- Email: [email protected]
sucrose glasses used for protein incorporation as studied by infrared and optical spectroscopy
ANALYTICAL BIOCHEMISTRY Analytical Biochemistry 307 (2002) 167–172 www.academicpress.com
Mixed trehalose/sucrose glasses used for protein incorporation as studied by infrared and optical spectroscopyq Wayne W. Wright, Juan Carlos Baez, and Jane M. Vanderkooi* Department of Biochemistry and Biophysics, Johnson Research Foundation, School of Medicine, University of Pennsylvania, Philadelphia, PA 19104, USA Received 21 March 2002
Abstract Evaporation of water from a 1/1 mixture of trehalose and sucrose gives rise to optically clear glasses that are transparent in the UV and visible ranges and do not crystallize when they are prepared at ambient temperatures. Two proteins, liver alcohol dehydrogenase and parvalbumin, and the tryptophan derivative N-acetyl-tryptophanamide were incorporated into the glasses. Infrared spectroscopy of the amide I band reveals that the proteins retain secondary structure in the glass over a temperature range of 20– 300 K. The amide II band of the protein and the HOH bending band of residual water in the glass shift with temperature changes, consistent with increased H-bonding strength as temperature is lowered. Phosphorescence of tryptophan can be seen from the proteins at room temperature, which shows the immobilization of the protein by the glass and the curbing of oxygen diffusion. It is suggested that using mixed sugars to form glasses is a way to immobilize proteins over a wide temperature range without distortions from solvent crystals. Ó 2002 Elsevier Science (USA). All rights reserved.
An outstanding problem of proteins, as related to folding specificity, dynamics, stability, and enzymic activity, is the role of solvent/peptide interactions. Changing the solvent dynamical characteristics provides a method to study the influences of solvent. One way to immobilize the protein environs is to incorporate proteins into sugar glasses. Such glasses provide the OH groups for surface H-binding, but are rigid. Trehalose, widely used in nature as a protector of proteins under dehydrating conditions [1], has been used to make glass suitable for protein incorporation [2–4]. Other sugars, such as sucrose, also form glasses which can accommodate proteins [5]. The physical characteristics of these glasses are currently being examined, with the view of understanding how sugars can stabilize proteins and for experimental situations where it is desirable to immobilize biomacromolecules [1,6–9]. An experimental limitation of the use of pure sugar to form glasses is that crystallization occurs when the sugar solution evaporates slowly at low or ambient temperatures, and many literq
This work was supported by NIH PO1 48130. Corresponding author. Fax: 215-573-2042. E-mail address: [email protected] (J.M. Vanderkooi).
ature procedures require that the glass be made at 60 °C or higher. We report that using a mixture of sugars circumvents this problem. Since crystals tend not to form, the glasses can be made at lower temperatures where the proteins are more stable. Furthermore, the glasses can be made over a variety of temperatures, and it is potentially possible to study temperature-induced fluctuations that are trapped when the exterior becomes rigid. In our work we use UV/visible absorption spectroscopy to demonstrate that the glass is optically transparent. IR spectroscopy was utilized to monitor the static structure of the mixed-sugar glass and the incorporated protein. Long-lived phosphorescence from trp, observed in many proteins at room temperature [10–12], was used to monitor the effect of the glass on protein dynamics.
Materials and methods Materials. D (+) trehalose (a-D -glucopyranosyl-a-D glucopyranoside), sucrose (a-D -glucopyranosyl-b-D fructofuranoside), alcohol dehydrogenase (LADH)1
*
1 Abbreviations used: LADH, liver alcohol dehydrogenase; NATA, N-acetyl-trytophanamide.
0003-2697/02/$ - see front matter Ó 2002 Elsevier Science (USA). All rights reserved. PII: S 0 0 0 3 - 2 6 9 7 ( 0 2 ) 0 0 0 3 4 - 9
168
W.W. Wright et al. / Analytical Biochemistry 307 (2002) 167–172
from equine liver, and N-acetyl-tryptophanamide (NATA) were obtained from Sigma Chemical (St. Louis, MO) Parvalbumin was prepared from frozen cod fish [13]. Spectroscopy. IR spectra were obtained with a Bruker IFS 66 Fourier transform IR instrument (Bruker, Brookline, MA). The sample compartment was purged with nitrogen to reduce the contribution from water vapor. The light levels were monitored using an HgCdTe (MCT) detector. The spectral resolution was 2 cm1 . The spectra were smoothed using a 9 Savitzky– Golay smoothing algorithm. The sample holder was obtained from Graseby Specac (Smyrna GA 30082). The sample temperature was maintained using an APD closed cycle Helitran cryostat (Advanced Research Systems, Allentown, PA). The cryostat sample chamber was filled with He gas at atmospheric pressure, which aids in the transfer of heat from the sample. The outer cryostat windows were made of CaF2 . The inner cryostat windows, which experience the temperature gradient, were 2 mm thick and were made of ZnSe (Janos Technology, Townsend, VT). A holder for these windows was constructed to minimize strain on the windows due to contraction at low temperature (Research Instrumentation Shop, University of Pennsylvania School of Medicine, Philadelphia, PA). The temperature was measured with a silicon diode near the sample and the temperature was controlled using a Model 9650 temperature controller (Scientific Instruments, West Palm Beach, FL). Cryogenic temperature profiles were carried out from high to low temperature, with the temperature being measured every 10 °C. Phosphorescence spectra and decay profiles were obtained using a Fluorolog 3–21 Jobin–Yvon Spex instrument equipped with a xenon flash lamp and phosphorimeter attachment. For 77 K measurement a cold-finger liquid N2 dewar (Kontes Glass, Vineland, NJ) was used. A Hitachi Perkin–Elmer absorption instrument was used to take the visible absorption spectra. Widths of peaks were determined using PeakFit (Jandel Scientific Software, San Rafael, CA). Glass formation. The glass was prepared as follows. Trehalose (300 mg) and sucrose (300 mg) was dissolved in 500 ml of distilled water to form the stock sugar solution. They solution was heated to 100 °C to insure complete dissolving of the sugar and removal of gases. The solution was cooled and 1 mg of protein or NATA was added. The glass was formed in several ways. In one, suitable for UV/vis, 600 ll of the stock sugar solution was pipetted to cover a 25-mm round quartz plate of 2 mm thickness. Quartz plates were obtained from Esco Products (Oak Ridge, NJ). For IR measurements, 10 ll of the sugar solution was diluted with water to 600 ml and the solution was plated on the CaF2 plate. CaF2 plates were obtained from Janos Technology. In the third method, a soda straw was split lengthwise, and
the solution was added. Upon drying, the semicircular glass rod, suitable for the cold-finger dewar, was removed from the straw. The samples were allowed to dry at 20 or 60 °C. For drying at room temperature the sample was held in a dessicator. During drying at 60 °C, the sample temperature was maintained using a VWR Scientific Products Heat Block. The resulting glass was hard to the touch and optically clear. Examination of the glass under crossed polarizers showed no indication of crystal formation.
Results UV/vis and IR absorption of the sucrose/trehalose glass. The UV/vis absorption spectrum of sucrose/trehalose glass, formed on a quartz plate, has no significant absorption over the range of 250 to 400 nm (Fig. 1). The absorption between 400 and 700 nm was likewise at the baseline (not shown). Fig. 1 also shows the absorption spectrum of NATA in the trehalose/sucrose glass. The absorption is typical of the indole ring, and it illustrates that the optical clarity of the glass allows for absorption spectroscopy even in the UV region. The mid-infrared spectrum of the sucrose/trehalose glass is shown in Fig. 2. The spectra resemble those reported for trehalose [14]. The HOH bending mode arising from residual water is clearly evident at 1651 cm1 at room temperature. This peak shifts by about 9 cm1 over a temperature range of 300 to 20 K
Fig. 1. UV/visible absorption of NATA in trehalose/sucrose glass. (A) Trehalose/sucrose glass on quartz plate; reference, quartz plate. (B) NATA in trehalose/sucrose on quartz plate; reference, trehalose/sucrose glass on quartz plate.
W.W. Wright et al. / Analytical Biochemistry 307 (2002) 167–172
169
Fig. 2. IR spectrum of trehalose/sucrose glass on a CaF2 plate. Temperature indicated by arrow. Peak at 1654 cm1 represents the HOH bending mode. The position of this peak is plotted as a function of temperature in the inset.
(Fig. 2, inset) and the temperature profile is continuous with no sharp transition that would indicate a phase transition or crystallization. In other experiments, the temperature was raised to 60 °C, and the glass remained stable. IR of a protein in the glass. The IR spectra of the amide region of parvalbumin in the sugar glass is shown in Fig. 3, where the spectrum of the sugar glass is subtracted from the glass containing the protein. The amide I region at 1650 cm1 , the carboxylate region at 1570 cm1 [15,16], and the amide II region at 1550 cm1 are seen. Similar IR spectra have been observed for proteins that are freeze-dried with trehalose [17] and for cytochrome c in a trehalose glass [18]. The amide I peak is characteristic of a a-helix. The amide I band is not significantly temperature dependent, showing that the ahelical structure is maintained over the temperature range of 300–20 K. The amide II peak shifts to higher frequency as temperature decreases. The frequencies as a function of temperature are shown in Fig. 3, inset. Phosphorescence from aromatic groups of proteins. NATA in the trehalose/sucrose glass that was formed at 60 °C shows phosphorescence at 20 °C (Fig. 4A). No
Fig. 3. IR spectrum of parvalbumin in trehalose/sucrose glass on a CaF2 plate. The position of the amide II band is plotted as a function of temperature in the inset.
deoxygenation is required to see the phosphorescence; this is in contrast to fluid samples [10]. The spectrum of the sample in trehalose shows significant sharpening at low temperature, but at 77 K the spectrum of NATA in trehalose/sucrose glass that was formed at 60 °C remains broader than NATA in glycerol/water glass (glass formation: )120 °C). The phosphorescence of the single tryptophan-containing protein parvalbumin is shown in Fig. 4B. At given temperatures, the spectra of parvalbumin are sharper than NATA in comparable glass, showing that the protein matrix is influencing the line widths. At 77 K, the 0,0 emission maximum of NATA in trehalose/glass occurs at 408.6 nm and its FWHM width is 8.5 nm. These respective values are 407 and 5.5 nm for parvalbumin in the trehalose/sucrose glass. The spectral bands of the 0,0 transition for the model compound and protein are narrower in glycerol/water at 77 K. The 0,0 band of NATA in glycerol/water is at 406.4 nm and its
170
W.W. Wright et al. / Analytical Biochemistry 307 (2002) 167–172
Fig. 4. Phosphorescence spectrum. (A) NATA, (B) parvalbumin; (C) LADH. Excitation, 285 nm; excitation band pass, 14 nm; emission, 3 nm; acquisition time, 500 ms; delay after flash, 1 ms. The spectra marked sugar refer to samples made in trehalose/sucrose glass as described under Material and methods. The spectra marked glycerol are the samples made in 50% glycerol/water. The temperature of measurement is shown.
FWHM width is 7.6 nm. These values are 407.7 and 5.3 nm, respectively, for parvalbumin in glycerol/water. Since the trehalose/sucrose forms the glass at higher temperature than does glycerol/water, a likely interpretation is that the conformational disorder arising from fluctuations at high temperature remain at low temperature. The spectrum of NATA, representing indole devoid of the polypeptide chain, shows the limiting case for exposed trp. The phosphorescence emission spectrum of LADH is shown in Fig. 4C. This protein has two trp/monomers. The sharp vibronic features of indole phosphorescence are evident, and the spectra line widths are seen to depend upon the solvent. We note that the LADH samples at low temperature show emission at 380 nm where tyr emits. LADH has four tyr/monomers. Decay of phosphorescence. The phosphorescence emission decay profiles of the NATA, paralbumin, and LADH samples were measured. For all samples in glass at ambient temperature the decays are not single exponential. The results are illustrated in Fig. 5. For all samples at 77 K the decay is long with a lifetime of 4.5 s. (This lifetime is somewhat shorter than the literature value of 5–6 s for tryptophan under the same conditions [19], but our instrument does not allow us to go to longer times. If the lifetime of trp phosphorescence is nonsingle exponential at low temperature, with close lifetimes of say 4–6 s, most fitting programs would not detect this, but our short time window,
Fig. 5. Phosphorescence decay of samples in trehalose/sucrose glass. d, NATA at 77 K; j, NATA at 20 °C; N, parvalbumin at 20 °C. Excitation wavelength, 285 nm, emission wavelength, 435 nm. For NATA: excitation band pass, 10 nm; emission, 10 nm; acquisition time window, 300 ms. For parvalbumin: excitation band pass, 14 nm; emission, 14 nm; acquisition time, window, 50 ms.
W.W. Wright et al. / Analytical Biochemistry 307 (2002) 167–172
relative to the lifetime, would select for the shorter component.)
Discussion The message of this paper is that evaporation of water from a mixture of two naturally occurring sugars in 1/1 ratio produces glass without evidence of crystal formation even when the sample is slowly dried at room temperature. The glasses are clear, nonabsorbing to UV/ visible light, and suitable for optical measurements. IR spectroscopy of the glass gives information on the nature of the glass. There is residual water in the glass as seen in the IR spectrum of the HOH bending mode. As temperature decreases, its frequency increases. An increase in the frequency of a bending mode is an indication of increased H-bonding [20], and it follows that as temperature decreases the H-bonding increases because rotational disorder decreases. We can take the change in frequency as an indication that water is rotationally mobile in the glass. The change in frequency is continuous with temperature, consistent with the glass nature of the sugar sample. The IR spectrum of parvalbumin in the trehalose/ sucrose glass indicates that the protein maintains its secondary helical structure in the glass. The amide stretch frequencies of the protein, like frequencies of the water modes, are affected by H-bonding [20]. We have previously interpreted an increase in absorption and a shift to low frequency of amide I for proteins in water to an increase in H-bonding strength to water as the temperature is lowered [21]. In contrast to protein samples in water, the amide I band is not (or very little) temperature dependent. Looking at the high frequency side of the amide I band, where the water contribution is small, there is no significant shift with temperature, and there was also none seen in cytochrome c [18]. The amide II band, however, decreases in absorbance (Fig. 2) and shifts significantly to higher frequency (Fig. 2, inset) as the temperature decreases. The amide II band arises predominately from an NAH bending mode, whereas amide I is predominately a C@O stretching mode [22]. The smaller size of the H atom may make the NAH bending mode more sensitive to H-bonding changes, but it is also the characteristic of glasses that all motions do not show the same behavior. The temperature dependence of protein fluctuations is a ramification of the nature of the forces holding proteins together. Both data of various experiments, such as diffusion of small molecules and neutron scattering, and computation suggest that hydrated proteins undergo a transition at around 220–250 K [23–25]. In the IR spectrum of the protein in a glass, we did not see this (Fig. 2). Lack of the transition was seen for dry proteins [26] and for protein in trehalose glass [27]. The
171
incorporation of proteins into glasses where the surrounding atoms are rigidified may be useful in elucidating whether a transition occurs in proteins where the surface is H-bonded to the OHs of the sugar and residual water but where the environs is rigid. In the case of parvalbumin, most residues are in a-helices and its small size means that many of the amide residues are H-bonded to the surface molecules. A potential use of the glass is to see whether large proteins and proteins with different folding motifs would show indication of temperature-related structural changes. We show that phosphorescence can be observed for the Trp derivative N-acetyl tryptophanamide and for proteins in mixed sugar glasses at room temperature. The lifetime of an excited triplet-state molecule is a function of the dynamics of the chromophore and its immediate environment [28] and, therefore, phosphorescence is a sensitive probe of protein dynamics [29–33]. Oxygen is a potent dynamical quencher of excited triplet states, and in fluid samples Trp phosphorescence can be seen only in the absence of oxygen. The observation that phosphorescence can be seen at room temperature for samples in the glass without deoxygenation is an indication that the glass is rigid enough to prevent the diffusion of O2 . The phosphorescence spectra of the samples in the trehalose/sucrose glass are broader at room temperature than at 77 K (Fig. 4), and the lifetimes are nonsingle exponential (Fig. 5). The data suggest that site heterogeneity contributes to the spectral features in the glass. The origin of the heterogeneity is of some interest, and one possibility is that some, but not all, residues are H-bonded to the residual water. Changing the water content can be achieved by equilibrating the sample at different humidity and monitoring the water content by IR. Heating the sample is another way to reduce residual water [14]. Long-lived phosphorescence at room temperature from Trp in proteins in fluid aqueous solution appears to be limited to buried residues in rigid parts of the protein [12]. An interesting feature of the phosphorescence emission of Trp for proteins in the glasses is that residues at the surface can also be seen, as indicated by the phosphorescence of NATA, which represents the indole ring without a protein. The active sites of proteins are usually in flexible regions, and therefore phosphorescence from Trp in these regions is usually not detected at room temperature. In the sugar glass this region would be immobilized, but would likely maintain the conformational heterogeneity at the glass formation temperature. Potential experiments include determining whether there are specific substrate–Trp interactions that can be detected by the phosphorescence of Trp. Another issue to be examined is how buried residues will be affected by immobilization of the surface. In the case of parvalbumin, the phosphorescence lifetime retained a short component (Fig. 5), suggesting that immobiliza-
172
W.W. Wright et al. / Analytical Biochemistry 307 (2002) 167–172
tion of the surface did not completely immobilize the interior. However, this needs further study. In summary, sugar glasses were prepared. Mixed sugar glasses can be formed without the formation of crystals at temperatures in the range of interest for proteins. They are transparent in the UV range and therefore they are suitable for the study of intrinsic chromophores in proteins. Proteins maintain their structure in these glasses. Phosphorescence measurements reveal that the glasses do not allow the diffusion of O2 and that the indole derivative is immobilized by incorporation into the glass.
References [1] L.M. Crowe, D.S. Reid, J.H. Crowe, Is trehalose special for preserving dry biomaterials?, Biophys. J. 71 (1996) 2087–2093. [2] D. Beece, L. Eisenstein, H. Frauenfelder, D. Good, M.C. Marden, L. Reinisch, A.H. Reynolds, L.B. Sorensen, K.T. Yue, Solvent viscosity and the dynamics of dioxygen binding to myoglobin, Biochemistry 19 (1980) 5147–5157. [3] L. Cordone, P. Galajda, E. Vitrano, A. Gassmann, A. Ostermann, F. Parak, A reduction of protein specific motions in CO-ligated myoglobin embedded in a trehalose glass, Eur. Biophys. J. 27 (1998) 173–176. [4] S.J. Hagen, J. Hofrichter, W.A. Eaton, Protein reaction kinetics in a room-temperature glass, Science 269 (1995) 959–962. [5] C.P. McCaul, R.D. Ludescher, Room temperature phosphorescence from tryptophan and halogenated tryptophan analogs in amorphous sucrose, Photochem. Photobiol. 70 (1999) 166–171. [6] S. Magazu, G. Maisano, P. Migliardo, V. Villari, Experimental simulation of macromolecules in trehalose aqueous solutions: a photon correlation spectroscopy study, J. Chem. Phys. 111 (1999) 9086–9092. [7] J.L. Green, C.A. Angell, Phase relations and vitrification in saccharide-water solutions and the trehalose anomaly, J. Phys. Chem. 93 (1989) 2880–2882. [8] C. Branca, S. Magazu, G. Maisano, P. Migliardo, a; a-Trehalosewater solutions. 3. Vibrational dynamics studies by inelastic light scattering, J. Phys. Chem. B 103 (1999) 1347–1353. [9] S. Magazu, G. Maisano, P. Migliardo, A.M. Musolino, V. Villari, Fragile-like behavior and H-bond interactions of the glassforming water–trehalose system, Philos. Mag. B 77 (1998) 655– 661. [10] J.M. Vanderkooi, D.B. Calhoun, S.W. Englander, On the prevalence of room temperature protein phosphorescence, Science 236 (1987) 568–569. [11] J.M. Vanderkooi, J.W. Berger, Excited triplet states used to study biological macromolecules at room temperature, Biochim. Biophys. Acta. 976 (1989) 1–27. [12] S. Papp, J.M. Vanderkooi, Tryptophan phosphorescence at room temperature as a tool to study protein structure and dynamics, Photochem. Photobiol. 49 (1989) 775–784. [13] K. Sudhakar, C.M. Phillips, C.S. Owen, J.M. Vanderkooi, Dynamics of parvalbumin studied by fluorescence emission and triplet absorption spectroscopy of tryptophan, Biochemistry 34 (1995) 1355–1363. [14] K. Akao, Y. Okubo, N. Askawa, Y. Inoue, M. Sukurai, Infrared spectroscopic study on the properties of the anhydrous form II of trehalose: implications for the functional mechanism of trehalose as a biostabilizer, Carbohydr. Res. 334 (2001) 233–241.
[15] M. Nara, M. Tasumi, M. Tanokura, T. Hiraoki, M. Yazawa, A. Tsutsumi, Infrared studies of interaction between metal ions and Ca2þ -binding proteins: marker bands for identifying the types of coordination of the side-chain COO- groups to metal ions in pike parvalbumin (pI ¼ 4.10), FEBS Lett. 349 (1994) 84–88. [16] M. Laberge, W.W. Wright, S. Sudhakar, P.A. Liebman, J.M. Vanderkooi, Conformational effects of calcium release from parvalbumin: comparison of computational simulations with spectroscopic investigations, Biochemistry 36 (1997) 5363– 5371. [17] J.F. Carpenter, J.H. Crowe, An infrared spectroscopic study of the interactions of carbohydrates with dried proteins, Biochemistry 28 (1989) 3916–3922. [18] N.V. Prabhu, S.D. Dalosto, K.A. Sharp, W.W. Wright, J.M. Vanderkooi, Optical spectra of FeII cytochrome c interpreted using molecular dynamics simulations and quantum mechanical calculations, J. Phys. Chem. 106 (2002) 5561–5571. [19] J.W. Longworth, Luminescence of polypeptides and Proteins, in: Excited States of Proteins and Nucleic Acids, 1971, pp. 319–484. [20] G.A. Jeffrey, An Introduction to Hydrogen Bonding, Oxford University Press, New York, 1997. [21] E.S. Manas, Z. Getahun, W.W. Wright, W.F. DeGrado, J.M. Vanderkooi, Infrared spectra of amide groups in a-helical proteins: evidence for hydrogen bonding to water, J. Am. Chem. Soc. 122 (2000) 9883–9890. [22] S. Krimm, J. Bandekar, Vibrational spectroscopy and conformation of peptides, polypeptides, and proteins, Adv. Protein Chem. 38 (1986) 181–364. [23] H. Frauenfelder, S.G. Sligar, P. Wolynes, The energy landscapes and motions of proteins, Science 254 (1991) 1598–1603. [24] W. Doster, S. Cusack, W. Petry, Dynamical transition of myoglobin revealed by ineleastic neutron scattering, Nature 337 (1989) 754–756. [25] K.E.S. Tang, K. Dill, Native protein fluctuations: the conformational-motion temperature and the inverse correlation of protein flexibility with protein stability, J. Biomol. Struct. Dyn. 16 (1998) 397–411. [26] A.M.U.T.J. Tsai, D.A. Neumann, The inverse relationship between protein dynamics and thermal stability, Biophys. J. 81 (2001) 2339–2343. [27] L. Cordone, M. Ferrand, E. Vitrano, G. Zaccai, Harmonic behavior of trehalose-coated carbon-monoxy-myoglobin at high temperature, Biophys. J. 76 (1999) 1043–1047. [28] G.B. Strambini, M. Gonnelli, The indole nucleus triplet-state lifetime and its dependence on solvent microviscosity, Chem. Phys. Lett. 115 (1985) 196–200. [29] G.B. Strambini, P. Cioni, R.A. Felicioli, Characterization of tryptophan environments in glutamate dehydrogenases from temperature-dependent phosphorescence, Biochemistry 26 (1987) 4968–4975. [30] D.B. Calhoun, S.W. Englander, W.W. Wright, J.M. Vanderkooi, Quenching of room temperature protein phosphorescence by added small molecules, Biochemistry 27 (1988) 8466–8474. [31] D.B. Calhoun, J.M. Vanderkooi, G.V. Woodrow III, S.W. Englander, Penetration of dioxygen into proteins studied by quenching of phosphorescence and fluorescence, Biochemistry 22 (1983) 1526–1532. [32] J.V. Mersol, D.G. Steel, A. Gafni, Quenching of Tryptophan Phosphorescence in Escherichia coli alkaline phosphatase by longrange transfer mechanisms to external agents in the rapiddiffusion limit, Biochemistry 30 (1991) 668–675. [33] C.J. Fischer, J.A. Schauerte, K.C. Wisser, A. Gafni, D.G. Steel, Hydrogen exchange at the core of Escherichia coli alkaline phosphatase studied by room-temperature tryptophan phosphorescence, Biochemistry 39 (2000) 1455–1461.
Mixed trehalose/sucrose glasses used for protein incorporation as studied by infrared and optical spectroscopyq Wayne W. Wright, Juan Carlos Baez, and Jane M. Vanderkooi* Department of Biochemistry and Biophysics, Johnson Research Foundation, School of Medicine, University of Pennsylvania, Philadelphia, PA 19104, USA Received 21 March 2002
Abstract Evaporation of water from a 1/1 mixture of trehalose and sucrose gives rise to optically clear glasses that are transparent in the UV and visible ranges and do not crystallize when they are prepared at ambient temperatures. Two proteins, liver alcohol dehydrogenase and parvalbumin, and the tryptophan derivative N-acetyl-tryptophanamide were incorporated into the glasses. Infrared spectroscopy of the amide I band reveals that the proteins retain secondary structure in the glass over a temperature range of 20– 300 K. The amide II band of the protein and the HOH bending band of residual water in the glass shift with temperature changes, consistent with increased H-bonding strength as temperature is lowered. Phosphorescence of tryptophan can be seen from the proteins at room temperature, which shows the immobilization of the protein by the glass and the curbing of oxygen diffusion. It is suggested that using mixed sugars to form glasses is a way to immobilize proteins over a wide temperature range without distortions from solvent crystals. Ó 2002 Elsevier Science (USA). All rights reserved.
An outstanding problem of proteins, as related to folding specificity, dynamics, stability, and enzymic activity, is the role of solvent/peptide interactions. Changing the solvent dynamical characteristics provides a method to study the influences of solvent. One way to immobilize the protein environs is to incorporate proteins into sugar glasses. Such glasses provide the OH groups for surface H-binding, but are rigid. Trehalose, widely used in nature as a protector of proteins under dehydrating conditions [1], has been used to make glass suitable for protein incorporation [2–4]. Other sugars, such as sucrose, also form glasses which can accommodate proteins [5]. The physical characteristics of these glasses are currently being examined, with the view of understanding how sugars can stabilize proteins and for experimental situations where it is desirable to immobilize biomacromolecules [1,6–9]. An experimental limitation of the use of pure sugar to form glasses is that crystallization occurs when the sugar solution evaporates slowly at low or ambient temperatures, and many literq
This work was supported by NIH PO1 48130. Corresponding author. Fax: 215-573-2042. E-mail address: [email protected] (J.M. Vanderkooi).
ature procedures require that the glass be made at 60 °C or higher. We report that using a mixture of sugars circumvents this problem. Since crystals tend not to form, the glasses can be made at lower temperatures where the proteins are more stable. Furthermore, the glasses can be made over a variety of temperatures, and it is potentially possible to study temperature-induced fluctuations that are trapped when the exterior becomes rigid. In our work we use UV/visible absorption spectroscopy to demonstrate that the glass is optically transparent. IR spectroscopy was utilized to monitor the static structure of the mixed-sugar glass and the incorporated protein. Long-lived phosphorescence from trp, observed in many proteins at room temperature [10–12], was used to monitor the effect of the glass on protein dynamics.
Materials and methods Materials. D (+) trehalose (a-D -glucopyranosyl-a-D glucopyranoside), sucrose (a-D -glucopyranosyl-b-D fructofuranoside), alcohol dehydrogenase (LADH)1
*
1 Abbreviations used: LADH, liver alcohol dehydrogenase; NATA, N-acetyl-trytophanamide.
0003-2697/02/$ - see front matter Ó 2002 Elsevier Science (USA). All rights reserved. PII: S 0 0 0 3 - 2 6 9 7 ( 0 2 ) 0 0 0 3 4 - 9
168
W.W. Wright et al. / Analytical Biochemistry 307 (2002) 167–172
from equine liver, and N-acetyl-tryptophanamide (NATA) were obtained from Sigma Chemical (St. Louis, MO) Parvalbumin was prepared from frozen cod fish [13]. Spectroscopy. IR spectra were obtained with a Bruker IFS 66 Fourier transform IR instrument (Bruker, Brookline, MA). The sample compartment was purged with nitrogen to reduce the contribution from water vapor. The light levels were monitored using an HgCdTe (MCT) detector. The spectral resolution was 2 cm1 . The spectra were smoothed using a 9 Savitzky– Golay smoothing algorithm. The sample holder was obtained from Graseby Specac (Smyrna GA 30082). The sample temperature was maintained using an APD closed cycle Helitran cryostat (Advanced Research Systems, Allentown, PA). The cryostat sample chamber was filled with He gas at atmospheric pressure, which aids in the transfer of heat from the sample. The outer cryostat windows were made of CaF2 . The inner cryostat windows, which experience the temperature gradient, were 2 mm thick and were made of ZnSe (Janos Technology, Townsend, VT). A holder for these windows was constructed to minimize strain on the windows due to contraction at low temperature (Research Instrumentation Shop, University of Pennsylvania School of Medicine, Philadelphia, PA). The temperature was measured with a silicon diode near the sample and the temperature was controlled using a Model 9650 temperature controller (Scientific Instruments, West Palm Beach, FL). Cryogenic temperature profiles were carried out from high to low temperature, with the temperature being measured every 10 °C. Phosphorescence spectra and decay profiles were obtained using a Fluorolog 3–21 Jobin–Yvon Spex instrument equipped with a xenon flash lamp and phosphorimeter attachment. For 77 K measurement a cold-finger liquid N2 dewar (Kontes Glass, Vineland, NJ) was used. A Hitachi Perkin–Elmer absorption instrument was used to take the visible absorption spectra. Widths of peaks were determined using PeakFit (Jandel Scientific Software, San Rafael, CA). Glass formation. The glass was prepared as follows. Trehalose (300 mg) and sucrose (300 mg) was dissolved in 500 ml of distilled water to form the stock sugar solution. They solution was heated to 100 °C to insure complete dissolving of the sugar and removal of gases. The solution was cooled and 1 mg of protein or NATA was added. The glass was formed in several ways. In one, suitable for UV/vis, 600 ll of the stock sugar solution was pipetted to cover a 25-mm round quartz plate of 2 mm thickness. Quartz plates were obtained from Esco Products (Oak Ridge, NJ). For IR measurements, 10 ll of the sugar solution was diluted with water to 600 ml and the solution was plated on the CaF2 plate. CaF2 plates were obtained from Janos Technology. In the third method, a soda straw was split lengthwise, and
the solution was added. Upon drying, the semicircular glass rod, suitable for the cold-finger dewar, was removed from the straw. The samples were allowed to dry at 20 or 60 °C. For drying at room temperature the sample was held in a dessicator. During drying at 60 °C, the sample temperature was maintained using a VWR Scientific Products Heat Block. The resulting glass was hard to the touch and optically clear. Examination of the glass under crossed polarizers showed no indication of crystal formation.
Results UV/vis and IR absorption of the sucrose/trehalose glass. The UV/vis absorption spectrum of sucrose/trehalose glass, formed on a quartz plate, has no significant absorption over the range of 250 to 400 nm (Fig. 1). The absorption between 400 and 700 nm was likewise at the baseline (not shown). Fig. 1 also shows the absorption spectrum of NATA in the trehalose/sucrose glass. The absorption is typical of the indole ring, and it illustrates that the optical clarity of the glass allows for absorption spectroscopy even in the UV region. The mid-infrared spectrum of the sucrose/trehalose glass is shown in Fig. 2. The spectra resemble those reported for trehalose [14]. The HOH bending mode arising from residual water is clearly evident at 1651 cm1 at room temperature. This peak shifts by about 9 cm1 over a temperature range of 300 to 20 K
Fig. 1. UV/visible absorption of NATA in trehalose/sucrose glass. (A) Trehalose/sucrose glass on quartz plate; reference, quartz plate. (B) NATA in trehalose/sucrose on quartz plate; reference, trehalose/sucrose glass on quartz plate.
W.W. Wright et al. / Analytical Biochemistry 307 (2002) 167–172
169
Fig. 2. IR spectrum of trehalose/sucrose glass on a CaF2 plate. Temperature indicated by arrow. Peak at 1654 cm1 represents the HOH bending mode. The position of this peak is plotted as a function of temperature in the inset.
(Fig. 2, inset) and the temperature profile is continuous with no sharp transition that would indicate a phase transition or crystallization. In other experiments, the temperature was raised to 60 °C, and the glass remained stable. IR of a protein in the glass. The IR spectra of the amide region of parvalbumin in the sugar glass is shown in Fig. 3, where the spectrum of the sugar glass is subtracted from the glass containing the protein. The amide I region at 1650 cm1 , the carboxylate region at 1570 cm1 [15,16], and the amide II region at 1550 cm1 are seen. Similar IR spectra have been observed for proteins that are freeze-dried with trehalose [17] and for cytochrome c in a trehalose glass [18]. The amide I peak is characteristic of a a-helix. The amide I band is not significantly temperature dependent, showing that the ahelical structure is maintained over the temperature range of 300–20 K. The amide II peak shifts to higher frequency as temperature decreases. The frequencies as a function of temperature are shown in Fig. 3, inset. Phosphorescence from aromatic groups of proteins. NATA in the trehalose/sucrose glass that was formed at 60 °C shows phosphorescence at 20 °C (Fig. 4A). No
Fig. 3. IR spectrum of parvalbumin in trehalose/sucrose glass on a CaF2 plate. The position of the amide II band is plotted as a function of temperature in the inset.
deoxygenation is required to see the phosphorescence; this is in contrast to fluid samples [10]. The spectrum of the sample in trehalose shows significant sharpening at low temperature, but at 77 K the spectrum of NATA in trehalose/sucrose glass that was formed at 60 °C remains broader than NATA in glycerol/water glass (glass formation: )120 °C). The phosphorescence of the single tryptophan-containing protein parvalbumin is shown in Fig. 4B. At given temperatures, the spectra of parvalbumin are sharper than NATA in comparable glass, showing that the protein matrix is influencing the line widths. At 77 K, the 0,0 emission maximum of NATA in trehalose/glass occurs at 408.6 nm and its FWHM width is 8.5 nm. These respective values are 407 and 5.5 nm for parvalbumin in the trehalose/sucrose glass. The spectral bands of the 0,0 transition for the model compound and protein are narrower in glycerol/water at 77 K. The 0,0 band of NATA in glycerol/water is at 406.4 nm and its
170
W.W. Wright et al. / Analytical Biochemistry 307 (2002) 167–172
Fig. 4. Phosphorescence spectrum. (A) NATA, (B) parvalbumin; (C) LADH. Excitation, 285 nm; excitation band pass, 14 nm; emission, 3 nm; acquisition time, 500 ms; delay after flash, 1 ms. The spectra marked sugar refer to samples made in trehalose/sucrose glass as described under Material and methods. The spectra marked glycerol are the samples made in 50% glycerol/water. The temperature of measurement is shown.
FWHM width is 7.6 nm. These values are 407.7 and 5.3 nm, respectively, for parvalbumin in glycerol/water. Since the trehalose/sucrose forms the glass at higher temperature than does glycerol/water, a likely interpretation is that the conformational disorder arising from fluctuations at high temperature remain at low temperature. The spectrum of NATA, representing indole devoid of the polypeptide chain, shows the limiting case for exposed trp. The phosphorescence emission spectrum of LADH is shown in Fig. 4C. This protein has two trp/monomers. The sharp vibronic features of indole phosphorescence are evident, and the spectra line widths are seen to depend upon the solvent. We note that the LADH samples at low temperature show emission at 380 nm where tyr emits. LADH has four tyr/monomers. Decay of phosphorescence. The phosphorescence emission decay profiles of the NATA, paralbumin, and LADH samples were measured. For all samples in glass at ambient temperature the decays are not single exponential. The results are illustrated in Fig. 5. For all samples at 77 K the decay is long with a lifetime of 4.5 s. (This lifetime is somewhat shorter than the literature value of 5–6 s for tryptophan under the same conditions [19], but our instrument does not allow us to go to longer times. If the lifetime of trp phosphorescence is nonsingle exponential at low temperature, with close lifetimes of say 4–6 s, most fitting programs would not detect this, but our short time window,
Fig. 5. Phosphorescence decay of samples in trehalose/sucrose glass. d, NATA at 77 K; j, NATA at 20 °C; N, parvalbumin at 20 °C. Excitation wavelength, 285 nm, emission wavelength, 435 nm. For NATA: excitation band pass, 10 nm; emission, 10 nm; acquisition time window, 300 ms. For parvalbumin: excitation band pass, 14 nm; emission, 14 nm; acquisition time, window, 50 ms.
W.W. Wright et al. / Analytical Biochemistry 307 (2002) 167–172
relative to the lifetime, would select for the shorter component.)
Discussion The message of this paper is that evaporation of water from a mixture of two naturally occurring sugars in 1/1 ratio produces glass without evidence of crystal formation even when the sample is slowly dried at room temperature. The glasses are clear, nonabsorbing to UV/ visible light, and suitable for optical measurements. IR spectroscopy of the glass gives information on the nature of the glass. There is residual water in the glass as seen in the IR spectrum of the HOH bending mode. As temperature decreases, its frequency increases. An increase in the frequency of a bending mode is an indication of increased H-bonding [20], and it follows that as temperature decreases the H-bonding increases because rotational disorder decreases. We can take the change in frequency as an indication that water is rotationally mobile in the glass. The change in frequency is continuous with temperature, consistent with the glass nature of the sugar sample. The IR spectrum of parvalbumin in the trehalose/ sucrose glass indicates that the protein maintains its secondary helical structure in the glass. The amide stretch frequencies of the protein, like frequencies of the water modes, are affected by H-bonding [20]. We have previously interpreted an increase in absorption and a shift to low frequency of amide I for proteins in water to an increase in H-bonding strength to water as the temperature is lowered [21]. In contrast to protein samples in water, the amide I band is not (or very little) temperature dependent. Looking at the high frequency side of the amide I band, where the water contribution is small, there is no significant shift with temperature, and there was also none seen in cytochrome c [18]. The amide II band, however, decreases in absorbance (Fig. 2) and shifts significantly to higher frequency (Fig. 2, inset) as the temperature decreases. The amide II band arises predominately from an NAH bending mode, whereas amide I is predominately a C@O stretching mode [22]. The smaller size of the H atom may make the NAH bending mode more sensitive to H-bonding changes, but it is also the characteristic of glasses that all motions do not show the same behavior. The temperature dependence of protein fluctuations is a ramification of the nature of the forces holding proteins together. Both data of various experiments, such as diffusion of small molecules and neutron scattering, and computation suggest that hydrated proteins undergo a transition at around 220–250 K [23–25]. In the IR spectrum of the protein in a glass, we did not see this (Fig. 2). Lack of the transition was seen for dry proteins [26] and for protein in trehalose glass [27]. The
171
incorporation of proteins into glasses where the surrounding atoms are rigidified may be useful in elucidating whether a transition occurs in proteins where the surface is H-bonded to the OHs of the sugar and residual water but where the environs is rigid. In the case of parvalbumin, most residues are in a-helices and its small size means that many of the amide residues are H-bonded to the surface molecules. A potential use of the glass is to see whether large proteins and proteins with different folding motifs would show indication of temperature-related structural changes. We show that phosphorescence can be observed for the Trp derivative N-acetyl tryptophanamide and for proteins in mixed sugar glasses at room temperature. The lifetime of an excited triplet-state molecule is a function of the dynamics of the chromophore and its immediate environment [28] and, therefore, phosphorescence is a sensitive probe of protein dynamics [29–33]. Oxygen is a potent dynamical quencher of excited triplet states, and in fluid samples Trp phosphorescence can be seen only in the absence of oxygen. The observation that phosphorescence can be seen at room temperature for samples in the glass without deoxygenation is an indication that the glass is rigid enough to prevent the diffusion of O2 . The phosphorescence spectra of the samples in the trehalose/sucrose glass are broader at room temperature than at 77 K (Fig. 4), and the lifetimes are nonsingle exponential (Fig. 5). The data suggest that site heterogeneity contributes to the spectral features in the glass. The origin of the heterogeneity is of some interest, and one possibility is that some, but not all, residues are H-bonded to the residual water. Changing the water content can be achieved by equilibrating the sample at different humidity and monitoring the water content by IR. Heating the sample is another way to reduce residual water [14]. Long-lived phosphorescence at room temperature from Trp in proteins in fluid aqueous solution appears to be limited to buried residues in rigid parts of the protein [12]. An interesting feature of the phosphorescence emission of Trp for proteins in the glasses is that residues at the surface can also be seen, as indicated by the phosphorescence of NATA, which represents the indole ring without a protein. The active sites of proteins are usually in flexible regions, and therefore phosphorescence from Trp in these regions is usually not detected at room temperature. In the sugar glass this region would be immobilized, but would likely maintain the conformational heterogeneity at the glass formation temperature. Potential experiments include determining whether there are specific substrate–Trp interactions that can be detected by the phosphorescence of Trp. Another issue to be examined is how buried residues will be affected by immobilization of the surface. In the case of parvalbumin, the phosphorescence lifetime retained a short component (Fig. 5), suggesting that immobiliza-
172
W.W. Wright et al. / Analytical Biochemistry 307 (2002) 167–172
tion of the surface did not completely immobilize the interior. However, this needs further study. In summary, sugar glasses were prepared. Mixed sugar glasses can be formed without the formation of crystals at temperatures in the range of interest for proteins. They are transparent in the UV range and therefore they are suitable for the study of intrinsic chromophores in proteins. Proteins maintain their structure in these glasses. Phosphorescence measurements reveal that the glasses do not allow the diffusion of O2 and that the indole derivative is immobilized by incorporation into the glass.
References [1] L.M. Crowe, D.S. Reid, J.H. Crowe, Is trehalose special for preserving dry biomaterials?, Biophys. J. 71 (1996) 2087–2093. [2] D. Beece, L. Eisenstein, H. Frauenfelder, D. Good, M.C. Marden, L. Reinisch, A.H. Reynolds, L.B. Sorensen, K.T. Yue, Solvent viscosity and the dynamics of dioxygen binding to myoglobin, Biochemistry 19 (1980) 5147–5157. [3] L. Cordone, P. Galajda, E. Vitrano, A. Gassmann, A. Ostermann, F. Parak, A reduction of protein specific motions in CO-ligated myoglobin embedded in a trehalose glass, Eur. Biophys. J. 27 (1998) 173–176. [4] S.J. Hagen, J. Hofrichter, W.A. Eaton, Protein reaction kinetics in a room-temperature glass, Science 269 (1995) 959–962. [5] C.P. McCaul, R.D. Ludescher, Room temperature phosphorescence from tryptophan and halogenated tryptophan analogs in amorphous sucrose, Photochem. Photobiol. 70 (1999) 166–171. [6] S. Magazu, G. Maisano, P. Migliardo, V. Villari, Experimental simulation of macromolecules in trehalose aqueous solutions: a photon correlation spectroscopy study, J. Chem. Phys. 111 (1999) 9086–9092. [7] J.L. Green, C.A. Angell, Phase relations and vitrification in saccharide-water solutions and the trehalose anomaly, J. Phys. Chem. 93 (1989) 2880–2882. [8] C. Branca, S. Magazu, G. Maisano, P. Migliardo, a; a-Trehalosewater solutions. 3. Vibrational dynamics studies by inelastic light scattering, J. Phys. Chem. B 103 (1999) 1347–1353. [9] S. Magazu, G. Maisano, P. Migliardo, A.M. Musolino, V. Villari, Fragile-like behavior and H-bond interactions of the glassforming water–trehalose system, Philos. Mag. B 77 (1998) 655– 661. [10] J.M. Vanderkooi, D.B. Calhoun, S.W. Englander, On the prevalence of room temperature protein phosphorescence, Science 236 (1987) 568–569. [11] J.M. Vanderkooi, J.W. Berger, Excited triplet states used to study biological macromolecules at room temperature, Biochim. Biophys. Acta. 976 (1989) 1–27. [12] S. Papp, J.M. Vanderkooi, Tryptophan phosphorescence at room temperature as a tool to study protein structure and dynamics, Photochem. Photobiol. 49 (1989) 775–784. [13] K. Sudhakar, C.M. Phillips, C.S. Owen, J.M. Vanderkooi, Dynamics of parvalbumin studied by fluorescence emission and triplet absorption spectroscopy of tryptophan, Biochemistry 34 (1995) 1355–1363. [14] K. Akao, Y. Okubo, N. Askawa, Y. Inoue, M. Sukurai, Infrared spectroscopic study on the properties of the anhydrous form II of trehalose: implications for the functional mechanism of trehalose as a biostabilizer, Carbohydr. Res. 334 (2001) 233–241.
[15] M. Nara, M. Tasumi, M. Tanokura, T. Hiraoki, M. Yazawa, A. Tsutsumi, Infrared studies of interaction between metal ions and Ca2þ -binding proteins: marker bands for identifying the types of coordination of the side-chain COO- groups to metal ions in pike parvalbumin (pI ¼ 4.10), FEBS Lett. 349 (1994) 84–88. [16] M. Laberge, W.W. Wright, S. Sudhakar, P.A. Liebman, J.M. Vanderkooi, Conformational effects of calcium release from parvalbumin: comparison of computational simulations with spectroscopic investigations, Biochemistry 36 (1997) 5363– 5371. [17] J.F. Carpenter, J.H. Crowe, An infrared spectroscopic study of the interactions of carbohydrates with dried proteins, Biochemistry 28 (1989) 3916–3922. [18] N.V. Prabhu, S.D. Dalosto, K.A. Sharp, W.W. Wright, J.M. Vanderkooi, Optical spectra of FeII cytochrome c interpreted using molecular dynamics simulations and quantum mechanical calculations, J. Phys. Chem. 106 (2002) 5561–5571. [19] J.W. Longworth, Luminescence of polypeptides and Proteins, in: Excited States of Proteins and Nucleic Acids, 1971, pp. 319–484. [20] G.A. Jeffrey, An Introduction to Hydrogen Bonding, Oxford University Press, New York, 1997. [21] E.S. Manas, Z. Getahun, W.W. Wright, W.F. DeGrado, J.M. Vanderkooi, Infrared spectra of amide groups in a-helical proteins: evidence for hydrogen bonding to water, J. Am. Chem. Soc. 122 (2000) 9883–9890. [22] S. Krimm, J. Bandekar, Vibrational spectroscopy and conformation of peptides, polypeptides, and proteins, Adv. Protein Chem. 38 (1986) 181–364. [23] H. Frauenfelder, S.G. Sligar, P. Wolynes, The energy landscapes and motions of proteins, Science 254 (1991) 1598–1603. [24] W. Doster, S. Cusack, W. Petry, Dynamical transition of myoglobin revealed by ineleastic neutron scattering, Nature 337 (1989) 754–756. [25] K.E.S. Tang, K. Dill, Native protein fluctuations: the conformational-motion temperature and the inverse correlation of protein flexibility with protein stability, J. Biomol. Struct. Dyn. 16 (1998) 397–411. [26] A.M.U.T.J. Tsai, D.A. Neumann, The inverse relationship between protein dynamics and thermal stability, Biophys. J. 81 (2001) 2339–2343. [27] L. Cordone, M. Ferrand, E. Vitrano, G. Zaccai, Harmonic behavior of trehalose-coated carbon-monoxy-myoglobin at high temperature, Biophys. J. 76 (1999) 1043–1047. [28] G.B. Strambini, M. Gonnelli, The indole nucleus triplet-state lifetime and its dependence on solvent microviscosity, Chem. Phys. Lett. 115 (1985) 196–200. [29] G.B. Strambini, P. Cioni, R.A. Felicioli, Characterization of tryptophan environments in glutamate dehydrogenases from temperature-dependent phosphorescence, Biochemistry 26 (1987) 4968–4975. [30] D.B. Calhoun, S.W. Englander, W.W. Wright, J.M. Vanderkooi, Quenching of room temperature protein phosphorescence by added small molecules, Biochemistry 27 (1988) 8466–8474. [31] D.B. Calhoun, J.M. Vanderkooi, G.V. Woodrow III, S.W. Englander, Penetration of dioxygen into proteins studied by quenching of phosphorescence and fluorescence, Biochemistry 22 (1983) 1526–1532. [32] J.V. Mersol, D.G. Steel, A. Gafni, Quenching of Tryptophan Phosphorescence in Escherichia coli alkaline phosphatase by longrange transfer mechanisms to external agents in the rapiddiffusion limit, Biochemistry 30 (1991) 668–675. [33] C.J. Fischer, J.A. Schauerte, K.C. Wisser, A. Gafni, D.G. Steel, Hydrogen exchange at the core of Escherichia coli alkaline phosphatase studied by room-temperature tryptophan phosphorescence, Biochemistry 39 (2000) 1455–1461.