Effect of halogenated fluorescent compounds on

Effect of halogenated fluorescent compounds on bioluminescent reactions

Analytical and Bioanalytical Chemistry ISSN 1618-2642 Volume 400 Number 2 Anal Bioanal Chem (2011) 400:343-351 DOI 10.1007/s00216-011-4716x

1 23

Your article is protected by copyright and all rights are held exclusively by SpringerVerlag. This e-offprint is for personal use only and shall not be self-archived in electronic repositories. If you wish to self-archive your work, please use the accepted author’s version for posting to your own website or your institution’s repository. You may further deposit the accepted author’s version on a funder’s repository at a funder’s request, provided it is not made publicly available until 12 months after publication.

1 23

Author's personal copy Anal Bioanal Chem (2011) 400:343–351 DOI 10.1007/s00216-011-4716-x

ORIGINAL PAPER

Effect of halogenated fluorescent compounds on bioluminescent reactions Tamara N. Kirillova & Marina A. Gerasimova & Elena V. Nemtseva & Nadezhda S. Kudryasheva

Received: 10 December 2010 / Revised: 24 January 2011 / Accepted: 24 January 2011 / Published online: 19 February 2011 # Springer-Verlag 2011

Abstract The paper investigates an application of luminescent bioassays to monitor the toxicity of organic halides. Effects of xanthene dyes (fluorescein, eosin Y, and erythrosin B), used as model compounds, on bioluminescent reactions of firefly Luciola mingrelica, marine bacteria Photobacterium leiognathi, and hydroid polyp Obelia longissima were studied. Dependence of bioluminescence quenching constants on the atomic weight of halogen substituents in dye molecules was demonstrated. Bacterial bioluminescence was shown to be most sensitive to heavy halogen atoms involved in molecular structure; hence, it is suitable for construction of sensors to monitor toxicity of halogenated compounds. Mechanisms of bioluminescence quenching—energy transfer processes, collisional interactions, and enzyme–dye binding—were considered. Changes of bioluminescence (BL) spectra in the presence of the dyes were analyzed. Interactions of the dyes with enzymes were studied using fluorescence characteristics of the dyes in steady-state and time-resolved experiments. The dependences of fluorescence anisotropy of enzyme-bound dyes, the average fluorescence lifetime, and the number of exponential components in fluorescence decay on the atomic weight of halogen substituents were demonstrated. The results are

Published in the special issue Analytical and Bioanalytical Luminescence with Guest Editor Petr Solich. T. N. Kirillova : E. V. Nemtseva : N. S. Kudryasheva (*) Institute of Biophysics SB RAS, Akademgorodok 50, 660036 Krasnoyarsk, Russia e-mail: [email protected] M. A. Gerasimova : E. V. Nemtseva : N. S. Kudryasheva Siberian Federal University, Svobodny Prospect 79, 660041 Krasnoyarsk, Russia

discussed in terms of “dark effect of heavy halogen atom” in the process of enzyme–dye binding; hydrophobic interactions were assumed to be responsible for the effect. Keywords Bioluminescent assay . Xanthene dyes . Effect of heavy halogen atom . Bioluminescent enzymes . Fluorescence . Lifetime . Anisotropy Abbreviations ATP Adenosine-5-triphosphate disodium salt BL Bioluminescence EDTA Ethylenediaminetetraacetic acid FMN Flavin mononucleotide NADH β-Nicotinamide adenine dinucleotide reduced Tris Tris(hydroxymethyl)aminomethane

Introduction Most living organisms contain traces of the heavy halogen atoms bromine and iodine which are essential for metabolic processes. In nature, iodine and bromine are rare elements. Redistribution of halides in the atmosphere, and in terrestrial and marine ecosystems [1] affects their content in living organisms. Iodine is the heaviest element utilized biologically. Halides can activate or inhibit biochemical processes, and it is generally known that excess or the lack of these elements leads to metabolic disorders in lower and higher organisms [2]. These are the reasons to elaborate on methods for monitoring the toxic effect of halogenated compounds. Bioluminescence (BL) assays are appropriate candidates for such investigations. BL-based sensors successfully combine high sensitivity of bioassays to toxic compounds

Author's personal copy 344

T.N. Kirillova et al.

and technical simplicity of recording the main response— luminescence intensity. BL reactions of luminous organisms can serve as a basis for construction of sensors to monitor the toxicity of halogenated compounds because of simple and rapid recording of rates of enzymatic processes [3–5]. As was demonstrated for bacteria-based BL assays [6, 7], a response of the assay system depends on structural peculiarities and physicochemical characteristics of the toxic compounds. In Refs. [8–11], sensitivity of BL reactions to halogen compounds was examined. The effects of potassium halides and halogen-substituted dyes on bacterial BL were studied in Refs. [8–10]. In Ref. [10], the quenching efficiency of potassium halides was compared for BL reactions of fireflies, marine bacteria, and coelenterates. “Effect of heavy halogen atom” i.e. dependence of BL quenching efficiency on halogen atomic weight was demonstrated. The term “heavy-atom effect” was originally introduced by physicists studying changes in the spectral characteristics of compounds under the influence of Br or I atoms. M. Kasha and S. McGlynn were the first to observe this phenomenon [12–16]. They found that heavy halogen atoms incorporated into molecular structure reduce fluorescence quantum yields of the dye molecules. They hypothesized that this effect results from substantial spin–orbit coupling in the molecules [17, 18]. The energy of the spin– orbit interaction is proportional to Z4, where Z is the effective charge of the nucleus. The spin–orbit coupling is promoted by halogens of high atomic weight with filled porbitals, and practically undergoes no changes in the presence of metal cations i.e. elements with vacant electron orbitals. Among heavy-atom-related events, mention should be made of:

These compounds can be regarded as fluorescent probes in the binary interactions. Changes of fluorescent characteristics of the compounds testify to the efficiency of the binding process. In our experiments we chose a series of chemically inert homologous compounds, xanthene dyes— fluorescein, eosin, erythrosin—with bromine and iodine atoms incorporated into the structures of eosin and erythrosin, respectively. The fluorescent dyes mentioned are widely used as bioanalytical tools [18–21] for investigation of the conformational dynamics of proteins, the location of enzyme active centers, and the interactions of macromolecules with intracellular structures, etc. The purpose of this research was to reveal the most sensitive BL reaction for monitoring the toxicity of halogenated organic compounds. The sensitivities of BL reactions of fireflies, marine bacteria, and coelenterates were compared. Mechanisms of effect of the dyes on the BL reactions were studied. The effects of model organic compounds, xanthene dyes, on BL reactions are complex, with energy transfer processes, collisional interactions, and enzyme–dye binding contributing to BL inhibition. Our investigation dealt with:

&

Materials and methods

&

increased intersystem crossing constants and extension of the population of the lowest triplet state; and increased the phosphorescence quantum yield and lifetime [17].

The “external” heavy-atom effect takes place in excited fluorescent compounds, emitters, in the presence of halogenated compounds in solution [18]. In Ref. [10], two mechanisms of BL quenching by potassium halides were considered: 1. the original physical mechanism of intramolecular energy transfer mentioned above; and 2. a biochemical mechanism of interactions of halogenated compounds with enzymes. A conclusion was reached that the biochemical mechanism contributes to bioluminescence quenching to a greater extent. Fluorescent organic compounds are convenient model molecules for studying the binding of halides by enzymes.

1. the effects of halogenated dyes on BL intensity and spectra; 2. the binding of dyes with the enzymes using fluorescent characteristics of the dyes under steady-state and timeresolved conditions; and 3. the correlation between halogen atomic weight, halide quenching efficiency in BL reactions, and characteristics of enzyme–dye binding interactions.

Chemicals Luciola mingrelica firefly luciferase (2.38 mg mL−1) and firefly luciferin (LH2) were obtained from the Department of Chemical Enzymology, Moscow State University, Russia [22, 23]. The recombinant Ca2+-regulated photoprotein obelin from marine hydroid polyp Obelia longissima (2.73 mg mL−1) was from the Photobiology Laboratory, Institute of Biophysics, SB RAS, Krasnoyarsk, Russia [24, 25]. The bacterial preparation involved a coupled enzyme system: NADH:FMN-oxidoreductase from Vibrio fischeri, 0.15 U mL−1, and luciferase from Photobacterium leiognathi, 0.5 mg mL−1 [26–28]. It was produced in the Bacterial BL Laboratory, Institute of Biophysics, SB RAS, Krasnoyarsk, Russia. Synthetic coelenterazine was purchased from Prolume, CaCl2 from Khimreaktiv, Russia, and FMN from Serva. Preparations of EDTA, NADH and disodium salts of xanthene dyes (fluorescein, eosin Y, and

Author's personal copy Effect of halogenated fluorescent compounds on bioluminescence

345

erythrosin B) were obtained from Sigma; Tris was from Fluka.

Professional) using linear and modified Stern–Volmer equations [18]:

Equipment

I0 ¼ 1 þ KSV C I

ð1Þ

I0 ¼ ð1 þ KSV C Þ expðKS C Þ I

ð2Þ

Absorption spectra were recorded with the Uvikon-943 double-beam spectrophotometer (Kontron Instruments, Italy). Fluorescence spectra and anisotropy were measured with an Aminco-Bowman, Series 2 luminescent spectrometer (Thermo Spectronic, USA) equipped with polarizers. The Fluorolog 3-22 spectrofluorimeter (Horiba Jobin Yvon, France) with option of single photon counting was used to measure the spectra, BL kinetics, and fluorescence lifetimes of the dyes. BL measurements To study BL spectra and kinetics, the following reaction mixtures were used: Firefly BL: 60 μL luciferase solution, 120 μL Tris-acetate buffer (0.1 mol L−1 Tris, 2 mmol L−1 EDTA, 10 mmol L−1 MgSO4, 0.1 mol L−1 CH3COOH, pH 7.8), 60 μL LH2, and 60 μL 1 mmol L−1 ATP. The total volume of the reaction mixture was 300 μL. Coelenterate BL: 20 μL obelin solution; 190 μL Tris-HCl buffer (20 mmol L−1 Tris, 5 mmol L−1 EDTA, pH 7.0), 10 μL 0.05 mol L−1 CaCl2. The total volume of the reaction mixture was 220 μL. Bacterial BL: 50 μL luciferase solution, 50 μL tetradecanal, 50 μL FMN, 200 μL phosphate buffer (pH 6.8), 200 μL NADH. The total volume of the reaction mixture was 550 μL The 5–10 μL dye solution was added to the BL solutions before the BL reaction was initiated by ATP, Ca2+, or NADH for reactions of firefly, coelenterate, and bacteria, respectively. The concentration of the dyes in solutions varied from 1 to 100 μmol L−1; BL spectra were recorded. All spectra were corrected according to BL kinetics and spectral sensitivity of devices. BL intensities in the presence and absence of the dyes (I and I0, respectively) were compared. The reliability of the measurements was confirmed by the invariable ratio of intensity decay slopes in the absence and presence of dyes. All measurements were performed at 25°C. Analysis of BL quenching The BL quenching effects of the dyes were treated by nonlinear regression analysis (Microcal Origin 7.0

where C is the dye concentration; KSV is the dynamic Stern–Volmer quenching constant; and KS is the association constant for complex formation. The standard errors were determined using the leastsquare analysis. The overall efficiency of the BL quenching by xanthene dyes was estimated as the average quenching constant KQ in the case of positive Stern–Volmer plot deviation, with dye concentrations ranging from 0 to 80 μmol L−1: P ð1 þ KSV Ci Þ expðKS Ci Þ 1 P ð3Þ KQ ¼ i Ci i

Fluorescence lifetime measurements Time-resolved fluorescence experiments were carried out using the time-correlated single-photon counting technique [29]. A real decay profile I′(t) represents a combination of the instrument response function E(t) and the fluorescence decay function I(t): I 0 ðtÞ ¼ EðtÞ IðtÞ

ð4Þ

Nanosecond lifetimes were recovered by a deconvolution procedure with the instrument response E(t) measured with a highly scattering solution (Ludox). The fluorescence intensity time decay I(t) was fitted by a sum of N discrete exponentials: IðtÞ ¼

N X

Ai expðt =t i Þ

ð5Þ

i¼1

where Ai is the amplitude and τi is the fluorescence lifetime of the ith component. The fluorescence lifetime decays were measured for dyes in buffer solutions in the absence and presence of the enzymes. The samples were excited using the 453 nm NanoLED. The detection wavelengths were 520, 540, and 550 nm for fluorescein, eosin, and erythrosin, respectively. The calibration time was 0.056 ns/channel. The average lifetime 〈τ〉 was calculated as: hti ¼

N X i¼1

Bi t i ;

N X i¼1

Bi ¼ 1

ð6Þ



where Bi is the fractional contribution of the ith component to the steady-state intensity. The values of 〈τ〉 were obtained for different concentrations of enzymes: 1.3– 14.1, 0.4–13.4, and 1.2–32.3 μmol L−1 for fireflies, marine bacteria, and apo-obelin, respectively. The error for lifetime measurements was 15%. Fluorescence anisotropy measurements The samples were excited with vertically polarized light, and the fluorescence intensity was measured at 90° in the parallel (Ij j ) or vertical (I? ) direction. The steady-state fluorescence anisotropy was defined as: Ij j G I? r¼ Ij j þ 2G I?

ð7Þ

where the G factor is the ratio of the sensitivities of the detection system for vertically and horizontally polarized light [18]. The anisotropy r of dye solutions (1 μmol L−1) was measured in the presence of enzymes. The concentrations of the enzymes were varied from 0.2 to 32.5 μmol L−1. Conditions for recording were: excitation wavelengths 460, 515, and 525 nm, and registration wavelengths 510, 540, and 550 nm for solutions of fluorescein, eosin, and erythrosin, respectively.

The apparent dissociation constant of a dye–protein complex, Kd0 , was estimated in accordance with Ref. [18]: Cp ð1 fb Þ fb

ð8Þ

Here Cp is the protein concentration, and fb and (1−fb) are the fractions of enzyme-bound and free dye molecules, respectively. The value of fb was calculated as: fb ¼

r rf rb rf

ð9Þ

where rf and rb are the anisotropies of free (in the absence of protein) and enzyme-bound dye molecules, respectively, and r is the fluorescence anisotropy of an equilibrated mixture of bound and free dyes at protein concentration Cp. The rb values were taken as the maximum anisotropy of rigidly bound enzyme–dye pairs and calculated using the Perrin equation [18]: rb ¼ rbmax ¼

r0 1 þ hti=qb

Results and discussion To compare sensitivities of the BL reactions to halogenated organic compounds, dependencies of BL intensities on dye concentration were analyzed using Stern–Volmer equations. The Stern–Volmer theory classifies quenching effects as collisional and associative (i.e. binding). Additionally, binding mechanisms were studied using fluorescence characteristics of the dyes on addition of the enzymes in steadystate and time-resolved experiments: fluorescence anisotropy of enzyme-bound dyes, the average fluorescence lifetime, and the number of exponential components in fluorescence decay were examined. Changes in BL spectra in the presence of the dyes were investigated and interpreted in terms of nonradiative intermolecular energy transfer. BL intensity in the presence of xanthene dyes

Calculation of dissociation constants

Kd0 ¼

the rotational correlation time of the rigidly bound fluorophore. Values of r0 were taken as 0.35 [18], 0.36, and 0.38 [30] for fluorescein, eosin, and erythrosin, respectively. The values of 〈τ〉 were determined in timeresolved fluorescence experiments in the presence of excess concentration of proteins. Values of θb, a function of the molecular weight of the proteins [18], were estimated as 8, 22.5, and 28 ns for apo-obelin (22 kDa), firefly (62 kDa), and bacterial (78 kDa) luciferases, respectively.

ð10Þ

Here r0 is the fundamental anisotropy, 〈τ〉 is the fluorescence lifetime of the enzyme-bound dye, and θb is

The firefly and obelin BL intensity in the presence of xanthene dyes was analyzed and compared with that of bacterial BL [9]. The Stern–Volmer plots of the firefly BL (Fig. 1a) had upward curvatures similar to those of bacterial BL described in Ref. [9]. The obelin BL plots were linear (Fig. 1b). It is known that linear dependences represent the dynamic type of fluorescence quenching, whereas positive deviations at higher dye concentration are a sign of both static and dynamic enzyme–dye interactions [18, 31–33]. The constants of dynamic quenching of obelin BL (KSV) were calculated by use of Eq. (1). Equation (2) was used to determine the association constants of the dye–enzyme complex (KS), and dynamic quenching constants (KSV) for bacterial and firefly BL spectra. The results are summarized in Table 1. For the series of dyes fluorescein–eosin–erythrosin, KSV remains unchanged for firefly BL, whereas KSV shows 1.4fold and 10-fold increase for obelin and bacterial BL, respectively. The values of KSV for bacterial BL exceed those for obelin and firefly BL (Table 1), thus showing that the dynamic quenching contributes substantially to bacterial BL inhibition. In bacterial and firefly BL, the KS values of erythrosin are larger than those of eosin by 6.5 and 1.6-fold, respectively. Values of KS for fluorescein could not be


347

Fig. 1 Stern–Volmer plots for firefly (a) and obelin (b) BL intensity on addition of fluorescein (1), eosin (2), and erythrosin (3)

determined in these two reactions because of linearity of Stern–Volmer plots. The effects of xanthene dyes on the firefly BL reaction were studied in detail by KrishnaMurthy et al. [34]. The authors obtained linear relationships in Stern–Volmer plots for BL quenching by fluorescein, eosin, and erythrosin, with KSV 2.9–3.5×104 mol L−1. These constants are in a good agreement with our data (Table 1). Nonlinearity of the plots in our case (Fig. 1a) is conditioned by a wider range of dye concentrations (up to 40–70 μmol L−1 vs. 20 μmol L−1 used in Ref. [34]). The difference in types of firefly luciferases (Luciola mingrelica vs. North American firefly in Ref. [34]) may also account for the difference. The average effective BL quenching constants KQ (Eq. 3) were used to evaluate the overall sensitivity of BL reactions to halogenated compounds. The relationships between KQ and the atomic weight of the substituent in dye molecules are shown in Fig. 2. It is seen that KQ of obelin BL does not depend on the atomic weight of the halogen substituents, but KQ values of firefly and, particularly, of bacterial BL are highly dependent. So the bacterial BL reaction could serve as most effective assay for monitoring the toxicity of halogenated organic molecules. The process of dynamic quenching contributes to a high degree to the bacterial BL inhibition by halogenated dyes (Table 1). BL spectra in the presence of xanthene dyes BL spectra were measured in the absence and presence of xanthene dyes. Emission maxima for firefly, bacterial, and Table 1 Results for BL quenching by xanthene dyes

Dye

Fluorescein Eosin Erythrosin

obelin BL spectra were 570 nm [22], 500 nm [9, 28], and 485 nm [24], respectively. In all buffer solutions applied, absorption maxima of free fluorescein, eosin, and erythrosin were 490, 516, and 526 nm, respectively, and fluorescence maxima were 515, 540, and 550 nm, respectively (Fig. 3). By way of example, the firefly BL spectra in the presence of eosin (Fig. 3a, curves 1–5) and obelin BL spectra in the presence of erythrosin (Fig. 3b, curves 1–4) are given. Changes of bacterial BL spectra on addition of halogenated fluoresceins [9] are similar to those of obelin BL. Three dyes quenched the BL intensity to a different extent; the efficiency of quenching increased in the order fluorescein–eosin–erythrosin. The profile of the firefly BL spectrum remained constant (Fig. 3a) but those of obelin BL (Fig. 3b) and bacterial BL [9] changed, i.e. additional intensity was observed in the wavelength range of dye fluorescence (Fig. 3b, curve 6). Thus, obelin BL spectra in the presence of xanthene dyes are characterized both by BL intensity quenching and enhancement of dye-sensitized fluorescence. Similar data were obtained for bacterial BL spectra in Ref. [9]. The results are interpreted in terms of nonradiative energy transfer from BL emitters to dye molecules. It should be noted that the blue shift of the obelin BL maximum in Fig. 3b is a result of BL absorption by the dye molecules. For further data treatment, the BL spectra were corrected for this effect. The fluorescent state energy diagram (Fig. 4) is given to look into possibility of intermolecular energy transfer from

Obelin BL

Firefly BL

Bacterial BL

KSV (mol−1 L)

KSV (mol−1 L)

KS (mol−1 L)

KSV (mol−1 L)

KS (mol−1 L)

(2.7±0.1)×104 (3.5±0.2)×104 (3.8±0.3)×104

(4.3±0.2)×104 (4.1±0.4)×104 (4.1±0.4)×104

– (4.5±0.4)×104 (7.2±0.6)×104

(2.9±0.2)×105 (1.5±0.1)×106 (2.9±0.3)×106

– (8.6±0.7)×103 (5.6±0.3)×104



Fig. 4 Energy of fluorescent states of obelin, bacterial, and firefly BL emitters, and xanthene dyes Fig. 2 Average effective BL quenching constant KQ vs. atomic weight of substituent in xanthene dyes: 1, obelin BL; 2, firefly BL; 3, bacterial BL. Substituents: empty squares, H; half-filled circles, Br; filled triangles, I

the BL emitters as donors to the xanthene dyes as acceptors. Fluorescence levels of BL emitters and dye molecules were determined from their spectra. Because the energy of firefly emitter is lower than the energies of xanthene dyes, one can predict the lowest efficiency of energy transfer in the firefly BL reaction. This can be supported by minimal overlap integrals of the firefly BL emitter with fluorescein (1.4× 10−15 mol−1 L cm3), and eosin (3.4×10−14 mol−1 L cm3) as against overlap integrals of obelin and bacterial emitters (1.1–2.3×10−13 mol−1 L cm3). Thus, sensitized fluorescence of the dyes was observed in obelin (Fig. 3b) and bacterial [9] BL only; it was not observed in the firefly BL reaction (Fig. 3a). It should be noted that the conditions enabling recording of the upper (nonfluorescent) electron excited states of BL emitters [7, 35–39] were not used in our experiments.

Fig. 3 (a) Firefly BL spectra at different concentrations of eosin: 1, 0; 2, 1.6 μmol L−1; 3, 5.5 μmol L−1; 4, 23.6 μmol L−1; 5, 38.4 μmol L−1. Absorption (6) and emission (7) spectra of eosin in Tris-acetate buffer solution (pH 7.8). (b) Obelin BL spectra at different concentrations of

Fluorescence lifetimes of xanthene dyes in the presence of enzymes To verify dye–enzyme interactions, fluorescence lifetimes of the dyes in the presence of BL enzymes were analyzed. With the time-resolved technique applied, the fluorescence intensity decays of the dyes were recorded. The decay of fluorescein was fitted by a single exponential component whereas those of eosin and erythrosin were fitted by two components. Addition of enzymes to dye solutions resulted in emergence of the third component in the decay of erythrosin, whereas fluorescein and eosin decays did not change the number of components. The multicomponent decay is known [18] to be conditioned by the involvement of the fluorescent molecules in various intermolecular interactions. The above mentioned result demonstrates that integration of the enzymes with erythrosin is stronger than that with fluorescein or eosin. Average fluorescence lifetimes of fluorescein, eosin, and erythrosin were calculated in the absence 〈τ0〉 and presence

erythrosin: 1, 0; 2, 45.4 μmol L−1; 3, 68.2 μmol L−1; 4, 159.1 μmol L−1. Absorption (5) and emission (6) spectra of erythrosin in Tris-HCl buffer solution (pH 7.0)


349

Table 2 Average lifetimes of xanthene dyes in the absence 〈τ0〉 and presence 〈τ〉 of the enzymes Dye


C (μmol L−1)

0.5 1 5

Apo-obelin

Firefly luciferase

Bacterial luciferase

〈τ0〉 (ns)

〈τ〉 (ns)

hti ht 0 i

〈τ0〉 (ns)

〈τ〉 (ns)

hti ht 0 i

〈τ0〉 (ns)

〈τ〉 (ns)

hti ht 0 i

3.99 1.33 0.08

3.99 1.34 0.33

1.0 1.0 4.1

4.04 1.38 0.09

4.04 1.48 0.22a

1.0 1.1 2.2

4.03 1.33 0.09

4.02 1.42 0.43a

1.0 1.1 4.8

Values of 〈τ〉 were measured at Cenzyme/Cdye =10 a

Cenzyme/Cdye =3

〈τ〉 of the enzymes by using Eq. (6). These values are presented in Table 2. It should be emphasized that the average lifetimes of free dyes depend on pH, and, as is seen from Table 2, 〈τ0〉 varies slightly in buffer solutions of different BL systems. Table 2 shows that 〈τ0〉 and 〈τ〉 decrease in the series of xanthene dyes: fluorescein, eosin, and erythrosin. So “the effect of heavy halogen atom” is observed not only for free dyes, but for enzyme-bound dyes, also. Here one can also see that on addition of enzyme the lifetime of fluorescein does not change whereas that of eosin seems to increase slightly, if at all, and that of erythrosin increases substantially. The average lifetime increase is known to by indicative of the intensification of intermolecular interactions [18]. Ratios 〈τ〉/〈τ0〉 (Table 2) describe these interactions more clearly for all enzymes. The data demonstrate the dependence of the efficiency of intermolecular interactions on the atomic weight of halogen substituents in homologous dye molecules, free and bound. Steady-state fluorescence anisotropy of xanthene dyes in the presence of enzymes Because fluorescence anisotropy measurements provide information on molecular size and shape, they can be used for studying interactions of small fluorophores with proteins [18]. The efficiency of enzyme–dye binding was evaluated from changes in the fluorescence anisotropy of xanthene dyes in the presence of excess concentrations of the enzymes. The effect of firefly luciferase on steady-state fluorescence anisotropy of xanthene dyes is depicted in Fig. 5. For all three enzymes, the characteristics of enzyme–dye binding are presented in Table 3. It is seen from Fig. 5 and Table 3 that anisotropy of the free dyes in the absence of enzymes, rf, grows as the atomic weight of halogen substituents increases. Additionally, rf values depend on fluorophore fluorescence lifetime: short fluorescence lifetime of erythrosin (Table 2) reduces the contribution of rotational diffusion to fluorescence depolar-

ization whereas longer lifetime of fluorescein (Table 2) increases it. It should be noted that self-association of xanthenes in dilute solutions is negligible [40] and does not contribute to rf values. Fluorescence anisotropy of the dyes was measured for different protein concentrations. The increase in anisotropy (Fig. 5) indicates binding of the dyes to the macromolecules [18]. The maximum anisotropy of rigidly bound enzyme– dye pairs, rbmax , calculated by use of Eq. (10) and the apparent dissociation constant of dye-protein complex, Kd0 , calculated by use of Eq. (8) are presented in Table 3. The following general trend is evident from Kd0 values (Table 3): binding efficiency increases in the series fluorescein–eosin–erythrosin for all three enzymes. The data demonstrate “the effect of heavy halogen atom” in the process of nonspecific binding of halogenated molecules with enzymes. It could be defined as “dark heavy-atom effect” because it is not directly affected by the population of electron-excited states of BL emitters; enzyme–dye interactions were followed by formation of the excited emitter.

Fig. 5 Fluorescence anisotropy r of fluorescein (1), eosin (2), and erythrosin (3) solutions (C= 1 μmol L−1) vs. firefly luciferase concentration



Table 3 The apparent dissociation constants (Kd0 ), and fluorescence anisotropy of free (rf) and rigidly bound (rbmax ) dyes Dye


Apo-obelin

Firefly luciferase

Bacterial luciferase

rf

rbmax

Kd0 (mol L−1)

rf

rbmax

Kd0 (mol L−1)

rf

rbmax

Kd0 (mol L−1)

0.006 0.033 0.194

0.235 0.308 0.365

(1.0±0.1)×10−3 (1.8±0.2)×10−4 (3.5±0.4)×10−5

0.005 0.029 0.197

0.296 0.338 0.370

(2.4±0.2)×10−4 (1.7±0.2)×10−5 (1.9±0.2)×10−6

0.005 0.031 0.195

0.306 0.343 0.374

(2.0±0.2)×10−3 (1.0±0.1)×10−4 (3.1±0.3)×10−5

Similar results were obtained for interaction of fluorescein and its halogenated derivatives with albumin microspheres [41]. The authors of Ref. [41] noticed that the highly polarizable halogen substituents bromine and iodine substantially change the characteristics of the xanthene ring. The octanol–water partition coefficients (log Kow) of fluorescein, eosin, and erythrosin are 4.05, 7.48, and 8.42, respectively [41], indicative of a hydrophobicity increase in this series. The enhanced self-association of halogenated fluoresceins in comparison with unsubstituted one confirms this assumption [40]. So, the trend mentioned might result from hydrophobic enzyme–dye interactions. To compare the hydrophobic properties of the enzymes, the accessible nonpolar surface of proteins was analyzed by use of VADAR software [42] and using PDB files for obelin (1SL9), firefly luciferase (2D1Q), and bacterial luciferase (1LUC) [43]. Nonpolar surface areas of firefly and bacterial luciferases were found to be approximately 13000Å2, and that of obelin approximately 5000Å2. This means that firefly luciferase, being smaller than the bacterial one, is the most hydrophobic protein. The result is in accordance with the second general trend that is evident from Kd0 (Table 3): efficiency of dye binding is maximum for firefly luciferase. Note that in terms of the Stern–Volmer approach, the hydrophobic dye–enzyme interactions are interpreted as static quenching of BL emitters. This is supported by similar values of the binding constant Kb0 (calculated as Kb0 ¼ 1=K 0 d ) and KS values (Tables 3 and 1, respectively).

Conclusion The paper develops the application of luminescent bioassays to monitoring of the toxicity of organic halides. It demonstrates the effect of heavy halogen atom on BL reactions: in the reactions of luminous organisms (fireflies, marine bacteria, and marine coelenterates), luminescence inhibition by xanthene dyes depends on the weight of the halogen atom incorporated into the dye structure. The bacterial BL reaction was shown to be most sensitive to heavy halogen atoms (Fig. 2); the process of dynamic quenching contributes substantially to bacterial

BL inhibition by halogenated dyes (Table 1). Hence, bacterial BL can be recommended for construction of BL sensors to monitor the toxicity of halogenated organic compounds. The obelin BL reaction is the least sensitive to heavy halogens involved in the molecule structure. Interaction of the dyes with enzymes was studied using the fluorescence characteristics of the dyes in steady-state and time-resolved experiments. Dependences of fluorescence anisotropy of enzyme-bound dyes, average fluorescence lifetime, and number of exponential components in fluorescence decay on atomic weight of halogen substituents were demonstrated. The effects were attributed to the “dark” process, namely, to the interaction of halogenated compounds with enzymes, followed by excited BL emitter formation; hydrophobic interactions were assumed to be responsible for this effect. Firefly luciferase was found to be the most effective enzyme in dye–enzyme binding interactions.

Acknowledgements The authors thank Professor L. Frank for obelin preparation and Professor A.G. Sizykh for discussion of the results. The work was supported by: the program “Molecular and Cellular Biology” of the RAS; the grant “Leading Scientific School” N 64987.2010.4, a Grant of the Ministry of Education and Science RF N 2.2.2.2/5309; the Federal Target Program “Research and scientificpedagogical personnel of innovation in Russia”, 2009–2013, contract N 02.740.11.0766.

References 1. Amachi S, Kamagata Y, Kanagawa T, Muramatsu Y (2001) Bacteria mediate methylation of iodine in marine and terrestrial environments. Appl Environ Microbiol 67(6):2718–2722 2. Ehrenshaft M, Mason RP (2006) Protein radical formation on thyroid peroxidase during turnover as detected by immuno-spin trapping. Free Radic Biol Med 41(3):422–430 3. Esimbekova EN, Torgashina IG, Kratasyuk VA (2009) Comparative study of immobilized and soluble NADH:FMN-oxidoreductaseluciferase coupled enzyme system. Biochem Moscow 74(6):695–700 4. Roda A, Pasini P, Mirasoni M, Michchelini E, Guardigli M (2004) Biotechnological application of bioluminescence and chemiluminescence. Trends Biotechnol 22(6):295–303 5. Girotti S, Ferri E, Fumo M, Maiolini E (2008) Monitoring of environmental pollutants by bioluminescent bacteria. Anal Chim Acta 608(1):2–29

Author's personal copy Effect of halogenated fluorescent compounds on bioluminescence 6. Kudryasheva NS (2006) Bioluminescence and exogenous compounds. Physico-chemical basis for bioluminescent assay. J Photochem Photobiol B 83(1):77–86 7. Nemtseva EV, Kudryasheva NS (2007) The mechanism of electronic excitation in bacterial bioluminescent reaction. Usp Khim 76(1):101–112, in Russian 8. Gerasimova MA, Kudryasheva NS (2002) Effects of potassium halides on bacterial bioluminescence. J Photochem Photobiol B 66:218–222 9. Gerasimova MA, Sizykh AG, Slyusareva EA (2009) The role of energy transfer in bioluminescence quenching by xanthene dyes. J Photochem Photobiol B 97:117–122 10. Kirillova TN, Kudryasheva NS (2007) Effect of heavy atoms in bioluminescent reactions. Anal Bioanal Chem 387:2009–2016 11. Rodionova NS, Petushkov VN (2006) Effect of different salts and detergents on luciferin-luciferase luminescence of the enchytraeid Fridericia heliota. J Photochem Photobiol B 83(2): 123–128 12. Kasha M (1952) Collisional perturbation of spin-orbital coupling and the mechanism of fluorescence quenching. A visual demonstration of the perturbation. J Chem Phys 20(1):71–74 13. McGlynn SP, Sunseri R, Christodouleas N (1962) External heavyatom spin-orbital coupling effect. I. The nature of the interaction. J Chem Phys 37(8):1818–1824 14. Nag-Chaudhuri J, Stoessell L, McGlynn SP (1963) External heavy-atom spin-orbital coupling effect. II. Comments on the effect of ferric acetylacetonate on the spectra of polyacenes. J Chem Phys 38(8):2027–2028 15. McGlynn SP, Daigre J, Smith FJ (1963) External heavy-atom spin-orbital coupling effect. IV. Intersystem crossing. J Chem Phys 39(3):675–679 16. McGlynn SP, Azumi T, Kasha M (1964) External heavy-atom spin-orbital coupling effect. V. Absorption studies of triplet states. J Chem Phys 40(2):507–515 17. Solov’ev KN, Borisevich EA (2005) Intramolecular heavy-atom effect in the photophysics of organic molecules. Usp Fiz Nauk 175(3):247–270, in Russian 18. Lakowicz JR (2006) Principles of fluorescence spectroscopy, 3rd edn. Springer, New York 19. Demchenko AP (2009) Introduction in fluorescence sensing. Springer, New York 20. Liu J, Liu Z, Hu X, Kong L, Liu S (2008) Fluorescence quenching method for the determination of bleomycins A5 and A2 with halofluorescein dyes. Luminescence 23(1):1–6 21. Kuijt J, Ariese F, Brinkman UATh, Gooijer C (2003) Room temperature phosphorescence in the liquid state as a tool in analytical chemistry. Anal Chim Acta 488(2):135–171 22. Ugarova NN (1989) Luciferase of Luciola mingrelica fireflies: kinetics and regulation mechanism. J Biol Chem 4(1):406–418 23. Devine JH, Kutuzova GD, Green VA, Ugarova NN, Baldwin TO (1993) Luciferase from the East European firefly Luciola mingrelica: cloning and nucleotide sequence of the cDNA, overexpression in Escherichia coli and purification of the enzyme. Biochim Biophys Acta 1173:121–132 24. Markova SV, Vysotski ES, Blinks JR, Burakova LP, Wang B-C, Lee J (2002) Obelin from the bioluminescent marine hydroid Obelia geniculata: cloning, expression, and comparison of some properties with those of other Ca2+-regulated photoproteins. Biochemistry 41(7):2227–2236

351 25. Liu ZJ, Vysotski ES, Chen CJ, Rose JP, Lee J, Wang B-C (2000) Structure of the Ca2+-regulated photoprotein obelin at 1.7 Å resolution determined directly from its sulfur substructure. Protein Sci 9:2085–2093 26. Kuznetsov AM, Rodicheva EK, Shilova EV (1996) Biotest based on lyophilized bacteria. J Biotechnol 9:57–61 27. Clarc AC, Raso SW, Sinclair JF, Ziegler MM, Chaffotte AF, Baldwin TO (1997) Kinetic mechanism of luciferase subunit folding and assembly. Biochemistry 36(7):1891–1899 28. Eckstein JW, Hastings JW, Ghisla S (1993) Mechanism of bacterial bioluminescence: 4a, 5-dihydroflavin analogs as models for luciferase hydroperoxide intermediates and the effect of substituents at the 8-position of flavin on luciferase kinetics. Biochemistry 32(2):404–411 29. O’Connor DV, Phillips D (1984) Time correlated single photon counting. Academic Press, London 30. Lettinga MP, Zuilhof H, van Zandvoort MAMJ (2000) Phosphorescence and fluorescence characterization of fluorescein derivatives immobilized in various polymer matrices. Phys Chem Chem Phys 2:3697–3707 31. Eftink MR, Ghiron CA (1981) Fluorescence quenching studies with proteins (Review). Anal Biochem 114(2):199–227 32. Thipperudrappa J, Biradar DS, Hanagodimath SM (2007) Simultaneous presence of static and dynamic component in the fluorescence quenching of Bis-MSB by CCl4 and aniline. J Lumin 124(1):45–50 33. Zelent B, Kusba J, Gryczynski I, Johnson ML, Lakowicz JR (1998) Time-resolved and steady-state fluorescence quenching of N-acetyl-L-tryptophanamide by acrylamide and iodide. Biophys Chem 73(1–2):53–75 34. KrishnaMurthy NV, Sudhaharan T, Ram Reddy A (2007) Dye induced quenching of firefly luciferase–luciferin bioluminescence. Spectrochim Acta A 68:851–859 35. Kudryasheva NS, Belobrov PI, Kratasyuk VA, Shigorin DN (1991) Electron-excited states in bioluminescence. Dokl AN SSSR 321(4):837–841, in Russian 36. Kudryasheva NS, Nemtseva EV, Meshalkin YuP, Sizykh AG (2001) Upper electron-excited states in bioluminescence: experimental indication. Luminescence 3:243–246 37. Kudryasheva NS, Nemtseva EV, Sizykh AG, Kratasyuk VA, Visser AJWG (2002) Estimation of energy of the upper electronexcited states of the bacterial bioluminescent emitter. J Photochem Photobiol B 68(2–3):88–92 38. Belogurova NV, Kudryasheva NS, Alieva RR (2009) Activity of upper electron-excited states in bioluminescence of coelenterates. J Mol Struct 924–926:148–152 39. Kudryasheva NS, Nemtseva EV, Kirillova TN (2004) Exogenous compounds in studying the mechanism of electron-excited states formation in bioluminescence. Biopolymers 74(1–2):100–104 40. De S, Das S, Girigoswami A (2005) Environmental effects on the aggregation of some xanthene dyes used in lasers. Spectrochim Acta A 61:1821–1833 41. Egbaria K, Friedman M (1992) Adsorption of fluorescein dyes on albumin microspheres. Pharm Res 9(5):629–635 42. Willard L, Ranjan A, Zhang H, Monzavi H, Boyko RF, Sykes BD, Wishart DS (2003) VADAR: a web server for quantitative evaluation of protein structure quality. Nucleic Acids Res 31 (13):3316–3319 43. http://www.pdb.org