. Faraci, W. S.; Pratt, R. F. J. Am. Chem. Soc. 1984, 106, 1489.
35. Russell, A. D.; Fountain, R. H. J. Bacteriol. 1971, 106, 65.
36. Boyd, D. B.; Lunn, W. H. W. J. Med. Chem. 1979, 22, 778.
37. Boyd, D. B. J. Org. Chem. 1985, 50, 886.
38. Page, M. L.; Proctor, P. J. Am. Chem. Soc. 1984, 106, 3820.
39. Grabowski, E. J.; Douglas, A. W.; Smith, G. B. J. Am.
Chem. Soc. 1985, 107, 267.
5. Mayer, H.; Bollag, W.; Hanni, R.; Ruegg, R. Experientia
1978, 34, 1105.
6. Tanaka, H.; Kagechika, H.; Kawachi, E.; Fukasawa, H.;
Hashimoto, Y.; Shudo, K J. Med. Chem. 1992, 35, 567.
7. Collins, S. J.; Gallo, R. C.; Gallagher, R. E. Nature (Lon-
don) 1977, 270, 347.
8. Koeffler, H. P. Blood 1983, 62, 709.
9. Sporn, M. B.; Roberts, A. B.; Goodman, D. S. The Reti-
noids; Academic: Orlando, FL, 1984.
10. Dawson, M. I.; Okamura, W. H. Chemistry and Biology of
Synthetic Retinoids; CRC: Boca Raton, FL, 1990.
11. Petkovich, M.; Brand, N. J.; Kurst, A.; Chambon, P.A
Nature (London) 1987, 330, 444.
40. Hwu, J. R.; Moshfegh, A. A.; Tsay, S.-C.; Lin, C.-C.;
Tseng, W. N.; Azaripour, A.; Mottaghian, H.; Hakimelahi,
G. H. J. Med. Chem. 1997, 40, 3434.
41. Bagshawe, K. D. Clin. Pharmacokinet. 1994, 27, 368.
42. (a) Rodrigues, M. L.; Carter, P.; Wirth, C.; Mullins, S.;
Lee, A.; Blackburn, B. K. Chem. Biol. 1995, 2, 223 and refer-
ences cited therein. (b) Rodrigues, M. L.; Presta, L. G.; Kotts,
C. E.; Wirth, C.; Mordenti, J.; Osaka, G.; Wong, W. L. T.;
Nuijens, A.; Blackburn, B.; Carter, P. Cancer Res. 1995, 55,
63. (c) Jungheim, L. N.; Shepherd, T. A.; Meyer, D. L. J. Org.
Chem. 1992, 57, 2334.
43. (a) Svensson, H. P.; Kadow, J. F.; Vrudhula, V. M.; Wal-
lace, P. M.; Senter, P. D. Bioconjugate Chem. 1992, 3, 176. (b)
Svensson, H. P.; Frank, I. S.; Berry, K. K.; Senter, P. D. J.
Med. Chem. 1998, 41, 1507.
12. Giguere, V.; Ong, E. S.; Segui, P.; Evans, R. M. Nature
(London) 1987, 330, 624.
13. Brand, N.; Petkovich, M.; Kurst, A.; Chambon, P.; de
The, H.; Marchio, A.; Tiollais, P.; Dejean, A. Nature (Lon-
don) 1988, 332, 850.
14. Hashimoto, Y.; Kagechika, H.; Kawachi, E.; Shudo, K.
Jpn. J. Cancer Res. (Gann) 1988, 79, 473.
15. Hashimoto, Y.; Petkovich, M.; Gaub, M. P.; Kagechika,
H.; Shudo, K.; Chambon, P. Mol. Endocrinol. 1989, 3, 1046.
16. Kurst, A.; Kastner, P.; Petkovich, M.; Zelent, A.; Cham-
bon, P. Proc. Natl. Acad. Sci. U.S.A. 1989, 86, 5310.
17. Hashimoto, Y.; Kagechika, H.; Shudo, K. Biochem. Bio-
phys. Res. Commun. 1990, 166, 1300.
18. Sporn, M. B.; Newton, D. L.; Smith, J. M.; Acton, N.;
Jacobson, A. E.; Brossi, A. Carcinogens: Identification and
Mechanism of Action; Raven: New York, 1979; p 441.
19. Hakimelahi, G. H.; Zakerinia, M.; Zamanpour, K. Med.
J. Iran 1989, 3, 187.
20. Hakimelahi, G. H.; Just, G.; Chopra, D. J. Sci. Tech. Iran
1985, 9, 111.
21. Hakimelahi, G. H.; Zarrinehzad, M.; Jarrahpour, A. A.;
Sharghi, H. Helv. Chim. Acta 1987, 70, 219.
22. Bard, D. R.; Lasnitzki, I. Br. J. Cancer 1977, 35, 115.
23. Gander, R. J.; Gurney, J. A. Belgian Patent 847 942, 1977;
Chem. Abstr. 1978, 88, 89892e.
44. Vrudhula, V. M.; Svensson, H. P.; Kennedy, K. A.; Sen-
ter, P. D.; Wallace, P. M. Bioconjugate Chem. 1993, 4, 334.
45. Hanessian, S.; Wang, J. Can. J. Chem. 1993, 71, 896.
46. Hudyma, T. W.; Bush, K.; Colson, K. L.; Firestone, R. A.;
King, H. D. Bioorg. Med. Chem. Lett. 1993, 3, 323.
47. Jungheim, L. N.; Shepherd, T. A.; Kling, J. K. Hetero-
cycles 1993, 35, 339.
48. Jungheim, L. N.; Shepherd, T. A. Chem. Rev. 1994, 94, 1553.
49. Albrecht, H. A.; Beskid, G.; Chan, K.-K.; Christenson,
J. G.; Cleeland, R.; Deitcher, K. H.; Georgopapadakou, N. H.;
Keith, D. D.; Pruess, D. L.; Sepinwall, J.; Specian, A. C.;
Then, R. L.; Weigele, M.; West, K. F.; Yang, R. J. Med.
Chem. 1990, 33, 77.
50. Bonjouklian, R.; Phillips, M. L. Tetrahedron Lett. 1981,
22, 3915.
51. (a) Sporn, M. B.; Dunlop, N. M.; Newton, D. L.; Hen-
derson, W. R. Nature (London) 1976, 263, 110. (b) Sporn,
M. B.; Clamon, G. H.; Dunlop, N. M.; Newton, D. L.; Smith,
J. M.; Saffiotti, U. Nature (London) 1975, 253, 47. (c) Clamon,
G. H.; Sporn, M. B.; Smith, J. M.; Saffiotti, U. Nature (Lon-
don) 1974, 250, 64.
52. Finney, D. J. Probit Analysis; Cambridge University:
Cambridge, 1971.
24. Bollag, W. Eur. J. Cancer 1974, 10, 731.
53. Helwig, J. T.; Council, K. A. SAS User’s Guide; SAS:
Raleigh, 1979.
25. Pawson, B. A.; Cheung, H. C.; Han, R. J. L.; Trown,
P. W.; Buck, M.; Hansen, R.; Bollag, W.; Ineichen, U.; Pleil,
H.; Ruegg, B.; Dunlop, N. M.; Newton, D. L.; Sporn, M. B. J.
Med. Chem. 1977, 20, 918.
54. Rai’c-Malic
A. M.; Pavelic, K.; DeClercq, E.; Mintas, M. J. Med. Chem.
1999, 42, 2673.
, S.; Hergold-Brundic, A.; Nagl, A.; Grdis,
´ ´
´
26. Mangelsdorf, D. J.; Umesons, K.; Evans, R. M. In The
Retinoids, Biology, Chemistry, and Medicine, 2nd ed.; Sporn,
M. B., Robert, A. B., Goodman, D. S., Eds.; Raven: New
York, 1994; pp 319–350.
27. Sporn, M. B.; Dunlop, N. M.; Newton, D. L.; Smith, J. M.
Fed. Proc., Fed. Am. Soc. Exp. Biol. 1976, 35, 1332.
28. Just, G.; Hakimelahi, G. H. US Patent 4 385 175, 1983.
55. Hong, W. K.; Itri, M. I. In The Retinoids, Biology, Chem-
istry, and Medicine, 2nd ed.; Sporn, M. B., Robert, A. B.; Good-
man, D. S., Eds.; Raven Press: New York, 1994; pp 597–630.
56. Meijer, D. K. F.; Jansen, R. W.; Molema, G. Antiviral
Res. 1992, 18, 215.
57. Sun, S.-Y.; Wan, H.; Yue, P.; Hong, W. K.; Lotan, R. J.
Biol. Chem. 2000, 275, 17149.