Selective fructose transport mediated by di-boronic acids derived from neopentyl glycol

Article abstract of DOI:10.1071/CH02267

The transport properties of diboronic acids linked to a neopentylglycol core through ether and carboxylate ester links was examined. Molecular modelling was performed on the pentaerythritol-derived di-boronic acid and its fructose boronate esters. It was

Full text of DOI:10.1071/CH02267

Rapid Communication
CSIRO PUBLISHING
www.publish.csiro.au/journals/ajc
Aust. J. Chem. 2003, 56, 17–21
Selective Fructose Transport Mediated by Di-Boronic Acids
Derived from Neopentyl Glycol
Peter J. DugganA,B and Melanie L. SzydzikA
A School of Chemistry, Monash University, Clayton 3800, Australia.
B Author to whom correspondence should be addressed (e-mail: peter.duggan@sci.monash.edu.au).
Manuscript received: 23 December 2002.
Final version: 9 January 2003.
Boronic acids that are able to selectively transport carbohy-
dratesacrosslipophilicmembraneshavepotentialapplication
in environmentally benign industrial sugar production.[1,2]
The accepted mechanism by which boronic acids are able
to facilitate the passage of sugars through organic mem-
branes is shown in Figure 1. In this so-called ‘tetrahedral
mechanism’,[1,2] a lipophilic quaternary ammonium salt
playsavitalpartinthetransportprocess, mainlybyimproving
the solubility of the boronate esters in the organic membrane.
With industrial applications in mind, we have been devel-
oping carriers that select for fructose over glucose and
sucrose; fructose being the sweetest of the naturally occur-
ring carbohydrates.[3] Sucrose is poorly transported through
lipophilic membranes by boronic acid carriers,[2,4] so, apart
from aiming for high fluxes and stable membranes, the
main challenge in this work is the improvement of the
ratio of fructose to glucose fluxes. We recently reported[5]
the enhanced fructose selectivity (Table 1) of a diboronic
acid (1) (Scheme 1), constructed on a pentaerythritol core.
This di-boronic acid is thought to form, within the mem-
brane, a 2 : 1 diboronate ester with fructose stabilized by
intramolecular hydrogen bonds (2).
cavitand boronic acids was found to be highly selective and
rapid compared with previously reported carriers.[6] Cavitand
(3) in particular is impressive in this regard (Table 1).
With our intention to further this work through the devel-
opment of polyvalent boronic acid based carriers, it became
important for us to explore the properties of the different
Table 1. Sugar fluxes through supported liquid membrane
A
Boronic acid
Flux (10−8 mol m−2 s−1
)
Ratio of fluxes
Fructose
Glucose
(1)B
(3)C
(7)
(11)
(15)
26.1 2.6
24.4 2.4
24.7 1.0
15.5 1.1
42.3 2.8
3.4 0.3
2.3 0.6
2.5 0.7
3.8 0.4
10.6 1.0
7.6 1.1
10.6 2.9
9.9 2.8
4.1 0.5
4.0 0.5
A Boronic acid concentration in membrane = 52 mM. Fluxes shown
are averages of 2–3 runs. T = 298 K.
B Data from ref. [5].
C Data from ref. [6].
OH
HO
O
O
O
H
(HO)2B
B(OH)2
O
O
O
The favourable transport properties of (1) suggested to us
that higher fructose selectivities and fluxes could be obtained
with carriers that bear more than two boronic acid groups pro-
jecting from the same side of a given scaffold. Consistent with
this prediction, the fructose transport promoted by a series of
O
H
O
O
O
B
B
O
O
O
O
O
O
O
O
O
O
O
O
OH
Ar
B
Q
X
sugar
HO
sugar
HO
OH
OH
OH
(2)
(1)
OH
OH
(HO)2B
H
H
B(OH)2
O
sugar
O
O
O
O
O
X
2H2O
O
X
2H2O
O
O
O
B
Q
OH
Ar
C5H11 C5H11
C5H11 C5H11
Aqueous Departure
Phase
Organic
Membrane
Aqueous Receiving
Phase
(3)
Scheme 1. Structures of di-boronic acids (1)–(3).
10.1071/CH02267 0004-9425/03/010017
Fig. 1. Accepted ‘tetrahedral’mechanism whereby boronic acids assist
membrane passage. Q⊕ is a quaternary ammonium cation.
© CSIRO 2003

18
P. J. Duggan and M. L. Szydzik
linking groups through which aromatic boronic acids might
be attached to a range of possible scaffolds.We report here the
results of our first investigation of this sort, in which we have
examined the transport properties of diboronic acids linked
to a neopentylglycol core though ether and carboxylate ester
links.
method used in the preparation of (1).[5] The pinacol pro-
tecting groups were removed by treatment with periodate[7]
which, although not high yielding, provided the di-boronic
acid (7) in a highly pure form.
Alkylation of neopentyl glycol (5) with p-bromobenzyl
bromide (8), Scheme 3, afforded the di-bromide (9) in good
yield. Palladium-catalyzed boronation of (9) with bis(pinaco-
lato)diboron provided the di-boronate ester (10), which was
deprotected to afford the di-benzyl ether linked di-boronic
acid (11).
Results and Discussion
Design and Synthesis of Carriers
Molecular modelling performed on the pentaerythritol-
derived di-boronic acid (1) and its fructose boronate esters
suggested that the high fructose selectivity shown by this car-
rier results from the favourable placement of its boronic acids,
such that stable diesters similar to (2) can readily form within
the membrane. Thus, (2) is thought to transport two fructose
molecules per molecule of boronic acid[5] for most of its pas-
sages across the membrane. We were interested to discover
the role played by the carboxylate ester links in this arrange-
ment, and if the analogous and potentially more stable benzyl
ether links would give the same outcome. For ease of synthe-
sis, we chose to use neopentyl glycol (5) as the core rather
than the closely analogous pentaerythritol. We expected that
the absence of the large lipophilic tails possessed by (1)
may result in some leaching of the neopentyl glycol-derived
boronic acids from the membrane during transport, but would
present fewer synthetic challenges. Thus (7) and (11) were
both prepared from neopentyl glycol as shown in Schemes 2
and 3.
In order to find evidence for intramolecular cooperativity
between boronic acids, it is important to compare the sugar-
transport properties of di-boronic acids with an appropriate
mono-boronic acid. Given that (7) and (11) are somewhat
smaller than many of the boronic acid carriers we have used
in the past, it seemed appropriate that a new mono-boronic
acid, similar in structure to (7) and (11), be prepared and
its transport properties used as a standard for the results
obtained with (7) and (11). The nitro compound (15) was
chosen for this purpose because it possesses a closely related
substitution pattern to (7) and (11). In addition, the inclu-
sion of a nitro substituent gives (15) physical properties
more compatible with ortho-nitrophenyloctyl ether (NPOE),
the commonly used lipophilic liquid membrane employed
in transport experiments.[1,2,4–6] The monoboronic acid (15)
was prepared in three steps, as shown in Scheme 4.
Transport Studies
An examination of Figure 1 reveals that sugar transport pro-
moted by boronic acid can be driven with a pH gradient
In the preparation of (7), Scheme 2, we first used a DCC/
DMAP activating system to acylate neopentyl glycol (5) to
produce the di-carboxylate ester (6) in excellent yield and
purity. This is a significant improvement on the alkylation
Br
Br
Br
NaH, NaI
HO
OH
O
O
O
O
DMF
(97%)
O
B
B
O
B
O
O
Br
HO
OH
DCC, DMAP
(9)
(8)
(5)
DCM
(98%)
O
O
O
O
O
B
O
O
O
[PdCl2(dppf)],
KOAc, DMSO
O
OH
B
(4)
(5)
(18%)
(6)
NaIO4, HCl,
THF / H2O
(51%)
O
O
O
O
(HO)2B
B(OH)2
(HO)2B
B(OH)2
B
B
NaIO4, HCl,
THF / H2O
(32%)
O
O
O
O
O
O
O
O
(7)
(10)
(11)
Scheme 2. Preparation of di-boronic acid (7). DCC = 1,3-dicyclo-
hexylcarbodiimide; DMAP = 4-(N,N-dimethyl-amino) pyridine.
Scheme 3. Preparation of di-benzyl ether linked di-boronic acid (11).

Selective Fructose Transport Mediated by Di-Boronic Acids
19
because, as in the tetrahedral mechanism, one equivalent of
hydroxide is transported through the membrane with each
equivalent of sugar. Thus, in the present study and in accord
with previous investigations,[4–6] the pH of the departure and
receiving phases were buffered at 11.3 and 6.0, respectively.
The fluxes of fructose and glucose through NPOE supported
on the porous polypropylene Accurel,[4–6] promoted by the
neopentylglycol-derived boronic acids (7), (11), and (15) and
the lipophilic quaternary ammonium salt Aliquat 336 in the
presence of the above-mentioned pH gradient, are shown in
Table 1.
Despite our concerns regarding the leaching of these
smaller, and potentially more water-soluble boronic acids,
plots of receiving phase sugar concentration versus time over
the 7 h transport experiments were observed to be linear. Only
one carrier, (11), was studied for a longer period, 32 h, and
the plot in that case still appeared to be linear.
In a broad sense, the sugar fluxes promoted by (7) and (11)
are quite similar to those of previously studied di-boronic
acids (Table 1), but the transport facilitated by the reference
mono-boronic acid (15) is the highest of any mono-boronic
acids studied thus far. One of the rate-limiting steps in the
membrane-transport process has been shown to be diffusion
of the sugar-boronate ester through the organic membrane,[1]
so the enhanced sugar flux produced by (15) most likely
relates to its compact nature.
The fructose to glucose transport selectivity shown by (15)
is similar to that observed for other mono-boronic acids,[4–6]
andreflectstheinherentpreferenceforfructosetransportpos-
sessed by boronic acids. Importantly, the fructose preference
of almost 10 : 1 shown by the di-carboxylate (7) is actually
superior to that previously observed for (1) and second only to
the cavitand di-boronic acid (3). This result supports the idea
that the di-boronic acid motif present in both (1) and (7) is
particularly well suited to binding and transporting fructose.
The ease of preparation of (7), combined with its impressive
sugar transport properties, makes it an ideal candidate for use
in future pilot-scale sugar-separation experiments.
The most striking result shown in Table 1, however, is the
low fructose selectivity shown by the di-benzyl ether (11).
The change from two carbonyl carbons to two methylene car-
bons in the linking groups, although seemingly minor, not
only reduces the angle at which the boronic acids project from
the neopentylglycol core but also introduces more degrees of
freedom into the links. These two features are likely to work
against the stabilizing effect identified for the di-fructose
di-boronate (2) and expected to be present in similar fructose
esters of (7).
Conclusions
Three new boronic acids based on a neopentyl glycol core
have been prepared, a di-boronic acid linked through car-
boxylate esters (7), a di-boronic acid linked through ether
links (11), and a reference mono-boronic acid (15). All three
compounds were found to be good carriers for fructose and
glucose through the lipophilic membrane NPOE, containing
Aliquat 336, in the presence of a pH gradient. No evidence
for carrier leaching was observed during the several-hour
time frame of the transport experiments. Relative to the
reference mono-boronic acid (15), the carboxylate-linked
di-boronic acid (7) showed a remarkable fructose : glucose
transport selectivity, almost 10 : 1, approaching the best thus
farreported.[6] Itisthoughtthat(7), analogousto(1), isableto
form stable di-fructose di-boronate esters similar to (2), and
that this is how most of the fructose is transported through
the NPOE by (7). In contrast, the ether linked di-boronic acid
(11) gave the slowest fructose transport of the three carriers
studied, and a fructose : glucose selectivity of 4 : 1, similar to
that typically observed for mono-boronic acids. Apparently,
the different angle at which the boronic acids project from the
neopentyl glycol core in (11), and the added flexibility of its
ether links, reduces the stability of its di-fructose di-boronate
esters, resulting in lower rates of fructose transport.
NO2
NO2
Ag2O
HO
OH
DCM
(43%)
O
OH
Br
(13)
(5)
(12)
DCC, DMAP,
DCM
(4)
(68%)
Experimental
General Methods
O
O
NO2
B
NO2
B(OH)2
All non-specialized starting materials were commercially available
research-grade chemicals and used without further purification. THF
was distilled from sodium wire / benzophenone and used directly. DMF
and DMSO were stored over activated 4 Å molecular sieves for at least
24 h. DryDCMwasobtainedbydistillationfromCaH2 anduseddirectly.
All other solvents were used as purchased. Thin-layer chromatography
was performed on silica-coated aluminium sheets (Merck, Silica gel
60 F254) and viewed using ultraviolet (254 nm) light. Compounds con-
taining boronic acids were visualized with the aid of diphenylcarbazone
stain. Melting points were determined with a Reichert hot-stage melting
point apparatus. Microanalyses were performed by Chemical and Micro
Analytical Services (Belmont,Victoria).The majority of IR spectra were
NaIO4, HCl,
THF / H2O
(95%)
O
O
O
O
O
O
(14)
(15)
Scheme 4. Preparation of mono-boronic acid (15).

20
P. J. Duggan and M. L. Szydzik
recorded with a Perkin–Elmer 1600 series instrument. Samples were
analysed as thin films (neat) or Nujol mulls mounted on NaCl plates.
The IR spectrum of (15) was recorded on a Bruker IFS 55 spectrometer
using a Specac single-reflectionATR system fitted with a single bounce
diamond top-plate. 1H NMR and 13C NMR spectra were recorded on a
Varian 300 MHz spectrometer and referenced to the residual protonated
solvent peak. Carbon atoms attached to boron were not observed due
to band broadening. Low-resolution mass spectra of methanol solutions
wererecordedwithaMicromassPlatformIIAPIQMSelectrospraymass
spectrometer operating in the positive ion mode. High resolution mass
spectra were recorded on a Bruker BioApex 47e Fourier Transform
mass spectrometer.
36.6 mmol) was added, followed by dropwiseaddition of 4-bromobenzyl
bromide (8) (9.75 g, 39.0 mmol) in DMF (10 mL). The reaction mix-
ture was then heated to 90◦C for 24 h then allowed to cool to room
temperature. Thereafter the mixture was diluted with water (70 mL)
and extracted with DCM (3 × 40 mL). The organic layers were com-
bined, washed with water (5 × 80 mL) and brine (1 × 50 mL), dried
over Na2SO4, and filtered. The filtrate was concentrated in vacuum
to give the crude product as a yellow oil (8.26 g, 97%). This com-
pound was used for the next step without further purification. νmax
(neat)/cm−1 3025 w, 2928 s, 2854 s, 1593 m, 1487 s, 1399 m, 1356 m,
1298 w, 1278 w, 1244 w, 1201 w, 1094 s, 1070 s, 1012 s, 828 m, 803 s,
734 m, 629 w; δH (CHCl3) 7.45 (4 H, d, J 8.6, m-ArH), 7.18 (4 H, d, J
8.6, o-ArH), 4.44 (4 H, s, Ar-CH2), 3.25 (4 H, s, CCH2O), 0.95 (6 H,
s, CH3); δC (CDCl3) 138.2, 131.6, 129.2, 121.4, 76.7, 72.7, 36.7, 22.2;
m/z 463.0 (100%, [M + Na+]).
2,2-Dimethyl–1,3-propanediyl bis[p-(4,4,5,5-tetramethyl-
1,3,2-dioxaborolan-2-yl) benzoate] (6)
Following a procedure adapted from that of Ward et al. for the
synthesis of an alanine ester derivative,[8] p-(4,4,5,5-tetramethyl-1,3,2-
dioxaborolan-2-yl)-benzoic acid (4)[5] (0.755 g, 3.25 mmol), neopentyl
glycol (0.156 g, 1.49 mmol), and DMAP (0.081 g, 0.66 mmol) were
dissolved in dry DCM (6.6 mL) at room temperature under N2. The
solution was cooled to 0◦C and DCC (0.715 g, 3.47 mmol) added. The
resulting mixture was stirred at 0◦C for 1 h, then allowed to warm to
room temperature and stirred for a further 20 h. Ethyl acetate (10 mL)
was added, and the precipitated urea was removed by suction filtra-
tion and washed with more ethyl acetate (2 × 2 mL). The combined
filtrate and washings were washed with 0.5 M citric acid (2 × 10 mL)
and saturated NaHCO3 (2 × 10 mL), then dried over Na2SO4 and
filtered. The filtrate was concentrated in vacuum to give the crude
product as an off-white solid (0.83 g, 98%). The title compound (6)
was the only component in the crude reaction mixture detectable by
1H NMR spectroscopy. A sample was further purified by recrystal-
lization from EtOH / H2O to give a white powder. M.p. 124–126◦C.
(Found: C, 65.85; H, 7.61. C31H42B2O8 requires C, 65.98; H, 7.50%).
νmax (Nujol)/cm−1 1721 s, 1399 s, 1361 s, 1264 m, 1144 m, 1112 m,
1098 m, 1018 m, 857 m, 711 m; δH (CDCl3) 8.01 (4 H, d, J 8.4, o-
ArH), 7.86 (4 H, d, J 8.4, m-ArH), 4.25 (4 H, s, CH2), 1.35 (24 H,
s, OC(CH3)2), 1.16 (6 H, s, (CH2)2C(CH3)2); δC(CDCl3) 166.5,
134.8, 132.3, 128.6, 84.3, 70.0, 35.6, 25.1, 22.3; m/z 587.5 (100%,
[M + Na+]), 628.6 (60).
1,1ꢀ-[(2,2-Dimethyl–1,3-propandiyl) bis(oxymethylene)] bis[p-
(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)] (10)
The procedure for the synthesis of (10) was adapted from method-
ology used by Malan and Morin.[9] Potassium acetate (1.70 g, 17.3
mmol), (9) (0.53 g, 1.20 mmol), and 1,1-bis(pinacolato)diboron (1.08 g,
4.25 mmol) were combined in DMSO (13 mL) and stirred for 4 h at room
temperature with a constant stream of Ar bubbling through the mixture.
Thereafter, [PdCl2(dppf)] (0.16 g, 0.20 mmol) was added and the reac-
tion was stirred for a further 20 h at 100◦C under Ar. Upon cooling,
the mixture was diluted with ethyl acetate (100 mL) and washed with
1 M HCl (20 mL) and water (3 × 20 mL). The aqueous washings were
extracted with ethyl acetate (1 × 30 mL) and the organic layers com-
bined, dried over Na2SO4, and filtered. The filtrate was concentrated
in vacuum to give a near-black oil. This was dissolved in acetonitrile
(30 mL) and extracted with cyclohexane / n-hexane (2 / 1, 2 × 50 mL).
Thehexanelayerswerecombined, washedwithacetonitrile(2 × 20 mL),
and then concentrated in vacuum to give a pale yellow oil which par-
tially crystallized on standing (4◦C). This partly crystalline mixture was
triturated with acetonitrile, filtered, and washed again with acetonitrile,
and the remaining white solid dried under high vacuum (∼2 mmHg)
overnight to give the title compound (10) as fine white crystals (0.82 g,
18%). M.p. 98–101◦C. (Found: M + Na+, 559.3445. C31H46B2O6.Na
requires 559.3378). νmax (Nujol)/cm−1 1616 m, 1518 m, 1362 s, 1323 s,
1275 s, 1143 s, 1100 s, 1020 m, 860 m, 825 m, 660 m; δH (CDCl3) 7.77
(4 H, d, J 8.1, m-ArH), 7.31 (4 H, d, J 8.1, o-ArH), 4.51 (4 H, s,
ArCH2), 3.25 (4 H, s, CCH2O), 1.35 (24 H, s, OC(CH3)2), 0.92 (6 H,
s, (CH2)2C(CH3)2); δC (CDCl3) 142.3, 134.8, 126.6, 83.8, 76.6, 73.3,
35.6, 25.1, 22.6; δB CDCl3) 31.1; m/z 559.6 (100%, [M + Na+]).
2,2-Dimethyl–1,3-propanediyl bis(p-borono benzoate) (7)
Following a procedure adapted from Falck et al.[7] NaIO4 (1.25 g,
5.84 mmol) was added to a solution of (6) (0.505 g, 0.894 mmol) in
THF / H2O (6.7 mL / 1.6 mL) and stirred at room temperature until the
reaction mixture was homogenous. HCl (2 M, 5.0 mL) was then added
and the reaction mixture stirred for a further 20 min. Thereafter, the
organic soluble material was extracted with ethyl acetate (2 × 15 mL).
The combined organics were washed with water (2 × 10 mL) and brine
(1 × 10 mL), dried over Na2SO4, then filtered. The filtrate was con-
centrated in vacuum to give an off-white coloured residue. This was
dissolved in DCM (3 mL), and toluene (1 mL) was added.The DCM was
removed in vacuum and cyclohexane (2 mL) added. Upon refrigeration
of this solution a white precipitate formed which was filtered, washed
with cyclohexane, and then dried under vacuum (∼2 mmHg) to give the
title compound (7) (0.18 g, 51%) as a white solid. M.p. 205–208◦C. νmax
(Nujol)/cm−1 3410 b, 1722 m, 1265 s, 1144 m, 1112 m 1098 m, 1018 m,
711 m; δH ([D6]acetone) 8.01 (4 H, d, J 8.4, o-ArH), 7.96 (4 H, d, J 8.4,
m-ArH), 4.27 (4 H, s, CH2), 1.19 (6 H, s, CH3); δC ([D6]acetone) 166.5,
134.9, 132.4, 128.9, 70.3, 36.0, 22.2; δB ([D6]acetone) 28.1; m/z 213.1
(70%), 225.2 (100), 249.2 (40), 437.4 (50, [monomethyl boronate +
Na+]), 451.4 (90, [dimethyl borate + Na+]), 559.6 (30).
1,1ꢀ-[(2,2-Dimethyl–1,3-propandiyl) bis(oxymethylene)]-
bis(p-borono benzene) (11)
Compound (10) (0.100 g, 0.18 mmol) was deprotected using NaIO4
(0.305 g, 1.42 mmol), THF (2.25 mL), water (0.40 mL), and HCl (2 M,
1.00 mL) according to a similar procedure to that used for the deprotec-
tion of (6). In this case, the reaction mixture was stirred for 2 h at room
temperature and following workup, the crude product was obtained as
an orange solid. This was purified using DCM (2 mL), toluene (1 mL),
and cyclohexane (2 mL), as described for (7), to give the title com-
pound (11) as a white solid (0.021 g, 32%). M.p. 188–190◦C. νmax
(Nujol)/cm−1 1613 m, 1408 s, 1340 s, 1304 s, 1133 m, 1084 m, 1021 m,
847 s, 728 m, 687 m; δH (CDCl3) 7.78 (4 H, d, J 7.8, m-ArH), 7.17 (4 H,
d, J 7.8, o-ArH), 4.51 (4 H, s,ArCH2), 3.31 (4 H, s, CCH2O), 0.98 (6 H,
s, CH3); δC (CDCl3) 143.2, 135.4, 126.4, 74.3, 72.5, 36.2, 22.8; m/z
409.4 (15%, [monomethyl boronate + Na+]), 423.5 (100, [dimethyl
boronate + Na+]).
1,1ꢀ-[(2,2-Dimethyl–1,3-propandiyl) bis(oxymethylene)]-
bis(p-bromobenzene) (9)
3-[(p-Nitrophenyl)methoxy]–2,2-dimethyl-1-propanol (13)
NaH (4.56 g, 114 mmol) was added to a solution of neopentyl gly-
col (2.02 g, 19.4 mmol) in dry DMF (240 mL) under an atmosphere
of N2 at 0◦C. The mixture was allowed to warm to room tempera-
ture and then stirred for 30 min. Upon cooling to 0◦C, NaI (5.48 g,
Following the procedure of Bouzide and Suave for the mono-alkylation
of diols,[10] Ag2O (1.02 g, 4.32 mmol) was added, under an atmosphere
of N2 to a stirred solution of neopentyl glycol (0.30 g, 2.88 mmol) and
p-nitrobenzyl bromide (12) (0.68 g, 3.17 mmol) in dry DCM (5 mL).

Selective Fructose Transport Mediated by Di-Boronic Acids
21
The resulting mixture was protected from light and stirred at room tem-
perature for 24 h then filtered through a short silica plug. This was
washed with DCM (200 mL) then ethyl acetate (200 mL). The ethyl
acetate fraction was collected and washed with water (3 × 20 mL) and
brine (1 × 20 mL), dried over Na2SO4, and filtered. The filtrate was
concentrated in vacuum to give the desired product as a pale yellow oil
(0.44 g, 43%). (Found: M + Na+, 262.1046. C12H17NO4.Na requires
262.1055). νmax (neat)/cm−1 3414 bs, 2960 s, 2871 s, 1605 m, 1522 s,
1495 w, 1477 m, 1403 w, 1347 s, 1201 w, 1175 w, 1105 s, 1050 s, 1016 m,
860 m, 843 m, 738 s; δH (CDCl3) 8.21 (2 H, d, J 8.7, m-ArH), 7.48 (2 H,
d, J 8.7, o-ArH), 4.61 (2 H, s, ArCH2), 3.49 (2 H, s, CH2OH), 3.36
(2 H, s, CH2OCH2), 1.61 (1 H, br, OH), 0.96 (6 H, s, CH3); δC (CDCl3)
147.4, 145.9, 127.6, 123.8, 79.4, 73.0, 71.6, 36.7, 22.0 (CH3 signals
overlapping); m/z 262.7 (100%, [M + Na+]).
336, and 2-nitrophenyloctyl ether (NPOE). The membrane solution was
generally prepared in duplicate by dissolving 25 µmol carrier and Ali-
quat 336 (1 equiv. per boronic acid) in a suitable solvent, dividing the
solution into two equal parts, and then adding NPOE (0.25 g) to each.
The solvent was removed in vacuum to give an oil, which was used to
coat the polymer support. The membrane was then stored under vacuum
(∼2 mmHg) for 15 h.
Enzyme Assay: At 30 min intervals, aliquots were taken from the
receiving phase and analyzed for sugar content. Fructose and glucose
assays were undertaken following a procedure adapted from Kinksy.[11]
An aliquot (initially 500 µL, but less as the sugar concentration in the
receiving phase increased) of the sugar solution was transferred into a
1 mL cuvette. To this, hexokinase and glucose–6-phosphate dehydro-
genase (2.5 units in 50 µL of NaH2PO4 (70 mM) and MgCl2 (4 mM)
at pH 7) and ATP (0.2 M, in 5 µL NaH2PO4 (70 mM), pH 7.5) were
added. NADP (0.1 M, in 5 µL NaH2PO4 (70 mM), pH 7.5) and, sub-
sequently, buffer (NaH2PO4 (0.10 M), pH 7.5) were added so that
the combined volume of sugar solution and buffer was 890 µL. The
absorbance of NADPH at 340 nm was monitored until a maximum was
reached. The total change in absorbance was used to calculate glucose
concentrations. Subsequent to stabilization of the absorbance curve,
phosphoglucose isomerase (14.3 units in 50 µL of NaH2PO4 (70 mM)
and MgCl2 (4 mM) at pH 7) was added to the cuvette and the second
absorbance change was used to calculate fructose concentrations.
Determination of Sugar Flux: Before each transport study, stock
solutions of fructose and glucose were used to produce standard curves
of fructose and glucose concentrations versus total change in absorbance
resulting from the enzymatic reactions. This standard curve was then
used to determine glucose and fructose concentrations in the subsequent
transport experiments for that day.This approach was found to give more
reliable data than simply using the literature extinction coefficient for
NADPHtodetermineitsand, hence, fructoseandglucoseconcentrations
in the receiving phase. Plots of fructose and glucose concentrations
in the receiving phase versus time were used to determine sugar flux,
3-[(p-Nitrophenyl)methoxy]–2,2-dimethyl-1-propanol-[p-(4,4,5,5-
tetramethyl-1,3,2-dioxaborolan-2-yl) benzoate] (14)
The pinacol boronate (14) was prepared from the alcohol (13) (0.14 g,
0.59 mmol) using DMAP (0.01 g, 0.12 mmol), (4) (0.16 g, 0.64 mmol),
DCC (0.15 g, 0.73 mmol), and dry DCM (2.1 mL) according the proce-
dure used for the preparation of (6). Ethyl acetate (5 mL) was used to
precipitate the DCU and, after workup, the crude product was obtained
as a yellow oil. This was refrigerated (4◦C) for 15 h, causing most
of the oil to crystallize. The solid was collected and recrystallized
from ethanol / H2O. The resulting crystals were washed with MeOH
to give the title compound (14) (0.19 g, 68%) as a white solid. M.p.
95–98◦C. νmax (Nujol)/cm−1 1704 s, 1604 m, 1521 s, 1400 m, 1360 s,
1344 s, 1268 s, 1125 m, 1104 m; δH (CDCl3) 8.10 (2 H, d, J 8.7, m-
ArH-p-NO2), 7.95 (2 H, d, J 8.1, m-ArH-p-BO2), 7.84 (2 H, d, J
8.1, o-ArH-p-BO2), 7.43 (2 H, d, J 8.7, o-ArH-p-NO2), 4.58 (2 H, s,
ArCH2), 4.19 (2 H, s, CO2CH2), 3.36 (2 H, s,ArCH2OCH2), 1.37 (12 H,
s, OC(CH3)2), 1.08 (6 H, s, (CH2)2C(CH3)2); δC (CDCl3) 166.3, 147.1,
146.1, 134.6, 132.4, 128.4, 127.5, 123.5, 84.2, 76.6, 72.1, 70.1, 36.0,
25.0, 22.3 (CH3 signals overlapping); m/z 492.4 (100%, [M + Na+]).
according to Equation (1), where F is the sugar flux [mol m−2 −1],
s
S the slope of the time-dependent concentration plot [M s−1], V the
volume in receiving phase [L], and r the radius of membrane exposed
to the departure phase [m2].
3-[ (p-Nitrophenyl)methoxy]–2,2-dimethyl-1-propanol-(p-borono
benzoate) (15)
The pinacol-protected boronate (14) (36 mg, 0.08 mmol) was depro-
tectedusingNaIO4 (60 mg, 0.28 mmol),THF(0.7 mL), water(0.14 mL),
and HCl (2 M, 0.2 mL) according to the procedure used for the deprotec-
tionof(6).Thereactionmixturewasstirredfor3.5 hatroomtemperature
and, following workup, the title compound (15) was obtained as a white
solid(28 mg, 95%). M.p. 126–129◦C. νmax (ATR)/cm−1 3493 br, 1708 s,
1606 m, 1518 s, 1402 m, 1340 s, 1264 s, 1125 m, 1101 s, 1019 m, 736 m;
δH ([D6]acetone/D2O) 8.14 (2 H, d, J 8.6, m-ArH-p-NO2), 7.95 (4 H,
br s, ArH-p-BO2), 7.43 (2 H, d, J 8.6, o-ArH-p-NO2), 4.67 (2 H, s,
ArCH2), 4.18 (2 H, s, CO2CH2), 3.44 (2 H, s, ArCH2OCH2), 1.08 (6 H,
s, CH3); δC ([D6]acetone/D2O) 66.6, 147.7, 146.4, 134.8, 132.2, 128.7,
127.5, 123.9, 76.8, 72.3, 70.4, 36.3, 22.3 (CH3 signals overlapping);
m/z 410.3 (32%, [M + Na+]), 424.2 (100, [monomethyl boronate +
Na+]).
F = S × V/πr2
(1)
Acknowledgments
This work was funded by Monash University.
References
[1] (a) B. D. Smith, Supramol. Chem. 1996, 7, 55–60; (b)
B. D. Smith, S. J. Gardiner, Adv. Supramol. Chem. 1999,
5, 157, and references therein.
[2] M. J. Karpa, P. J. Duggan, G. J. Griffin, S. J. Freudigmann,
Tetrahedron 1997, 53, 3669.
[3] L. M. Hanover, J. S. White, Am. J. Clin. Nutr. 1993, 58, 724S.
[4] S. J. Gardiner, B. D. Smith, P. J. Duggan, M. J. Karpa,
G. J. Griffin, Tetrahedron 1999, 55, 2857.
[5] S. P. Draffin, P. J. Duggan, S. A. M. Duggan, Org. Lett. 2001,
3, 917.
[6] T. M. Altamore, E. S Barrett, P. J. Duggan, M. S. Sherburn,
M. L. Szydzik, Org. Lett. 2002, 4, 3489.
[7] J. Flack, M. Bondlela, S. Venkataraman, D. Srinivas, J. Org.
Chem. 2001, 66, 7148.
[8] D. Ward, G. Yuanzhu, R. Lazny, M. Pedres, J. Org. Chem.
2001, 66, 7832.
[9] C. Malan, C. Morin, J. Org. Chem. 1998, 63, 8019.
[10] A. Bouzide, G. Sauve, Tetrahedron Lett. 1997, 38, 5945.
[11] S. Kinsky, Methods Enz. 1974, 32, 501.
Transport Experiments
Transport Apparatus: Transport studies followed a procedure similar
to one previously reported.[5] The transport apparatus consisted of two
water-jacketedhalf-cells, clampedtogetherandseparatedbyasupported
liquid membrane. The cells were maintained at 298 K. The half-cell
volume (32.5–34.0 mL) was specific for each half-cell and measured
independently for each transport study. Each half-cell was stirred by
an internally mounted magnetic stirrer and paddle (200–210 rpm). The
departure phase consisted of a freshly prepared solution of fructose
(0.30 M) and glucose (0.30 M) in Na2CO3 (0.10 M, buffered at pH
11.3). The receiving phase consisted of a solution of NaH2PO4 (0.10 M,
buffered at pH 6.0).
Membrane Preparation: The supported liquid membrane consisted
of a polymer support (flat sheet of Accurel 1E membrane (thickness
0.1 mm, pore size 0.1 µm)) coated in a solution of the carrier, Aliquat