Self-assembly of β-glucosidase and D-glucose-tethering zeolite crystals into fibrous aggregates

Article abstract of DOI:10.1021/ja0028222

β-Glucosidase and D-glucose-tethering micrometer-sized zeolite crystals self-assemble into thin (2-20 μm) and very long (>1 cm) fibrous aggregates in water. The process proceeds at a faster rate in a buffer solution of pH 4.8 at which the enzymatic activity is highest. The zeolite and enzyme remain intact within the fibrous material. Furthermore, the enzymatic activity of β-glucosidase is preserved even after they are kept in water for more than 6 months at room temperature. With the zeolite to enzyme weight ratio of 5, all the zeolite crystals are buried within the round fibrils which consist of either a single strand or helical double strands. Upon increasing the ratio to 10, clusters of unburied zeolite crystals appear on the exterior of the fibrils, while narrow flat fibers with smooth surfaces are formed upon decreasing the ratio to 2.5. The process is proposed to initiate by the tight binding between the zeolite-bound D-glucose moieties and β-glucosidase followed by crystallization of the enzyme over the zeolite-bound enzyme monolayer. This report thus reveals a novel behavior of β-glucosidase and demonstrates an unprecedented phenomenon that an enzyme and its substrate-tethering inorganic crystals self-assemble into structured aggregates.

Full text of DOI:10.1021/ja0028222

J. Am. Chem. Soc. 2000, 122, 12151-12157
12151
Self-Assembly of â-Glucosidase and D-Glucose-Tethering Zeolite
Crystals into Fibrous Aggregates
Goo Soo Lee, Yun-Jo Lee, So Yeun Choi, Yong Soo Park, and Kyung Byung Yoon*
Contribution from the Center for Microcrystal Assembly and Department of Chemistry, Sogang UniVersity,
Seoul 121-742, Korea
ReceiVed July 31, 2000
Abstract: â-Glucosidase and D-glucose-tethering micrometer-sized zeolite crystals self-assemble into thin
(2-20 µm) and very long (>1 cm) fibrous aggregates in water. The process proceeds at a faster rate in a
buffer solution of pH 4.8 at which the enzymatic activity is highest. The zeolite and enzyme remain intact
within the fibrous material. Furthermore, the enzymatic activity of â-glucosidase is preserved even after they
are kept in water for more than 6 months at room temperature. With the zeolite to enzyme weight ratio of 5,
all the zeolite crystals are buried within the round fibrils which consist of either a single strand or helical
double strands. Upon increasing the ratio to 10, clusters of unburied zeolite crystals appear on the exterior of
the fibrils, while narrow flat fibers with smooth surfaces are formed upon decreasing the ratio to 2.5. The
process is proposed to initiate by the tight binding between the zeolite-bound D-glucose moieties and
â-glucosidase followed by crystallization of the enzyme over the zeolite-bound enzyme monolayer. This report
thus reveals a novel behavior of â-glucosidase and demonstrates an unprecedented phenomenon that an enzyme
and its substrate-tethering inorganic crystals self-assemble into structured aggregates.
Introduction
special attention has been directed to those that involve
biologically important interactions such as DNA base pairing
and antigen-antibody complexation to gain insights into the
natural biological processes.1-7 However, the enzyme-substrate
complexation, which is one of the most important biological
interactions, has not yet been explored. In particular, the
examples of self-assembly of substances into structured ag-
gregates are very rare.1-3 Stemming from our interests in
organizing zeolite crystals over the surfaces by covalent
linkages,21-26 we have discovered a novel phenomenon that
â-glucosidase and the D-glucose-tethering zeolite crystals readily
self-assemble into thin and long fibrous aggregates. Interestingly,
the enzymatic activity of â-glucosidase in the fibrous composite
material is preserved even after they were stored in water for 6
months at room temperature. This paper reports the unprec-
edented, highly intriguing phenomenon of self-assembly of an
enzyme and its substrate-tethering micrometer-sized inorganic
crystals into structured entities.
Self-assembly is a process in which atoms, molecules, and
systems of molecules arrange themselves into structured func-
tioning entities without human intervention and the most
elaborate expression of the process is demonstrated in life
systems.1,2 Studies have shown that self-assembly occurs among
numerous substances of all length scales.1-20 Among these,
(1) Whitesides, G. M. Sci. Am. 1995, 273, 146-149.
(2) Ozin, G. A. Chem. Commun. 2000, 419-432.
(3) Boal, A. K.; Ilhan, F.; DeRouchey, J. E.; Thurn-Albrecht, T.; Russell,
T. P.; Rotello, V. M. Nature 2000, 404, 746-748.
(4) Shenton, W.; Davis, S. A.; Mann, S. AdV. Mater. 1999, 11, 449-
451.
(5) Li, M.; Wong, K. K. W.; Mann, S. Chem. Mater. 1999, 11, 23-26.
(6) Alivisatos, A. P.; Johnsson, K. P.; Peng, X.; Wilson, T. E.; Loweth,
C. J.; Bruchez, M. P., Jr.; Schultz, P. G. Nature 1996, 382, 609-611.
(7) (a) Taton, T. A.; Mucic, R. C.; Mirkin, C. A.; Letsinger, R. L. J.
Am. Chem. Soc. 2000, 122, 6305-6306. (b) Storhoff, J. J.; Mirkin, C. A.
Chem. ReV. 1999, 99, 1849-1862. (c) Mirkin, C. A.; Letsinger, R. L.;
Mucic, R. C.; Storhoff, J. J. Nature 1996, 382, 607-609. (d) Elghanian,
R.; Storhoff, J. J.; Mucic, R. C.; Letsinger, R. L.; Mirkin, C. A. Science
1997, 277, 1078-1081.
Experimental Section
(8) Li, M.; Schnablegger, H.; Mann, S. Nature 1999, 402, 393-395.
(9) Murray, C. B.; Kagan, C. R.; Bawendi, M. G. Science 1995, 270,
1335-1338.
Materials. â-Glucosidase (lot number 37H0947, recombinant) and
p-nitrophenyl â-D-glucopyranoside (PNPG) were purchased from
Sigma. D-Glucose, D-cellobiose, ω-undecylenyl bromide, 4-hydroxy-
(10) Andres, R. P.; Bielefeld, J. D.; Henderson, J. I.; Janes, D. B.;
Kolagunta, V. R.; Kubiak, C. P.; Mahoney, W. J.; Osifchin, R. G. Science
1996, 273, 1690-1693.
(19) Bowden, N.; Choi, I. S.; Grzybowski, B. A.; Whitesides, G. M. J.
Am. Chem. Soc. 1999, 121, 5373-5391.
(11) Sun, S.; Murray, C. B.; Weller, D.; Folks, L.; Moser, A. Science
2000, 287, 1989-1992.
(20) Choi, I. S.; Bowden, N.; Whitesides, G. M. J. Am. Chem. Soc. 1999,
121, 1754-1755.
(12) Stupp, S. I.; LeBonheur, V.; Walker, K.; Li, L. S.; Huggins, K. E.;
Keser, M.; Amstutz, A. Science 1997, 276, 384-389.
(13) Petit, C.; Taleb, A.; Pileni, M.-P. AdV. Mater. 1998, 10, 259-262.
(14) Motte, L.; Billoudet, F.; Lacaze, E.; Pileni, M.-P. AdV. Mater. 1996,
8, 1018-1020.
(21) Kulak, A.; Lee, Y.-J.; Park, Y. S.; Yoon, K. B. Angew. Chem., Int.
Ed. 2000, 39, 950-953.
(22) Choi, S. Y.; Lee, Y.-J.; Park, Y. S.; Ha, K.; Yoon, K. B. J. Am.
Chem. Soc. 2000, 122, 5201-5209.
(23) Ha, K.; Lee, Y.-J.; Lee, H. J.; Yoon, K. B. AdV. Mater. 2000, 12,
1114-1117.
(15) Brust, M.; Bethell, D.; Schiffrin, D. J.; Kiely, C. J. AdV. Mater.
1995, 7, 795-797.
(24) Lee, G. S.; Lee, Y.-J.; Ha, K.; Yoon, K. B. Tetrahedron 2000, 56,
6965-6968.
(16) Wang, Z. L. AdV. Mater. 1998, 10, 13-30.
(17) Bowden, N.; Terfort, A.; Carbeck, J.; Whitesides, G. M. Science
1997, 276, 233-235.
(25) Ha, K.; Lee, Y.-J.; Jung, D.-Y.; Lee, J. H.; Yoon, K. B. AdV. Mater.
2000, 12, 1614-1617.
(18) Weck, M.; Choi, I. S.; Jeon, N. L.; Whitesides, G. M. J. Am. Chem.
Soc. 2000, 122, 3546-3547.
(26) Kulak, A.; Park, Y. S.; Lee, Y.-J.; Chun, Y. S.; Ha, K.; Yoon, K.
B. J. Am. Chem. Soc. 2000, 122, 9308-9309.
10.1021/ja0028222 CCC: $19.00 © 2000 American Chemical Society
Published on Web 11/22/2000

12152 J. Am. Chem. Soc., Vol. 122, No. 49, 2000
Lee et al.
Preparation of 1,2,3,4-Tetra-O-acetyl-â-D-glucose (TAG, 3).
Anhydrous R-D-glucose (12.0 g, 67 mmol) and triphenylmethyl chloride
(19.3 g 69 mmol) were dissolved in anhydrous pyridine (50 mL) by
warming on a steam bath. The yields depend on the dryness of the
reagents and containers. Acetic anhydride (30 mL) was poured in one
portion into the pyridine solution without cooling. The solution was
allowed to cool to room temperature and kept at that temperature for
an additional 12 h. The above mixture was then introduced in a thin
stream into the mixture of ice water (950 mL) and acetic acid (50 mL)
while rapidly swirling the container. After 2 h of vigorous agitation
the precipitate was isolated by filtration. To remove residual pyridine,
the gummy filter cake was immediately introduced into fresh ice water
(1 L) and the mixture was stirred for an additional 10 min. White
precipitate was isolated by filtration and subsequently washed with
copious amounts of fresh water. As a means to preferentially remove
the R-isomer, which has higher solubility in diethyl ether, the crude,
air-dried product was then introduced into diethyl ether (50 mL). The
less soluble â-isomer was isolated and recrystallized from hot 95%
ethanol. Upon cooling, fine needle forms of crystalline 1,2,3,4-tetra-
O-acetyl-6-O-triphenylmethyl-â-D-glucose were liberated. C33H34O10
590.62, yield 17 g (43%), mp 166-167 °C. The above tritylated
compound (7.4 g, 12.5 mmol) was dissolved in glacial acetic acid (35
mL) by warming on a water bath. The solution was cooled to 10 °C,
then the glacial acetic acid solution of hydrogen bromide (33%, 2.7
mL) was added and the mixture was shaken for 45 s. The liberated
triphenylmethyl bromide was immediately removed by suction filtration,
and the filtrate was poured in one portion into ice water (150 mL).
The produced glucose tetraacetate (TAG) was extracted with chloroform
(4 × 40 mL). The extract was washed with cold water (15 mL × 4) to
remove residual acetic acid and then dried over anhydrous magnesium
sulfate. The solvent in the extract was stripped off in vacuo. Anhydrous
diethyl ether (15 mL) was introduced onto the syrupy residue.
Crystallization began immediately upon rubbing the syrup using a glass
rod. The compound slowly decomposes above -10 °C. C14H20O10
348.30, yield 3.4 g (78%), white solid, mp 128-129 °C, 1H NMR
(CDCl3, δ/ppm) 5.74 (d, 1H), 5.32 (t, 1H), 5.10 (m, 2H), 3.77 (m,
1H), 3.57-3.69 (m, 2H), 2.12 (s, 3H), 2.08 (s, 3H), 2.05 (s, 3H), 2.04
(s, 3H).
Scheme 1
benzaldehyde, sodium chlorite, and hydrogen hexachloroplatinate were
purchased from Aldrich and were used as received. Trimethoxysilane
(Aldrich) was distilled and kept in a Schlenk storage flask under argon.
All the organic solvents were purified according to the well-known
literature procedures prior to use. Reflectoquant, a qualitative test strip
for D-glucose, was purchased from Merck. A buffer solution of pH 4.8
was prepared from a 1:1 mixture of acetic acid (0.01 M) and sodium
acetate (0.01 M).
Zeolite-A (Na+ form) and ZSM-5 (Si/Al ) 100) were synthesized
according to the literature procedures.27 The average size of zeolite-A
used in this study was ∼0.4 µm and that of ZSM-5 was ∼1.5 × 1.1 ×
0.6 µm3. The template ions such as tetramethylammonium (TMA+)
and tetrapropylammonium (TPA+) employed during synthesis of
zeolite-A and ZSM-5, respectively, were removed by calcining at 550
°C for 12 h under flowing oxygen prior to treatment of the zeolite
crystals with the following silylating reagent (vide infra).
Preparation of 1,2,3,4-Tetra-O-acetyl-6-O-{4-[11-(trimethoxy-
silyl)undecyloxy]benzoyl}-â-D-glucose (TAG-BU-TMS). As a means
to tether D-glucose to zeolite crystals, TAG-BU-TMS was prepared
according to Scheme 1 as described in detail in the following steps.
Preparation of 4-(10-Undecenyloxy)benzaldehyde (1). 11-Bromo-
1-undecene (2.19 mL, 2.33 g, 10 mmol), 4-hydroxybenzaldehyde (1.22
g, 10 mmol), tetrabutylammonium iodide (catalytic amount, 0.10 g),
and potassium carbonate (6.90 g, 50 mmol) were introduced into acetone
(50 mL). After refluxing for 6 h, the mixture was allowed to cool to
room temperature and the solids were filtered off. The filtrate was
concentrated in vacuo and the crude product of 1 was purified by
column chromatography (ethyl acetate:hexane ) 1:4). C18H26O2 274.40,
Preparation of 1,2,3,4-Tetra-O-acetyl-6-O-[4-(10-undecenyloxy)-
benzoyl]-â-D-glucose (4). The acid 2 (1.91 g, 6.6 mmol), TAG 3 (2.30
g, 6.6 mmol), dicyclohexylcarbodiimide (DCC, 2.72 g, 13.2 mmol),
and 4-(N,N-dimethylamino)pyridine (DMAP, 0.40 g, 3.30 mmol) were
dissolved in dichloromethane (60 mL), and the solution was stirred at
room temperature for 3 h. Acetic acid (5 mL) and methanol (5 mL)
were successively added into the mixture to destroy the excess DCC.
The mixture was additionally stirred at 45 °C for 30 min. After cooling
to room temperature, the mixture was filtered through a silica pad. The
filtrate was concentrated in vacuo and the crude product was purified
by column chromatography (ethyl acetate:hexane ) 1:4). The product
was further purified by recrystallization from methanol. C32H44O12
1
1
yield 2.33 g (85%), colorless oil, H NMR (CDCl3, δ/ppm) 9.87 (s,
620.69, yield (2.59 g, 63%), white solid, mp 99.5-100 °C, H NMR
1H), 7.82 (d, 2H), 6.98 (d, 2H), 5.80 (m, 1H), 4.96 (m, 2H), 4.03 (t,
2H), 2.03 (m, 2H), 1.81 (m, 2H), 1.31-1.48 (12H).
(CDCl3, δ/ppm) 7.99 (d, 2H), 6.91 (d, 2H), 5.84 (m, 1H), 5.77 (d, 1H),
5.13-5.33 (m, 3H), 4.95 (m, 2H), 4.32-4.49 (m, 2H), 4.01 (t, 2H),
3.98 (m, 1H), 2.09 (s, 3H), 2.02 (s, 3H), 2.00 (s, 6H), 1.78 (m, 2H),
1.30-1.44 (14H). 13C NMR (CDCl3, δ/ppm) 169.99, 169.21, 169.17,
168.80, 165.75, 163.20, 139.11, 131.79, 121.58, 114.13, 114.06, 91.71,
72.87, 72.78, 70.31, 68.16, 61.86, 33.71, 29.40, 29.32, 29.25, 29.02,
28.84, 25.90, 20.70, 20.48.
Preparation of 4-(10-Undecenyloxy)benzoic Acid (2). The alde-
hyde 1 (2.00 g, 7.29 mmol) and resorcinol (1.05 g, 9.50 mmol) were
dissolved in tert-butyl alcohol (140 mL). Independently, sodium chlorite
(3.80 g) and sodium dihydrogenphosphate (3.04 g) were dissolved in
water (30 mL). The aqueous solution was added in a dropwise manner
into the tert-butyl alcohol solution over a 10-min period. The pale
yellow reaction mixture was then stirred at room temperature overnight.
Volatile components were removed in vacuo and the residue was
dissolved in water (100 mL). The aqueous solution was acidified to
pH 3 by adding 1 N aqueous HCl. The liberated white precipitate was
isolated, washed successively with water and hexane, and dried in the
air. C18H26O3 290.40, yield 1.97 g (93%), white solid, mp 79.5-80
°C, 1H NMR (CDCl3, δ/ppm) 8.05 (d, 2H), 6.92 (d, 2H), 5.80 (m, 1H),
4.96 (m, 2H), 4.02 (t, 2H), 2.05 (m, 2H), 1.81 (m, 2H), 1.31-1.46
(12H).
Preparation of TAG-BU-TMS. Anhydrous toluene (5 mL) was
introduced into a round-bottomed flask containing the olefin 4 (620
mg, 1 mmol) and hydrogen hexachloroplatinate (catalytic amount, ∼0.1
mg) under argon. Trimethoxysilane (0.3 mL, 293 mg, 2.4 mmol) was
introduced into the mixture using a hypodermic syringe and the mixture
was stirred at 42 °C for 15 h. After cooling to room temperature, the
mixture was immediately filtered through double pads of silica and
charcoal by applying argon pressure. Upon concentrating the filtrate,
TAG-BU-TMS was obtained as an extremely hygroscopic solid.
C35H54O15Si 742.88, quantitative yield, 1H NMR (CDCl3, δ/ppm) 7.98
(d, 2H), 6.91 (d, 2H), 5.75 (d, 1H), 5.13-5.33 (m, 3H), 4.33-4.48
(m, 2H), 4.00 (t, 2H), 3.97 (m, 1H), 3.56 (s, 9H), 2.10 (s, 3H), 2.03 (s,
3H), 2.01 (s, 6H), 1.79 (m, 2H), 1.28-1.42 (14H). 13C NMR (CDCl3,
(27) (a) Zhu, G.; Qui, S.; Yu, J.; Sakamoto, Y.; Xiao, F.; Xu, R.; Terasaki,
O. Chem. Mater. 1998, 10, 1483-1486. (b) U.S. Patent 3702886, 1972.
(c) U.S. Patent 4061724, 1977.

â-Glucosidase and D-Glucose-Tethering Zeolite Crystals
J. Am. Chem. Soc., Vol. 122, No. 49, 2000 12153
Scheme 2
δ/ppm) 170.07, 169.28, 169.24, 168.88, 165.83, 163.28, 131.85, 121.62,
114.19, 114.06, 91.78, 72.93, 72.84, 70.36, 68.23, 61.93, 50.45, 33.10,
29.56, 29.52, 29.47, 29.35, 29.22, 29.10, 25.97, 22.58, 20.76, 20.53,
20.52, 9.15.
allowed to proceed for 30 min at 30 °C. A Na2CO3/NaHCO3 buffer
solution (5 mL, 50 mM, pH ) 10.2) was subsequently introduced into
the PNPG solution to quench the activity of the enzyme. The released
p-nitrophenol was monitored spectrophotometrically at 400 nm using
the molar extinction coefficient of 14 130 M-1cm-1 29
For the long-
.
Attachment of TAG Groups on Zeolite Surface (Step A in Scheme
2). Dry toluene (50 mL) was introduced into a round-bottomed Schlenk
flask containing freshly dried zeolite-A (100 mg) or ZSM-5 (100 mg)
under a counter flow of argon. The toluene slurry was sonicated for 10
min to disperse the zeolite crystals into the solution. Independently,
dry toluene (10 mL) was introduced into a round-bottomed Schlenk
flask containing TAG-BU-TMS (150 mg) under an argon flow. The
TAG-BU-TMS solution was transferred to the sonicated toluene slurry
under argon. The mixture was refluxed for 1 h under argon atmosphere.
After cooling to room temperature, the TAG-BU-tethering zeolite
powders were collected by rapid filtration in the atmosphere over a
filter paper. The filtered powders were washed successively with fresh
toluene (100 mL) and ethanol (100 mL). The collected powders were
dried in an oven at 120 °C for 30 min.
Deacetylation of TAG Groups Tethered to Zeolite Crystals. Dry
benzene (100 mL) was introduced into a round-bottomed Schlenk flask
containing the TAG-tethering zeolite-A or ZSM-5 (100 mg) under
argon. 1,8-Diazabicyclo[5.4.0]undec-7-ene (DBU, 0.3 mL), a reagent
known to selectively cleave the acetyl groups from glucose,28 was added
to the zeolite slurry in benzene and the mixture was refluxed for 4 h
under argon. After cooling to room temperature, the glucose-coated
zeolite powders were collected by rapid filtration in the atmosphere
over a filter paper. The filtered powders were washed successively with
fresh toluene (100 mL) and ethanol (100 mL). The collected powders
were dried in an oven at 120 °C for 30 min (step B in Scheme 2).
Formation of Fibrous Zeolite-â-Glucosidase Composite Material.
â-Glucosidase (2 mg) was introduced into distilled deionized water
(20 mL) dispersed with TAG-tethering zeolite-A (10 mg). The slurry
was stirred with the aid of a small magnetic stirrer at room temperature.
A Teflon slab with four supporting legs was introduced into the reaction
vessel to collect the precipitating fibrous aggregates upon it. The added
amounts of enzyme were also varied from 2 to 1 and 4 mg to see the
effect of the zeolite/enzyme weight ratio. The fibrillation was also tested
in an acetate buffer solution of pH 4.8 to check the effect of the pH,
at which the enzymatic activity of â-glucosidase is known to be highest,
on the rate of fibrillation.
term activity tests, the composite fibers were stored in fresh water and
the activity was occasionally tested by employing the more convenient
Reflectoquant strips. Since the composite fiber is highly fluffy, the
volume of even 1 mg was large enough to weigh accurately using a
microbalance.
Instrumentation. The scanning electron microscope (SEM) images
of zeolites, crystalline â-glucosidase, and the zeolite-â-glucosidase
composite fibers were obtained from a FE-SEM (Hitachi S-4300) at
an acceleration voltage of 10 to 20 kV. On top of the samples platinum/
palladium alloy (in the ratio of 8 to 2) was deposited with a thickness
of about 15 nm. The X-ray diffraction patterns for the identification of
the zeolite crystals were obtained from a Rigaku diffractometer (D/
MAX-1C) with the monochromated beam of Cu KR. The UV-vis
spectra of the samples were recorded on a Shimadzu UV-3101PC. The
diffuse reflectance UV-vis spectra of solid samples were obtained using
an integrating sphere. FT-IR spectra were recorded on a JASCO FT/
1
IR-620. The solution H and 13C NMR spectra were recorded on a
Varian Jemini 300 NMR spectrometer. 13C, 27Al, and 29Si MAS solid-
state NMR spectra were recorded on a Varian 300 Unity Inova NMR
spectrometer. Freeze-drying of the fibrous composite materials was
conducted on a LABCONCO Freezone 6.
Results and Discussion
I. Confirmation of D-Glucose Tethered to Zeolite Crystals.
The ready attachment of TAG via the BU-SiO spacer to the
zeolite crystals was confirmed by the close matching between
the intense diffuse-reflectance UV-vis band of the TAG-BU-
TMS-treated zeolite crystals centered at 260 nm (Figure 1) and
the corresponding absorption of TAG-BU-TMS in ethanol at
259 nm (shown in the inset), due to the presence of a phenyl
moiety in both cases. The appearance of a carbonyl stretching
band at 1762 cm-1 in the diffuse-reflectance FT-IR spectrum
of the zeolite crystals treated with TAG-BU-TMS further
confirmed the ready attachment of TAG groups to the zeolite
crystals via siloxane linkages. Consistent with this, the TAG-
tethering zeolite crystals became very hydrophobic and conse-
quently floated on water.
Test of Enzymatic Activity of Fibrous Composite Fiber. The
composite fiber (1 mg) and D-cellobiose (10 mg) were introduced into
water (10 mL) and the heterogeneous mixture was allowed to equilibrate
for 24 h. A strip of Reflectoquant was dipped into the solution for 10
s. After removal from the solution, the strip was dried for a visual
color test. Alternatively, the composite fiber (1 mg) was added into 1
mL of the citrate-phosphate buffer solution (100 mM, pH 5) dissolved
with PNPG (15 mM).29 The fibril-induced hydrolysis of PNPG was
Restoration of hydroxyl groups on the glucose units was first
confirmed by the appearance of broad multiple hydroxyl
stretching bands in the 3000-3600 cm-1 region of the diffuse-
(29) Riccio, P.; Rossano, R.; Vinella, M.; Domizio, P.; Zito, F.;
Sansevrino, F.; D’Elia, A.; Rosi, I. Enzyme Microb. Technol. 1999, 24,
123-129.
(28) Baptistella, L. H. B.; Santos, J. F.; Ballabio, K. C.; Marsaioli, A. J.
Synthesis 1989, 436-438.

12154 J. Am. Chem. Soc., Vol. 122, No. 49, 2000
Lee et al.
Figure 1. Diffuse-reflectance UV-vis spectra of zeolite-A tethered
with TAG (s), D-glucose (- - -), and none (‚‚‚). The inset shows the
spectrum of TAG-BU-TMS in ethanol.
reflectance FT-IR spectrum of the DBU-treated zeolite powders
(Figure 2C). The BU-SiO spacer being intact after treatment
with DBU was confirmed from the presence of the 260-nm band
(due to the phenyl moiety) in the diffuse reflectance UV-vis
spectrum (dashed curve in Figure 1). The positive periodate test
for vicinal diols and the Benedict’s test for formyl groups further
confirmed the presence of glucose units on the surface.
Consistent with the restoration of hydroxyl groups, the DBU-
treated zeolite powders turned hydrophilic and as a result they
readily dispersed into water.
II. Formation of Fibrous Zeolite-â-Glucosidase Composite
Material. The aqueous slurry of D-glucose-tethering zeolite-A
powders remained turbid for an unlimited period of time as long
as the solution was kept stirred. Upon addition of â-glucosidase,
however, the turbid solution gradually turned clear while a
fibrous material slowly developed in the solution, in particular,
over the immersed Teflon support. After 2 days of stirring, the
solution turned completely transparent and the produced fibrous
material became visually apparent as shown in Figure 3.
The presence of Teflon support was not necessary for the
above fibrillation to occur. However, it helped the fibrils grow
into a larger aggregate without being chopped into smaller
fragments by the underlying spinning magnetic bar. Even the
magnetic stirring bar was not necessary for the fibrillation to
proceed. Any methods to maintain dispersion of the micro
zeolite crystals into the solution are equally effective. The nature
of the container (glass or plastic) also did not affect the
fibrillation process and the fibrils did not grow on the walls of
the containers, indicating that the container walls do not serve
as the seeds or anchoring spots for the fibrils to grow. The
density of the resulting fibrils was apparently larger than that
of water hence readily submerged onto the bottom of the
containers.
Figure 2. Diffuse-reflectance FT-IR spectra of solid TAG-BU-TMS
(A) and the zeolite crystals tethering TAG (B) and D-glucose (C) units,
respectively, via BU-SiO spacers.
plain (TAG-free) zeolite-A crystals are employed. This result
suggests that the presence of D-glucose units on the zeolite
surfaces is essential for the structure-forming process. Thermal
denaturation of the enzyme31 induced by boiling the aqueous
solution of the enzyme for 2 h prior to introduction to the
aqueous slurry of D-glucose-tethering zeolite-A also did not lead
to fibrillation. Chemical denaturation of the enzyme induced
by addition of sodium dodecyl sulfate (50 mg per 2 mg of
enzyme), a well-known chemical denaturant,32 into the mixture
also effectively prevented fibrillation. These results underscore
that the preservation of the enzyme in its originally active state
is essential for the fibrillation.
The fibrillation process proceeded at a much faster rate in an
acetate buffer solution of pH 4.8. Thus, while it took 48 h in
distilled deionized water, only 8 h was enough to complete the
fiber-forming process in the buffer solution. This indicates an
about 6-fold increase in the rate. Considering the fact that the
activity of â-glucosidase is highest at pH 4.8,30 this result
indicates that the enzymatic activity plays a crucial role in the
fibrillation.
In strong contrast, the fiber-forming process did not proceed
at all even after much longer periods of stirring (>1 month) if
the acetyl groups are not removed from the TAG units nor if
(31) Wong, C.-H.; Whitesides, G. M. Enzymes in Synthetic Organic
Chemistry; Pergamon Press: 1994; pp 17-18.
(32) Boyer, R. Concepts in Biochemistry; Brooks/Cole: London, 1999;
pp 125-126.
(30) Tanaka, T.; Oi, S. Agric. Biol. Chem. 1985, 49, 1267-1273.

â-Glucosidase and D-Glucose-Tethering Zeolite Crystals
J. Am. Chem. Soc., Vol. 122, No. 49, 2000 12155
Figure 5. A typical SEM image of the cross section of a cylindrical
zeolite-A-â-glucosidase fibrous material showing the lumpy enzyme-
coated zeolite-A crystals.
Figure 3. A photographic image showing the fibrous nature of the
zeolite-A-â-glucosidase composite material.
sectional view of SEM images of the cylindrical fibers revealed
that they consist of lumps that look like zeolite-A crystals
covered with thick enzyme layers as typically shown in Figure
5. The presence of zeolite-A crystals within the fibrils was
apparent from the identical magic angle spinning solid-state 27Al
and 29Si NMR spectra of the fibrils and the plain, unmodified
zeolite-A crystals (Figure 6, panels A and B). The fact that the
cubic morphology of zeolite-A crystals remains intact within
the fibrous composite was apparent from the SEM image of
the remained zeolite crystals after heating the composite fibers
at 520 °C under flowing oxygen to burn off the organic enzyme
(Figure 7). The presence of â-glucosidase as the single organic
species within the fibrous material was also confirmed from
the two identical solid-state 13C NMR spectra of the neat enzyme
and the composite fiber (Figure 6C). The above results hence
clearly verify that the fibrils consist of only zeolite-A crystals
and â-glucosidase and the zeolite crystals are buried within the
fibrils. A similar fibrous material also readily developed from
the aqueous mixture of â-glucosidase and the larger, prismatic
ZSM-5 crystals tethered with D-glucose, indicating that the
morphology of the inorganic crystal does not affect that of the
fiber.
Upon increasing the zeolite to â-glucosidase weight ratio to
10, unburied clusters of zeolite-A crystals appeared on the
exterior of most of the fibrils (Figure 8, panels A and B). Upon
decreasing the ratio to 2.5, all the zeolite crystals were again
buried within the fibers but the fibers became flat narrow bands
resembling the flakelike morphology of the crystalline enzyme
(Figure 9). This result showed that the zeolite to enzyme weight
ratio sensitively affects the cross-sectional shape of the fibrous
aggregates.
Figure 4. A typical SEM image showing the fibrous zeolite-A-â-
glucosidase composite material obtained from an aqueous mixture of
D-glucose-tethering zeolite-A and the enzyme in the weight ratio of 5.
It should also be noted that the highly water soluble
â-glucosidase alone, either in the presence or absence of a large
amount (100 mM) of plain, unmodified D-glucose, does not
spontaneously crystallize into such fibrils as long as the glucose-
tethering zeolite-A crystals is left out. On the basis of the above
facts, we propose that the enzyme-substrate interaction between
â-glucosidase and D-glucose units tethered to zeolite surface is
essential for the above fibrillation to occur. In strong support
to this, the fibrillation of â-glucosidase and D-glucose-tethering
zeolite crystals did not proceed in a highly concentrated (100
mM) solution of D-glucose, due to the inhibiting effect of free
D-glucose molecules on the fibrillation.
III. Morphology of the Fibrous Composite. The scanning
electron microscope (SEM) images of the freeze-dried fibrous
material clearly showed the fibrous morphology of the ag-
gregated material as typically shown in Figure 4. The cross-
sectional diameters of the cylindrical fibers range from 2 to 20
µm but more commonly from 5 to 15 µm. The closer looks of
the terminals of the thick cylindrical fibers revealed that they
split into several thinner fibers. As indicated by arrows in Figure
4, many fibers also exist in the helical forms as two strands
twisted together. The lengths of some of intermingled fibers
even exceed a few centimeters. The fibers resumed growing
upon addition of fresh D-glucose-tethering zeolite-A crystals and
â-glucosidase into the transparent solution containing the fibrils.
Interestingly, the presence of zeolite crystals within the fibers
was not apparent from the SEM images. However, the cross-
IV. Insights into the Mechanism of the Fibrillation. To
gain insight into the fibrillation process, a few drops of the
aqueous slurries were periodically taken out of reaction vessels
at early and intermediate stages of the fibrillation process and
freeze-dried for SEM analyses. As shown in Figure 10A, it was
found that crystalline â-glucosidase starts surrounding the
zeolite-A crystals. The material that surrounds zeolite crystals
being crystalline â-glucosidase was obvious not only from the
spectrum of solid-state 13C NMR but also from the fact that the
flat morphology of the surrounding material resembles that of
â-glucosidase shown in Figure 9. It was also revealed that the
enzyme binds several zeolite crystals into randomly aggregated
clusters by crystallizing between the zeolite crystals (Figure
10B). Some of SEM images also revealed that the randomly
aggregated zeolite clusters are further interwoven by intercon-

12156 J. Am. Chem. Soc., Vol. 122, No. 49, 2000
Lee et al.
Figure 7. The aggregates of cubic zeolite-A crystals remaining after
burning off the organic enzyme from the zeolite-A-â-glucosidase
composite fiber.
Figure 8. SEM images showing the fibers covered with unburied
clusters of zeolite-A crystals on the external surface that resulted with
the zeolite to enzyme weight ratio of 10 (A, B) and the smooth flat
fibers that resulted with the corresponding ratio of 2.5 (C, D), at two
different magnifications, respectively.
Figure 6. MAS solid-state NMR spectra of the fibrous zeolite-A-â-
glucosidase composite material and its corresponding reference material
(as indicated): 27Al (A), 29Si (B), and 13C (C).
necting crystalline â-glucosidase crystallites (Figure 10C). In
strong contrast, phase separation of the turbid mixture into piles
of zeolite crystals and the crystalline enzyme results as shown
in Figure 10D if the zeolite crystals are not tethered with
D-glucose. This result indicates that drying artifact is not the
reason for the observed intermediate stages of fibrillation (panels
A, B, and C).
Based on the facts presented so far, we propose that the
fibrillation occurs according to Scheme 3. First, â-glucosidase
binds to each D-glucose unit tethered to zeolite crystals via
enzyme-substrate complexation (step A). The surface-covering
monolayers of enzymes then act as pseudocrystalline surfaces
that in turn serve as seed surfaces for the enzymes to start
crystallizing over the surfaces (step B). At the present moment,
however, the question about the nature of the driving force to
attract such a highly water-soluble enzyme from the aqueous
solution onto the substrate-tethering surface remains open. The
zeolite crystals covered with crystalline â-glucosidase are then
Figure 9. The SEM image showing the flat flakelike morphology of
the crystalline â-glucosidase.
interconnected into clusters by further intergrown crystalline
â-glucosidase (step C). We propose that continual crystallization
of â-glucosidase between the zeolite clusters eventually leads
to zeolite-enzyme composite fibers (step D). However, it
remains to be elucidated why the organic-inorganic composite
adopts this particular morphology. Nevertheless, it is certain
that the entrapped zeolite crystals play crucial roles in shaping
the overall structure of the composite fiber from the fact that

â-Glucosidase and D-Glucose-Tethering Zeolite Crystals
J. Am. Chem. Soc., Vol. 122, No. 49, 2000 12157
Scheme 3
Figure 10. SEM images showing the initial and intermediate stages
of the fibrillation process from a mixture of â-glucosidase and
D-glucose-tethering zeolite-A in distilled deionized water (A, B, and
C). Panel D represents that of a mixture of a pile of individual plain
zeolite-A crystals and chunky crystalline â-glucosidase that results from
phase separation of the turbid aqueous slurry of unmodified, plain
zeolite-A crystals and â-glucosidase stirred for 24 h into its components
upon drying.
rather flat fibers result upon increasing the relative amount of
enzyme. The enzymes may transmit a recognition effect between
the opposite faces of the zeolite crystals causing the growth
process to be favored in an axial direction.
V. Enzymatic Activity of â-Glucosidase in the Fibrils.
Interestingly, the light yellow strips of Reflectoquant im-
mediately turned intense green upon dipping into the equilibrated
solution of D-cellobiose and fibrils. This strongly indicates that
D-cellobiose is hydrolyzed to D-glucose in the presence of the
composite fiber. Consistent with the color test, the composite
fiber also readily hydrolyzed PNPG into D-glucose and p-
nitrophenol. The above results therefore clearly indicate that
the enzymatic activity of â-glucosidase within the composite
fiber is preserved. We propose that the enzymes existing in the
periphery or at the outermost external surfaces of the fibrils
provide the catalytic sites for the hydrolysis reactions. To our
surprise, the enzymatic activity of the composite fibers persisted
for more than 6 months even if they were kept in distilled
deionized water at ambient temperatures. This indicates that the
enzymes are very stable in the composite fiber as if they were
stored in the crystalline state. This result raises the possibility
to devise a novel method to preserve and immobilize the highly
expensive enzyme on solid supports by modifying the proce-
dures described in the Experimental Section.
VI. Concluding Remarks. Overall, this report demonstrates
for the first time the self-assembly of an enzyme and its
substrate-tethering inorganic crystals into a fibrous composite
material. This result opens the possibility that enzymes may
serve as structure-forming agents in living organisms while
retaining their enzymatic activities. From the fact that enzymes
consist of proteins, the resulting material can be considered as
a type of protein-inorganic composite material which plays a
vital role in living organisms as structural material. This result
also elucidates the remarkable behavior of â-glucosidase in the
presence of the surface-bound D-glucose.
Acknowledgment. We thank the Ministry of Science and
Technology (MOST), Korea, for supporting this work through
the Creative Research Initiatives (CRI) program.
JA0028222