Luminescent and substrate binding activities of firefly luciferase N-terminal domain

Biochimica et Biophysica Acta 1649 (2003) 183 – 189 www.bba-direct.com

Luminescent and substrate binding activities of firefly luciferase N-terminal domain Tamotsu Zako1, Keiichi Ayabe, Takahide Aburatani, Noriho Kamiya 2, Atsushi Kitayama 3, Hiroshi Ueda *, Teruyuki Nagamune* Department of Chemistry and Biotechnology, School of Engineering, University of Tokyo, 7-3-1 Hongo, Bunkyo, Tokyo 113-8656, Japan Received 2 January 2003; received in revised form 14 April 2003; accepted 30 April 2003

Abstract Firefly luciferase catalyzes highly efficient emission of light from the substrates luciferin, Mg-ATP, and oxygen. A number of amino acid residues are identified to be important for the luminescent activity, and almost all the key residues are thought to be located in the N-terminal domain (1 – 437), except one in the C-terminal domain, Lys529, which is thought to be critical for efficient substrate orientation. Here we show that the purified N-terminal domain still binds to the substrates luciferin and ATP with reduced affinity, and retains luminescent activity of up to 0.03% of the wild-type enzyme (WT), indicating that all the essential residues for the activity are located in the N-terminal domain. Also found is low luminescence enhancement by coenzyme A (CoA), which implies a lower product inhibition than in the WT enzyme. These findings have interesting implications for the light emission reaction mechanism of the enzyme, such as reaction intermediates, product inhibition, and the role of the C-terminal domain. D 2003 Elsevier B.V. All rights reserved. Keywords: Truncated firefly luciferase; The N-terminal domain; Luminescent activity; ATP binding

1. Introduction Basic research on firefly bioluminescence has been largely focused on the North American firefly Photinus pyralis to provide an understanding of how light is produced by fireflies [1,2]. Including the structural studies of firefly luciferase and its closely related enzyme gramicidin S synthase by Brick et al. [3 –5], and subsequent mutagenesis studies by Branchini et al. [6 –8], considerable advances Abbreviations: C-domain, C-terminal domain of Photinus pyralis luciferase; CoA, coenzyme A; GST, glutathione S-transferase; IPTG, isopropyl h-D-thiogalactoside; LH2, D-luciferin; N-domain, N-terminal domain of P. pyralis luciferase; PBS, phosphate buffered saline; PCR, polymerase chain reaction; WT, wild-type P. pyralis luciferase * Corresponding authors. Tel.: +81-3-5841-7328; fax: +81-3-58418657. E-mail addresses: [email protected] (H. Ueda), [email protected] (T. Nagamune). 1 Present address: Department of Physics, School of Science and Engineering, Waseda University, Tokyo, Japan. 2 Present address: Department of Applied Chemistry, Graduate School of Engineering, Kyushu University, Fukuoka, Japan. 3 Present address: Division of Morphogenesis, Department of Developmental Biology, NIBB, Aichi, Japan. 1570-9639/03/$ - see front matter D 2003 Elsevier B.V. All rights reserved. doi:10.1016/S1570-9639(03)00179-1

have been made in the understanding of the structure – function relationships of the enzyme and enzyme-catalyzed light emission. Firefly luciferase catalyzes a sequence of reactions that convert firefly luciferin into an electronically excited oxyluciferin product that then emits light [9,10]. First, firefly luciferase catalyzes the formation of an enzyme-bound luciferyl adenylate. Next, a proton is pulled out from the C-4 carbon of the adenylate by a presumed basic amino acid of the enzyme, molecular oxygen then adds to the newly formed anion, and an electronically excited state oxyluciferin molecule and CO2 are produced from a highly reactive dioxetanone intermediate. Based on the molecular modeling studies using crystal structures of firefly luciferase without bound substrates [3] or the homologous enzyme gramicidin S synthetase (PheA) in a complex with AMP and L-phenylalanine, whose Nterminal adenylation subunit is very similar to that of firefly luciferase [4], an active site model was proposed [7]. This model is, so far, in good agreement with the newer model proposed from the crystal structure of firefly luciferase containing two molecules of bromoform [5], a general anesthetic and luciferase inhibitor. In these active site

184

T. Zako et al. / Biochimica et Biophysica Acta 1649 (2003) 183–189

2.2. General methods

Fig. 1. (a) The N-domain. The upper figure shows the two domains of firefly luciferase based on the crystal structure (PDB 1LCI). The figure was drawn with Rasmol [27]. The figure below shows the primary structure with residue numbers. The N- and C-domains are shown in dark and light gray, respectively. (b) SDS-PAGE of the purified N-domain and wild-type (WT) firefly luciferases. Lanes: 1, molecular weight marker; 2, N-domain (49 kDa); 3, WT (62 kDa). Proteins were stained with Coomassie brilliant blue.

models, all the key amino acid residues, except Lys529, are located in the N-terminal domain (N-domain, 1 – 437 in amino acids, Fig. 1a). It was suggested that Lys529 is a critical residue for effective substrate orientation, and provides favorable polar interactions important for transition state stabilization leading to efficient adenylate production [8]. If these explanations are right, it might be deduced that the N-domain in itself has substrate binding and catalytic activities, albeit very weak. However, Sung and Kang [11] and Sala-Newby and Campbell [12] showed that deletion mutants of firefly luciferase from N terminus and C terminus, respectively, lose their luminescent activities, suggesting that all the amino acid residues are essential for the light generating activity. In this study, we demonstrate that the purified N-domain has luminescence activity by itself, and binding activities to substrates ATP and luciferin (LH2). Also, its preferences for Mg ion and coenzyme A (CoA) concentrations, and the luminescent spectral property were examined and compared to those of the wild-type enzyme (WT).

The bioluminescence activity was determined using integration-based light assays. The standard activity assays were performed by using MicroLumat plus (Perkin-Elmer, Shelton, CT) and Luminescencer JNR (Atto, Tokyo, Japan) at 25 jC, as previously reported [13,14]. A fluorescencefree microcuvette was filled with luciferin and ATP reagents. Addition of the luciferase started the light emitting reaction. The reaction mixture contained 100 mM tricine, 10 mM MgSO4, 300 AM LH2, 100 AM CoA and 10 mM ATP, pH 8.0 unless otherwise indicated. Enzyme concentrations were 3 nM for WT and 1 AM for the N-domain luciferases, respectively, unless otherwise indicated. For the time course experiment in Fig. 2, MicroLumat plus with fast mixing equipment was used. Each data point was acquired with MicroLumat control software (Perkin-Elmer) with a sampling rate of 0.1 s. For the measurement of bioluminescence activity shown in Figs. 4– 6, Luminescencer JNR was used. Data were acquired with Luminescencer JNR control software (Atto) with a sampling rate and an integration time of 0.02 and 20 s, respectively. 2.3. Cloning of the luciferase N-terminal domain DNA fragment encoding of the N-terminal domain (1– 437 in amino acids) of P. pyralis luciferase was prepared from the WT enzyme plasmid pGEX-Ppy, which was made from pGEX-6P-2 (Amersham Bioscience, Tokyo, Japan) and pGEM-luc (Promega, Tokyo, Japan) as described previously [15]. The DNA fragment was amplified by standard polymerase chain reaction (PCR) using KOD DNA poly-

2. Material and methods 2.1. Materials The following materials were obtained from indicated sources: MgSO4, ATP, LH2, and CoA (Wako, Osaka, Japan); KOD DNA polymerase (Toyobo, Osaka, Japan); DNA ligation kit and restriction endonucleases (Takara, Kyoto, Japan). All other chemicals are of analytical grade and solutions were made up in ultra pure water.

Fig. 2. Time course of luminescence from the N-domain and WT luciferases. Data were acquired with a 1-min time interval and a sampling rate of 0.1 s. Data points show the luminescence in cps/nmol protein. The reaction mixture contained 100 mM Tricine, 10 mM MgSO4, 300 AM LH2, 100 AM CoA and 10 mM ATP, pH 8.0. Enzyme concentration was 1 AM for the N-domain and 0.0034 AM for WT. Symbols: closed circle, N-domain; open triangle, WT.

T. Zako et al. / Biochimica et Biophysica Acta 1649 (2003) 183–189

merase and the following primers: 5V-TATCGGATCCATGGGAAGACGCCAAAAAC-3V and 5V-CCTATATCTCCTTCTGCGGCCGCTTAGCGGTCAACTATGAAGAAGTG-3V (cleavage sites by BamHI and NotI are shown in bold letters, and an underline represents an anti-stop codon), digested with BamHI and NotI, and cloned to pGEX-6P-2 digested with the same. The plasmid was confirmed to have the inserted sequence using a DNA sequencer SQ-5500 (Hitachi, Tokyo, Japan) and Thermosequenase sequencing kit (Amersham Bioscience), designated pGEX-PpyN.

185

2.6. Spectral analysis The luminescent spectrum of the N-domain was taken under various pH conditions (pH 6.9, 7.2, 7.7, and 8.2) with an IMUC-7000 L multichannel analyzer (Otsuka Electronics, Osaka, Japan). The protein concentration was 0.55 mg/ml and substrate concentrations were 300 nM for LH2 and 10 mM for ATP, respectively. Each spectrum from 450 to 700 nm was drawn after integration for 40 min.

3. Results 2.4. Expression and purification of the proteins 3.1. Overexpression and purification of luciferases The glutathione S-transferase (GST) fusion proteins for WT and the N-domain luciferases were expressed in E. coli XL-10 Gold (Stratagene, La Jolla, CA). The cells from a single colony were cultured at 27 jC in 200 ml of LB medium supplemented with 100 Ag/ml ampicillin to a mid log phase (A600 = 0.5), induced with 0.1 mM IPTG and cultured at 20 jC overnight. The cells were harvested by centrifugation at 4 jC and stored frozen at 80 jC. Cell pellets were resuspended in 6-ml phosphate buffered saline (PBS; 1.48 g/l Na2HPO4, 0.43 g/l KH2PO4, 7.2 g/l NaCl, pH 7.2), and lysed by sonication. The whole cell extract was isolated by centrifugation at 20,000  g for 15 min at 4 jC. WT and the N-domain luciferases were purified from the whole cell extracts by affinity chromatography using Glutathione Sepharose 4B (Amersham Bioscience) according to the manufacturer’s instructions. The luciferases were cleaved on the gel from GST by incubation with PreScission protease (Amersham Bioscience) in PreScission protease cleavage buffer CB (50 mM Tris – HCl, 150 mM NaCl, 1 mM EDTA, 1 mM DTT, pH 7.0) for 4 h at 4 jC. Proteins were eluted with CB, and stored at 20 jC until use. Luciferase protein concentrations were determined using a BCA protein assay kit (Pierce, Rockford, IL) with bovine serum albumin as a standard. SDS-PAGE was performed according to the method of Laemmli, and the gel was stained with Coomassie brilliant blue R-250.

Full-length P. pyralis luciferase (WT) and its N-terminal domain (N-domain), each having the additional N-terminal peptide GPLGS, were expressed in E. coli as GST-fusion proteins, and purified with a glutathione column followed by on-column cleavage with PreScission protease. Average yields of purified WT and the N-domain luciferases were 10 and 3 mg/l culture, respectively. All the proteins were purified to homogeneity as judged by SDS-PAGE (Fig. 1b). As judged from the calculated molecular weights of WT (61.2 kDa) and the N-domain (49.0 kDa) luciferases, both were thought to be full-length products. 3.2. Bioluminescence activity of the N-domain The luminescence activity of the N-domain in the presence of excess substrates was recorded with a luminometer, and its time course compared with that of WT. Unlike WT, which showed peak light emission at 0.5 s after the reaction started, the N-domain derived luminescence increased very slowly and reached its maximum at around 4 min after the reaction started (Fig. 2 and Table 1). After the peak, the luminescence from the N-domain decreased more slowly than the luminescence from WT, and its decay rate was less than one-tenth of that of WT (Table 1). The relative Table 1 Bioluminescence activity of wild-type (WT) and the N-terminal domain (Ndomain) luciferases

2.5. Steady state kinetics Steady state kinetic constants were deduced from luminescence assays in which the maximum light intensity was taken as an estimate of initial velocity. The concentration of one substrate was maintained at saturation (300 AM LH2 or 20 mM ATP, respectively) and the other varied as follows: 0.5 –60 AM LH2 and 0.005 –1.0 mM ATP for WT, and 0.5 – 200 AM LH2 and 0.1 – 20 mM ATP for the N-domain, respectively. Peak height data were collected and kinetic constants were calculated by the nonlinear least square method using KaleidaGraph 3.0 (Synergy software, Reading, PA).

WT N-domain

Relative activity (peak)a

Relative activity (integrated)b

Rise time (s)c

Decay rate (min 1)d

100 0.0018

100 0.031

0.5 250

0.26 0.022

a Peak-based activity was measured with time interval of 0.167 s and a sampling rate of 0.1 s for WT and with a time interval of 1 min and a sampling rate of 0.1 s for the N-domain, respectively. WT value was defined as 100.0, which was equivalent to 1.2  1010 photons s 1 nmol 1. b Integrated for 60 min with a time interval of 1 min and a sampling rate of 0.1 s. WT value was defined as 100.0, which was equivalent to 3.5  1010 photons s 1 nmol 1. c Time to reach peak value from the initiation of the reaction is shown. d Bioluminescence decay rates are calculated from the single exponential fitting to the data in Fig. 2.

186

T. Zako et al. / Biochimica et Biophysica Acta 1649 (2003) 183–189

Fig. 4. Effect of Mg2 + ion on the light emission of the N-domain and WT luciferases. The maximum integrated light intensity was recorded with a luminometer in the presence of ATP and CoA. The Mg2 + ion concentration was varied from 0 to 100 mM for the N-domain and from 0 to 300 mM for WT. Symbols: closed circle, N-domain; open triangle, WT.

larger than those of WT for both LH2 and ATP; ca. 3.6 times and 30 times, respectively. 3.4. Effects of Mg2+ ion and CoA

Fig. 3. Michaelis – Menten diagrams for the N-domain as a function of luciferin (a) and ATP (b) concentrations. The maximum light intensity (20 s integration) was taken as an estimate of initial velocity. Data points show the luminescence in 107 cpm/nmol protein. The reaction mixture contained 100 mM Tricine, pH 8.0, 10 mM MgSO4, 100 AM CoA, 1 AM enzyme and various concentrations of LH2 and ATP.

The preferences for Mg2 + and CoA concentrations in the luminescent reaction of WT and the N-domain luciferases were examined. As shown in Fig. 4, Mg2 + ion preferences for WT and the N-domain enzymes were almost the same, with an optimal concentration of 10 mM. However, CoA preferences were markedly different between the WT and the N-domain enzymes (Fig. 5). For WT, the addition of

integrated-activity of the N-domain in this condition was about 0.03% of WT. Essentially the same results were obtained with different batches of the N-domain preparation derived from different single colonies, ruling out the possibility of WT contamination (data not shown). 3.3. Substrate binding activities To evaluate the kinetic behavior of the N-domain, steadystate kinetic parameters for the luminescent reactions of WT and the N-domain luciferases were determined (Fig. 3 and Table 2). The peak light emission from the N-domain and the initial light emission for WT were adopted to evaluate the luminescent activity. Km values of the N-domain were Table 2 Michaelis – Menten constants for the N-domain and WT of firefly luciferase

N-domain WT

Km (AM) for luciferin

Km (mM) for ATP

26 F 1.7 7.2 F 0.57

6.9 F 1.3 0.23 F 0.023

Fig. 5. Effect of CoA on the light emission of the N-domain and WT luciferases. The maximum integrated light intensity was recorded with a luminometer in the presence of ATP. The CoA concentration was varied from 0 to 3.0 mM. Symbols: closed circle, N-domain; open triangle, WT.

T. Zako et al. / Biochimica et Biophysica Acta 1649 (2003) 183–189

187

C-domain was suggested to be critical for effective substrate orientation and important for transition state stabilization leading to efficient adenylate production [8]. Our study, however, clearly shows that not only K529 but also the entire C-domain is dispensable for the luminescent reaction. Though our data seem to contradict the previous results of inactive C-terminal deletion mutants [12], it is possible that they dismissed the activity because they measured luminescence only for 10 s from the reaction start. Our data show that the luminescence of the N-domain increases far more slowly than the WT enzyme. 4.1. Binding of the N-domain to substrates luciferin and ATP

Fig. 6. Luminescent spectra of the N-domain. Each spectrum was integrated for the luminescence for 40 min from the reaction start. The reaction mixture contained 200 mM each of MOPS [pH 6.9 (bold solid line)], HEPES [pH 7.2 (solid line)], or Tricine [pH 7.7 (dotted line) and pH 8.2 (dashed line)] and standard assay buffer (10 mM MgSO4, 300 AM LH2, 100 AM CoA and 10 mM ATP). Enzyme concentration was 10 AM.

CoA enhanced the luminescent activity about twofold, and the highest activity was observed at 1.2 mM CoA. For the N-domain, however, the effect of CoA was modest (1.1 times enhancement) and the highest activity was observed at as low as 10 AM CoA. The addition of CoA at higher concentrations inhibited the luminescent activity. 3.5. Luminescent spectral property of the N-domain It has been shown that firefly luciferases, including that of P. Pyralis, exhibit two different light colors under different pH conditions, i.e. a f 620-nm peak at acidic pH and a f 560-nm peak at neutral and basic pH. To test whether this is the case for the N-domain, the bioluminescent spectra from the concentrated enzyme were taken at different pH conditions. Due to the low specific activity and the slow increase in luminescence, the spectra were numerically integrated for 40 min. Fig. 6 shows that for the N-domain, only red luminescence, peaking at 620 nm, was observed, and no prominent pH-dependent luminescent peak shift was observed in the pH range of 6.9 to 8.2.

4. Discussion The results shown here clearly indicate that the Ndomain enzyme has luminescent activity by itself. Also, larger but measurable Michaelis –Menten constants for LH2 and ATP show that the N-domain enzyme retains WT substrate binding activities. These findings would support the idea that all the amino acids essential for bioluminescent activity are located in the N-domain. The role of K529 in the

The Michaelis –Menten constants obtained for LH2 and ATP indicate that the N-domain retains binding activities for these substrates, although the affinities were lower than those of WT. As for LH2, the result seems acceptable since all the amino acids, except K529, which are involved in the LH2 binding in the active site model are located in the Ndomain. These amino acids would essentially be sufficient for LH2 binding, and the C-domain, including K529, only helps the orientation of LH2 captured in the active site, which is not contradictory to the previous results. As for ATP binding, the decrease in the binding activity for the Ndomain was more prominent, where a 30-fold reduction in affinity was observed. This result shows that K529 and/or other residue(s) in the C-domain are also involved in ATP binding and play more important roles for ATP binding than LH2 binding. From the mutational study by Branchini et al. [8], the Km value of the K529A mutant for ATP was about 7.5 times higher than that of WT. Although the major part of the Km increase could be explained by the lack of the K529 side chain in the N-domain, the difference in Km values between the N-domain and WT was larger than that between K529A and WT. This indicates that the main chain of K529 or other amino acid residues in the C-domain also participate in ATP binding. Our result is not contradictory to the two ATP binding site model [16,17]. Steghens et al. proposed the hypothesis of the existence of two ATP binding sites with low and high affinities. Our results imply that the low affinity binding site is located in the N-domain and the other in the N- and/or C-domain. In our calculation, maximum light intensity was taken as an estimate of initial velocity. No correction was made for the difference in rise time to reach maximum intensity, as in the case of the previous calculation by Branchini et al. [7] for the mutant and WT luciferases. The relative peak activity shown in Table 1 corresponds to the relative kcat values of WT and the N-domain. As shown in Table 1, the peak activity of the N-domain is about 0.002% of that of WT, which indicates that the enzymatic turn-over of the Ndomain is much slower than that of WT. Positions of amino acid residues necessary for the light emission of the Ndomain would not be optimized compared with that of WT.

188

T. Zako et al. / Biochimica et Biophysica Acta 1649 (2003) 183–189

It is possible that in WT the C-terminal domain supports the structural arrangement of essential amino acid residues for maximum activity. 4.2. Effect of CoA Our result suggests that CoA is not effective in the enhancement of light production from the N-domain. On the contrary, CoA stimulates light production from WT by promoting the release of the product oxyluciferin, thereby releasing free enzyme [18,19]. We think that this is because the affinity between oxyluciferin and the N-domain is weaker than that of WT, which leads to weaker product inhibition. The larger Km value of the N-domain for LH2 obtained in this study indirectly supports this idea. 4.3. The slow increase in the luminescent activity of the N-domain The most remarkable feature of the N-domain is the gradual increase of light emission in the presence of excess substrates. The time required to yield maximum emission was about a thousandfold longer than that of WT. It is clear that the C-domain supports the fast light increase (flash) observed in WT. Considering the dramatically reduced turnover of the N-domain, it is possible that this gradual increase is due to a slower adenylation reaction of LH2, on the assumption that a slower subsequent reaction allows for the time-dependent accumulation of a luciferyl-AMP intermediate. The C-domain should be important for adenylation since gramicidin S synthetase, which catalyzes the adenylation of L-phenylalanine, has a similar tertiary structure with luciferase, including the C-domain [4,7]. Our data, however, indicate that the C-domain is not essential for adenylation. The C-domain would help the amino acid residues of the N-domain assume the proper tertiary structure for the reaction. 4.4. Luminescent color of the N-domain It has been shown that the WT firefly luciferases show two (green and red) luminescent peaks, especially under low pH conditions [9,10]. In the case of the N-domain, only a longer wavelength peak could be observed, irrespective of pH, with no observed luminescent peak shift. Apparently, the presence of the C-domain is necessary for the generation of green luminescence under physiological pH conditions. There are three mechanisms that are generally proposed to explain the luminescent color shift [20]. The first is the equilibration ratio between two tautomerization forms (ketoand enolate-) of the electronically exited singlet state of oxyluciferin [21]. The second is the polarity of the oxyluciferin binding site [22]. The third is the conformation of oxyluciferin regarding rotation around the C2 –C2V bond [23 –25]. These three possibilities are thought to not be

mutually exclusive, on the assumption that the excited oxyluciferin bound in the active site of luciferase emits the light. From the current data, it is difficult to determine which is the main reason for the dominant red light emission from the N-domain. However, the possible reasons may be: (1) The C-domain has the residues necessary for the peak shift. The residue(s) can be necessary either for the LH2 deprotonation or induction of the C2 – C2V rotation, or (2) a subtle structural change of the luciferase active site, which lies on the surface of the N-domain. The absence of the Cdomain abrogates the conformational change necessary for the green light emission, which is usually accompanied by the binding of the C-domain. This possibility cannot be dismissed because in the recent mutational studies, including Ref. [6], only a point mutation in the N-domain diminished the green light emission. Also, the absence of the C-domain would surely affect the polarity of the oxyluciferin binding site, while it is assumed to be mechanically less dominant [18,19]. In other words, according to Viviani et al. [26], certain amino acids are considered to deprotonate the substrate luciferin in order to make the ‘‘enolate’’ state, which produces yellow-green light. Deletion of the C-domain in this study, as well as mutations in the N-domain, would alter the position of these residues, so that only the ‘‘keto’’ state is present. Alternatively, McCapra et al. [23], McCapra [24] and Branchini et al. [25] proposed that the color change is associated with the excited state oxyluciferin conformations related to the rotation about the C2 – C2V bond. In this model, the red emission color is attributed to a minimum energy conformation of the excited state oxyluciferin, and green to a higher energy conformer. It is also possible that deletion of the C-domain alters the position of the amino acid residues necessary for the rotation reaction, and/or decreases the stability of the active site so that the relaxed or distorted active site allows for an increase in energetically lower oxyluciferin conformers, which can then lead to an increased emission of red light. This may also explain the previous observations of red light emission from the enzymes under harsh conditions. In conclusion, we showed that the N-terminal domain of firefly luciferase retains luminescent activity and retains binding activity to the substrates luciferin and ATP. This indicates that all the residues essential for the activity are involved in the N-terminal domain. Though further study is needed to clarify the reason for slow/delayed kinetics, the truncated, smaller luciferase may find applications as a smaller bioluminescent reporter or enzyme label, especially in vivo.

Acknowledgements We thank Dr. T. Nakatsu in RIKEN, Harima, Hyogo, Japan for his helpful discussions and comments.

T. Zako et al. / Biochimica et Biophysica Acta 1649 (2003) 183–189

References [1] T. Wilson, J.W. Hastings, Bioluminescence, Annu. Rev. Cell Dev. Biol. 14 (1998) 197 – 230. [2] K.V. Wood, Marker proteins for gene expression, Curr. Opin. Biotechnol. 6 (1995) 50 – 58. [3] E. Conti, N.P. Franks, P. Brick, Crystal structure of firefly luciferase throws light on a superfamily of adenylate-forming enzymes, Structure 4 (1996) 287 – 298. [4] E. Conti, T. Stachelhaus, M.A. Marahiel, P. Brick, Structural basis for the activation of phenylalanine in the non-ribosomal biosynthesis of gramicidin S, EMBO J. 16 (1997) 4174 – 4183. [5] N.P. Franks, A. Jenkins, E. Conti, W.R. Lieb, P. Brick, Structural basis for the inhibition of firefly luciferase by a general anesthetic, Biophys. J. 75 (1998) 2205 – 2211. [6] B.R. Branchini, R.A. Magyar, M.H. Murtiashaw, S.M. Anderson, L.C. Helgerson, M. Zimmer, Site-directed mutagenesis of firefly luciferase active site amino acids: a proposed model for bioluminescence color, Biochemistry 38 (1999) 13223 – 13230. [7] B.R. Branchini, R.A. Magyar, M.H. Murtiashaw, S.M. Anderson, M. Zimmer, Site-directed mutagenesis of histidine 245 in firefly luciferase: a proposed model of the active site, Biochemistry 37 (1998) 15311 – 15319. [8] B.R. Branchini, M.H. Murtiashaw, R.A. Magyar, S.M. Anderson, The role of lysine 529, a conserved residue of the acyl-adenylate-forming enzyme superfamily, in firefly luciferase, Biochemistry 39 (2000) 5433 – 5440. [9] E.H. White, E. Rapaport, H.H. Seliger, T.A. Hopkins, The chemi- and bioluminescence of firefly luciferin: an efficient chemical production of electronically excited states, Bioorg. Chem. 1 (1971) 92 – 122. [10] E.H. White, M.G. Steinmetz, J.D. Milano, P.D. Wildes, R. Morland, Chemiluminescence and bioluminescence of firefly luciferin, J. Am. Chem. Soc. 102 (1980) 3199 – 3208. [11] D. Sung, H. Kang, The N-terminal amino acid sequences of the firefly luciferase are important for stability of the enzyme, Photochem. Photobiol. 68 (1998) 749 – 753. [12] G.B. Sala-Newby, A.K. Campbell, Stepwise removal of the C-terminal 12 amino acids of firefly luciferase results in graded loss of activity, Biochim. Biophys. Acta 1206 (1994) 155 – 160. [13] T. Zako, H. Deguchi, A. Kitayama, H. Ueda, T. Nagamune, Refolding of firefly luciferase immobilized on agarose beads, J. Biochem. 127 (2000) 351 – 354. [14] T. Zako, K. Harada, T. Mannen, S. Yamaguchi, A. Kitayama, H.

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22] [23]

[24]

[25]

[26]

[27]

189

Ueda, T. Nagamune, Monitoring the refolding process for immobilized firefly luciferase with a biosensor based on surface plasmon resonance, J. Biochem. 129 (2001) 1 – 4. A. Kitayama, H. Yoshizaki, Y. Ohmiya, H. Ueda, T. Nagamune, Creation of a themostable firefly luciferase with pH insensitive luminescent color, Photochem. Photobiol. 77 (2003) 333 – 338. M. DeLuca, W.D. McElroy, Two kinetically distinguishable ATP sites in firefly luciferase, Biochem. Biophys. Res. Commun. 123 (1984) 764 – 770. J.P. Steghens, K.L. Min, J.C. Bernengo, Firefly luciferase has two nucleotide binding sites: effect of nucleoside monophosphate and CoA on the light-emission spectra, Biochem. J. 336 (1998) 109 – 113. M. Pazzagli, J.H. Devine, D.O. Peterson, T.O. Baldwin, Use of bacterial and firefly luciferases as reporter genes in DEAE-dextran-mediated transfection of mammalian cells, Anal. Biochem. 204 (1992) 315 – 323. S.R. Ford, L.M. Buck, F.R. Leach, Does the sulfhydryl or the adenine moiety of CoA enhance firefly luciferase activity? Biochim. Biophys. Acta 1252 (1995) 180 – 184. K.V. Wood, The chemical mechanism and evolutionary development of beetle bioluminescence, Photochem. Photobiol. 62 (1995) 662 – 673. E.H. White, B. Branchini, Modification of firefly luciferase with a luciferin analog. A red light producing enzyme, J. Am. Chem. Soc. 97 (1975) 1243 – 1245. M. Deluca, Hydrophobic nature of the active site of firefly luciferase, Biochemistry 8 (1969) 160 – 166. F. McCapra, D.J. Gilfoyle, D.W. Young, N.J. Church, P. Spencer, in: A.K. Campbell, L.J. Kricka, P.E. Stanley (Eds.), Bioluminescence and Chemiluminescence, Wiley, Chichester, UK, 1994, pp. 387 – 391. F. McCapra, Chemical generation of excited states: the basis of chemiluminescence and bioluminescence, Methods Enzymol. 305 (2000) 3 – 47. B.R. Branchini, M.H. Murtiashaw, R.A. Magyar, N.C. Portier, M.C. Ruggiero, J.G. Stroh, Yellow-green and red firefly bioluminescence from 5,5-dimethyloxyluciferin, J. Am. Chem. Soc. 124 (2002) 2112 – 2113. V.R. Viviani, E.J. Bechara, Y. Ohmiya, Cloning, sequence analysis, and expression of active Phrixothrix railroad-worms luciferases: relationship between bioluminescence spectra and primary structures, Biochemistry 38 (1999) 8271 – 8279. R.A. Sayle, E.J. Milner-White, RASMOL: biomolecular graphics for all, Trends Biochem. Sci. 20 (1995) 374.