Novel lithocholaphanes: Syntheses, NMR, MS, and molecular modeling studies

Article abstract of DOI:10.1016/j.molstruc.2007.01.030

Novel head-to-head lithocholaphanes 6 and 11 have been synthesized via precursors 1-5 and 7-10 with overall good yields, and characterized by ¹H, ¹³C, and ¹⁵N NMR spectroscopy, ESI-TOF mass spectrometry, thermal analysis,

Full text of DOI:10.1016/j.molstruc.2007.01.030

Available online at www.sciencedirect.com
Journal of Molecular Structure 846 (2007) 65–73
www.elsevier.com/locate/molstruc
Novel lithocholaphanes: Syntheses, NMR, MS,
and molecular modeling studies
Arto Valkonen a,¤, Elina Sievänen a, Satu Ikonen a, Nikolai V. Lukashev b,
Pavel A. Donez b, Alexej D. Averin b, Manu Lahtinen a, Erkki Kolehmainen a
a Department of Chemistry, University of Jyväskylä, P.O. Box 35, FIN-40014, University of Jyväskylä, Finland
b Department of Chemistry, Lomonosov Moscow State University, Leninskie Gory, Moscow 119992, Russian Federation
Received 14 November 2006; accepted 15 January 2007
Available online 24 January 2007
Abstract
Novel head-to-head lithocholaphanes 6 and 11 have been synthesized via precursors 1–5 and 7–10 with overall good yields, and char-
acterized by 1H, 13C, and 15N NMR spectroscopy, ESI-TOF mass spectrometry, thermal analysis, and molecular modeling. In addition,
the binding abilities of 6 and 11 towards alkali metal cations have been investigated via competitive complexation studies using equimolar
mixtures of Li+, Na+, K+, and Rb+-cations, and cholaphanes 6 and 11. The formation of cation–cholaphane adducts was detected by ESI-
TOF mass spectrometry. The trends in these comparative binding studies are nicely reproduced theoretically with PM3 energetically opti-
mized structures of 6 and 11 and their interaction energies with alkali metal cations calculated by molecular mechanics. Cholaphane 11
possessing a peptoid type structural fragment, –(CH2CONHCH2CH2)2O–, as a coordination sphere, shows binding tendency towards
lithium and sodium cations, whereas 6 possessing an ester type, –(CH2OCOCH2)2O–, moiety and a bigger cavity size than 11, shows
merely a tendency towards bigger alkali metal cations, potassium and rubidium.
© 2007 Elsevier B.V. All rights reserved.
Keywords: Lithocholaphane; Synthesis; NMR; ESI-TOF MS; Molecular modeling studies
1. Introduction
The most abundant naturally occurring bile acids in
higher vertebrates are derivatives of cholanic acid, cyclo-
pentanoperhydrophenanthrene ring-containing steroid
consisting of 24 carbon atoms. Bile acids are end-products
of cholesterol metabolism in the liver. The steroidal
nucleus of cholesterol undergoes several hydroxylations
followed by a loss of an isopropyl group from the side
chain resulting in primary bile acids. Primary bile acids,
among which cholic acid (3ꢀ,7ꢀ,12ꢀ-trihydroxy-5ꢁ-cho-
lan-24-oic acid) and chenodeoxycholic acid (3ꢀ,7ꢀ-dihy-
droxy-5ꢁ-cholan-24-oic acid) are the most important
ones, are further transformed to secondary bile acids, such
as deoxycholic (3ꢀ,12ꢀ-dihydroxy-5ꢁ-cholan-24-oic acid),
hyodeoxycholic (3ꢀ,6ꢀ-dihydroxy-5ꢁ-cholan-24-oic acid),
and lithocholic (3ꢀ-hydroxy-5ꢁ-cholan-24-oic acid) acids
[2]. Synthesis of bile acids takes place in the liver, from
where they are secreted into bile, stored in gallbladder,
There exists a constantly growing need for new molecu-
lar frameworks with applications in the Weld of supramo-
lecular chemistry, driving researchers to development of
new host–guest structures. One approach is to design
novel macrocyclic synthetic receptors with molecular cav-
ities able to serve as model compounds for more complex
biological systems. Typically macrocyclic host molecules
bind substrates either in their deWned cavity or above their
plane [1]. Bile acids are widely used as building blocks for
such supramolecular entities.
*
Corresponding author. Tel.: +358 14 260 2670; fax: +358 14 260 2501.
E-mail address: arto.valkonen@jyu.W (A. Valkonen).
0022-2860/$ - see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.molstruc.2007.01.030

66
A. Valkonen et al. / Journal of Molecular Structure 846 (2007) 65–73
and emptied into the small intestine, where they are reab-
sorbed and carried to liver via portal vein. This cycle,
called enterohepatic circulation, happens 6–15 times per
day. Bile acids have an essential role in digestion and
absorption of lipids and lipid-soluble vitamins. In addi-
tion, they prevent cholesterol from precipitating in the
gall bladder [3]. Most prominent characteristics of bile
acids include their large, rigid, and curved steroidal skele-
ton, chemically diVerent hydroxyl groups, enantiomeric
purity, and unique amphiphilicity. Furthermore, they are
low-priced and easily available. Owing to these properties,
bile acids are ideal building blocks in construction of
novel molecular and supramolecular assemblies for
molecular recognition.
Cholaphanes, bile acid-derived macrocycles, which con-
sist of steroidal units joined together by various spacer
groups, have been continuously in a focus of supramolecu-
lar chemistry owing to their versatile properties [4–16]. One
way to tailor a bile acid derivative suitable for cation recog-
nition is, to attach a crown ether type moiety either directly
to a hydroxy group of a bile acid [17] or by bridging
together two steroidal hydroxyls by a polyalkoxy chain
[18]. Moreover, various crown ethers and bile acid deriva-
tives are used as building blocks in synthetic ion channels
and pores [19]. One future objective to synthesize ionophor-
ous bile acid macrocycles could also be their suitability for
preparation of ion-selective electrodes [20]. Recently, chola-
phanes containing pyridine moieties are reported to have
an ability to bind Xavin analogues [21]. More generally,
many bile acid derivatives are valuable compounds in
supramolecular and pharmaceutical chemistry, as reviewed
by us [22,23]. As a continuation of our previous studies on
various cholaphanes [6,7,9,12,24] we now wish to report
syntheses, spectral characterization (NMR and ESI-TOF
MS), molecular modeling, and alkali metal cation binding
properties of two novel cholaphanes, 6 and 11, with aro-
matic and alkoxy/aminoalkoxy building blocks oVering
proper binding sites, for example, for alkali and earth alka-
line metal cations (see Fig. 1). These structures resemble
also, for example, cyclopeptides [25–27] and ionophorous
antibiotics [28], which show interesting receptor properties.
Scheme 1. Preparation of cholaphane 6.
2. Experimental
2.1. Syntheses and analytical data
The synthetic steps leading to cholaphanes 6 and 11 are
illustrated in Schemes 1 and 2.
3ꢀ-Hydroxy-5ꢁ-cholan-24-oic acid (lithocholic acid) (1)
(97% by titration). Compound 1 was purchased from Sigma
Fig. 1. Structures and numbering of cholaphanes 6 and 11.

A. Valkonen et al. / Journal of Molecular Structure 846 (2007) 65–73
67
CDCl3, ppm): ꢂ 71.87 (C-3), 63.57 (C-24), 56.54 (C-14),
56.22 (C-17), 42.72 (C-13), 42.13 (C-5), 40.48 (C-9), 40.22
(C-12), 36.48 (C-4), 35.88 (C-8), 35.58 (C-20), 35.37 (C-1),
34.58 (C-10), 31.85 (C-22), 30.56 (C-2), 29.45 (C-23), 28.29
(C-16), 27.21 (C-6), 26.44 (C-7), 24.22 (C-15), 23.36 (C-19),
20.84 (C-11), 18.64 (C-21), 12.04 (C-18).
24-Triphenylmethoxy-5ꢁ-cholan-3ꢀ-ol (3). In a vial with a
screw cap the mixture of 2 (90mg, 0.25mmol), triphenylm-
ethylchloride (tritylchlorideDTrCl) (104mg, 0.37mmol), 4-
(dimethylamino)pyridine (DMAP) (5mg, 0.04mmol), and
triethylamine (Et3N) (84ꢂL, 0.6mmol) in dichloromethane
(DCM) (1.5mL) was heated on an oil bath (60°C) for 15h.
The solvent was removed under reduced pressure and the
residue was subjected to column chromatography (DCM,
Rf D0.30) to give 3 as a white crystalline solid (119mg, 83%);
mp 93–94°C; Anal. found: C, 85.61; H, 9.44. Calcd for
1
C43H56O2: C, 85.38; H, 9.33%; H NMR (CDCl3, 400MHz,
ppm): ꢂ 7.44–7.19 (m, 15H, OTr), 3.61 (m, 1H, 3ꢁ-H), 3.00 (m,
2H, 24-CH2), 0.90 (s, 3H, 19-CH3), 0.88 (d, 3H, 21-CH3), 0.60
(s, 3H, 18-CH3).
Bis(24-triphenylmethoxy)-3ꢀ,3Јꢀ-(isophthaloyloxy)-bis
(5ꢁ-cholane) (4). An argon-Xushed two-necked Xask
equipped with a condenser, dropping funnel, and a mag-
netic stirrer was charged with 3 (120 mg, 0.2 mmol), DMAP
(7mg, 0.06 mmol), Et3N (44 ꢂL, 0.3mmol), and isophthaloyl
dichloride (20 mg, 0.1mmol) in anhyd. toluene (2 mL) in the
order given at rt. The mixture was stirred at 100 °C for 15 h.
After cooling to rt the mixture was diluted with DCM and
washed twice with sat. aq. NaHCO3 solution. The organic
layer was dried (MgSO4) and the solvent evaporated under
reduced pressure. The product was puriWed by column
chromatography (petroleum ether/EtOAc, 10:1, Rf D0.50)
to give 4 as a white crystalline solid (112mg, 85%); mp 150–
152 °C; Anal. found: C, 84.18; H, 8.85. Calcd for C94H114O6:
Scheme 2. Preparation of cholaphane 11.
and used without further puriWcation. Isophthalic and tere-
phthalic acids and other reagents used in synthetic steps
were analytical grade reagents and used as such. All sol-
vents used in chromatography and recrystallization were
also analytical grade reagents and used without further
puriWcation.
1
C, 84.26; H, 8.58%; H NMR (CDCl3, 400MHz, ppm): ꢂ
8.69 (m, 1H, 29-H), 8.19 (m, 2H, 30-H, 30Ј-H), 7.51–7.18 (m,
16H, 31-H, OTr), 4.98 (m, 2H, 3ꢁ-H, 3Јꢁ-H), 3.00 (m, 4H,
24-CH2, 24Ј-CH2), 0.95 (s, 6H, 19-CH3, 19Ј-CH3), 0.89 (d,
6H, 21-CH3, 21Ј-CH3), 0.62 (s, 6H, 18-CH3, 18Ј-CH3).
5ꢁ-Cholane-3ꢀ,24-diol (2). Compound 2 has been pre-
pared as described before [29]. An argon-Xushed two-
necked Xask equipped with a condenser, dropping funnel,
and a magnetic stirrer was charged with the solution of
lithocholic acid, 1, (300mg, 0.796mmol) in anhyd. THF
(8mL) and cooled to 0 °C. Suspension of LiAlH4 (LAH)
(90 mg, 2.39 mmol) in 2 ml of THF was added. The mixture
was stirred for 30 min at rt and then reXuxed for 2 h. The
cold (0 °C) reaction mixture was carefully quenched by
dropwise addition of a 10% HCl solution (10 mL) and
swirled at rt until the reaction mixture became transparent.
The product was extracted with ether (3 £ 30 mL), the ethe-
real solution was dried (MgSO4), and evaporated to give 2
(289 mg, 100%) as a white crystalline solid, which was used
without further puriWcation since its mp corresponded to
that reported in the literature: 172–173 °C from acetone. 1H
NMR (CDCl3, 500 MHz, ppm): ꢂ 3.65–3.55 (m, 3H, 3ꢁ-H
and 24-CH2), 0.92 (d, J D 5.6Hz, 3H, 21-CH3), 0.92 (s, 3H,
19-CH3), 0.64 (s, 3H, 18-CH3); 13C NMR (126 MHz,
3ꢀ,3Јꢀ-(Isophthaloyloxy)-bis(5ꢁ-cholan-24-ol) (5). Com-
pound 4 (106mg, 0.08 mmol) was dissolved in the mixture
of MeOH (6 mL) and DCM (2 mL). To this solution aque-
ous 48% HBr (200 ꢂL) was added and the mixture was reX-
uxed for 3h. The product was extracted with DCM
(2£ 20 mL), the organic layer was washed with water, dried
(MgSO4), and concentrated. The resulting oil was subjected
to column chromatography (DCM/MeOH, 20:1, Rf D0.20)
to give 5 as a white crystalline solid (50 mg, 74%); mp 188–
190 °C; Anal. found: C, 78.64; H, 10.35. Calcd for C56H86O6:
1
C, 78.64; H, 10.14%; H NMR (CDCl3, 400 MHz, ppm): ꢂ
8.68 (m, 1H, 29-H), 8.19 (m, 2H, 30-H, 30Ј-H), 7.49 (m, 1H,
31-H), 4.98 (m, 2H, 3ꢁ-H, 3Јꢁ-H), 3.60 (m, 4H, 24-CH2, 24Ј-
CH2), 0.95 (s, 6H, 19-CH3, 19Ј-CH3), 0.92 (d, 6H, 21-CH3,
21Ј-CH3), 0.65 (s, 6H, 18-CH3, 18Ј-CH3).
Cyclic 3ꢀ,3Јꢀ-(isophthaloyloxy)-bis(5ꢁ-cholan-24-yloxy)-
3-oxaglutarate (6). In a vial with a screw cap the mixture of 5
(68mg, 0.08mmol), DMAP (3mg, 0.02mmol), 2,2Ј-oxydiacetyl

68
A. Valkonen et al. / Journal of Molecular Structure 846 (2007) 65–73
chloride (14mg, 0.08 mmol), and Et3N (0.5 mL) in toluene
(8mL) was heated on an oil bath (100 °C) until TLC analy-
sis revealed the reaction was no longer proceeding (15–
20 h). The solvent was removed under reduced pressure
and the residue was subjected to column chromatography
(DCM, Rf »0.2–0.25) to give 6 as a white crystalline solid
3Јꢁ-H), 2.43–2.22 (m, 4H, 23-CH2, 23Ј-CH2), 0.95 (s, 6H, 19-
CH3, 19Ј-CH3), 0.90 (d, 6H, 21-CH3, 21Ј-CH3), 0.59 (s, 6H,
18-CH3,18Ј-CH3).
3ꢀ,3Јꢀ-(Terephthaloyloxy)-bis(5ꢁ-cholan-24-oic
acid)
(9). To the mixture of 8 (2.045 g, 1.92 mmol) and 10% Pd/C
(61 mg, 3%) in argon atmosphere, 20 mL of anhydrous
THF was added. Then the Xask was purged with H2 for 1–
2 s, connected to a balloon of H2, and stirred for 3 h at rt.
The mixture was Wltered through a short plug of silica gel to
remove the catalyst. The Wltrate was evaporated to dryness
1
(14mg, 15%); Tm D314.6 °C; H NMR (CDCl3, 500 MHz,
ppm): ꢂ 8.45 (t, J D 1.6 Hz, 1H, 29-H), 8.26 (dd, 2H, J D7.8,
1.7 Hz, 30-H, 30Ј-H), 7.54 (t, 1H, J D7.8Hz, 31-H), 4.95 (m,
2H, J D7.3, 4.3 Hz, 3ꢁ-H, 3Јꢁ-H), 4.25-4.12 (m, 4H, 26-CH2,
26Ј-CH2), 4.20 and 4.24 (ABq, 4H, JAB D 16.5Hz, 24-CHab,
24Ј-CHab), 0.99 (s, 6H, 19-CH3, 19Ј-CH3), 0.93 (d, 6H,
J D6.5 Hz, 21-CH3, 21Ј-CH3), 0.68 (s, 6H, 18-CH3, 18Ј-
CH3); 13C NMR (CDCl3, 126 MHz, ppm): ꢂ 169.9 (C-25, C-
25Ј), 165.8 (C-27, C-27Ј), 134.3 (C-30, C-30Ј), 131.3 (C-28,
C-28Ј), 128.8 (C-29), 128.6 (C-31), 76.4 (C-3, C-3Ј), 68.5 (C-
26, C-26Ј), 65.3 (C-24, C-24Ј), 56.5 (C-14, C-14Ј), 56.4 (C-17,
C-17Ј), 42.8 (C-13, C-13Ј), 42.3 (C-5, C-5Ј), 40.6 (C-9, C-9Ј),
40.2 (C-12, C-12Ј), 35.9 (C-8, C-8Ј), 35.2 (C-1, C-1Ј), 35.1 (C-
20, C-20Ј), 34.8 (C-10, C-10Ј), 32.4 (C-4, C-4Ј), 32.1 (C-22,
C-22Ј), 28.3 (C-16, C-16Ј), 27.2 (C-6, C-6Ј), 26.8 (C-2, C-2Ј),
26.5 (C-7, C-7Ј), 25.3 (C-23, C-23Ј), 24.3 (C-15, C-15Ј), 23.4
(C-19, C-19Ј), 20.9 (C-11, C-11Ј), 18.6 (C-21, C-21Ј), 12.0
(C-18, C-18Ј); MS m/z ESI-TOF+ found 975.8177 [M+Na]+
(100%) and 991.7653 [M+K]+(42.5%), C60H88O9Na requires
975.6326 and C60H88O9K 991.6066, respectively.
Benzyl 3ꢀ-hydroxy-5ꢁ-cholan-24-oate (7). To a stirred
solution of lithocholic acid, 1, (500 mg, 1.33mmol) in anhy-
drous DCM (3 mL), DMAP (32mg, 0.27 mmol) and benzyl
alcohol (216 mg, 2 mmol) were added. Then dicyclohexyl
carbodiimide, DCC (357mg, 1.73mmol), was added to the
mixture. After stirring overnight in a closed Xask the mix-
ture was diluted with DCM and Wltered. The Wltrate was
washed with 5% aq. HCl, sat. NaHCO3, and H2O, then
dried (MgSO4), and evaporated to dryness. The crude prod-
uct was puriWed by column chromatography (petroleum
ether/EtOAc, 4:1, Rf D0.28) and recrystallized from hexane/
EtOAc. The yield of 7 was 432 mg (74%); 1H NMR (CDCl3,
400MHz, ppm): ꢂ 7.33 (m, 5H, Ph), 5.09 (m, 2H, PhCH2),
3.58 (m, 1H, 3ꢁ-H), 2.42-2.21 (m, 2H, 23-CH2), 0.88 (s and d,
6H, 19-CH3, 21-CH3), 0.59 (s, 3H, 18-CH3). These corre-
spond to those reported previously [30].
1
to give 9 as a white crystalline solid (1.637 g, 96%). H
NMR (CDCl3, 400 MHz, ppm): ꢂ 11.56 (br s, 2H, 24-
COOH, 24Ј-COOH), 7.98 (s, 4H, 29-H, 29Ј-H), 4.86 (m, 2H,
3ꢁ-H, 3Јꢁ-H), 2.24-2.02 (m, 4H, 23-CH2, 23Ј-CH2), 0.88 (s,
6H, 19-CH3,19Ј-CH3), 0.84 (d, 6H, 21-CH3, 21Ј-CH3), 0.57
(s, 6H, 18-CH3,18Ј-CH3).
Bis(pentaXuorophenyl)
3ꢀ,3Јꢀ-(terephthaloyloxy)-bis
(5ꢁ-cholan-24-oate) (10). To a stirred solution of 9 (1.637g,
1.84mmol) in anhydrous DCM (50mL) pentaXuorophenol
(PFPOH) (846 mg, 4.60 mmol) and N,NЈ-dicyclohexyl car-
bodiimide (DCC) (1.140 g, 5.53 mmol) was added in the
order given. After being stirred overnight the mixture was
diluted with DCM and Wltered. The Wltrate was washed
with sat. NaHCO3, and H2O, then dried (Na2SO4), and
evaporated to dryness. The crude product was puriWed by
column chromatography (petroleum ether/DCM, 1:1,
1
Rf D0.30). The yield of 10 was 1.888 g (84%). H NMR
(CDCl3, 400 MHz, ppm): ꢂ 8.07 (s, 4H, 29-H, 29Ј-H), 4.98
(m, 2H, 3ꢁ-H, 3Јꢁ-H), 2.73-2.53 (m, 4H, 23-CH2, 23Ј-CH2),
0.97 (s, d, s and d, 12H, 19-CH3, 21-CH3, 19Ј-CH3, 21Ј-
CH3), 0.67 (s, 6H, 18-CH3, 18Ј-CH3); MS m/z ESI-TOF+
found 1237.7355 [M+MeOH+Na]+ (100%) and 1253.7007
[M+MeOH+K]+ (46.3%), C68H80F10O6CH3OHNa requires
1237.5955 and C68H80F10O6CH3OHK 1253.5695, respectively.
Cyclic N,NЈ-(3-oxa-1,5-pentanediyl) 3ꢀ,3Јꢀ-(terephtha-
loyloxy)-bis(5ꢁ-cholan-24-amide) (11). To a stirred solu-
tion of 10 (73 mg, 0.06 mmol), DMAP (22 mg, 0.18 mmol)
and Et3N (25 ꢂL, 0.18 mmol) in anhydrous DCM (6 mL),
and 2,2Ј-oxybis(ethylamine) (7 mg, 0.07 mmol) was added
at once. After being stirred in a stoppered Xask for 12–
15 h the mixture was diluted with DCM and washed with
10% aq. KOH and with water. The organic layer was dried
(MgSO4) and concentrated. The crude product was puri-
Wed by column chromatography (DCM/MeOH, 30:1,
Rf » 0.2–0.3). The yield of 11 was 21 mg (37%);
Dibenzyl 3ꢀ,3Јꢀ-(terephthaloyloxy)-bis(5ꢁ-cholan-24-oate)
(8). To the terephthaloyl dichloride (90mg, 0.443 mmol) in
anhyd. toluene (3 mL), DMAP (36mg, 0.29 mmol), Et3N
(0.5mL), and 7 (430 mg, 0.98 mmol) were added at rt. The
mixture was stirred in argon atmosphere at rt for 1h and
then at 100°C for 12 h. After cooling to rt the mixture was
diluted with DCM and washed twice with sat. NaHCO3
solution. The organic layer was dried over MgSO4 and the
solvent evaporated under reduced pressure. The product
was puriWed by column chromatography (DCM, Rf D0.55)
to give 8 as a white crystalline solid (401 mg, 86%, based on
terephthaloyl dichloride); mp 132–134°C; Anal. found: C,
1
Tm D 352.8 °C (decomposition during melting); H NMR
(CDCl3, 500 MHz, ppm): ꢂ 8.10 (s, 4H, 29-H, 29Ј-H), 5.74
(t, 2H, NH), 5.06 (m, 2H, J D 5.8, 5.3, 4.9 Hz, 3ꢁ-H, 3Јꢁ-H),
3.62-3.34 (m, 8H, 26-H, 26Ј-H, 25-H, 25Ј-H), 0.96 (s, 6H,
19-CH3, 19Ј-CH3), 0.95 (d, 6H, J D 6.7 Hz, 21-CH3, 21Ј-
CH3), 0.68 (s, 6H, 18-CH3, 18Ј-CH3); 13C NMR (CDCl3,
126 MHz, ppm): ꢂ 173.8 (C-24, C-24Ј), 165.1 (C-27, C-
27Ј), 134.6 (C-28, C-28Ј), 129.4 (C-29, C-29Ј), 74.7 (C-3,
C-3Ј), 69.7 (C-26, C-26Ј), 57.1 (C-14, C-14Ј), 55.2 (C-17,
C-17Ј), 42.7 (C-13, C-13Ј), 41.7 (C-5, C-5Ј), 40.7 (C-9, C-
9Ј), 40.5 (C-12, C-12Ј), 39.0 (C-25, C-25Ј), 35.6 (C-8, C-8Ј),
35.1 (C-20, C-20Ј), 34.9 (C-1, C-1Ј), 34.4 (C-10, C-10Ј),
1
78.99; H, 9.05. Calcd for C70H94O8: C, 79.06; H, 8.91%; H
NMR (CDCl3, 400MHz, ppm): ꢂ 8.07 (s, 4H, 29-H, 29Ј-H),
7.34 (m, 10H, Ph), 5.09 (m, 4H, PhCH2), 4.97 (m, 2H, 3ꢁ-H,

A. Valkonen et al. / Journal of Molecular Structure 846 (2007) 65–73
69
32.2 (C-23, C-23Ј), 32.1 (C-4, C-4Ј), 31.8 (C-22, C-22Ј),
28.3 (C-16, C-16Ј), 26.8 (C-6, C-6Ј), 26.6 (C-2, C-2Ј), 26.2
(C-7, C-7Ј), 24.0 (C-15, C-15Ј), 23.2 (C-19, C-19Ј), 21.0 (C-
11, C-11Ј), 18.4 (C-21, C-21Ј), 12.0 (C-18, C-18Ј); MS m/z
ESI-TOF+ found 973.8069 [M+Na]+ (100%) and 989.7914
[M+K]+ (5%), C60H90N2O7Na requires 973.6646 and
C60H90N2O7K 989.6385, respectively.
mental procedure was analogous to that described for
adduct formation experiment with alkali metal chlorides.
Adduct formation between the ligands and selected tet-
raalkyl- or dialkyldiarylammonium salts (tetramethyl-,
dimethyldipropyl-, diethyldihexyl-, dimethyldibenzyl-, and
dimethyldioctylammonium bromides, respectively) was
tested without success. Finally, adduct formation experi-
ment between the ligands and pyridinium cation was per-
formed. 0.5mM solutions of pyridine in MeCN/HCOOH
(for experiment with 6) as well as in MeOH (for experiment
with 11) were prepared and equal volumes of pyridine solu-
tion and ligand solution mixed and stirred thoroughly. For
11 no adduct with pyridinium cation was observed to form.
For 6, however, a signal with a rather weak intensity (»3%)
of [M+Py+H]+ ion was detected. [M+Na]+ ion gave the
most intense signal in the spectrum, even though no alkali
metal salts were added to the sample solution.
Electrospray mass spectrometric measurements were
performed by using LCT time of Xight (TOF) mass spec-
trometer with electrospray ionization (ESI; Micromass
LCT). Controlling the LCT as well as acquiring and pro-
cessing the data were performed with a MassLynx NT soft-
ware system. In the experiments aimed at molecular weight
determinations a Xow rate of 80ꢂL/min was used for the
sample solution and the sample droplets were dried with
nitrogen gas. Positive ions were chosen for detection. For 6
the potentials of 150 and 6V for the sample and extraction
cones were applied. RF lens was set at a potential of 1100 V
and the potential in the capillary at 4500V. For 11 the
potentials for sample and extraction cones were set at 100
and 6V, respectively. Potentials of 1000 and 4000 V for RF
lens and capillary were applied. For 10 potentials of 195
and 6V for sample cone and extraction cone, respectively,
were applied. The potential at capillary was set at 4200V
and that of RF lens at 1200V. For all of the three com-
pounds the desolvation temperature was set at 120 °C and
the source temperature at 80°C.
2.2. DiVerential scanning calorimetry
Melting transitions (Tm) of 6 and 11 were determined
on power compensation type Perkin-Elmer PYRIS DIA-
MOND DSC. The measurements were carried out under
nitrogen atmosphere (Xow rate 50 mL min¡1) using 50 ꢂL
sealed aluminum sample pans with pin holes. The sealing
was made by using a 30 ꢂL aluminum pan with pinholes
as cover-pan to ascertain good thermal contact between
the sample and pan, and to minimize the free volume
inside the pan, as the sample volume is gently squeezed
by the cover-pan. The temperature calibration was
carried out using two standard materials (n-decane, In)
and energy calibration by an indium standard
(28.45 J g¡1). As small sample weights were used, the heat-
ing rates of 20 and/or 40 °C were used to enhance sensitiv-
ity of the measurements. Samples were heated from 25 to
400 °C and the sample weights used were about 0.15–
1 mg. The sample weights were checked afterwards to see
potential changes in the weights that may have occurred
during heating. Uncertainty for temperature is less than
0.8 °C.
2.3. Mass spectrometry
In order to ascertain the molecular weights of the
compounds, ESI-TOF MS measurements were per-
formed. For the MS measurement the NMR sample of 6
(in CDCl3) was diluted to a concentration of »0.5 mmol/
L by using acetonitrile (HPLC grade)–formic acid (0.1%)
mixture. The NMR sample of 11 was diluted to a con-
centration of »0.05 mmol/L by using methanol (HPLC
grade).
The adduct formation experiments with alkali metal
chlorides were performed by preparing 0.05 mM solutions
of LiCl, NaCl, KCl, and RbCl (Wrst dissolved in H2O and
then diluted with acetonitrile or methanol, respectively).
The ligands were Wrst dissolved in CHCl3 and then diluted
to a concentration of 0.05 mM by using acetonitrile (6) or
methanol–formic acid (0.1%) mixture (11). An equal vol-
ume of each individual alkali metal solution was taken,
mixed thoroughly, and the same volume of the ligand
solution added resulting in a molar ratio of 1:1 (cation/
ligand). After 15 min the measurement for the equimolar
mixture of the alkali metal chlorides and the ligand was
performed. For 11 adduct formation with earth alkaline
metal chlorides was tested as well. MgCl2, CaCl2, and
SrCl2 were Wrst dissolved in H2O and then diluted to a
concentration of 0.05 mM by using MeOH. The experi-
In the adduct formation experiment of 6 with alkali
metal chlorides a Xow rate of 80 ꢂL/min was used for the
sample solution. The potentials for sample and extraction
cones were set at 75 and 5 V and those of 750 and 3800 V
for RF lens and capillary were applied, respectively. In the
experiments aimed at investigation of adduct formation of
11 with alkali and earth alkaline metal chlorides, the values
of 100 and 80 and 5V for sample and extraction cones were
applied. The potential at the capillary was set at 3800V and
that of RF lens at 2000 V. A Xow rate of 40ꢂL/min of the
sample solution for both of the experiments was used. In
each of the adduct formation experiments the desolvation
temperature was set at 120 °C and the source temperature
at 80°C, respectively.
2.4. NMR spectroscopy
1H, 13C, 13C DEPT-135, PFG (pulsed Weld gradient)
1
DQF (double quantum Wltered) H,1H COSY (correlation
1
spectroscopy) [31,32], PFG H,13C HMQC (heteronuclear

70
A. Valkonen et al. / Journal of Molecular Structure 846 (2007) 65–73
multiple quantum coherence) [33,34], and PFG 1H,X
HMBC (heteronuclear multiple quantum correlation,
13
XD C or 15N) [35] spectra of 2, 6, and 11 were recorded in
dilute CDCl3-solutions at 30°C with Bruker Avance DRX
500 FT NMR spectrometer equipped with an inverse
detection dual probehead (BBI) and z-gradient accessory
working at 500.13 MHz for proton, at 125.77MHz for
carbon-13, and at 50.69MHz for nitrogen-15, respectively.
1H NMR spectra of the precursors 1, 3–5, and 7–10 have
been measured with Bruker Avance DPX 400 FT NMR
spectrometer. The 1H and 13C chemical shifts are referenced
to the signal of internal TMS (ꢂ D 0.0ppm) and the 15N
shift to the N-15 signal of an external CH3NO2
(ꢂ D0.0 ppm) capillary inserted coaxially inside the 5 mm
diameter NMR tube. The complete lists of the Bruker
Avance DRX 500 acquisition and processing parameters
are available from E.K. on request.
2.5. Calculations
Fig. 2. PFG 1H,13C HMQC spectrum of 6.
The molecular frameworks were drawn in HyperChem
molecular modeling software [36], after which optimization
of the geometries was performed using molecular mechan-
ics MM+ force Weld [37]. In order to locate the possible glo-
bal energy minimum on the potential surface, 10 simulated
annealing runs in MM+ force Weld were performed for
both cholaphanes. Each run started from 0 K, followed by
5 ps heating to 300 K in 25 K temperature steps, 10ps simu-
lation at 300K, and 5 ps cooling close to 0K, all in 0.0001 ps
time steps. After each run the obtained structure was fur-
ther optimized using molecular mechanics and then sub-
jected to semiempirical PM3 calculations [38]. The
calculations were performed on Dell Optiplex GX 280 PC
(3 GHz Intel® Pentium® 4 CPU, 2 GB). In order to locate
the most probable interaction site of the alkali metal, the
cation was introduced to three diVerent locations inside the
cavity as well as to the close vicinity of both spacer groups
outside the cavity in relation to the PM3-optimized chola-
phane conformation lowest in energy, given a charge of +1,
and the resulting complex optimized using molecular
mechanics.
Fig. 3. PFG 1H,13C HMQC spectrum of 11.
3. Results and discussion
by PFG 1H,15N HMBC experiment using a 100ms evolution
delay and referenced to the signal of an external CH3NO2
sample in a 1mm diameter capillary tube inserted coaxially
inside the 5mm NMR-tube. The value of ¡272.0ppm of 11 is
very typical for carboxylic acid amides [40]. The complete 1H
NMR spectral assignment of the steroidal skeleton is not
possible due to seriously overlapping signals. The H NMR
chemical shifts are referenced to the resonance of the residual
CHCl3, 7.26ppm from an internal (CH3)4Si (ꢂD0.0ppm).
3.1. NMR
The 13C NMR chemical shift assignments of 6 and 11 and
their precursors 1–5 and 7–10 are based on the large refer-
ence data collected in our review on several bile acid deriva-
tives [39], related terephthalate containing cholaphanes [12],
1
and 13C DEPT-135, PFG DQF H,1H COSY, PFG H,13C
1
1
HMQC, as well as PFG 1H,13C HMBC experiments of 6 and
11. In Figs. 2 and 3 are depicted the PFG H,13C HMQC
1
spectra of 6 and 11, respectively. The 13C NMR chemical
shifts are referenced to the center peak of the solvent CDCl3,
77.0ppm from the signal of an internal (CH3)4Si
(ꢂD0.0ppm). 15N NMR chemical shift of 11 is determined
3.2. MS
In order to ascertain the molecular weights of the ligands
6 and 11 ESI-TOF mass spectrometric measurements were

A. Valkonen et al. / Journal of Molecular Structure 846 (2007) 65–73
71
performed. An experiment for 6 resulted in MS m/z ESI-
3.3. Molecular modeling
TOF+ found 975.8177 [M+Na]+ (100%) and 991.7653
[M+K]+ (42.5%), C60H88O9Na requires 975.6326 and
C60H88O9K 991.6066. The corresponding results for 11 were
MS m/z ESI-TOF+ found 973.8069 [M+Na]+ (100%) and
989.7914 [M+K]+ (5%), C60H90N2O7Na requires 973.6646
and C60H90N2O7K 989.6385.
For 11 the simulated annealing runs led to three confor-
mations, which were further optimized semiempirically. For
6 only one conformation was found using molecular
dynamics simulations. The PM3-optimized conformations,
lowest in energy for both cholaphanes 6 and 11, are pre-
sented in Fig. 4.
The ability of the ligands to form adducts with alkali
metal cations was investigated by mixing equimolar
amounts of LiCl, NaCl, KCl, and RbCl, respectively, with
the ligand. The resulting solution was thoroughly mixed
and the measurement performed after 15 min. Similarly, as
reported recently for ionophorous lacalosid-derived antibi-
otic [28] as well as maleonitrile thiacrown ethers [41], alkali
metal adducts with 6 were also detected in competitive ESI-
TOF+ MS binding studies. [6+M]+-adducts appeared at m/z
values of 959.6570 [M+Li]+, 975.6213 [M+Na]+, 991.6106
[M+K]+, and 1037.5194 [M+Rb]+, respectively. The most
intense ion was the adduct ion with potassium (100%). The
second most intense ion was the adduct ion with sodium
(61.3%), the third most intense ion the adduct ion with
rubidium (42.5%), and the least intense ion that with lith-
ium (18.1%). Thus it seems as 6 does not show very signiW-
cant selectivity towards any of the alkali metal cations
investigated. On the other hand, for 11, the most intense ion
in the spectrum was the adduct ion formed with lithium
(100%; MS m/z ESI-TOF+ found 957.6642). The other
observed adduct ions were those with sodium (46.9%; MS
m/z ESI-TOF+ found 973.6648), with potassium (6.9%; MS
m/z ESI-TOF+ found 989.6382), and with rubidium (0.9%,
MS m/z ESI-TOF+ found 1035.6111), respectively.
As can be seen, the diameter of the cavity of 11 is smaller
compared to the cavity of 6. The width of the cavities is
more or less the same order of magnitude, whereas the cav-
ity of 6 is almost 2.5 Å longer than that of 11. The geome-
tries for the alkali metal ion adducts, lowest in energy for
Fig. 4. The conformations lowest in energy for 6 (on the left) and for 11
(on the right), respectively, optimized at the semiempirical PM3 level of
theory.
Even though adduct ions with all alkali metal cations with
cholaphane 11 were detected, the ligand, however, seemed to
possess some selectivity towards cations with smaller ionic
radii. In order to ensure this observation, adduct formation
tendency with some earth alkaline metal cations of the ligand
was conducted as well. Equimolar amounts of MgCl2, CaCl2,
and SrCl2 were thoroughly mixed, the ligand 11 added into
the mixture, and the measurement performed after 15min.
Ions at m/z values of 951.6459 (29.4%), 973.6121 (100%),
989.5849 (6.3%), and 1037.5742 (2.5%) were observed. Those
were interpreted to correspond the adduct ions [M+H]+,
[M¡H+Mg]+, [M¡H+Ca]+, and [M¡H+Sr]+, respectively.
The most intense ion in this series was the adduct ion
[M¡H+Mg]+. The ionic radius of Mg2+ is ca. 65pm being
close to that of Li+, which is approximately 60pm. The sec-
ond most intense ion was [M+H]+, and the third most intense
one the adduct ion [M¡H+Ca]+. The intensity of the adduct
ion formed with strontium was practically negligible, as was
the one formed with rubidium in the alkali metal cation
series. These results thus suggest that the cavity of 11 is
smaller than that of 6 thus preventing the binding of cations
with larger ionic radii and, on the other hand, favoring the
binding of cations with relatively small ionic radius (60–
65pm). Pyridinium, tetraalkyl-, and dialkyldiarylammonium
cations did not show signiWcant adduct formation with 6 and
11 in ESI-TOF+ MS experiments.
Fig. 5. The optimized geometries of the adducts formed by 6 with the
alkali metal ions Li+, Na+, K+, and Rb+, respectively. The optimizations
were performed using molecular mechanics and PM3-optimized confor-
mation of 6.

72
A. Valkonen et al. / Journal of Molecular Structure 846 (2007) 65–73
the cavity, whereas K+ and Rb+ are excluded outside. Most
probably the arrangement with the cation inside the cavity
corresponds to an adduct detected by mass spectrometric
measurements. The tendency of 6 to form adducts with cat-
ions located outside the cavity according to the present
molecular modeling calculations could be explained by the
favorable coordination environment of two carbonyl and
ether oxygen of cholaphane 6. Cholaphane 11 provides
only one carbonyl oxygen for electrostatic interaction for
cations sited outside the cavity, resulting in a remarkably
more instable adduct, respectively.
4. Conclusions
Two novel head-to-head cholaphanes with diVerent cav-
ity sizes have been designed and synthesized via several
steps with overall good yields. These cholaphanes show
diVerent alkali metal cation binding properties, one favor-
ing the lighter ones, and the other one the heavier ones,
which possess larger ionic radii, respectively These Wndings
can be further utilized in designing other bile acid-derived
cholaphanes for molecular recognition, antibiotic drug sub-
stances, and other medicinal as well as analytical uses.
Fig. 6. The optimized geometries of the adducts formed by 11 with the
alkali metal ions Li+, Na+, K+, and Rb+, respectively. The optimizations
were performed using molecular mechanics and PM3-optimized confor-
mation of 11.
Acknowledgements
Russian authors are grateful to the RFBR Grant N 04-03-
32995 for Wnancial support. Dr. Sara Busi (Dept. Chem.,
Univ. Jyväskylä, Finland) is gratefully acknowledged for pro-
viding the tetraalkyl and dialkyldiarylammonium salts. E.S.
wishes to sincerely thank Academy of Finland, project no.
7105950 for Wnancial support. S.I. gratefully acknowledges
the Wnancial support from the Finnish Ministry of Education
(Graduate School of Bio-organic and Medicinal Chemistry).
both cholaphanes 6 and 11, are shown in Figs. 5 and 6,
respectively.
In the case of 6 the alkali metal cations with smaller ionic
radius (Li+ and Na+, respectively) resettled outside the chola-
phane cavity. The carbonyl oxygens of the carboxymethoxy-
acetyl spacer seem to form a “tweezer-like” arrangement,
which pinches the cations between the partially negatively
charged oxygen atoms. On the other hand, K+ and Rb+
thrive inside the cholaphane cavity. The carbonyl oxygens of
the carboxymethoxyacetyl spacer seem to be responsible for
the electrostatic attraction of the positively charged ions in
these cases as well, except that they are oriented towards the
cavity instead of pointing outwards from it.
References
[1] D.A. Bell, E.V. Anslyn, in: J. L Atwood, J.E.D. Davies, D.D. Macnicol,
F. Vögtle (Eds.), Comprehensive Supramolecular Chemistry, vol. 2,
Elsevier, Oxford, 1996, pp. 439–475.
[2] D. Kritchevsky, P.P. Nair, in: P.P. Nair, D. Kritchevsky (Eds.), The
Bile Acids: Chemistry, Physiology, and Metabolism, vol. 1, Plenum,
New York, 1971, pp. 3–9.
[3] A. Ehnsen, W. Kramer, G. Wess, Drug Discov. Today 3 (1998) 409.
[4] A.D. Averin, E.R. Ranyuk, N.V. Lukashev, I.P. Beletskaya, Chem.
Eur. J. 11 (2005) 7030.
[5] S. Ghosh, A.R. Choudhury, T.N.G. Row, U. Maitra, Org. Lett. 7
(2005) 1441.
In the case of 11, Li+ and Na+ ions exhibiting smaller
ionic radii nicely Wt in a partially negatively charged
“pocket” formed by the carbonyl and ether oxygens of the
(2-amidoethoxy)ethylamido spacer. The alkali metal cat-
ions with larger ionic radius (K+ and Rb+, respectively), for
one, move outside the cholaphane cavity resettling to the
close vicinity of the carbonyl oxygen of the (2-amidoeth-
oxy)ethylamido spacer.
[6] E. Virtanen, J. Tamminen, M. Haapala, P. Mänttäri, R. Kauppinen,
E. Kolehmainen, Magn. Reson. Chem. 41 (2003) 567.
[7] E. Virtanen, J. Koivukorpi, J. Tamminen, P. Mänttäri, E. Kolehmai-
nen, J. Organomet. Chem. 668 (2003) 43.
The molecular modeling results support the mass spec-
trometric ones, according to which cholaphane 6 formed
adducts with all of the four alkali metal cations willingly,
whereas cholaphane 11 excluded the ions with larger ionic
radii. Based on molecular modeling results, in the case of 6
K+ and Rb+ are located inside the cavity, whereas Li+ and
Na+ are sited outside the cavity. Cholaphane 11 behaves
oppositely; the cations with smaller ionic radii are inside
[8] P.S. Pandey, R. Rai, R.B. Singh, J. Chem. Soc. Perkin Trans. 1
(2002) 918.
[9] J. Tamminen, E. Kolehmainen, M. Haapala, J. Linnanto, Synthesis
(2000) 1464.
[10] K.M. Bhattarai, A.P. Davis, J.J. Perry, C.J. Walter, S. Menzer,
D.J. Williams, J. Org. Chem. 62 (1997) 8463.
[11] P.S. Pandey, R.B. Singh, Tetrahedron Lett. 38 (1997) 5045.
[12] E. Kolehmainen, J. Tamminen, K. Lappalainen, T. Torkkel, R. Sep-
pälä, Synthesis (1996) 1082.

A. Valkonen et al. / Journal of Molecular Structure 846 (2007) 65–73
73
[13] R.P. Bonar-Law, A.P. Davis, B.J. Dorgan, Tetrahedron 49 (1993)
9855.
[29] S.R. Choon, W.C. Sang, W.C. Jae, Bull. Korean Chem. Soc. 21 (2000)
342.
[14] R.P. Bonar-Law, A.P. Davis, Tetrahedron 49 (1993) 9845.
[15] R.P. Bonar-Law, A.P. Davis, Tetrahedron 49 (1993) 9829.
[16] A.P. Davis, Chem. Soc. Rev. 22 (1993) 243, and references cited
therein.
[30] Y. Li, J.R. Dias, Synthesis (1997) 425–430.
[31] M. Rance, O.W. Sørensen, G. Bodenhausen, G. Wagner, R.R. Ernst,
K. Wüthrich, Biochem. Biophys. Res. Commun. 117 (1984) 479.
[32] A. Derome, M. Williamson, J. Magn. Reson. 88 (1990) 177.
[33] A. Bax, R.H. GriVey, B.L. Hawkins, J. Magn. Reson. 55 (1983) 301.
[34] A. Bax, S. Subramanian, J. Magn. Reson. 67 (1986) 565.
[35] A. Bax, M.F. Summers, J. Am. Chem. Soc. 108 (1986) 2093.
[36] HyperChem™, Release 7.0 for Windows, Hypercube, Inc., Gaines-
ville, USA, 2002.
[37] (a) N.L. Allinger, J. Am. Chem. Soc. 99 (1977) 8127;
(b) N.L. Allinger, Y.H. Yuh in: U. Burkert, N.L. Allinger (Eds.),
Quantum Chemistry Program Exchange; Bloomington, Indiana; Pro-
gram No. 395, Molecular Mechanics, ACS Monograph 177, Ameri-
can Chemical Society; Washington, DC 1982.
[17] U. Maitra, S. Nath, ARKIVOC iii (2005) 133.
[18] P. Babu, U. Maitra, Proc. Ind. Acad. Sci. Chem. Sci. 115 (2003) 607.
[19] S. Matile, A. Som, N. Sorde, Tetrahedron 60 (2004) 6405.
[20] P. Bühlmann, E. Pretsch, E. Bakker, Chem. Rev. 98 (1998) 1593.
[21] P. Chattopadhyay, P.S. Pandey, Tetrahedron 62 (2006) 8620.
[22] J. Tamminen, E. Kolehmainen, Molecules 6 (2001) 21.
[23] E. Virtanen, E. Kolehmainen, Eur. J. Org. Chem. (2004) 3385.
[24] M. Haapala, E. Kolehmainen, J. Tamminen, R. Kauppinen, J. Linn-
anto, E. Virtanen, R. Suontamo, P. Vainiotalo, Mater. Sci. Eng. C 18
(2001) 21.
[25] H. Huang, L. Mu, J. He, J.-P. Cheng, Tetrahedron Lett. 43 (2002)
2255.
[26] K.S. Kim, C. Cui, S.J. Cho, J. Phys. Chem. B 102 (1998) 461.
[27] P. Xie, M. Diem, J. Am. Chem. Soc. 117 (1995) 429.
[28] R. Pankiewicz, G. Schroeder, B. Brzezinski, J. Mol. Struct. (2007) in
press.
[38] J.J.P. Stewart, J. Comput. Chem. 10 (1989) 209.
[39] J.R. Dias, H. Gao, E. Kolehmainen, Spectrochim. Acta. A 56 (2000) 53.
[40] S. Berger, S. Braun, H.-O. Kalinowski, NMR Spectroscopy of the
Non-Metallic Elements, John Wiley & Sons, Chichester, 1997, p. 142.
[41] I. Starke, S. Fürstenberg, H. Müller, H.-J. Holdt, E. Kleinpeter, Rapid
Commun. Mass Spectrom. 20 (2006) 897.