Synthesis and antiproliferative activity of (2R,3R)-disubstituted tetrahydropyrans. Part 2: Effect of side chain homologation

Article abstract of DOI:10.1016/j.bmcl.2006.10.066

In this study, we synthesized a series of enantiomerically pure (2R,3R)- and (2R,3S)-disubstituted tetrahydropyrans bearing a CH₂O spacer group on the side chain at position 2 of the heterocyclic ring. The in vitro antiproliferative activities

Full text of DOI:10.1016/j.bmcl.2006.10.066

Bioorganic & Medicinal Chemistry Letters 17 (2007) 780–783
Synthesis and antiproliferative activity of (2R,3R)-disubstituted
tetrahydropyrans. Part 2: Effect of side chain homologationq
a
a,b
a,c,
a
*
´
´
´
´ ´
Vıctor S. Martın
Romen Carrillo, Leticia G. Leon, Tomas Martın,
a,b,
*
´
´
and Jose M. Padron
aInstituto Universitario de Bio-Orga´nica ‘Antonio Gonza´lez’ (IUBO-AG), Universidad de La Laguna,
C/Astrofı´sico Francisco Sa´nchez 2, 38206 La Laguna, Spain
bBioLab, Instituto Canario de Investigacio´n del Ca´ncer (ICIC), C/Astrofı´sico Francisco Sa´nchez 2, 38206 La Laguna, Spain
c
´
Instituto de Productos Naturales y Agrobiologꢀıa, Consejo Superior de Investigaciones Cientıficas (CSIC),
C/Astrofı´sico Francisco Sa´nchez 3, 38206 La Laguna, Spain
Received 3 October 2006; revised 23 October 2006; accepted 24 October 2006
Available online 27 October 2006
Abstract—In this study, we synthesized a series of enantiomerically pure (2R,3R)- and (2R,3S)-disubstituted tetrahydropyrans
bearing a CH2O spacer group on the side chain at position 2 of the heterocyclic ring. The in vitro antiproliferative activities of
the compounds were examined in the human solid tumor cell lines A2780 (ovarian cancer), SW1573 (non-small cell lung cancer),
and WiDr (colon cancer). Overall, the results point out the relevance for antiproliferative activity of the distance between the
heterocycle and the unsaturated group.
Ó 2006 Elsevier Ltd. All rights reserved.
Marine drugs with cyclic ether scaffolds continue to at-
OTBS
OTBS
O
3
2
tract considerably the attention of researchers in the
search for novel antitumor compounds.1 In our group,
we have developed diverse methodologies for the synthe-
sis of these structurally challenging molecules.2
6
2
R
O
O
O
R
O
R
H
H
O
spacer
cis- or trans-THP
GI50 6−38 μM
GI50 18−24 μM
R: CO2Me, CH2OH
R: CH=CH2
We have reported on the antiproliferative activity of a
series of (2R,3S)-disubstituted tetrahydropyrans (trans-
THPs),3 (2R,3R)-disubstituted tetrahydropyrans (cis-
THPs),4 and (2R,6S)-disubstituted tetrahydropyrans5
against a representative panel of human solid tumor cell
lines (Fig. 1). The structure–activity relationship (SAR)
study revealed that in general cis-THPs are more active
than trans-THPs. In addition, we found that an unsatu-
ration (either a,b-unsaturated ester or vinyl group) on
the side chain at position 2 of the THP ring appears rel-
evant for the biological activity. The latter was observed
also for 2,6-disubstituted THPs.
Figure 1. General structure of antiproliferative cis- and trans-THPs.
In this article, we explore the biological activity of a ser-
ies of cis- and trans-THPs, which show a CH2O spacer
group between the heterocycle and the side chain at po-
sition 2 of said THP ring (Fig. 1). The choice of the
CH2O spacer group was done with two ideas in mind.
On the one hand, the synthesis of the novel cis- and
trans-THPs could be achieved in a limited number of
steps from the readily available THPs 1 and trans-1,
respectively. On the other hand, the oxygen atom pro-
vides an additional point of diversity on the functional
groups that can be prepared in order to perform a
SAR study. The antiproliferative activity was assessed
against the panel of three representative human solid
tumor cell lines A2780 (ovarian cancer), SW1573
(non-small cell lung cancer), and WiDr (colon cancer).
Keywords: Marine products; Anticancer drugs; Cyclic ethers; Solid
tumors; Structure–activity relationship.
q
Part 1, see Ref. 4.
*
Corresponding authors. Tel.: +34 922 318 580; fax: +34 922 318
571(J.M.P); e-mail addresses: tmartin@ipna.csic.es; jmpadron@
ull.es
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.10.066

R. Carrillo et al. / Bioorg. Med. Chem. Lett. 17 (2007) 780–783
781
The results are compared to those obtained previously
for cis- and trans-THPs bearing shorter side chains.
OTBS
O
O
7
a
The synthesis of the cis-THP derivatives (Scheme 1)
starts from diol 1, which can be obtained from the com-
mercially available tri-O-acetyl-D-galactal.6 Silyl protec-
tion of 1 followed by selective cleavage of the exocyclic
silyl ether leads to the cis-alcohol 2.7 Two-carbon
homologation of the primary alcohol of compound 2
with sodium iodoacetate provides the carboxylic acid 3
in excellent yields. Esterification of the carboxylic acid
3 with benzyl alcohol gives the benzyl ester 4. Further
removal of the tert-butyl dimethylsilyl (TBS) ether af-
fords the hydroxy ester 5. Finally, the catalytic hydroge-
nation of compound 5 followed by lactonization using
OTBS
O
b
2
O
8
CO2Bn
OTBS
c
OTBS
d
O
O
OH
OH
O
9
CO2Me
O
10
e
e
OTBS
O
OTBS
O
f
the 2-chloro-1,3-dimethylimidazolinium chloride,
a
O
CO2Me
O
12
commercially available dehydrating agent,8 leads to
the bicycle 6. When applying the above-described
synthetic methodology to the commercially available
tri-O-acetyl-D-glucal,9 the series of corresponding
trans-THP derivatives is obtained (Scheme 1).
11
Scheme 2. Reagents and conditions: (a) NaH, allyl bromide, NBu4I
(cat.), THF, 97%; (b) benzyl propiolate, Bu3P, CH2Cl2, rt, 89%; (c)
methyl propiolate, Bu3P, CH2Cl2, rt, 93%; (d) DIBAL-H, THF,
–20 °C, 42%; (e) H2, Pd(C) (cat.), AcOEt, 96% for 11, 40% for 12;
(f) DIBAL-H, THF, 0 °C, 91%.
Further derivatization of compound 2 to obtain a vari-
ety of alkyloxy substituents with an unsaturation on the
side chain is shown in Scheme 2. Thus, alkylation of the
alcohol 2 with allyl bromide provides the allyl ether 7.
The hetero-Michael addition of alcohol 2 to the appro-
priate propiolate ester allows preparing benzyl ester 8
(from benzyl propiolate) and methyl ester 9 (from meth-
yl propiolate). The reaction is catalyzed by tributylphos-
phine10 and proceeds in high yields. The reduction of the
methyl (E)-a,b-unsaturated ester 9 with DIBAL-H in
THF at ꢀ20 °C provides the (Z) allylic alcohol 10 in
low yields. The catalytic hydrogenation of the vinyl
ether 10 gives the saturated alcohol 12 in modest yields
(17% yield from 9). On the contrary, the catalytic hydro-
genation of compound 9 gives saturated methyl ester 11,
which after subsequent reduction with DIBAL-H in
THF at 0 °C allows obtaining saturated alcohol 12 in
87% yield from 9.
O
Ph
O
O
a
b
c
13
O
O H
O
1
O
O
H O
14
H
O
H
O
N
15
O
O
O
O
Scheme 3. Reagents and conditions: (a) PhCH(OCH3)2, CSA (cat.),
˚
CH2Cl2, 92%; (b) terephthalaldehyde, p-TsOH (cat.), CH2Cl2, 4 A MS,
D, 70%; (c) 2,6-pyridinedicarboxaldehyde, 2,2-dimethoxypropane,
p-TsOH (cat.), toluene (0.25 M), D, 71%.
A final set of aromatic acetals is prepared from cis-al-
cohol 3 and it is shown in Scheme 3. Condensation of
diol 1 with benzaldehydedimethyl acetal using catalytic
amounts of CSA affords the benzylidene acetal 13 as
sole stereoisomer.11 The bis-acetal 14 is obtained by
a double condensation of two molecules of alcohol 1
with terephthalaldehyde. This reaction is catalyzed
O
O
O
OH
OH
OTBS
OH
OTBS
O
OH
f, g
c
c
a, b
a, b
e
e
CO2R
O
CO2Bn
CO2Bn
O
O
O
1
O
2
O
O
5
O
6
3 (R= H)
4 (R= Bn)
d
d
O
OH
OH
OTBS
OH
OTBS
OH
f, g
O
CO2R
O
O
O
O
O
O
trans-1
trans-2
trans-5
trans-6
trans-3 (R= H)
trans-4 (R= Bn)
Scheme 1. Reagents and conditions: (a) TBSCl, imidazole, CH2Cl2; (b) TFA/THF/H2O (1:1:1), 76% yield from 1, 78% yield from trans-1; (c) NaH,
ICH2CO2Na, THF, 94% for 3, 95% for trans-3; (d) BnOH, DMAP, CSA, DCC, CH2Cl2, 95% for 4, 95% for trans-4; (e) HF, CH3CN, 85% for 5, 90%
for trans-5; (f) H2, Pd(OH)2 (cat.), AcOEt; (g) 2-chloro-1,3-dimethylimidazolinium chloride, NaH, DMAP, CH2Cl2, 0 °C ! rt, 83% yield from 6,
87% yield from trans-6.

782
R. Carrillo et al. / Bioorg. Med. Chem. Lett. 17 (2007) 780–783
by p-TsOH. Finally, the condensation of diol 1
with 2,6-pyridinedicarboxaldehyde provides the bis-
acetal 15.12
From the analysis of the growth inhibition parameters
we obtain the following SAR. Consistently with our pre-
vious findings,3,4 all active products bear a TBS ether
group at position 3 of the THP ring, whilst the corre-
sponding free hydroxyl derivatives were inactive (4 vs
5, and trans-4 vs trans-5). In addition, the presence of
the TBS ether group is not sufficient for antiproliferative
activity as shown with inactive derivatives 2, trans-2,3,
trans-3, and 10–12. Overall, saturated benzyl esters 4
and trans-4 together with allylic ether 7 and a,b-unsatu-
rated esters 8–9 induce growth inhibition. On the con-
trary, allylic alcohol 10, saturated methyl ester 11,
saturated alcohol 12, and acetals 13–15 are inactive in
this assay.
The in vitro antiproliferative activity was evaluated
using the National Cancer Institute (NCI) protocol.13
We screened growth inhibition and cytotoxicity against
the panel of human solid tumor cell lines A2780,
SW1573, and WiDr after 48 h of drug exposure using
the sulforhodamine B (SRB) assay.4 In addition to the
biological activity, for each compound the lipophilicity
expressed as ClogP14 was calculated to correlate lipo-
philicity with antitumor activity.15 The ClogP values
together with the growth inhibition data are listed in
Table 1. From the results it is possible to divide the com-
pounds in two groups. The group of active products
comprises five products, which show ClogP values in
the range 3.5–5.3 and GI50 values in the range 15–
50 lM. The remaining derivatives are not active against
the panel of cell lines. This observation can be explained
in part by the decreased lipophilicity that these THPs
show (ClogP < 3.3).
A direct comparison of the growth inhibition parame-
ters between THPs with (4, 8–9) and without (16–19)
the CH2O spacer group reveals the following consider-
ations. An improvement in antiproliferative activity is
observed for the new b-alkoxy-a,b-unsaturated esters 8
and 9 when compared to the reported corresponding
a,b-unsaturated esters 17 and 18, respectively. This is
particularly evident for the colon cancer cells WiDr
(Table 1). A slight increase in biological activity evident
from the TGI values happens to analog 4 when
compared to its counterpart 19. The preparation of
b-alkoxy-a,b-unsaturated esters leads exclusively to
compounds with (E) geometry of the double bond.
Therefore, we were unable to perform SAR studies to
determine the influence of the stereochemistry of the
unsaturated functional group in this new type of
derivatives.
With the exception of compounds 7 and 9, all active
derivatives exhibit a similar antiproliferative activity
against the three cell lines (GI50 values in the range
17–37 lM). Thus, WiDr colon cancer cells and
SW1573 NSCLC cells were less sensitive to cis-THPs 7
and 9, respectively. When considering TGI and LC50
values it is possible to observe that compounds trans-4
and 7 are the most active of the series against the ovar-
ian cancer and the NSCLC cells.
Table 1. Lipophilicity and in vitro antiproliferative activity of THPs against human solid tumor cellsa
Compound
ClogPb
A2780
TGI
SW1573
TGI
WiDr
TGI
GI50
LC50
GI50
LC50
GI50
LC50
2
2.34
2.76
4.68
1.29
–0.58
2.34
4.68
1.29
–0.58
3.73
5.27
3.55
2.59
3.30
2.72
1.32
0.56
–0.94
3.79
5.26
3.52
4.97
2.38
>100
>100
>100
>100
>100
>100
3
4
28 ( 7.3)
>100
>100
81 ( 30)
36 ( 2.6)
94 ( 11)
29 ( 9.4)
>100
>100
85 ( 26)
37 ( 8.7)
>100
>100
88 ( 23)
91 ( 13)
5
6
trans-2
trans-4
trans-5
trans-6
7
>100
>100
>100
17 ( 1.8)
>100
>100
74 ( 3.2)
68 ( 2.9)
17 ( 4.2)
>100
>100
37 ( 7.6)
39 ( 3.2)
79 ( 13)
85 ( 6.4)
28 ( 1.3)
>100
>100
15 ( 1.6)
26 ( 5.3)
30 ( 0.6)
>100
32 ( 1.8)
82 ( 4.7)
18 ( 1.8)
26 ( 7.0)
50 ( 31)
>100
88 ( 18)
37 ( 9.3)
36 ( 3.7)
>100
8
9
10
11
12
>100
>100
>100
>100
>100
>100
13
14
>100
>100
>100
>100
>100
>100
15
16
>100
>100
>100c
>100
86 ( 13)c
32 ( 2.7)c
98 ( 2.6)c
23 ( 4.5)c
18 ( 11)d
>100c
90 ( 14)c
>100c
19 ( 1.6)c
>100
17
18
37 ( 5.7)c
>100c
19
20
30 ( 2.9)c
24 ( 5.0)c
a Values are given in lM and are means of two to four experiments, standard deviation is given in parentheses. TGI and LC50 values are given only if
they are less than 100 lM, which is the maximum concentration test.
b Ref. 15.
c Ref. 4.
d Ref. 2c.

R. Carrillo et al. / Bioorg. Med. Chem. Lett. 17 (2007) 780–783
783
´
Martın, T.; Martın, V. S. Tetrahedron Lett. 2000, 41, 2503;
´
OTBS
OTBS
CO2Bn
´
(h) Martın, T.; Soler, M. A.; Betancort, J. M.; Martın, V.
´
´
S. J. Org. Chem. 1997, 62, 1570; (i) Ramırez, M. A.;
Padron, J. M.; Palazon, J. M.; Martın, V. S. J. Org. Chem.
O
R
O
19
O
20
O
´
´
´
H
H
16 R = H
17 R = CO2Bn
18 R = CO2Me
O
´
´
1997, 62, 4584; (j) Betancort, J. M.; Martın, V. S.; Padron,
J. M.; Palazon, J. M.; Ramırez, M. A.; Soler, M. A.
´
´
J. Org. Chem. 1997, 62, 4570.
´
3. (a) Donadel, O. J.; Martın, T.; Martın, V. S.; Villar, J.;
´
Figure 2. Structure of previously reported THPs without spacer group.
´
Padron, J. M. Bioorg. Med. Chem. Lett. 2005, 15, 3536; (b)
´
Padron, J. M.; Donadel, O. J.; Leon, L. G.; Martın, T.;
´
´
Martın, V. S. Lett. Drug Des. Discov. 2006, 3, 29.
An interesting result is obtained for allylic ether 7. The
three-atom distance of the vinyl group from the THP
ring induces a positive effect on the biological activity,
when compared to the previously reported analog 16
(Fig. 2). We speculate on a steric hindrance effect of
the TBS group to account for the loss of activity of
the vinyl group in 16. This assumption is made consid-
ering a previous result obtained for THP 20,5 which
shows an antiproliferative effect due to the vinyl group,
and similar to that observed for derivative 7.
´
´
4. Carrillo, R.; Leon, L. G.; Martın, T.; Martın, V. S.;
´
´
Padron, J. M. Bioorg. Med. Chem. Lett. 2006, 16,
´
doi:10.1016/j.bmcl.2006.08.094.
´
5. Crisostomo, F. R. P.; Carrillo, R.; Leon, L. G.; Martın,
´
´
T.; Padron, J. M.; Martın, V. S. J. Org. Chem. 2006, 71,
´
´
2339.
6. Snatzke, G.; Raza, Z.; Habus, I.; Sunjic, V. Carbohydr.
Res. 1988, 182, 179.
7. (a) Carrillo, R.; Martın, V. S.; Martın, T. Tetrahedron
´
´
´
Lett. 2004, 45, 5215; (b) Carrillo, R.; Martın, V. S.; Lopez,
´
´
M.; Martın, T. Tetrahedron 2005, 61, 8177.
In summary, we have synthesized a series of enantiomer-
ically pure cis- and trans-THPs bearing a CH2O spacer
group on the side chain at position 2 of the ring. The
growth inhibition parameters were determined against
a panel of three representative human solid tumor cell
lines. Overall, the results show in these derivatives the
relevance of the distance to the THP ring of the unsatu-
ration located at the side chain.
8. (a) Isobe, T.; Ishikawa, T. J. Org. Chem. 1999, 64, 5832;
(b) Isobe, T.; Ishikawa, T. J. Org. Chem. 1999, 64, 6984;
(c) Isobe, T.; Ishikawa, T. J. Org. Chem. 1999, 64, 6989;
(d) Fujisawa, T.; Mori, T.; Fukumoto, K.; Sato, T. Chem.
Lett. 1982, 1891.
9. Nicolaou, K. C.; Hwang, C. K.; Marron, B. E.; DeFrees,
S. A.; Couladouros, E. A.; Abe, Y.; Carroll, P. J.; Snyder,
J. P. J. Am. Chem. Soc. 1990, 112, 3040.
10. Inanaga, J.; Baba, Y.; Hanamoto, T. Chem. Lett. 1993, 22,
241.
11. The orientation of the phenyl group was confirmed by
NOE experiments. The stereochemical control in this
reaction is a consequence of the formation of the most
stable cis-decaline, which has the phenyl group on an
equatorial orientation.
Acknowledgments
This research was supported by the EU INTERREG
IIIB-MAC initiative (05/MAC/2.5/C14 BIOPOLIS),
12. Heard, P. J.; King, P. M.; Bain, A. D.; Hazendonk, P.;
Tocher, D. A. J. Chem. Soc., Dalton Trans. 1999,
4495.
´
the Ministerio de Educacion y Ciencia of Spain, co-
financed by the European Regional Development Fund
(CTQ2005-09074-C02-01/BQU), and the Canary Islands
Government. R.C. thanks the Spanish MEC for an FPU
fellowship.
13. In this method, for each drug a dose–response curve is
generated and three levels of effect can be calculated, when
possible. The effect is defined as percentage of growth
(PG), where 50% growth inhibition (GI50), total growth
inhibition (TGI), and 50% cell killing (LC50) represent the
drug concentration at which PG is +50, 0, and ꢀ50,
respectively Skehan, P.; Storeng, P.; Scudeiro, D.; Monks,
A.; McMahon, J.; Vistica, D.; Warren, J. T.; Bokesch, H.;
Kenney, S.; Boyd, M. R. J. Natl. Cancer Inst. 1990, 82,
1107.
References and notes
1. Mayer, A. M. S.; Gustafson, K. R. Eur. J. Cancer 2006,
42, 2241, and references therein.
´
2. (a) Miranda, P. O.; Padron, J. M.; Padron, J. I.; Villar, J.;
Martın, V. S. ChemMedChem. 2006, 1, 323; (b) Crisosto-
´
14. In a comparative study, the computer program ClogPÒ
appeared the most accurate predictor of ClogP values
Machatha, S. G.; Yalkowsky, S. H. Int. J. Pharm. 2005,
294, 185.
´
´
´
´
´
mo, F. R. P.; Padron, J. M.; Martın, T.; Villar, J.; Martın,
V. S. Eur. J. Org. Chem. 2006, 1910; (c) Crisostomo, F. R.
P.; Martın, T.; Martın, V. S. Org. Lett. 2004, 6, 565; (d)
´
´
´
Betancort, J. M.; Martın, T.; Palazon, J. M.; Martın, V. S.
´
´
J. Org. Chem. 2003, 68, 3216; (e) Dıaz, D.; Martın, T.;
15. Software-predicted lipophilicity of the products was cal-
culated with the program ClogPÒ accessible via Internet
(www.daylight.com/daycgi/clogp) working with the Han-
sch-Leo’s ‘fragment constant’ method.
´
´
´
´
Martın, V. S. Org. Lett. 2001, 3, 3289; (f) Garcıa, C.;
Martın, T.; Martın, V. S. J. Org. Chem. 2001, 66, 1420; (g)
´
´