M. Drag-Zalesinska et al. / Bioorg. Med. Chem. Lett. 19 (2009) 4814–4817
4815
prodrug in vivo it will still fall apart into the desired triterpene
(drug) and the non-hazardous amino acids (Fig. 1).
Proliferation profiles of human gastric carcinoma parental
(EPG85-257P) and its daunorubicin resistant (EPG85-257RDB) cell
lines, and of human pancreatic carcinoma parental (EPP85-181P)
and its daunorubicin resistant (EPP85-181RDB) cell lines were
evaluated in multiple experiments after 24 h of treatment with
the compounds using SRB tests for all the compounds tested.12
Usually, most experiments aiming to determine cytotoxicity of bet-
ulinic acid or betulin in the literature are performed for at least
72 h.5,13 In our case, this time of incubation was much too long
for both tested parental cell lines. The most active compounds
(2c, 2d, 2f and 2g) completely destroyed the cells (none of the cell
types could be detected using comet assay performed after 72 h),
making final data not very reliable (not shown here). Additionally,
the response of daunorubicin resistant cell lines after 24 h was not
as good as of parental cell lines.
On the other hand, this observation demonstrates that besides
achieving better water solubility, the mode of action of the tested
compounds is much faster than for betulinic acid and betulin. The
choice of the tested cell lines was based on two major criteria. Tu-
mor models in this study correspond to human organs (stomach
and pancreas), which potentially could be easily ‘reached’ by the
tested compounds after oral administration. Keeping in mind the
known problems with the delivery of betulinic acid to the target
cells this argument should not be neglected. The second criteria
corresponds to the multidrug resistance of several known cancer
cells, therefore we have tested sensitive and daunorubicin resistant
cell lines as well, to additionally check the selectivity of our
compounds.14
Our preliminary attempts in the synthesis of esters of betulin
and betulinic acid with
L
-Boc protected amino acids using DCC
(N,N0-dicyclohexylcarbodiimide)/DMAP
(4-dimethylaminopyri-
dine) in THF were only partially successful (the usual yield after
5–10 days of coupling was only in the range of 10–15%). There-
fore, we decided to use the much stronger coupling reagent
CDI (1,10-carbonyldiimidazole) (Scheme 1). In the patent litera-
ture there is one example of coupling Boc-Gly to betulinic acid
using this method.8 We found that this protocol in every case
yields the desired esters in high yields even after only 24–48 h
of coupling (Table 1).9 In the case of betulinic acid, esters were
obtained at carbon C-3, while in the case of betulin only at car-
bon C-28 (Fig. 1) as determined using 1H NMR by shift of the CH-
OH proton at carbon C-3 (the very characteristic resonance signal
from an unsubstituted derivatives is usually at around 3.0–
3.2 ppm, while substitution of carbon C-3 shifts this signal down-
field by at least 1.0 ppm). In the case of the disubstituted deriv-
atives, the first esterification takes place at carbon C-28 and than
at carbon C-3 as determined by 1H NMR, as well. Interestingly,
we have never observed monosubstituted betulin derivatives at
carbon C-3. Deprotection of the Boc group was achieved using
3 M solution of HCl in dioxane for 30 min.10 The yield of this step
was between 64% and 82%. All the compounds were in the form
of white crystalline solids and were characterized using 1H NMR
and HR-MS.
Water solubility was evaluated by dissolving 1 mg of the com-
pound in 100
l
l of DMSO as described earlier by Jeong for other
For both pancreatic cell lines we have found that the most ac-
betulinic acid derivatives.11 10
ll aliquots were treated with dis-
tive compound is the lysine diester of betulin 2g (IC50 = 3.8
and only slightly weaker is alanine diester of betulin 2d
(IC50 = 4.6 M). However, for both compounds after 24 h of treat-
lM),
tilled water to achieve dilutions equal to 5ꢀ, 10ꢀ, 20ꢀ, 50ꢀ,
100ꢀ. Subsequent analysis of the presence of the precipitate or
gel-like suspension was performed in order to classify all the com-
pounds synthesized. The lysine disubstituted derivative of betulin
(2g) had excellent solubility in aqueous media. Disubstituted
derivatives of betulin and monosubstituted esters of betulinic acid
with lysine or alanine also improved solubility to a certain extent
(Table 2). While derivatives with bulky and hydrophobic side
chains (2e, 2f and 4d) were even less soluble, however still signif-
icantly better when compared to betulin or betulinic acid. Alterna-
tive assay conditions (at pH 7.4 in 0.05 M Tris buffer and fixed 5%
DMSO concentration) confirmed our water solubility results ob-
tained in previous experiment as well.
In order to determine the chemical stability of the compounds
at physiological pH (7.4) we have performed RP-HPLC assays for
compound 2g. We have found that 2g is stable for at least 6 days
of incubation in phosphate buffer (pH 7.4, 0.02 M) at room temper-
ature. This suggests that the triterpene ester is stable throughout
the cell culture in vitro.
l
ment, the activity towards the sensitive cell line was only around
four times better than compared to the resistant one. Proliferation
profiles of the other tested betulin esters were much less impres-
sive. Considering derivatives of betulinic acid, substitutions with
amino acids we have found that lysine ester 4a was about twice
better (IC50 = 14
lM) from L-a-aminobutanoic acid ester 4b
(IC50 = 26 M), while molecules with bulky, hydrophobic side
l
chains had practically no effect on the cells. Control compounds,
betulinic acid and betulin were around 10 times weaker than the
best compounds 2d and 2g.
In case of the human pancreatic carcinoma cell lines after 24 h
of incubation we have observed significant effects only for the
parental (EPG85-257P) cell line. Also, for this cell line the most ac-
tive compound was the lysine diester of betulin 2g (IC50 = 4.1
Around twice weaker was the lysine ester of betulinic acid 4a
(IC50 = 9.7 M), while around four time weaker was the alanine
diester of betulin 2d (IC50 = 17 M). Betulin and betulinic acid were
lM).
l
l
Table 1
Numbering, structure and yield of the betulin and betulinic acid protected and deprotected esters with amino acids
Compds
R1
R2
Yield (%)
Compds
Yield (%)
HR-MS (calcd)
HR-MS (exp.)
1a
1b
1c
1d
1e
1f
1g
3a
3b
3c
3d
H
Gly
H
Gly
Gly
Ala
Ala
Phe
Met
Lys
—
85
91
83
72
2a
2b
2c
2d
2e
2f
2g
4a
4b
4c
4d
68
81
65
68
80
71
68
64
77
69
82
500.4104
557.4318
514.4260
585.4631
737.5258
705.4698
699.5789
585.4631
542.4210
528.4053
604.4366
500.4109
557.4297
514.4241
585.4628
737.5261
705.4673
699.5792
585.4628
542.4531
528.4059
604.4304
Ala
Phe
Met
Lys
Lys
Abu
Ala
Phe
71/4a
66/7a
64/5a
62
—
—
—
78
73
65
Additionally, in the case of deprotected derivatives (2, 4) mass of the compound obtained using high-resolution mass spectroscopy in ESI mode (HR-MS ESI) is presented.
a
Values showing yield of the disubstituted and monosubstituted derivative, respectively.