Synthesis of silica-bound amylose by phosphorolytic elongation of immobilised maltoheptaosyl hydrazides

Article abstract of DOI:10.1016/S0040-4039(02)01293-5

Maltoheptaoside-alkoxysilane anchor molecules were synthesised by fusing aliphatic ω-Si(OEt₃) hydrazide linkers with maltoheptaose. After immobilisation of the primers on porous silica, support-bound amylose was synthesised by phosphorolytic synthesis. The hydrazone linkage as a pre-formed cleavage site allowed removal and subsequent characterisation of immobilised amylose, which showed a broad molecular weight distribution. Under HPLC conditions, amylose assumed a non-helical conformation, making surface interactions and not complexation the primary separation mechanism.

Full text of DOI:10.1016/S0040-4039(02)01293-5

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 6127–6131
Synthesis of silica-bound amylose by phosphorolytic elongation
of immobilised maltoheptaosyl hydrazides
Hans-Georg Breitinger*
Institut fu¨r Organische Chemie und Makromolekulare Chemie II, Heinrich-Heine-Universita¨t Du¨sseldorf, Universita¨tsstraße 1,
D-40225 Du¨sseldorf, Germany
Accepted 2 July 2002
Abstract—Maltoheptaoside-alkoxysilane anchor molecules were synthesised by fusing aliphatic v-Si(OEt3) hydrazide linkers with
maltoheptaose. After immobilisation of the primers on porous silica, support-bound amylose was synthesised by phosphorolytic
synthesis. The hydrazone linkage as a pre-formed cleavage site allowed removal and subsequent characterisation of immobilised
amylose, which showed a broad molecular weight distribution. Under HPLC conditions, amylose assumed a non-helical
conformation, making surface interactions and not complexation the primary separation mechanism. © 2002 Elsevier Science Ltd.
All rights reserved.
1. Introduction
ing for maximal conformational flexibility of the immo-
bilised polysaccharide chain. Immobilised amylose
could be removed via breaking of a pre-formed cleav-
age site, allowing analysis of the synthetic process.
Properties of the new phases in high-pressure liquid
chromatography (HPLC) were tested.
Amylose is a linear polysaccharide composed of a-1–4
linked glucose moieties, which, together with the highly
branched amylopectin forms the polysaccharide compo-
nent of starch. Native starch,1–3 derivatives of amylose
and cellulose,4–6 as well as cyclodextrins7 have been
used for chromatographic enantioseparation. Amylose
can adopt a helical conformation which is able to form
inclusion complexes with various guest compounds.8–12
Similar complexation mechanisms are known for
cyclodextrins,13 which have long been used for chiral
recognition and encapsulation of compounds.14–16 Car-
bamate derivatives of polysaccharides, adsorbed on or
covalently linked to solid supports are frequently used
for chiral separation.17,18 Silica-bound amylose carba-
mate was prepared by enzymatic synthesis followed by
immobilisation and derivatisation.19 However, inclusion
complexes of amylose have not yet been exploited for
chromatography. Here, synthesis and immobilisation of
oligosaccharide primers on porous silica,20 followed by
solid-phase enzymatic chain elongation, is reported.
The synthetic route ensured that silica-bound amylose
was attached to the support via its reducing end, allow-
2. Results and discussion
Bifunctional maltooligosaccharide-silane linkers were
synthesised from v-unsaturated carboxylic acids by
conversion to the corresponding hydrazides (Fig. 1).21
Keywords: immobilised amylose; phosphorolytic synthesis; mal-
tooligosaccharides; silica gel; silane linkers; hydrazides.
* Present address: Institut fu¨r Biochemie, Emil-Fischer-Zentrum,
Friedrich-Alexander-Universita¨t Erlangen-Nu¨rnberg, Fahrstraße
17, D-91054 Erlangen, Germany. Tel.: +49-(0)9131-852-6206; fax:
+49-(0)9131-852-2485; e-mail: hgb@biochem.uni-erlangen.de
Figure 1. Synthesis of v-unsaturated hydrazide linkers.
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)01293-5

6128
H.-G. Breitinger / Tetrahedron Letters 43 (2002) 6127–6131
2.1. Solid-phase amylose synthesis
Enzymatic synthesis of amylose using potato phospho-
rylase (EC 2.4.1.1) and glucose-1-phosphate allows gen-
eration of amylose polymers with a narrow molecular
weight distribution,28,29 a method which can also be
used for the generation of amylose-containing copoly-
mers by heterogeneous phosphorolytic synthesis.30,31
Similar to anionic polymerisation, a desired molecular
weight of the polymer can be defined via the ratio of
starter to monomer.32,28 Since the enzyme requires a
maltooligosaccharide starter of at least 4 glucose units
in length (Glc4), immobilised maltoheptaose was indeed
an efficient starter for polymer synthesis. Enzymatic
chain elongation of immobilised maltoheptaose primers
afforded silica-bound amylose (Fig. 3A).32
Variation of reaction time and amount of enzyme used
for synthesis gave an optimal duration of enzymatic
synthesis of 15 min, using 0.29 units33 of potato phos-
phorylase per mmol of immobilised maltoheptaose
starter. The total amount of immobilised polysaccha-
ride after phosphorolytic synthesis was ca. 6%, irrespec-
tive of the spacer length (Fig. 3B).34 The relative
increase in amylose content of the modified silica sup-
port was higher for the -(CH2)11- spacer, indicating that
although fewer immobilised primers were present on
the surface, these were better accessible for enzymatic
synthesis. Variation of the ratio of monomer (Glc-1-P)
to starter (immobilised maltoheptaose) revealed a satu-
ration behaviour, reaching a maximum at a theoretical
degree of polymerisation of ca. 100 for -(CH2)6-, and
ca. 250 for the -(CH2)11- spacer (Fig. 3B). This is
consistent with a more efficient amylose synthesis from
maltoheptaose molecules that are bound to the support
via the longer spacer. The saturation behaviour also
indicates that an equilibrium of amylose synthesis and
removal from the support by a disproportionation reac-
tion of the enzyme32 was being reached, regardless of
the amount of available monomer.
Figure 2. Synthesis and immobilisation of maltooligosaccha-
ride linkers.
2.2. Characterisation of immobilised amylose
The hydrazide linker had been introduced as a pre-
formed cleavage site to monitor the progress of solid
phase amylose synthesis. Under the conditions of syn-
thesis, coupling and deprotection, the hydrazide bond
was stable, but synthesised amylose could be cleaved
from the support by shaking the material in 0.2 M
citrate buffer at pH 4.0 at 60°C for 4 h (Fig. 3A). The
supernatant was directly tested by iodine complexation
and HPLC (Waters Ultrahydrogel 250, 0.2 M citrate
buffer pH 6.0, 0.5 ml/min, room temperature). Syn-
thetic amylose standards10,35 were used for calibration.
Solid-phase synthesised amylose was characterised by a
broad molecular weight distribution (Fig. 3C). Reduced
accessibility of immobilised primers, leading to varia-
tions in the rate of chain elongation on different sites
on the support, and beginning disproportionation may
account for the observed chain length distribution.
High molecular weight of support-generated amylose,
and the presence of unreacted primer indicated that
only few of the immobilised primer molecules were
utilised by the enzyme.
b-Cyclodextrin (b-CD) was cleaved by acid hydrolysis
to give linear maltoheptaose,22 which could be coupled
to the v-unsaturated hydrazides (Fig. 2A).23 OH-
groups were protected with trimethylsilyl chloride
(TMS-Cl) and hexamethyl disilazane,24 and the double
bond hydrosilylated25 using H-Si(OEt3) and bis-
cyclopentadienyl platinum chloride.26 These primers
were coupled to porous silica, TMS was removed by
treatment with dilute acid in methanol/water to give
silica-bonded primers for phosphorolytic amylose syn-
thesis (Fig. 2A).27
The (CH2)3-spacer resulted in inefficient amylose syn-
thesis (not shown), so only the (CH2)6-, and (CH2)11-
spacers were further tested. Among these, (CH2)6-
resulted in a more efficient immobilisation of the malto-
heptaose anchor group, giving a maximum spacer den-
sity of 72 mmol/g, compared to 22 mmol/g for the
11-C-spacer (Fig. 2B).

H.-G. Breitinger / Tetrahedron Letters 43 (2002) 6127–6131
6129
2.3. Chromatography36
Table 1. Chromatographic properties of modified silica
phases
Amylose-modified silica was suspended in water/0.2%
I2/KI solution added to induce formation of the helical
amylose–iodine inclusion complex prior to column
packing. Water/methanol (95:5, v/v) was used as
medium for packing and HPLC. Under these condi-
tions, complex formation ability of amylose was fully
preserved, while swelling of the material during column
Column material S 250 (base mat.)
S 250-(CH2)11amylose
Amylose (wt%)
Theor. plates
–
2700
6.2
2100
Compound
Capacity factor
Fenchone
2.16
1.94
1.94
1.51
1.02
1.73
2.10
2.18
1.39
1.09
D
-Menthol
L
-Menthol
2-Hexanone
Phenol
packing was minimised. The observed differences in the
elution sequence for complexands of starch and
cyclodextrin between base material and amylose-
modified silica indicated that bound amylose dominated
the separation process (Table 1). In the case of
-menthol, enantioseparation was observed. The chiral
D- and
L
separation factor of h=1.04, however, was not suffi-
cient for baseline separation. Note also that known
complexands of amylose were not retarded as strongly
as expected. Apparently, no inclusion complexes were
formed between analytes and immobilised amylose.
When amylose phases were removed from the columns
and probed with iodine solution, the typical blue colour
only became visible after 1–2 min, compared to instan-
taneous staining with materials that had not been used
in HPLC.
Thus, the conditions of HPLC column packing and use
(>25 bar) induced a conformational change of the
amylose, abolishing the low-density, helical V-confor-
mation. Retardation, therefore, was most likely due to
interactions on the carbohydrate surface and not com-
plex formation between analyte and amylose helices.
Amylose-grafted chromatography materials should thus
be useful in low and medium pressure chromatographic
applications, where helical inclusion should be the pre-
dominant separation mechanism.
Acknowledgements
Supported by the Federal Ministry of Research and
Technology, program ‘Renewable Resources’ and Eri-
dania-Begin-Say. Support and helpful discussions by
Professor Dr. G. Wulff are gratefully acknowledged.
References
1. Krebs, H.; Rasche, R. Z. Anorg. Allg. Chem. 1954, 276,
236.
2. Krebs, H.; Wagner, J. A.; Diewald, J. Chem. Ber. 1956,
89, 1875.
3. Hess, H.; Burger, G.; Musso, H. Angew. Chem. 1978, 90,
Figure 3. Generation and analysis of support-bound amylose.
645.

6130
H.-G. Breitinger / Tetrahedron Letters 43 (2002) 6127–6131
until pyridine-free.
4. Yashima, E.; Kasashima, E.; Okamoto, Y. Chirality
D
-Maltoheptaosyl-10-undecenoyl
1997, 9, 63.
hydrazone 5: Yield 94%, mp 158–200°C (dec.). Anal.
calcd for C52H93N2O36·H2O: C, 46.60; H, 7.09; N, 2.12.
Found: C, 46.77; H, 7.34; N, 2.25%.
5. Willems, J. G. H.; Duchateau, A. L. L.; Zwanenburg, B.
Chirality 1997, 9, 727.
6. Kubota, T.; Yamamoto, C.; Okamoto, Y. Chirality 2002,
14, 372.
7. Schurig, V.; Nowotny, H.-P. Angew. Chem., Int. Ed.
Engl. 1990, 29, 939.
D
-Maltoheptaosyl-5-hexenoyl hydrazone 6: Yield 98%,
mp 158–200°C (dec.). Anal. calcd for
C48H83N2O36·0.4N2H4: C, 45.15; H, 6.63; N, 3.07. Found:
C, 45.13; H, 7.09; N, 3.01%.
8. Yamashita, Y. J. Polymer Sci. A 1965, 3, 3251.
9. Winter, W. T.; Sarko, A. Biopolymers 1974, 13, 1461.
10. Wulff, G.; Kubik, S. Makromol. Chem. 1992, 193, 1071.
11. Wulff, G.; Breitinger, H.-G.; Kubik, S. Inklusionskom-
plexe von Sta¨rke und chemische Stabilisierung von
helikalen Strukturen der Amylose. In Nachwachsende
Rohstoffe-Polysaccharid Forschung, Ergebnisbericht zum
Abschluß des BMFT-Forschungsverbunds Polysaccharid
Forschung 1987–1993; BMFT: Bonn, 1993.
12. Kubik, S.; Ho¨ller, O.; Steinert, A.; Tolksdorf, M.; Wulff,
G. Macromol. Symp. 1995, 99, 93.
D
-Maltoheptaosyl-methacryloyl hydrazone 7: Yield 85%,
mp 170–220°C (dec.). Anal. calcd for C46H79N2O36: C,
44.70; H, 6.44; N, 2.27. Found: C, 44.80; H, 6.87; N,
2.31%.
24. Havill, T.; Joffe, L.; Post, L. J. Org. Chem. 1948, 13, 280.
Per-O-trimethylsilyl-D-maltoheptaosyl-10-undecenoyl
hydrazide: Yield 90%, mp 80–83°C, H NMR (CDCl3): l
4.93–5.32 (m, 9H), 3.19–4.21 (m, broad, 42H), 1.18–2.27
(m, broad, 16H), 0.14 (s, Si-CH6 3). Anal. calcd for
C122H277N2O36Si23: C, 48.93; H, 9.32; N, 0.94. Found: C,
48.62; H, 9.54; N, 0.93%.
1
13. Wenz, G.; Wolf, F.; Wagner, M.; Kubik, S. New J.
Per-O-trimethylsilyl-D-maltoheptaosyl-5-hexenoyl hydra-
Chem. 1993, 17, 729.
14. Breslow, R.; Zhang, B. J. Am. Chem. Soc. 1996, 118,
8495.
15. Breslow, R.; Dong, S. D. Chem. Rev. 1998, 98, 1997.
16. Zhang, B.; Breslow, R. J. Am. Chem. Soc. 1997, 119,
1676.
17. Ward, T. M. Anal. Chem. 2000, 72, 4521.
18. Szymura-Oleksiak, J.; Bojarski, J.; Aboul-Enein, H. Y.
Chirality 2002, 14, 417.
zide: Yield 84%, mp 91–94°C, 1H NMR (CDCl3): l
4.82–5.27 (m, 9H), 3.19–4.21 (m, broad, 42H), 1.82–2.47
(m, broad, 6H), 0.14 (s, Si-CH6 3). Anal. calcd for
C117H267N2O36Si23: C, 48.05; H, 9.20; N, 0.96. Found: C,
47.75; H, 9.24; N, 1.30%.
Per-O-trimethylsilyl-D-maltoheptaosyl-methacryloyl
hydrazide: Yield 79%, mp 123–128°C, H NMR (CDCl3):
l 4.95–5.23 (m, 9H), 3.26–4.18 (m, broad, 42H), 2.1 (2,
1
2H), 0.14 (s, Si-CH6 3, theor. 484H). Anal. calcd for
C115H263N2O36Si23: C, 47.69; H, 9.15; N, 0.97. Found: C,
48.13; H, 9.60; N, 1.05%.
19. Enomoto, N.; Furukawa, S.; Ogasawara, Y.; Akano, H.;
Kawamura, Y.; Yashima, E.; Okamoto, Y. Anal. Chem.
1996, 68, 2798.
20. Blaschke, G.; Fraenkl, W.; Bro¨ker, W. Angew. Chem.
25. Djuric, S.; Venit, J.; Magnus, P. Tetrahedron Lett. 1981,
22, 1787. Appearance of the O-CH2-CH3 signal was
monitored by NMR.
1986, 98, 808.
21. Purgett, M. D.; Xie, S.; Bansleben, D. A.; Vogl, O. J.
Pol. Sci. A 1988, 26, 665. Melting points (uncorrected):
Bu¨chi 510; NMR: Varian EM 390, TMS as internal
standard; CHN analyses: Pharmazeutical Institut, U.
Du¨sseldorf.
Per-O-trimethylsilyl-D-maltoheptaosyl-ꢀ-(triethoxysilyl)-
n-undecanoyl hydrazide 8: mp 80–83°C, 1H NMR
(CDCl3): l 4.93–5.32 (m, 9H), 3.19–4.21 (m, broad, 42H),
3.68 (q, -O-CH2
6
), 1.18–2.27 (m, broad, 16H), 1.17 (t,
). Anal. calcd for
1
10-Undecenoylhydrazide 1: Yield: 97%, mp 86–87°C, H
O-CH2-CH3), 0.14 (s, Si-CH3
6
6
NMR (CDCl3): l 6.92–7.22 (s, 1H, broad, removed with
D2O), 5.54–6.05 (m, 1H), 4.80–5.11 (m, 2H), 3.60–3.93 (s,
2H, broad, removed with D2O), 1.82–2.26 (m, 6H), 1.10–
1.83 (m, 12H). Anal. calcd for C11H22N2O: C, 66.62; H,
11.18; N, 14.13. Found: C, 66.96; H, 11.75; N, 14.11%.
5-Hexenoylhydrazide 2: Yield: 81%, mp 46–47°C, 1H
NMR (CDCl3): l 7.25–7.63 (s, 1H, broad, removed with
D2O), 5.51–6.02 (m, 1H), 4.77–5.13 (m, 2H), 3.66–4.06 (s,
2H, broad, removed with D2O), 1.55–2.31 (m, 6H). Anal.
calcd for C6H12N2O: C, 56.32; H, 9.44; N, 21.86. Found:
C, 56.99; H, 10.03; N, 20.85%.
C128H293N2O39Si24: C, 48.67; H, 9.35; N, 0.89. Found: C,
48.14; H, 9.16; N, 0.99%.
Per-O-trimethylsilyl-D-maltoheptaosyl-ꢀ-(triethoxysilyl)-
1
n-hexanoyl hydrazide 9: mp 91–94°C, H NMR (CDCl3):
l 4.82–5.27 (m, 9H), 3.18–4.21 (m, broad, 42H), 3.65 (q,
-O-CH2
0.14 (s, Si-CH3
0.91. Found: C, 46.56; H, 9.20; N, 1.02.
6
), 1.85–2.43 (m, broad, 6H), 1.19 (t, O-CH2-CH3
6 ),
6
); C123H283O39N2Si24: C, 47.83; H, 9.24; N,
Per-O-trimethylsilyl-D-maltoheptaosyl-ꢀ-(triethoxysilyl)-
(1-methyl)-n-propionyl hydrazide 10: mp 123–128°C, 1H
NMR (CDCl3): l 4.95–5.23 (m, 9H), 3.26–4.18 (m,
22. Maltoheptaose 4: 125
g of b-cyclodextrin (Avebe,
broad, 42H), 3.69 (q, -O-CH2
6
), 1.30–2.10 (m, broad,
). Anal. calcd
Krefeld, Germany) were heated to reflux for 2 h in 500 ml
of 0.01 N HCl. After neutralisation, 1 ml 1 M phosphate
buffer pH 7.0 was added and the solution stored at 4°C
overnight. Unreacted b-cyclodextrin was removed as p-
xylene complex. Repeated precipitation from water/etha-
nol and precipitation with acetone gave 4, RF 0.19
(n-butanol/methanol/water 4:3:3).
16H), 1.17 (t, O-CH2-CH3), 0.14 (s, Si-CH3
6
6
for C121H279N2O39Si24: C, 47.49; H, 9.19; N, 0.92. Found:
C, 46.80; H, 9.23; N, 1.34%.
26. (a) Chatt, J.; Vallarino, L. M.; Venzani, L. M. J. Chem.
Soc. 1957, 2496; (b) Kharash, M. S.; Ashford, T. A. J.
Am. Chem. Soc. 1936, 58, 1733.
27. Silica 250 K (Amicon, Hamburg, Germany) was neu-
tralised with HNO3. 1 mmol of anchor per m2 silica
surface was added and the suspension gently shaken at
60°C for 72 h. TMS was removed by repeatedly shaking
23. 5 mmol of maltoheptaose, 4, were dissolved in 40 ml of
dry pyridine; 15 mmol of v-alkenyl hydrazide were added
and the mixture kept at 60°C for 48 h. Solvent was
removed and the solid residual washed with ethyl acetate

H.-G. Breitinger / Tetrahedron Letters 43 (2002) 6127–6131
6131
modified silica in 10 ml/g of 0.5 % acetic acid in
methanol/water (1:1) for 1 h.
Schupp, E. Makromol. Chem. 1958, 86, 181.
34. Hodge, J. E.; Hofreiter, B. T. Anthrone test. In Methods
in Carbohydrate Chemistry; Whistler, R. L., Ed.; Wiley:
New York, 1962; Vol. 1; pp. 380. E625 was corrected for
turbidity due to dispersed silica particles.
35. Wulff, G.; Kubik, S.; Breitinger, H.-G. In Nachwachsende
Rohstoffe; Eggersdorfer, M.; Warwel, S.; Wulff, G., Eds.
Charakterisierung von Amylosekomplexen und chemis-
che Stabilisierung von helikalen Strukturen der Amylose;
VCH: Weinheim, 1993.
36. HPLC columns (0.4×25 cm) were packed using a Knauer
pneumatic dual piston pump (H. Knauer GmbH, Berlin,
Germany), and Knauer Vertex cartridges. The HPLC
setup consisted of a Waters 6000 pump, Latek 7125
Rheodyne injector, Waters R401 refractory index detec-
tor (Waters GmbH, Eschborn, Germany), and a Hewlett-
Packard 3390 A recorder/integrator (Hewlett-Packard,
Ratingen, Germany); flow rate 0.5 ml/min, pressure 25
bar.
28. Saenger, W.; Noltemeyer, N.; Manor, P. C.; Hingerty, B.;
Klar, B. Bioorg. Chem. 1976, 5, 187. Crude enzyme
extract of 20–30 U/ml could be stored at 4°C under
toluene for up to 6 months without significant loss of
activity.
29. Kadokawa, J.-I.; Kaneko, Y.; Nakaya, A.; Tagaya, H.
Macromolecules 2001, 34, 6536.
30. Pfannemu¨ller, B.; Potratz, C. Starch-Staerke 1976, 29, 73.
31. Akiyoshi, K.; Kohara, M.; Ito, K.; Kitamura, S.;
Sunamoto, J. Macromol. Rapid Commun. 1999, 20, 112.
32. Pfannemu¨ller, B.; Burchard, W. Makromol. Schem. 1969,
121, 1. 1 g of maltoheptaose-modified silica was sus-
pended in 6 ml of 0.2 M citrate buffer at pH 6.0, 15 ml
of Glc-1-P solution and 10 units of phosphorylase were
added. The mixture was shaken at 40°C for 15 min,
washed with water, ethanol, acetone, and dried.
33. Husemann, E.; Fritz, B.; Lippert, R.; Pfannemu¨ller, B.;