Chem. Mater. 1999, 11, 3671-3679
`
`3671
`
`Effects of Encapsulation in Sol-Gel Silica Glass on
`Esterase Activity, Conformational Stability, and
`Unfolding of Bovine Carbonic Anhydrase II
`
`Jovica D. Badjie and Nenad M. KostK*
`
`Department of Chemistry, [owa State Universit~v, Ames, [owa 50010
`
`Received August 6, 1999
`
`Bovine carbonic anhydrase II retains its overall conformation when encapsulated in silica
`monoliths by the sol-gel method. Upon gradual heating the enzyme in solution precipitates
`at 64 °C, whereas the encapsulated enzyme does not; it unfolds, with the nominal melting
`temperature of 51 + 3 °C. Even at 74.0 °C, the encapsulated enzyme is only ~77% unfolded,
`but it does not refold upon cooling. Upon treatment with guanidinium chloride, the degree
`of enzyme unfolding is 100% in solution but only ~83% within the silica matrix. Again, the
`enzyme does not refold upon removal of the denaturing agent from the glass. The glass
`matrix constrains the motions of the encapsulated protein molecules and prevents both their
`full unfolding and refolding. The former may be taken for "stabilization", the later for
`"destabilization" of the native conformation. Evidently, the effect of the glass on the
`encapsulated protein cannot be described in these general terms. The encapsulated enzyme
`obeys the Michaelis-Menten kinetic law as it catalyzes hydrolysis of !~nitrophenyl acetate.
`The apparent Michaelis constant (K’M) is practically the same in the glass and in solution,
`but the apparent turnover number (k’ca0 and specific activity for the encapsulated enzyme
`are only 1-2% of these values for the enzyme in solution. Because the substrate diffuses
`slowly into silica pores, most of the catalysis is due to the enzyme embedded near the surfaces
`of the glass monolith. Effect of encapsulation on structure and activity of proteins should be
`studied in quantitative detail before protein-doped glasses can be used as reliable biosensors,
`heterogeneous catalysts, and composite biomaterials.
`
`Introduction
`
`Controlled hydrolysis of alkoxides and polymerization
`of the resulting oxyacids, the sol-gel method, is very
`useful in many branches of chemistry and materials
`science. This method is used for the making of sensors,
`catalyst supports, optical elements, coatings, and special
`polymers.1-4 Enzymes, catalytic antibodies, and other
`proteins may be entrapped in robust silica glasses under
`mild conditions, so that they retain their chemical activ-
`ity. Because the resulting bioceramics can be fabricated
`as monoliths, thin films, powders, and fibers, these re-
`cent achievements open many possibilities for research
`and application in bioanalytical chemistry, biocatalysis,
`biotechnology, and environmental technology.5-1z Hy-
`drogels and xerogels are often referred to simply as sol-
`gel glasses and, colloquially, as sol-gels.
`
`Despite widespread interest in various applications
`of protein-doped glasses, relatively little is known about
`their fundamental properties and, especially, about
`interactions between the protein molecules and the
`matrix in which they are embedded. These interactions
`govern the reactivity of the embedded protein, and they
`must be understood before useful devices can be made.
`Particularly important are the possible effects of en-
`capsulation on conformation of enzymes, their interac-
`tions with substrates, and their catalytic ability.
`Research in our laboratory showed that proteins and
`also small molecules may behave quite differently in
`traditional solutions and in the pores of a hydrogel.13,14a
`We took advantage of transparency of silica monoliths
`to study photoinduced chemical reactions. The triplet
`state of zinc-substituted heme proteins cytochrome c and
`
`(1) Avnir, D.; Braun, S.; Lev, O.; Ottolenghi, M. Chem. Matez: 1994,
`6, 1605 14.
`(2) Avnir, D. Acc. Chem. Rex 1995, 28, 328 34.
`(3) Rosenfeld, A.; Blum, J.; Avnir, D. J. Catal. 1996, 164, 363
`368.
`(4) Wojcik, A. B.; Klein, L. C. AppI. Oz’gar~omet. Chem. 1997, 11,
`129 135.
`(5) Livage, J. C. R. Acad. Sci., Set: II." Mec., Phys., Chim., Astron.
`1996, 322, 417 27.
`(6) Dunn, B.; Zink, J. I. Chem. Matet~ 1997, 9, 2280 2291.
`(7) Braun, S.; Rappoport, S.; Shtelzer, S.; Zusman, R.; Druckmann,
`S.; Avnir, D.; Ottolenghi, M. Desigr~ and properties of enzymes im
`mobilized ir~ sol gel glass matrices; Kamely, D. C., Ananda M, Ed.;
`Kluwer: Boston, Mass, 1991; pp 205 18.
`
`(8) Reisfeld, R.; Joergensen, C. K. Srrucr. Bor~dir~g (t3etTir~) 1992,
`77, 207 56.
`(9) Braun, S.; Shtelzer, S.; Rappoport, S.; Avnir, D.; Ottolenghi, M.
`J. Nor~ C~st. Solids 1992, 14~148, 739 43.
`(10) Miller, J. M.; Dunn, B.; Valentine, J. S.; Zink, J. I. J. Non
`Ctysr. Solids 1996, 202, 279 289.
`(11) Zheng, L.; Reid, W. R.; Brennan, J. D. Anal Chem. 1997, 69,
`3940 3949.
`(12) Lan, E. H.; Dunn, B.; Valentine, J. S. V.; Zink, J. I. J. Sol Gel
`Sci. Tectmol. 1996, 7, 109 116.
`(13) Shen, C.; KostK, N. M. J. Elecrt’oar~al. Chem. 1997, 438, 61
`65.
`(14) (a) Shen, C.; KostK, N. M. J. Am. Chem. Soc. 1997, 119, 1304
`1312. (b) Kadnikova, E. N.; Kosti~, N. M. Unpublished work.
`
`10.1021/cm990507c CCC: $18.00 © 1999 American Chemical Society
`Published on Web 11/24/1999
`
`Akermin, Inc.
`Exhibit 1009
`Page 1
`
`

`

`3672 Chem. Mater., Vol. 11, No. 12, 1999
`
`Badji~ and Kosti~
`
`myoglobin, designated Zn(heme), is oxidatively quenched
`by charged metal complexes and neutral organic com-
`pounds, designated Q, as in
`
`We chose a well-characterized enzyme, isozyme II of
`bovine carbonic anhydrase, and a simple reaction,
`hydrolysis of p-nitrophenyl acetate:22-25
`
`3Zn(heme) ÷ Q ~ Zn(heme)+ ÷ Q
`
`(1)
`
`~ ~0 CA
`
`Because the protein was encapsulated and the quench-
`ers were dialyzed in, both reactants were already inside
`the glass before the light pulse. Therefore, the problem
`of macroscopic diffusion was avoided, and precise kinetic
`experiments were possible. They revealed interesting
`electrostatic and structural properties of the glass
`interior, which make silica matrixes quite different from
`pure solvent as a reaction medium. The same pair of
`reactants behaves differently when confined in the
`matrix pores and when free in solution. We also
`examined the ability of peroxidases encapsulated in
`silica and in alkylated silica to catalyze oxidation of
`various sulfide compounds,14b as in
`
`RSR’ + H202
`
`catalyst
`
`in silica
`
`O
`RiSeR, + R--s--R’II + H20
`
`(2)
`
`Our studies13,14a of the first kind disproved the
`popular notion that because the pores in silica are
`relatively large, small molecules diffuse freely in and
`out of the glass matrix. In fact, even prolonged soaking
`of silica monoliths in electrolyte solutions may not
`ensure equal partitioning of a given ion between the
`monolith and solution. At pH value at which the pore
`
`walls are negatively charged, anions, such as [Fe(CN)6] 3-,
`are only partially taken up, whereas cations, such as
`
`[Ru(NH3)6] 3+, are preferentially taken up. In either case,
`internal and external concentrations will remain un-
`equal after the equilibrium is reached. Since many
`analytes are ions, such interactions must be understood
`before reliable biosensors for these analytes may be
`designed.
`
`Several elegant, recent studies found that proteins are
`stabilized by encapsulation in silica.15-17 But the notion
`of "stability" has various meanings. We decided to
`examine one aspect of protein stability, namely propen-
`sity for controlled unfolding and refolding. The few
`kinetic studies of encapsulated enzymes to date all have
`been done with powdered glasses, to minimize the
`problem of substrate diffusion.18-zl We decided to study
`this diffusion in glass monoliths. An advantage of
`monoliths over powders is transparency, which permits
`the use of spectrophotometry.
`
`(15) Chen, Q.; Kenausis, G. L.; Heller, A. d. Am. Chem. Soc. 1998,
`120, 4582 4585.
`(16) Heller, J.; Heller, A. ~. Am. Chem. Soc. 1998, 120, 4586 4590.
`(17) Shtelzer, S.; Rappoport, S.; Avnir, D.; Ottolenghi, M.; Braun,
`S. BiotechnoL AppI. Biochem. 1992, 15, 227 35.
`(18) Yamanaka, S. A.; Nishida, F.; Ellerby, L. M.; Nishida, C. R.;
`Dunn, B.; Valentine, J. S.; Zink, J. I. Chem. Matez~ 199Z, 4, 495 7.
`(19) Yamanaka, S. A.; Dunn, B.; Valentine, J. S.; Zink, J. I. d-. Am.
`Chem. Soc. 1995, 117, 9095 6.
`(20) Braun, S.; Rappoport, S.; Zusrnan, R.; Avnir, D.; Ottolenghi,
`M. Mate~ Letb 1990, 10, 1 5.
`(21) Dosoretz, C.; Arrnon, R.; Starnsvetzky, J.; Rothschild, N. d-. Sol
`Gel Sci. TechnoL 1996, 7, 7 11.
`
`O2N~~--~ ~ O--C-CH3 + H20
`
`O2N
`
`O--H
`
`+
`
`OH--C-CH3
`
`(3)
`
`By comparing the protein unfolding behavior, and also
`the kinetics of the reaction in eq 3 inside the glass and
`in solution, one can learn about the unexpected and
`important differences between the glass matrix and
`solvent as reaction media.
`
`Experimental Procedures
`
`Chemicals. Bovine carbonic anhydrase II, ultrapure guani-
`dinium chloride, 3-(N-morpholino)propanesulfonic acid (MOPS),
`and its sodium salt were obtained from Sigma Chemical Co.
`Tetramethyl orthosilicate (also called tetramethoxysilane and
`designated TMOS), p-nitrophenyl acetate, p-nitroanisole, and
`acetonitrile were obtained from Aldrich Chemical Co. The salts
`NaH2PO4"H20 and Na2HPO4.THaO were obtained from Fischer
`Chemical Co. The salt [Ru(NH3)sC1]Cla was obtained from
`Strem Chemical Co. Distilled water was demineralized to
`electrical resistivity greater than 17 Mr2 cm. A MOPS buffer
`having the ionic strength 11.6 mM was prepared by dissolving
`1.70 g of the acid and 0.97 g of its sodium salt in 400 mL of
`water and adjusting the pH to 7.00 with a 0.100 M solution of
`NaOH. Phosphate buffer with ionic strength of 10.0 mM was
`prepared by dissolving 0.050 g of NaHaPO4.HaO and 0.86 g of
`Na2HPO4.7 HaO in 1.000 L of water and adjusting the pH to
`8.00 with a 0.100 M solution of NaOH. Ionic strength, not
`concentration, is specified in this article. The Accumet 925 pH
`meter was equipped with a glass electrode. Glass monoliths
`(SLABS) intended for different experiments were prepared
`with, and finally kept in, different buffers. For the monoliths
`intended for studies of carbonic anhydrase conformation and
`unfolding, 11.6 mM MOPS buffer at pH 7.00 was used; for the
`monoliths intended for enzymatic catalysis and diffusion of
`substrate analogues, 10.0 mM phosphate buffer at pH 8.00
`was used.
`The Sol-Gel Process. Silica sol was prepared by a
`published procedure.18 A mixture of 15.25 g of Si(OCH3)4, 3.38
`g of water, and 0.300 g of 0.040 M HC1 was ultrasonicated in
`an ice bath for 30 min. A 4.64-mL portion of the resulting sol
`was mixed with 5.64 mL of the appropriate buffer (see above)
`and kept in an ice bath. A stock solution of carbonic anhydrase
`II containing a variable amount (2.0-14 mg) of the enzyme
`in 3.4 mL of the appropriate buffer (see above) was added to
`the sol, and the clear liquid was transferred to polystyrene
`cuvettes sized 10 × 10 × 40 mm. The cuvettes were sealed
`with Parafilm and kept at 4 °C. During the 14 days of aging
`the monoliths were washed twice a day with the same buffer
`with which they had been made. Then the Parafilm was
`removed for partial drying, still at 4 °C, and in additional 14
`days the monoliths shrunk to 8 × 8 × 27 mm. The final volume
`was ~50% of that of the aged monoliths. The partially dried
`monoliths were stored in the same buffer with which they had
`been prepared, depending on the purpose.
`Determining the Enzyme Concentration. Concentration
`of bovine carbonic anhydrase II in free solution and in solution
`
`(22) Pocker, Y.; Meany, J. E. ]~iochemistzy 1967, 6, 239 46.
`(23) Pocker, Y.; Storm, D. R. Biochemistr~ 1968, 7, 1202 14.
`(24) Anderson, J.; Byrne, T.; Woelfel, K. J.; Meany, J. E.; Spyridis,
`G. T.; Pocker, Y. J. Chem. Educ. 1994, 71, 715 18.
`(25) Thorslund, A.; Lindskog, S. Eun ~. Biochem. 1967, 3, 117
`23.
`
`Akermin, Inc.
`Exhibit 1009
`Page 2
`
`

`

`t~ncapsulation in Sol-Gel Silica Glass
`
`Chem. Mater., Vol. 11, No. 12, 1999 3673
`
`confined in the pores of silica monoliths was determined on
`the basis of absorbance at 280 nm and the plot of that
`absorbance versus the enzyme concentration in a 11.6 mM
`MOPS buffer at pH 7.00. For accuracy, liquids were measured
`by weight rather than volume; because the buffer solutions
`were dilute, their density was taken to be unity. This ap-
`proximation did not introduce a significant error in subsequent
`experiments. The enzyme concentration in the stock solution
`was 3.84 mg/mL; from it, by dilution, were made six more
`solutions, and their absorbances were measured. The slope of
`the plot gave the conversion factor of 1.48 mL/mg. Monolith
`doped with the enzyme was soaked in the buffer appropriate
`for the subsequent experiments, as explained in the preceding
`subsection. Both the doped monolith (in the sample beam) and
`the undoped monolith of the same dimensions (in the reference
`beam) were held in standard cuvettes, filled with the buffer.
`The ultraviolet absorption spectrum was recorded, and the
`concentration was calculated with the conversion factor, taking
`into account the path length of 0.80 cm. Concentration so
`determined always equaled that calculated from the amount
`of the enzyme encapsulated, considering the increase in con-
`centration caused by shrinkage of the partially dried monolith.
`Circular Dichroism Spectra. These spectra were recorded
`with the instrument J-710 by JASCO, equipped with a peltier
`thermoelectric cell. When the samples were solutions, the
`sensor was placed inside the cuvette, and temperature was
`kept within 4- 0.05 °C; when the samples were silica monoliths
`immersed in solution, the sensor had to be placed next to the
`cuvette, and temperature was kept within 4-0.10 °C. In
`experiments not concerning thermal unfolding and refolding,
`temperature was 20.0 °C, and monoliths were kept at it for
`30 min before the measurements were started. Far-UV CD
`spectra, in the region 200-240 nm, were recorded in 1-mm
`cuvettes, with silica monoliths (1 mm thickness) and free
`solutions that were both 12 ,uM in carbonic anhydrase. Near-
`UV CD spectra, in the region 240-340 nm, were recorded in
`10-mm cuvettes; the enzyme concentration was 28 ,uM in silica
`and 25 ,uM in free solution. The soaking buffer was always
`11.6 mM MOPS at pH 7.00. In calculation of molar residue
`ellipticity,26 the molecular mass of the enzyme was 28 900 Da;
`the path length was 0.80 cm in the case of doped silica
`monoliths (because of their shrinkage); and the enzyme
`concentration was determined as explained above.
`Unfolding of Carbonic Anhydrase in Free Solution.
`In studies of thermal denaturation, the solution of the enzyme
`in 11.6 mM MOPS buffer at pH 7.00 had a concentration of
`0.761 mg/mL. Elfipticity at 245 nm in a 10-mm cuvette was
`recorded continuously, during slow heating at the rate of 0.50
`°C/min. In addition, the CD spectrum in the range 240-340
`nm was recorded at 20.0, 40.0, 55.0, and 75.0 °C. A stock
`solution of guanidinium chloride (GdmC1) was prepared care-
`fully, by dissolving 15.1591 g of this denaturant in 25.0251 g
`of the 11.6 mM MOPS buffer at pH 7.00.26 Concentration was
`also checked by measuring the refraction index of the solution.
`A stock solution of bovine carbonic anhydrase II was prepared
`by adding 1.0704 g of NaC1 to 6.0227 g of the solution
`containing 3.84 mg of this enzyme per milliliter of 11.6 mM
`MOPS buffer at pH 7.00. The protein concentration was
`checked by measuring absorbance at 280 nm. To 0.5000 g of
`the stock solution of the protein (and NaC1) were added
`required weights of the GdmC1 stock solution and of the MOPS
`buffer, to obtain a solution that was 0.600 M in NaC1 and
`between 0 and 3.488 M in GdmC1. A series of these solutions
`differing only in the denaturant concentration was equilibrated
`at room temperature for 12 h before their near-UV CD spectra
`were recorded. The denaturation was reversed (i.e., the protein
`was refolded) by exhaustive dialysis with Slyde-A-Lyzer 10K
`cassettes, obtained from Pierce Chemical Co. A 3.00-mL
`sample of the solution after the CD spectroscopic measure-
`ments was dialyzed against 400 mL of the 11.6 mM MOPS
`buffer at pH 7.00 that was 0.100 M in NaC1.~7 Then the near-
`UV CD spectrum was recorded again.
`
`(26) Giletto, A.; Pace, C. N. Proteir~ Stabilitj/; Price, N. C., Ed.;
`Academic: San Diego, CA, 1996; pp 233 239.
`
`Unfolding of Encapsulated Carbonic Anhydrase by
`Heating. A transparent silica monolith sized 8 × 8 × 27 mm
`containing 28 ,uM bovine carbonic anhydrase II was soaked in
`a 11.6 mM MOPS buffer at pH 7.00 inside a quartz cuvette
`sized 10 × 10 × 40 mm. The temperature of the cell holder
`was raised from 20.0 to 74.0 °C in 3.0-deg steps. After a 30-
`min period of equilibration, the near-UV CD spectrum was
`recorded (for an additional 20 min). The mean residue elfip-
`ticities at 245 and 269 nm were plotted against temperature.
`In control experiments performance of the CD instrument was
`verified by recording near-UV spectra of the monolith not
`doped with the protein at 20.0 and 74.0 °C; the heating was
`done gradually, with the doped monolith. The doped monolith
`was then cooled, at a rate of 3.0 °C/h, back to 20.0 °C, and the
`CD spectrum was recorded again. The apparent melting
`(denaturation) temperature of the protein, T~, was obtained
`by fitting the plots at 245 and 269 nm to the simple model of
`direct conversion of the folded state into an unfolded state.~8,~9
`Unfolding of Encapsulated Carbonic Anhydrase with
`Guanidinium Chloride. Because the denaturing agent in
`this salt is the cation, we first determined the conditions under
`which cations can be introduced into sol-gel silica glass. In
`these preliminary experiments we used the salt [Ru(NH3)~C1]-
`Cl~ because its cation is colored and because its charge is
`higher than that of the Gdm+ ion. A silica monolith sized 8 ×
`8 × 27 mm was soaked, for 24 h at 25 °C, in a 5.2 × 10 4 M
`solution of the complex salt in a 11.6 mM MOPS buffer at pH
`7.00 that was 0.600 M in NaC1. Diffusion was assisted by
`gentle shaking. The UV-vis spectrum of the monolith infused
`with [Ru(NH3)~C1]2+ cations was recorded against an "empty"
`monolith in the reference beam and compared with the
`spectrum of the fresh soaking solution. Concentration of the
`complex cation was determined on the basis of the absorptivity
`~3~6 : 1850 M lcm 1 and the optical path length of the
`monolith of 0.80 cm. Transparent silica monoliths sized 8 × 8
`× 27 mm containing 28 ,uM bovine carbonic anhydrase II were
`each soaked in 10.0 mL of 11.6 mM MOPS buffer at pH 7.00
`that was 0.600 M in NaC1 and between 0 and 4.35 M in
`GdmC1. Diffusion of the denaturant and NaC1 into the doped
`glass was assisted by gentle shaking for 24 h at 25 °C. Circular
`dichroism spectra at 240-340 nm of all the monoliths were
`recorded, and molar residue ellipticity at 269 nm was plotted
`against the GdmC1 concentration. The monolith containing
`4.35 M GdmC1 was soaked in 400 mL of the 11.6 mM MOPS
`buffer at pH 7.00 that was 0.600 M in NaC1, for 24 h at 4 °C.
`This same procedure was repeated with a fresh buffer. Near-
`UV CD spectra were recorded at 24 and 48 h, after each
`soaking.
`Degree of Unfolding of Carbonic Anhydrase. This
`degree, designated f~, varies from 0.0 to 1.0 and is proportional
`to the molar residue elfipticity at 269 nm. The limiting values
`correspond to the protein dissolved in a 11.6 mM MOPS buffer
`containing no GdmC1 and 3.488 M GdmC1.
`This denaturant completely unfolds bovine carbonic anhy-
`drase.3° The same f~ scale was used for the enzyme in free
`solution and encapsulated in silica.
`Diffusion of Small Molecules into Silica Glass. The
`substrate for carbonic anhydrase (acting as an esterase) is
`p-nitrophenyl acetate. An unreactive analogue of it is p-
`nitroanisole. The solvent was a 9:1 mixture by volume of a 10
`mM phosphate buffer at pH 8.00 and acetonitrfle. Solutions
`were made in volumetric flasks. Their concentrations were 1.00
`mM in the substrate and 0.14 mM in its analogue. Silica
`monoliths were sized 8 × 8 × 27 mm. Three monoliths doped
`with 28 ,uM carbonic anhydrase II were each soaked in 30.0
`mL of the substrate solution. The yellow color of p-nitrophe-
`noxide anion advanced into the monoliths for 10, 60, and 600
`
`(27) Wong, K. P.; Tanford, C. J. Biol. Chem. 1973, 248, 8518 23.
`(28) Biltonen, R. L.; Lumry, R. J. Am. Chem. Soc. 1969, 91, 4256
`64.
`(29) Lee, J. C.; Timasheff, S. N. J. Biol. Chem. 1981, 256, 7193
`201.
`(30) Uversky, V. N.; Ptitsyn, O. B. J. Mol. Biol. 1996, 255, 215
`28.
`
`Akermin, Inc.
`Exhibit 1009
`Page 3
`
`

`

`3674 Chem. Mater., Vo1. 11, No. 12, 1999
`
`J~adji~ and Kosti~
`
`min. From each soaked monolith a 8 × 8 × 7 mm slice was
`obtained by cutting off 10-mm pieces from both ends, and this
`slice was photographed with a Canon T-60 camera equipped
`with a macro lens having FD of 50 mm and 1:3:5 ratio and
`with an extension tube having FD of 25 mm. An undoped
`(enzyme-free) monolith sized 8 × 8 × 27 mm was soaked first
`in the solvent (the aforementioned 9:1 mixture) and then in
`20.0 mL of the p-nitroanisole solution for 18 h. Occasionally
`the monolith was removed from the solution, and its absor-
`bance at 316 nm was recorded against an unsoaked monolith
`in the reference beam. Absorbance at 316 nm was plotted
`versus the soaking time.
`Kinetics of Catalysis by the Encapsulated Carbonic
`Anhydrase. These experiments were done with 32 very
`similar silica monoliths sized 8 × 8 × 27 mm, chosen from a
`set of 64 monoliths prepared under identical conditions. These
`32 were divided into four sets of 8; members of each set were
`doped with the same concentration of carbonic anhydrase II:
`10, 31, 56, or 79 ,aM. Each monolith was first soaked in 100
`mL of a 10 mM phosphate buffer at pH 8.00 to establish the
`pH value of the interior. After this equilibration, each monolith
`was soaked in 2.0 mL of the solvent in which the hydrolysis
`was to be done, a 9:1 mixture by volume of this buffer and
`acetonitrile; the containers polystyrene spectrophotometric
`cuvettes sized 10 × 10 × 40 mm, were thermally equilibrated
`in a water bath at 25 °C for 15 min. The reaction was started
`by addition to each cuvette of a small volume of a solution of
`p-nitrophenyl acetate in the aforementioned solvent. The
`substrate concentrations in the reaction mixture spanned the
`range 1.0 to 10.0 mM. Absorbance at 400 nm of p-nitrophe-
`noxide anion, a sum of contributions from the monolith and
`the surrounding solution, was measured over time. Absorp-
`tivity of this product, ~4oo = 16 840 M 1 cm l, was determined
`by measuring absorbances of a series of solutions in the same
`solvent having concentrations in the relevant range, 1.0-10.0
`mM. After the incubation period of 50 s, dependencies of
`absorbance at 400 nm (A) on time (z) were fitted to
`
`A = V0t+ c
`
`(4)
`
`with the linear least-squares method embodied in the program
`SigmaPlot 1.0. In similar experiments with undoped silica
`monoliths, initial rates for "background" hydrolysis at the same
`concentration of p-nitrophenyl acetate were found to be less
`than 10% of the rates of hydrolysis catalyzed by the enzyme-
`doped monoliths. After the subtraction of the "background"
`rates, the initial rates (slopes Vo in eq 4) for the enzymatic
`hydrolysis of the substrate at concentration S were fitted to
`the Lineweaver-Burk equation
`
`1_ ~M 1
`Vo
`
`(5)
`
`and the following kinetic parameters were obtained: apparent
`Michaelis constant, K’M; maximum velocity, Vm~x; and apparent
`turnover number, k’cat.31 The primed quantities are termed
`apparent because the concentration of the enzyme doing
`catalysis is not known exactly.
`Specific activity of carbonic anhydrase is 10 3 times the
`number of milligrams of the substrate (p-nitrophenyl acetate)
`hydrolyzed by 1.0 mg of the enzyme per 1.0 min.
`
`Results and Discussion
`
`Encapsulation of Carbonic Anhydrase. The bo-
`vine isozyme II is a well-characterized protein, with the
`molecular mass of 28 900 Da and the isoelectric point
`p[ = 5.9. Because none of its 259 amino acid residues is
`cysteine, chemical and spectroscopic studies are not
`complicated by oxidation and formation of disulfide
`
`bridges.27,3°,32-34 Because the sol-gel method yields
`silica glass that is transparent at wavelengths as low
`as 250 nm, concentration of the encapsulated protein
`was accurately determined on the basis of the absor-
`bance at 280 nm; shrinkage upon partial drying of the
`glass monoliths was taken into account in determining
`the optical path length. Concentration of the enzyme
`in the monolith was 0.80 mg/mL on the basis of the UV
`absorption spectrum and 0.79 mg/mL on the basis of
`the amount of the enzyme used in the encapsulation
`experiments. Equality of these values within the error
`margins, seen in Figure S1 in the Supporting Informa-
`tion, proves that frequent rinsings of the monoliths over
`2 weeks during the aging of silica does not leach the
`trapped enzyme. Indeed, the enzyme was not detected
`by UV absorption in the rinsing buffer.
`
`Monitoring of Carbonic Anhydrase Unfolding.
`Although diffusion of relatively small molecules through
`the porous sol-gel silica glass has been studied be-
`fore,35,36 subtler kinds of motions, such as conforma-
`tional changes of encapsulated molecules, have only
`begun to be investigated.37 These investigations are
`needed, because conformational states of proteins de-
`termine their biological activity and chemical reactivity.
`Carbonic anhydrase is well-suited for these studies in
`the glass because its unfolding in solution has been
`examined. Because gradual heating causes irreversible
`precipitation at 64 °C and higher temperatures, this
`denaturation could not be monitored by spectroscopic
`methods; a more involved method had to be used.38
`Denaturation by guanidinium chloride (GdmC1) in solu-
`tion, however, is a well-characterized, reversible process,
`conveniently followed by circular dichroism (CD) spec-
`troscopy.3°
`
`The CD bands of bovine carbonic anhydrase II in the
`far-UV region are relatively weak, because a-helical
`content is low. A minimum in ellipticity at 217 nm is
`diagnostic of the high content of antiparallel f!-sheet.
`In the near-UV region, 240-320 nm, the spectrum is
`complex, with multiple Cotton effects. There are major
`bands at 245 and 270 nm and two minor ones above
`280 nm. Because isozymes II of human and bovine
`carbonic anhydrases have very similar near-UV CD
`spectra, the same chain length, the same tryptophan
`residues (nos. 5 and 16 in one of the two clusters of
`aromatic residues, and nos. 97, 123, 192,209, and 245
`in the central f!-sheet region), and many other similari-
`ties, reported properties of the mutants of the human
`enzyme are relevant to the bovine enzyme that we
`used.39 All the tryptophan residues make positive
`
`(31) Segel, I. H. Fnzynw Kinetics." Behavior and Analysis o£Rapid
`Fquilibrium and Steady State ~nzyme Systems; Wiley Interscience:
`New York, NY, 1975.
`(32 Wong, K. P.; Harnlin, L. M. BiochenHstzy 1974, 13, 2678 83.
`(33 Christianson, D. W.; Fierke, C. A. Acc. Chem. t~es. 1996, 29,
`331 339.
`(34 Silverrnan, D. N.; Lindskog, S. Acc. Chem. Rex 1988, 21, 30
`
`6.
`
`(35 Koone, N.; Shao, Y.; Zerda, T. W. J. Phys. Chem. 1995, 99,
`16976 81.
`(36 Sieminska, L.; Zerda, T. W. J. Phys. Chem. 1996, 100, 4591
`
`7.
`
`(37 Edrniston, P. L.; Warnbolt, C. L.; Smith, M. K.; Saavedra, S.
`S. d. Colloid Interface Sci. 1994, 163, 395 406.
`(38 McCoy, L. F., Jr.; Wong, K. P. t3iochemistzy 1981, 20, 3062
`
`7.
`
`(39 Freskgaard, P. 04 Maartensson, L. G.; Jonasson, P.; Jonsson,
`B.H. Carlsson, U. Biochemistr~ 1994, 33, 14281 8.
`
`Akermin, Inc.
`Exhibit 1009
`Page 4
`
`

`

`Encapsulation in Sol-Gel Silica Glass
`
`Chem. Mater., Vol. 11, No. 12, 1999 3675
`
`2
`
`-4
`200
`
`250
`
`200
`
`150
`
`100
`
`5o
`
`0
`40
`
`-100
`
`-150
`24O
`
`A
`
`240
`
`B
`
`210
`
`230
`220
`Wavelength(nm)
`
`260
`
`300
`280
`Wavelength (nm)
`
`320
`
`34C
`
`Figure 1. Far UV (A) and near-UV (B) circular dichroism
`spectra of bovine carbonic anhydrase II dissolved in a MOPS
`buffer having ionic strength 11.6 mM and pH 7.00 (dashed or
`dotted line) and encapsulated in silica glass (solid line). The
`protein concentration and path length were 5 ,aM and 1.0 mm
`(A) or 28 ,aM and 8.0 mm (B). Molar ellipticity is normalized
`to the mean residue value.
`
`contributions to the CD in the region 230-250 nm, but
`tyrosine residues make a minor negative contribution.
`Tryptophan residues nos. 5, 123, 209, and 245 contrib-
`ute mainly to the band at 269 nm. Because the relatively
`many aromatic residues are distributed in various
`regions of the carbonic anhydrase molecule, its CD
`spectrum reflects the conformation of the whole protein,
`not only a part of it. (This advantage is absent in studies
`that use only one chromophore or fluorophore for
`monitoring of conformational changes.) Our CD spectra
`of the enzyme in solution agree with those reported
`earlier.2r As Figure 1 shows, encapsulation in silica
`causes only small changes in band intensities and
`shapes. Similar small changes were found upon encap-
`sulation of bacteriorhodopsin and zinc cytochrome c in
`silica glasses.~4,4° Studies of these two proteins showed
`that they are not significantly perturbed in the glass,
`and we conclude that neither is carbonic anhydrase.
`Both the backbone (far-UV) and the relevant side chains
`(near-UV) seem to retain upon encapsulation the gen-
`eral structure that they had in solution.
`We monitored unfolding by measuring molar residue
`ellipticity, 0, at 245 and 269 nm, wavelengths at which
`the spectral changes were largest. The corresponding
`plots obtained at these two wavelengths were always
`very similar because each reflected the changes in the
`tertiary structure of the whole enzyme molecule.
`Electrostatic Effect of the Silica Glass on Up-
`take of Ionic Species. When a monolith of sol-gel
`
`silica is immersed in a solution, the solute will diffuse
`into the monolith. Even though the solute molecules
`may be much smaller than the pores in the glass, when
`an equilibrium is reached after long soaking concentra-
`tions of the solute inside the glass and in the surround-
`ing solution may not be equal. This inequality will
`persist if the solute is an ionic species and the surface
`of the glass pores is charged. Since the isoelectric point
`of silica is 2.1, the pores of sol-gel glass are negatively
`charged at pH values 7.00 and 8.00, used in our
`experiments.4~ Research in this laboratory showed
`partial uptake of the solute when the charges of the
`solute and the glass are like and excessive uptake when
`the charges are opposite. These electrostatic effects can
`be abolished by raising ionic strength of the solution.
`In the presence of 0.600 M NaC1, at pH 7.00, concentra-
`tions of [Fe(CN)6]3- ion and other ionic solutes "inside"
`the monolith and in the solution "outside" do become
`equal after sufficient time.~a,~4a We confirmed this result
`in experiments with [Ru(NH3)5C1]C12, a salt containing
`a colored complex cation. As Figure $2 in Supporting
`Information shows, concentration of this doubly charged
`cation inside and outside the silica monolith became
`equal in the buffer at pH 7.00 that was also 0.600 M in
`NaC1.
`In studies of unfolding of encapsulated carbonic
`anhydrase with guanidinium chloride, concentration of
`the colorless Gdm+ cation in the glass could be known
`only if it were kept equal to that in the external solution.
`The 0.600 M NaC1, which proved sufficient even for the
`multiply charged ions, was kept in all solutions. To
`permit correct comparisons of the unfolding process in
`the glass and in solution, 0.600 M NaC1 was always
`used, although this salt was not necessary for precisely
`knowing the concentration of GdmC1 in free solution.
`Unfolding of Carbonic Anhydrase in Solution.
`A previous study38 found that heating of the enzyme
`solution to 64 °C and higher temperatures causes
`precipitation that cannot be reversed by cooling. Dena-
`turation had to be followed by monitoring of pH as a
`function of temperature.38 The enzyme unfolds in the
`interval 55.0-70.0 °C; precipitation begins at 55.0 °C,
`and the precipitate remains up to 95.0 °C. The melting
`temperature Tm = 64.3 °C.38 As Figure $3 in the
`Supporting Information shows, our experiments confirm
`that onset of precipitation at 64 °C is accompanied by a
`large change of ellipticity and an increase in the spectral
`noise.
`Unfolding by guanidinium chloride at room temper-
`ature proceeds via two intermediates. At the denaturant
`con