Synthesis and biological evaluation of gemcitabine-lipid conjugate (NEO6002)

Article abstract of DOI:10.1016/j.bmcl.2005.03.046

A novel gemcitabine-lipid conjugate 5 was synthesized and tested for its in vivo efficacy and toxicity. Compound 5 was tested in BxPC-3 human pancreatic tumor model in SCID mice and exhibited promising activity and lower toxicity when compared with Gemzar

Full text of DOI:10.1016/j.bmcl.2005.03.046

Bioorganic & Medicinal Chemistry Letters 15 (2005) 2571–2574
Synthesis and biological evaluation of gemcitabine–lipid
conjugate (NEO6002)
Shoukath M. Ali, Abdul R. Khan, Moghis U. Ahmad, Paul Chen,
Saifuddin Sheikh and Imran Ahmad*
NeoPharm Inc., Research and Development Facility, 1850 Lakeside Drive, Waukegan, IL 60085, USA
Received 19 January 2005; revised 8 March 2005; accepted 14 March 2005
Available online 11 April 2005
Abstract—A novel gemcitabine–lipid conjugate 5 was synthesized and tested for its in vivo efficacy and toxicity. Compound 5 was
tested in BxPC-3 human pancreatic tumor model in SCID mice and exhibited promising activity and lower toxicity when compared
with Gemzarꢀ.
ꢁ 2005 Elsevier Ltd. All rights reserved.
Gemcitabine (20,20-difluorodeoxyribofuranosylcytosine,
dFdC) 1 is a difluorinated analogue1 of deoxycytidine
and currently marketed as Gemzarꢀ for the treatment
of non-small cell lung and pancreatic cancer. It is a po-
tent antitumor agent in murine and human xenograft
solid tumor models.2 The drug acts as antimetabolite,
inhibiting ribonucleotide reductase and DNA synthesis.
It is activated inside the cell by deoxycytidine kinases to
its active forms, the diphosphate and triphosphate of
gemcitabine (dFdCDP, dFdCTP).3 However, the com-
pound has a narrow therapeutic index due to rapid
deamination by an endogenous enzyme, deoxycytidine
deaminase, to its corresponding inactive uracil deriva-
tive (dFDU).4 The other limitations include the cellular
mechanism of anticancer drug resistance, inadequate
uptake of drug by lymphoid and hematopoietic tissues,
toxicity, oral bioavailability, short-drug half-life, and
prevention of extracellular drug metabolism.5
Alexander et al.7 reported the development of lipid–
nucleoside conjugates, which comprise an alkyl
phospholipid moiety covalently conjugated with nucleo-
side drugs such as gemcitabine 3, and demonstrated to
have cytotoxicity against most cell lines (Fig. 1).
GemMP[10],8 a novel multimeric prodrug of gemcita-
bine monophosphate, is a potent cytotoxic agent that
serves to induce apoptosis in association with increased
FAS expression in cultured thyroid cancer cell lines.
GemMP[10] has shown to decrease tumor cell growth
at concentrations ranging from 1 to 50 nM.
To improve the half-life and reduce the toxicity of gem-
citabine, we chose to conjugate gemcitabine with ether
analogue of cardiolipin. Cardiolipin (CL) 4 constitutes
a class of complex phospholipids that occur mainly in
heart and skeletal muscles and that usually associated
with membranes of subcellular fractions showing high
metabolic activity, for example, mitochondria.9 CL
plays an important role in many biological processes
and is currently utilized in many areas of biochemical
and biomedical research.10 Our novel approach was to
conjugate the ether analogue of synthetic cardiolipin
with gemcitabine via a succinate linker as prodrug. We
speculate that the succinate ester group would be hydro-
lyzed in vivo by the action of esterases which are preva-
lent in cells. In this manner there will be a prolonged
release of the drug.
Methods, that have been employed to circumvent the
problems associated with gemcitabine and other anti-
cancer nucleoside-based drugs, include the development
of lipid–nucleosides conjugates, prodrugs, and liposome
preparations. Some of these methods include the synthe-
sis of derivatives6a of gemcitabine as prodrugs such as
elaidic acid (50)-gemcitabine ester 2a and elaidic acid
(N4)–gemcitabine amide 2b and amino acid esters.6b
Keywords: Gemcitabine; Lipid–nucleoside conjugate; Cardiolipin;
Prodrug.
*
In this letter, we report the synthesis of gemcitabine–
lipid conjugate (Fig. 1) 5(NEO6002) and the evaluation
Corresponding author. Tel.: +1 847 8870800; fax: +1 847 8879281;
e-mail: Imran@neophrm.com
0960-894X/$ - see front matter ꢁ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.03.046

2572
S. M. Ali et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2571–2574
NH2.HCl
N
NH2
N
C12H25
S
O
C10H21
O
N
O
N
O
HO
O
P
O
O
O
OH
F
F
F
F
1
OH
OH
NHR2
3
N
OR
OR
RO
RO
O
O
OH
N
O
R1O
O
P
O
O
O
P
O
OH
OH
F
F
OH
4
Cardiolipin (R = Fatty acid chain)
NH2
2a. R1 = C18:1 Fatty acid chain, R2 = H
2b. R1 = H, R2 = C18:1 Fatty acid chain
N
O
N
O
F
O
O
HO
F
O
OC6H13
OC6H13
C6H13
O
O
O
O
O
C6H13
O
P
O
O
P
O
OMe
OMe
5
Figure 1. Structures of gemcitabine, gemicatabine derivatives, cardiolipin and gemcitabine–lipid conjugate.
of its biological activity of tumor growth inhibition in
vivo, and prolongation of animals survival. The syn-
thetic methodology (Scheme 1), which we have devel-
oped for the synthesis of cardiolipin, involves the
application of phosphoramidite approach.11 Alkylation
of commercially available (R)-(+)-3-benzyloxy-1,2-pro-
panediol 6 with 1-bromohexane in the presence of so-
dium hydride (60% in oil dispersion) in dimethyl
formamide (DMF) followed by debenzylation via cata-
lytic hydrogenolysis (H2/Pd–C) provided 1,2-di-O-hex-
yl-sn-glycerol 8 in 89% yield. The next step involves
the reaction of 1,2-di-O-hexyl-sn-glycerol 8 with the
bifunctional phosphitylating reagent N,N-diisopropyl-
methyl phosphonamidic chloride in the presence of
N,N-diisopropylethylamine (DIPEA) in dichlorometh-
ane at room temperature to give 9, which was not
isolated but subsequently reacted with 2-benzyloxy-1,3-
propanediol in the presence of 1H-tetrazole to provide
phosphite triester 10. In situ oxidation of phosphite
triester 10 with m-chloroperbenzoic acid (mCPBA)
at ꢀ40 ꢂC afforded 2-O-benzyl-1,3-bis-(1,2-di-O-hexyl-
sn-glycero-3-phosph-oryl)glycerol dimethyl ether 11 as
colorless oil in 79% yield after purification on silica
gel column (hexane–ethyl acetate, 8:2–6:4). Hydrogenol-
ysis of compound 11 with H2/Pd–C at 50 psi for
2 h furnished 1,3-bis-(1,2-di-O-hexyl-sn-glycero-3-phos-
phoryl)glycerol dimethyl ether 12 as colorless oil in
96% yield. The central hydroxyl functionality of cardio-
lipin analogue 12 was reacted with succinic anhydride in
the presence of triethylamine and 4-dimethylamino pyr-
idine (DMAP) in 1,2-dichloroethane to provide 1,3-bis-
(1,2-di-O-hexyl-sn-glycero-3-phosphoryl)-2-succinyl-
glycerol dimethyl ether12 13 in 79% yield. Coupling of 13
with 4-N-30-O-bis(tert-butoxycarbonyl)gemcitabine13 14
in the presence of N,N-dicyclohexyl carbodiimide
(DCC) and DMAP in dichloromethane at room temper-
ature for 8 h followed by purification on silica gel col-
umn (hexane–ethyl acetate, 7:3–4:6) afforded 15 in
84% yield. The protecting groups were removed using
trifluoroacetic acid (TFA) in dichloromethane at room
temperature. The reaction solution was neutralized with
5% aqueous sodium bicarbonate at 0 ꢂC, extracted with
dichloromethane and concentrated. Purification of the
crude compound on silica gel column (chloroform–
methanol, 98:2–96:4) afforded pure 50-O-succinyl-[2-O-
1,3-bis-(1,2-di-O-hexyl-sn-glycero)-3-phosphorylglycerol
dimethyl ether]gemcitabine 5 as colorless viscous oil in
80% yield. The product was characterized14 by 1H
NMR, 13C NMR, IR, and Mass spectrometry. The pur-
ity was checked by HPLC and elemental analysis.
Toxicity and efficacy of compound 5 were evaluated and
compared with Gemzarꢀ in mice. In order to perform in
vivo studies, compound 5 was formulated into a co-sol-
vent system composed of 5% ethanol and 5% dextrose at
a concentration 0.5, 1.0, 1.5, and 2.0 mg/ml.
Multiple dose toxicity of compound 5 and Gemzarꢀ on
non-diseased animals was conducted using CD2F1 mice.
Mice were treated with compound 5 or Gemzarꢀ for five

S. M. Ali et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2571–2574
2573
C6H13
O
O
C6H13
O
O
HO
HO
C6H13
O
O
a
c
b
C6H13
C6H13
N
C6H13
OH
O
P
OBn
OBn
OMe
6
7
8
9
d
OC6H13
OC6H13
C6H13
O
O
OC6H13
OC6H13
C6H13
O
O
OBn
O
e
O
C6H13
OBn
11
C6H13
O
P
O
O
O
P
O
P
O
P
O
O
OMe
OMe
OMe
OMe
10
OH
O
f
O
OC6H13
OC6H13
C6H13
O
O
OC6H13
OC6H13
C6H13
O
O
O
O
O
P
O
O
P
g
OH
C6H13
C6H13
O
O
P
O
O
h
P
O
O
O
O
OMe
OMe
OMe
OMe
13
12
NHBoc
N
O
N
NH2
NHBoc
N
HO
O
F
N
14
N
O
N
O
O
BocO
F
O
F
F
O
O
O
HO
F
O
BocO
O
F
O
OC6H13
OC6H13
C6H13
C6H13
O
O
O
O
OC6H13
OC6H13
O
C6H13
C6H13
O
O
i
O
O
O
O
P O
O
P O
O
O
P
O
P
O
OMe
OMe
OMe
OMe
5
15
Scheme 1. Reagents and conditions: (a) 1-bromohexane, NaH, DMF, 60 ꢂC, 24 h; (b) H2/Pd–C, ethyl acetate, 4 h; (c) N,N-diisopropylmethyl
phosphonamidic chloride, DIPEA, CH2Cl2, rt, 2 h; (d) 2-benzyloxy-1,3-propanediol, 1H-tetrazole, CH2Cl2, rt, 6 h; (e) mCPBA, ꢀ40 ꢂC, 1 h; (f) H2,
Pd/C, 50 psi, 2 h; (g) succinic anhydride, Et3N, DMAP, CH2Cl2, rt, 5 h; (h) 14, DCC, DMAP, CH2Cl2, rt 8 h; (i) TFA, CH2Cl2, rt, 6 h; aq NaHCO3.
or six daily doses and mortality of the mice were re-
corded accordingly. The dose, schedule, and animal
death are summarized in Table 1. No mortality was seen
in groups injected with 18 lmol/kg or 27 lmol/kg of
compound 5. However, all animals were moribund when
they were injected with Gemzarꢀ at a dose of 27 lmol/
kg or 36 lmol/kg, while only 25% mice died at dose of
36 lmol/kg of compound 5. This suggested that com-
pound 5 is less toxic at equimolar dosage when com-
pared with Gemzarꢀ.
The in vivo efficacy was evaluated with human pancre-
atic (BxPC-3) tumor model in SCID mice. Tumor cells
(5 · 106) were implanted subcutaneously in each SCID
mice. When the tumors reached a volume of 80–
160 mm3, they were randomly divided into six groups.
Each of the four groups were administered with 9, 18,
and 36 lmol/kg of compound
5 formulation, 9,
18 lmol/kg of Gemzarꢀ and equal volume of control
vehicle, respectively, on days 1, 7, 15, 22, 26, 29, and
33. The tumor volumes were measured with Vernier cali-
pers. Percentage of tumor growth (% growth) was calcu-
lated as 100 · (Vt/V0), where Vt and V0 represent the
tumor volume at any given day and on the first day
when the treatment started. The study was monitored
up to 50 days where groups of mice treated with all
doses of compound 5 and Gemzarꢀ showed significant
tumor growth inhibition compared to those in control
group (Table 2).
Table 1. Mortality of non-diseased CD2F1 mice treated with com-
pound 5 or Gemzarꢀ
Compds
Dose
(lmol/kg)
Schedule
(day)
% Mortalitya
Day 8
D5E5Wb
(N/A)
18
27
1–6
1–5
1–6
1–6
1–5
1–6
1–6
0
0
Gemzarꢀ
Gemzarꢀ
Gemzarꢀ
100
100
0
In summary, we have designed and synthesized a novel
gemcitabine–lipid conjugate that showed lower toxicity
and promising efficacy in comparison to Gemzarꢀ. This
class of compounds with low toxicity may be further ex-
plored to circumvent the problems related with Gem-
zarꢀ. Efficacy studies using compound 5 in other
tumor models, mechanism of action and the
36
18
5(NEO6002)
5(NEO6002)
5(NEO6002)
27
36
0
25
a Values are of eight animals in each group (N/A = not applicable).
b 5% dextrose–5% ethanol.

2574
S. M. Ali et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2571–2574
Table 2. Inhibition of BxPC-3 human pancreatic tumor growth by
treatment with compound 5 or Gemzarꢀ
5. Kucera, L. S.; Fleming, R. A.; Ishaq, K. S.; Kucera, G. L.
Morris-Natscheke, S. L.; U.S. Pat. Appl. 2002/0082242
A1, 2002. Chem. Abstr. 2002, 137, 57593.
Compds
Dose (lmol/kg)
% Growtha Day 50
6. (a) Myhren, F.; Borretzen, B.; Dalen, A.; Sandvold, M.
PCT Intl. Appl. WO 98/32762 A1, 1998. Chem. Abstr.
1998, 129, 149189; (b) Kim, I.; Song, X.; Vig, B. S.; Mittal,
S.; Shin, H.-C.; Lorenzi, P. J.; Amidon, G. L. Mol.
Pharmaceutics 2004, 1, 117.
7. Alexander, R. L.; Morris-Natschke, S. L.; Ishaq, K. S.;
Fleming, R. A.; Kucera, G. L. J. Med. Chem. 2003, 46,
4205.
D5E5Wb
(N/A)
9
2884 278
1942 438
2007 422
1387 354
1393 477
1022 357
Gemzarꢀ
Gemzarꢀ
18
9
5(NEO6002)
5(NEO6002)
5(NEO6002)
18
36
a Values are means of eight animals in each group; standard deviations
are given in parentheses (N/A = not applicable).
b 5% dextrose–5% ethanol.
8. Kotchetkov, R.; Groschel, B.; Gmeiner, W. H.; Krivtchik,
A. A.; Trump, E.; Bitoova, M.; Cinatl, J.; Kornhuber, B.;
Cinatl, J. Anticancer Res. 2000, 20, 2915.
9. (a) Ioannou, P.; Golding, B. T. Prog. Lipid Res. 1979, 17,
279–318; (b) Semin, B. K.; Saraste, M.; Wikstrom, M.
Biochim. Biophys. Acta 1984, 769, 15.
pharmacokinetic studies will be the subject of further
investigation.
10. (a) Fernandez, J. A.; Kojima, K.; Petaja, J.; Hackeng, T.
M.; Griffin, J. H.; Beutler, E. Blood, Cells, Mol. Dis. 2000,
26, 115; (b) Chauhan, A.; Ray, I.; Chauhan, V. P.
Neurochem. Res. 2000, 25, 423; (c) Chen, Q. P.; Li, Q. T.
Arch. Biochem. Biophys. 2001, 389, 201.
Acknowledgements
We thank the bio-analytical group of PSE department at
NeoPharm for Mass spectral analysis.
11. (a) Krishna, U. M.; Ahmad, M. U.; Ahmad, I. Tetrahe-
dron Lett. 2004, 45, 2077; (b) Lin, Z.; Ahmad, M. U.; Ali,
S. M.; Ahmad, I. Lipids 2004, 39, 285.
12. All new compounds were fully characterized by 1H NMR,
1
Mass spec, and TLC. Selected data of compound 13: H
References and notes
NMR (300 MHz, CDCl3): d 4.26–4.14 (m, 8H); 4.09–4.07
(m, 2H); 3.79 and 3.77 (m, 6H); 3.63, 3.57, 3.52 and 3.46
(each m, 13H), 2.66 (m, 4H), 1.55 (m, 8H), 1.29 (m, 24H);
0.89 (t, J = 12 Hz, 12H).
1. Plunkett, W.; Huang, P.; Xu, Y. Z.; Heinemann, V.;
Grunewald, R.; Gandhi, V. Semin. Oncol. 1995, 22, 3.
2. (a) Lund, B.; Hansen, O. P.; Theilade, K.; Hansen, M.;
Neijt, J. P. J. Natl. Cancer Inst. 1994, 86, 1530–1533; (b)
Casper, E. S.; Green, M. R.; Kelsen, D. P.; Heelan, R. T.;
Brown, T. D.; Flombaum, C. D.; Trochanowski, B.;
Tarassoff, P. G. Invest. New Drugs 1994, 12, 29.
3. (a) Johnson, P. G.; Takimoto, C. H.; Grem, J. L.;
Chabner, B. A.; Allegra, C. J.; Chu, E. In Cancer
Chemotherapy and Biological Response Modifiers; Pinedo,
H. M., Longo, D. L., Chabner, B. A., Eds., Annual 16;
Elsevier Science, 1996, Chapter 1, pp 1–27; (b) Huang, P.;
Plunkett, W. Semin. Oncol. 1995, 22, 19.
13. Guo, Z.-w.; Gallo, J. M. J. Org. Chem. 1999, 64, 8319.
14. Compound 5, 1H NMR (500 MHz, CDCl3): d 7.55 (m,
1H, H); 6.43 (d, J = 7.5 Hz, 1H); 5.93 (m, 1H); 5.22 (m,
1H); 4.71 (m, 1H); 4.54 and 4.51 (dd, J = 12 Hz and 5 Hz,
1H); 4.56–4.39 (m, 2H); 4.26–4.12 (m, 8H); 4.08–4.02 (m,
2H); 3.81–3.76 (m, 6H); 3.62–3.42 (m, 13H); 2.72 (m, 4H);
1.55 (m, 8H); 1.25 (m, 24H); 0.88 (t, J = 12.0 Hz, 12H);
13C NMR (125 MHz, CDCl3): d 171.53, 171.38, 165.77,
155.58, 141.11, 124.14, 122.07, 120.00, 95.51, 83.98, 78.69,
77.25, 77.07, 76.99, 71.77, 70.62, 69.34, 67.61, 64.76, 61.84,
54.60, 31.59, 29.87, 29.49, 25.67, 25.61, 22.55, 22.54, 13.97.
IR (Neat, cmꢀ1): 3330, 3210, 2955, 2930, 2858, 1743, 1651.
MS (ESI) m/z 1111 (M+1), 1133 (M+Na); Anal. Calcd for
C48H87F2N3O19P2: C, 51.93; H, 7.90; F, 3.42; N, 3.79.
Found: C, 51.88; H, 7.94; F, 3.69; N, 3.76.
4. (a) Storniola, A.; Allerheiligen, S.; Pearce, H. Semin.
Oncol. 1997, 24, s7-2–s7-7; (b) Peters, G.; Schornagel, J.;
Milano, G. In Cancer Surveys; Workman, P., Graham, M.
A., Eds.; Cold Spring Harbor Laboratory: New York,
1993; Vol. 17, pp 123–156.