J. Phys. Chem. B 2000, 104, 4791-4798
4791
Effects of Threonine 203 Replacements on Excited-State Dynamics and Fluorescence Properties of the Green Fluorescent Protein (GFP) Andreas D. Kummer,† Jens Wiehler,‡ Hermann Rehaber,† Christian Kompa,† Boris Steipe,‡ and Maria Elisabeth Michel-Beyerle*,† Institut fu¨ r Physikalische und Theoretische Chemie, Technische UniVersita¨ t Mu¨ nchen, 85748 Garching, Germany, and Genzentrum der Ludwig-Maximilians-UniVersita¨ t Mu¨ nchen, Feodor-Lynen-Strasse 25, 81377 Mu¨ nchen, Germany ReceiVed: December 3, 1999; In Final Form: February 22, 2000
We report a comparative study of wild-type green fluorescent protein (GFP) and single-site mutants in which threonine at position 203 has been replaced by aliphatic and aromatic residues, i.e., by valine (V), isoleucine (I), phenylalanine (F), tyrosine (Y), and histidine (H). Steady-state absorption spectra reveal changes that reflect different charge distributions in the mutants as compared to wild-type GFP. While the absorption peak of the protonated fluorophore, RH, undergoes only a small red shift in all T203 mutants, a pronounced red shift is observed for the deprotonated form R-, ca. 1000 cm-1 for the aliphatic mutants T203V and T203I, ca. 1200 cm-1 for T203F, and 1360 cm-1 for T203Y. Thus, we conclude that a ground-state conformation higher in energy than the wild-type R- state is the predominant origin of the red shift in all the T203 mutants investigated. Furthermore, mutant-dependent changes in the ground-state equilibria of RH and R- result from at least two modes of electrostatic stabilization, one resting on hydrogen bonding as in T203 and the other one on π-π-stacking as in T203F and T203Y. Surprisingly, the deprotonation dynamics of RH* is only weakly affected by the mutations at position 203. Only in the most red-shifted mutant T203Y an additional ultrafast (1.7 ps) excited-state decay channel of RH* has been observed. The identical kinetics of both processes, decay of RH* and ground-state recovery of RH in T203Y, is discussed in terms of two mechanisms: (i) rate-determining electron transfer from the protonated (or deprotonated) tyrosyl 203 residue to RH* followed by considerably faster recombination processes, which cannot occur in T203F for energetic reasons, and (ii) internal conversion in RH* favored by rotational motion around the exocyclic double bond.
1. Introduction In recent years, the green fluorescent protein (GFP) from the jellyfish Aequorea Victoria gained widespread interest in biochemistry and cell biology.1 Owing to the fact that its fluorophore forms internally from the tripeptide Ser65-Tyr66Gly67 and no external cofactor is required,2,3 GFP can be used as a noninvasive marker for gene expression and protein localization in intact cells and organisms. Crystallographic studies revealed that the chromophore is tightly packed in a canlike tertiary structure.4,5 This so-called β-can not only provides the environment necessary for chromophore formation6 but is also responsible for the high fluorescence quantum yield: denatured GFP and model compounds of the fluorophore are nonfluorescent at room temperature7,8 but become highly fluorescent at low temperatures, e.g., at 77 K,8 suggesting that restriction of the molecular motion of the chromophore within the protein favors GFP fluorescence. On the basis of spectroscopic investigations and random mutagenesis experiments, the two bands at 397 and 477 nm in the absorption spectrum of wild-type GFP were attributed to two different states of the chromophore, one (RH) carrying a neutral phenol and the other (R-) a phenolate group, respectively.2,3 Both species are interconnected by a ground-state * Corresponding author. Email:
[email protected]. Fax: +4989-289-13026. † Technische Universita ¨ t Mu¨nchen. ‡ Genzentrum der Ludwig-Maximilians-Universita ¨ t Mu¨nchen.
equilibrium that is affected by environmental parameters such as protein concentration, ionic strength, pH, temperature, and cryoprotectors, e.g., glycerol.9,11 Excitation in either band leads to green fluorescence with almost identical emission spectra. Since phenols in their excited states are much more acidic than in the ground state, it was proposed that independent of the excitation wavelength R-* is the emitting species.3 Indeed, timeresolved spectroscopic measurements showed convincingly that upon excitation of RH, the chromophore undergoes rapid excited-state deprotonation RH* f R-* followed by the characteristic green emission and reprotonation in the ground state.10,11 The two intermediates R-* and R- in this so-called Fo¨rster cycle, however, are formed in a protein environment adapted to the protonated chromophore RH, which is different from that of the R- species present in the ground-state equilibrium. This conclusion is derived from a small difference in the emission peak wavelength at room temperature and more evident differences in the fluorescence spectra at low temperatures.10,11 Brejc et al.12 attributed these two different protein environments of the R- chromophore mainly to distinct orientations of threonine 203 (Figure 1A). While in the R- species present in the ground-state equilibrium the T203 hydroxy group is hydrogen bonded to the anionic phenolate, thereby stabilizing the negative charge, it is turned away from the protonated chromophore RH as well as from the R- species formed by excited-state proton transfer (ESPT). In this conformation the
10.1021/jp9942522 CCC: $19.00 © 2000 American Chemical Society Published on Web 04/22/2000
4792 J. Phys. Chem. B, Vol. 104, No. 19, 2000
Kummer et al.
A)
B)
Figure 1. (A, top) Main features of the mechanism for the photoisomerization of wild-type GFP based on ref 12. (B, bottom) Immediate chromophore environment of YFP.17
carbonyl oxygen of the backbone of T203 moves closer and accepts a hydrogen bond from the water molecule W22 (for a more detailed discussion of the structure see also ref 13). Whenever in the following the difference between the two protein conformations needs to be emphasized, we will use the notations RI- and RB- for the deprotonated chromophore, the subscript I referring to the intermediate protein conformation after ESPT and the subscript B to the ground-state equilibrium of wild-type GFP. (The structure in Figure 1A has recently been
doubted by FTIR measurements14 not showing the change in charge state of Glu-222 proposed to occur upon interconversion between RH and R-. However, it should be kept in mind that the FTIR signals of RH were compared to the ones of R-, the latter being generated in UV photoconversion. This detail seems to be important since it might point to a special conformation of R- characterized by a reduced quantum yield relative to RH. On the other hand, in the situation presented here the loss of quantum yield clearly predominates in RH. Thus, the interpreta-
Effects of Threonine 203 Replacements on GFP tion of our observations is based on the commonly accepted change in orientation of Thr-203 described above, not further discussing the charge state of Glu-222.) One motivation in GFP mutagenesis studies is to engineer variants with altered fluorescence excitation and emission spectra allowing, for example, simultaneous localization of two or more fusion proteins or FRET (fluorescence resonance energy transfer) experiments while maintaining the high quantum yield of wild-type GFP. Of particular interest in this context are mutants, the so-called “yellow fluorescent proteins” (YFPs), which carry the aromatic tyrosine or phenylalanine instead of threonine at position 203.5,15,16,17 These multiple mutants have the most red-shifted absorption (of ground-state R-) and fluorescence emission peaks achieved so far in Aequorea Victoria GFP variants, reaching peak positions of 516 and 529 nm, respectively.5 On the basis of the finding that the introduced tyrosine is coplanar and in close proximity (ca. 3.3 Å) to the tyrosyl group of the chromophore (Figure 1B), π-π interaction between the two residues has been invoked to be responsible for the red shift.17 However, it remains unclear why this postulated π-π interaction has only little effect on the spectral features of the RH species.17 Moreover, quantumchemical calculations of ground- and excited-state dipole moments18 suggest even a blue shift for R- in this π-stacked configuration. There are also few reports on the spectroscopic phenomenology of nonaromatic T203 mutants. Two similar T203I mutants were shown to shift the chromophore’s ground-state equilibrium toward RH.3,20 Owing to the missing hydroxy group of T203, there is less stabilization of R- and the corresponding absorption band is almost entirely lost. Another variant containing T203I in a multiple mutation environment (Ala1b/F64L/S65T/Q80R/ V163A/T203I/K238N) was reported to have both a RH and a R- peak, the latter red-shifted to 507 nm.21 This large red shift was assumed to arise not from any of the single mutations alone but rather from a cooperative effect.21 While YFPs exhibit the desirable shift of the fluorescence band from green to yellow, some of them have an unfavorable low fluorescence quantum yield, especially when excited in the RH band.15,17 In a previous paper, we attributed the corresponding fast loss channel to enhanced internal conversion due to increased degrees of rotational freedom of the chromophore.15 This interpretation seems somewhat problematic in view of the dense packing revealed in the X-ray structure (Figure 1B).17 However, Weber et al. argue that the YFP class characteristic stacking of an aromatic residue at position 203 with the chromophore favors a more complex and less space-consuming pattern of rotational motion enhancing internal conversion.18 Furthermore, as shown in Figure 1B, Glu-222 is proposed to be hydrogen bonded to the heterocyclic nitrogen of the chromophore originating from Tyr-66,17 indicating that also other chromophore states such as the cation and the zwitterion might be involved that are supposed to be more prone to fast radiationless losses.18,19 In view of the open questions related to the origin of red shifts in absorption spectra and of small quantum yield in green emission in mutations involving T203, the goal of this work is a systematic analysis of spectroscopic effects of packing, π-π interactions, and hydrogen-bonding effects based on a set of single-site aliphatic and aromatic mutations at the position 203. The primary focus will be on excited-state dynamics investigated by picosecond/femtosecond time-resolved spectroscopy.
J. Phys. Chem. B, Vol. 104, No. 19, 2000 4793 2. Materials and Methods 2.1. Protein Expression, Purification, and Characterization. GFP used in this study was based on the pGFP gene,22 which was prepared as previously described.23 All proteins carry the mutation S2G and an additional C-terminal Gly, followed by an His6-Tag. The mutants T203V, T203I, T203F, T203Y, and T203H were constructed by site-directed mutagenesis24 and confirmed by DNA sequencing. The proteins were expressed in Escherichia coli BL21 DE3 at 25 °C. Expression was induced with 0.5 mM IPTG, and cells were harvested after 4 h. After sonification, total cytoplasmic protein was applied to a NiNTA Agarose column (Quiagen). Columns were washed with 6 volumes of 50 mM imidazole, 300 mM NaCl, and 50 mM Na2HPO4, pH 8.0, and eluted with 3.5 column volumes by increasing the imidazole concentration to 0.3 M in one step. Owing to small yields of soluble protein, some GFP variants were contaminated with E. coli protein. In this case purification was repeated once or twice with a small size Ni-NTA column. Purity was indicated via SDS-PAGE and silver staining. Protein yields after the first purification step were 0.15-31 mg per gram bacterial wet weight. The samples were solubilized in PBS buffer, pH 7.4 (4 mM KH2PO4, 16 mM Na2HPO4, 115 mM NaCl). 2.2. Spectroscopic Measurements. 2.2.1. Steady-State Absorption and Fluorescence Measurements. Steady-state absorption and fluorescence spectra were measured as previously described with e2.0 nm resolution.11 A 1 mm path-length quartz cuvette was used in both spectrometers. Fluorescence was detected in a front-face geometry. 2.2.2. Picosecond Time-ResolVed Fluorescence Measurements. Fluorescence kinetics were measured using a synchroscan streak camera (Hamamatsu C1578). Excitation light pulses at 400 nm with an average excitation power of 60 µW were obtained by frequency-doubling the output of a Ti:sapphire laser (Coherent MIRA) using a BBO crystal. Using standard deconvolution procedures, averaged lifetimes shorter (down to a third) than the fwhm of the apparatus response function (in this instance 6 ps) were considered to have an ambiguity of a factor of 2. All other averaged lifetimes carry an approximate error of λ) and k2 being activationless, thereby maximizing the electron-transfer rate for a given electronic coupling. For an overall energy gap of 3.1 eV given by the S0 f S1 excitation energy of RH, such a situation could only be realized for a relatively large value of λ of the order of 1 eV. Such a large reorganization energy typical for electron-transfer processes in polar environment might be conceivable for the GFP fluorophore, which is surrounded by polarizable amino acid residues and water molecules. On the other hand, previous lowtemperature experiments on a related triple mutant (T203Y/ S65T/F64L) revealed a dramatic increase in the lifetime of RH* of a factor of 40 upon cooling to 80 K.15 However, owing to nuclear quantum effects in the inverted region only a weak temperature dependence of the electron-transfer rate is expected and in general observed.29 Thus, the dramatic increase of the lifetime of RH* at low temperatures reflecting a high activation energy for the recombination process is difficult to reconcile with electron transfer in the inverted Marcus regime. (ii) Ultrafast Internal ConVersion in RH*. An alternative mechanism explaining the fast decay pathway of RH* in T203Y is internal conversion induced by increasing the motional degrees of freedom around the exocyclic C-C bonds of the chromophore. The central difficulty with this mechanism is the compactness of the structure in the vicinity of the chromophore as revealed by X-ray structural data on the T203Y multiple
4798 J. Phys. Chem. B, Vol. 104, No. 19, 2000 mutant.17 Quantum chemical calculations18,19 indicate that the spatial requirements and activation barriers for rotational motion around the exocyclic C-C bonds depend on the nature of the excited state, which may be cationic or neutral in the case of RH* and anionic or zwitterionic in case of RI-*. According to ref 18 the spatial requirements for rotational motion are minimized in the case of the so-called hula-twist motion, i.e., a concerted and simultaneous rotation around both exocyclic C-C bonds. Assuming similar structures for T203Y and T203F, the essential difference between the two mutants might be the degree of protonation of the heterocyclic nitrogen (N2) (Figure 1B). Protonation might be more extensive in the presence of tyrosine, either in the ground state or only in the excited state of the chromophore. In quantum chemical calculations internal conversion is in fact favored for the cation,19 which carries a proton at the site N2. To explain the absence of the fast decay channel in the phenylalanine mutant T203F, the protonation of the heterocyclic nitrogen has to be related to the presence of the hydroxy group of tyrosine 203 in its protonated or deprotonated state. This hypothesis will be tested in resonance Raman experiments. 5. Conclusions Single mutants of GFP, in which threonine at the position 203 has been replaced by aliphatic and aromatic amino acids, have been investigated in steady-state and femtosecond/ picosecond time-resolved spectroscopy. The results are summarized as follows: (1) In contrast to a weak red shift of the absorption peak of the protonated chromophore (RH), the peak position of the deprotonated form R- undergoes a pronounced red shift of approximately 1000 cm-1 in the aliphatic mutants T203V and T203I and 1200 and 1360 cm-1 in the aromatic mutants T203F and T203Y, respectively. These observations point to a less relaxed protein environment around R- in these mutants and contradict a major role of π-π-interaction in the red shift. (2) The ground-state equilibrium between the protonated and deprotonated forms RH and R- is affected by at least two modes of stabilization, one resting on the hydrogen bond network as in T203 and the other one on π-π-interaction between the chromophore Y66 and aromatic residues F203 and Y203. (3) In the aliphatic mutants, T203V and T203I, the excitedstate proton-transfer rate is slower by a factor of 4, while it is constant within a factor of 2 in the aromatic mutant T203Y as compared to wild-type. (4) Although electron transfer from Y203 to the tyrosyl residue Y66 of the chromophore cannot be excluded, the qualitative discussion of the ultrafast decay of RH* in T203Y favors internal conversion. This process is assumed to be induced by rotational motion around the exocyclic C-C bonds, preferentially motions with minimum spatial requirements. On the basis of quantum chemical calculations,19 the cationic species (RH2+)*, where the heterocyclic nitrogen is protonated, would be favored. If this mechanism holds, selective protonation not occurring in the mutant T203F would be responsible for the ultrafast internal conversion in T203Y. Nonexponential kinetics would be consistent with the extreme sensitivity of rotational motion on both spatial freedom and the degree of protonation of the heterocyclic nitrogen. In summary, replacement of threonine 203 by valine, isoleucine, phenylalanine, and histidine yields red-shifted fluores-
Kummer et al. cence while maintaining high quantum yields after excitation of either the protonated or the deprotonated state of the chromophore. These features also hold for the tyrosine mutant with the exception that excitation of the protonated chromophore opens a fast decay channel winning over excited-state deprotonation. Acknowledgment. We are highly indebted to Dr. Alexander Voityuk and Prof. N. Ro¨sch for stimulating discussions and critical reading of the manuscript. Last but not least, we thank our referees for very interesting comments and suggestions. Financial support from the Deutsche Forschungsgemeinschaft (SFB 377) is gratefully acknowledged. References and Notes (1) Tsien, R. Y. Annu. ReV. Biochem. 1998, 67, 509. (2) Cubitt, A. B.; Heim, R.; Adams, S. R.; Boyd, A. E.; Gross, L. A.; Tsien, R. Y. TIBS 1995, 20, 448. (3) Heim, R.; Prasher, D. C.; Tsien, R. Y. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 12501. (4) Yang, F.; Moss, L. G.; Phillips, G. N., Jr. Nat. Biotechnol. 1996, 14, 1246. (5) Ormo¨, M.; Cubitt, A. B.; Kallio, K.; Gross, L. A.; Tsien, R. Y.; Remington, S. J. Science 1996, 273, 1392. (6) Dopf, J.; Horiagon, T. M. Gene 1996, 173, 39. (7) Ward, W. W.; Cody, C. W.; Hart, R. C.; Cormier, M. J. Photochem. Photobiol. 1980, 31, 611. (8) Niwa, H.; Inouye, S.; Hirano, T.; Matsuno, T.; Kojima, S.; Kubota, M.; Ohashi, M.; Tsuji, F. I. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 13617. (9) Ward, W. W.; Prentice, H. J.; Roth, A. F.; Cody, C. W.; Reeves, S. C. Photochem. Photobiol. 1982, 35, 803. (10) Chattoraj, M.; King, B. A.; Bublitz, G. U.; Boxer, S. G. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 8362. (11) Lossau, H.; Kummer, A.; Heinecke, R.; Po¨llinger-Dammer, F.; Kompa, C.; Bieser, G.; Jonsson, T.; Silva, C. M.; Yang, M. M.; Youvan, D. C.; Michel-Beyerle, M. E. Chem. Phys. 1996, 213, 1. (12) Brejc, K.; Sixma, T. K.; Kitts, P. A.; Kain, S. R.; Tsien, R. Y.; Ormo¨, M.; Remington, S. J. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 2306. (13) Helms, V.; Straatsma, T. P.; McCammon, J. A. J. Phys. Chem. B 1999, 103, 3263. (14) Van Thor, J. J.; Pierik, A. J.; Nugteren-Roodzant, I.; Xie, A.; Hellingwerf, K. J. Biochemistry 1998, 37, 16915. (15) Kummer, A. D.; Kompa, C.; Lossau, H.; Po¨llinger-Dammer, F.; Michel-Beyerle, M. E.; Silva, C. M.; Bylina, E. J.; Coleman, W. J.; Yang, M. M.; Youvan, D. C. Chem. Phys. 1998, 237, 183. (16) Dickson, R. M.; Cubitt, A. B.; Tsien, R. Y.; Moerner, W. E. Nature 1997, 388, 355. (17) Wachter, R. M.; Elsliger, M. A.; Kallio, K.; Hanson, G. T.; Remington, S. J. Structure 1998, 6, 1267. (18) Weber, W.; Helms, V.; McCammon, J. A.; Langhoff, P. W. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 6177. (19) Voityuk, A. A.; Michel-Beyerle, M. E.; Ro¨sch, N. Chem. Phys. Lett. 1998, 296, 269. (20) Ehrig, T.; O’Kane, D. J.; Prendergast, F. G. FEBS Lett. 1995, 367, 163. (21) Palm, G. J.; Wlodawer, A. Methods Enzymol. 1999, 302, 378. (22) Chalfie, M.; Tu, Y.; Euskirchen, G.; Ward, W. W.; Prasher, D. C. Science 1994, 263, 802. (23) Jung, G.; Wiehler, J.; Go¨hde, W.; Tittel, J.; Basche´, T.; Steipe, B.; Bra¨uchle, C. Bioimaging 1998, 6, 54. (24) Kunkel, T. A. Proc. Natl. Acad. Sci. U.S.A. 1985, 82, 488. (25) Po¨llinger, F.; Musewald, C.; Heitele, H.; Michel-Beyerle, M. E.; Anders, C.; Futscher, M.; Voit, G.; Staab, H. A. Ber. Bunsen-Ges. Phys. Chem. 1996, 100, 2076. (26) Voityuk, A. A.; Michel-Beyerle, M. E.; Ro¨sch, N. Chem. Phys. 1998, 231, 13. (27) Helms, V.; Winstead, C.; Langhoff, P. W. J. Mol. Struct.: THEOCHEM, in press. (28) Creemers, T. M. H.; Lock, A. J.; Subramaniam, V.; Jovin, T. M.; Vo¨lker, S. Nat. Struct. Biol. 1999, 6, 557. (29) Electron TransfersFrom Isolated Molecules to Biomolecules, (Jortner, J., and Bixon, M., Eds.); John Wiley: New York, 1999; Part 1, 69-78.