A Fluorescence Perspective on the Differential Interaction of Riboflavin and Flavin AdenineDinucleotide with Cucurbit[7]uril

Sharmistha Dutta Choudhury,* Jyotirmayee Mohanty, Achikanath C. Bhasikuttan, andHaridas Pal*Radiation & Photochemistry DiVision, Bhabha Atomic Research Centre, Trombay, Mumbai 400 085, India

ReceiVed: May 7, 2010; ReVised Manuscript ReceiVed: July 2, 2010

The interaction of the macrocyclic host, cucurbit[7]uril (CB7), with riboflavin (RF) and its derivative, flavinadenine dinucleotide (FAD), has been investigated using absorption and steady-state and time-resolvedfluorescence measurements. Interestingly, in the presence of CB7, the fluorescence intensity of RF is quenched,whereas the fluorescence intensity of FAD is enhanced. It is proposed that the fluorescence quenching of RFresults from the tautomerization of its isoalloxazine moiety from the lactam to the lactim forms, upon bindingto CB7. Such a tautomerization can be brought about since the two lactim forms have higher dipole momentsthan the lactam form of RF, and thus experience much stronger dipole-dipole interactions (and hence greaterbinding affinities) with CB7 in the former cases than in the latter. This tautomerization in the presence ofCB7 leads to a significant reduction in the observed radiative decay rate and hence a reduction in thefluorescence intensity of RF. Binding of CB7 with RF is confirmed by an increase in the rotational correlationtime of RF in the presence of CB7. Geometry optimization studies indicate the formation of an exclusioncomplex between CB7 and RF, possibly stabilized by H-bonding interactions, as also suggested by thecharacteristic red shift in the absorption spectra of the CB7-RF system. In the case of FAD, both theisoalloxazine ring and the adenine moiety can interact with the CB7 host. In aqueous solutions, a good fractionof FAD molecules exists in a “closed” conformation with the adenine and isoalloxazine rings stacked together,thus leading to very efficient fluorescence quenching due to the ultrafast intramolecular electron transferfrom adenine to the isoalloxazine moiety. Binding of the adenine and/or the isoalloxazine moiety of FADwith CB7 inhibits the stacking interaction and changes the “closed” conformation to the “open” conformation,wherein the adenine and isoalloxazine moieties are widely separated, thus prohibiting the electron transferprocess. This reduces the inherent fluorescence quenching of FAD molecule and results in the observedfluorescence enhancement. As observed for RF, the interaction of CB7 with the isoalloxazine ring of FADshould cause fluorescence quenching due to the lactam to lactim tautomerization process. However, in theinterplay between the above two opposing effects, the fluorescence enhancement due to the modulation inthe conformational dynamics of FAD by the CB7 host predominates. The conformational change is in factsupported by the observation of a long lifetime component in the fluorescence decay of FAD in the presenceof CB7. Moreover, at acidic pH, when FAD is already present mainly in the “open” form, the conformationaldynamics no longer plays any major role and the fluorescence of FAD is quenched by CB7, as expected, dueto the tautomerization at the isoalloxazine moiety.

1. Introduction

Supramolecular host-guest interaction is rapidly emergingas the method of choice for tuning molecular properties to meetdesired applications like drug delivery vehicles, sensors, design-ing molecular architectures, catalysis chambers, biomimeticsystems, and so on.1-8 Presently, the repertoire of host systemsavailable is quite vast, the most commonly studied macrocyclichosts being the different homologues and derivatives of cyclo-dextrins, cucurbiturils, and calixarenes.9-11 Different types ofinvestigations with these systems have led to a good understand-ing about the mode of their interactions with a variety of guestmolecules, the primary driving forces responsible for thecomplexation, and about the prospective applications of suchhost-guest interactions.9-15 Several groups, including ours, havedemonstrated many of the aforementioned interesting phenom-ena through host-guest interactions like changes in acid-base

behavior of encapsulated guests,5,16 guest relocation from hostcavity to biomolecular binding pocket,17 supramolecular archi-tecture formation,2,6,18 biological catalysis,3 sensors,19,20 andfluorescent capsule formation.21

In a previous communication, we reported on the modulationof the intramolecular electron transfer process in flavin adeninedinucleotide, upon interaction with the �-cyclodextrin (�-CD)host.22 Flavin mononucleotide (FMN) and flavin adenine di-nucleotide (FAD), which consist of a heterocyclic isoalloxazinemoiety tethered to a ribityl phosphate or ribityl adeninediphosphate chain, respectively, are the most commonly occur-ring flavins in flavoproteins (Scheme 1a). These flavin cofactorsare derivatives of riboflavin (RF), a compound better known asvitamin B2 (Scheme 1b).23 Because of their chemical versatility,flavoproteins are ubiquitous and participate in a broad spectrumof biological activities.24,25 Flavoproteins are the ideal systemsfor studies of intraprotein electron transfer and conformationaldynamics of biomacromolecules, not only because the flavin(isoalloxazine) moiety is a redox-active group suitably located

* Corresponding authors. E-mail: sharmidc@barc.gov.in (S.D.C.);hpal@barc.gov.in (H.P.). Fax: 91-22-25505151/25519613.

J. Phys. Chem. B 2010, 114, 10717–10727 10717

10.1021/jp1041662 2010 American Chemical SocietyPublished on Web 08/04/2010

in the heart of the active site but also due to the fact that it isa fluorescent chromophoric group, which makes it amenablefor various fluorescence studies.26-29 The fluorescence spectralcharacteristics as well as the fluorescence quantum yield offlavins strongly depend on environmental factors like solventpolarity and refractive index.30-32 Although RF and FMN havereasonably high fluorescence quantum yields (Φf ) 0.26) inaqueous solutions, FAD is very weakly fluorescent (Φf )0.03).33 The remarkably low fluorescence yield of FAD com-pared to RF or FMN was first reported by Weber, and it is nowwell understood that the reduction in the fluorescence quantumyield of FAD results from both static and dynamic quenchingof the flavin fluorescence due to the photoinduced electrontransfer from the adenine moiety to the isoalloxazine moiety.34-40

On the basis of different studies, like circular dichroism,41

NMR,42 ultraviolet resonance Raman spectroscopy,43 and MDsimulations,40 it is confirmed that, in solution, FAD exists intwo conformations: an extended or “open” conformation in

which the isoalloxazine and the adenine moieties are largelyseparated from each other and a “closed” conformation in whichthe two aromatic rings are in close proximity and stackedtogether.34-40 The “closed” conformation is preferred in aqueoussolutions, and is stabilized by the combined effect of the π-πinteraction between the isoalloxazine ring and the adeninemoiety and the intramolecular H-bonding interactions along thephosphate sugar backbone.35 Sequence-structure relationshipsof several FAD binding proteins have revealed that, in most ofthese proteins, the FAD cofactor is bound in an extended mannerexcept in some members of the ferredoxin reductase family andin the DNA photolyase enzyme, where the cofactor adopts abent conformation.44 Although there is no explicit correlationbetween the conformation of FAD and the activity of theflavoproteins, there have been some indications of a functionalimplication of the FAD conformation.45,46 Quantum chemicalcalculations on the FAD cofactor in DNA photolyase, which isan enzyme that catalyzes photorepair of UV damaged DNA by

SCHEME 1: Chemical Structures of (a) FAD, (b) RF (Different Tautomeric Forms), and (c) CB7a

a Major dimensions of interest for these systems are also indicated.

10718 J. Phys. Chem. B, Vol. 114, No. 33, 2010 Dutta Choudhury et al.

an electron transfer mechanism, suggest that, due to the bentconformation of FAD, the electron transfer between flavin andthe thymine bases takes place indirectly with the adenine moietyacting as an intermediate.46 Femtosecond time-resolved studies,on the other hand, rule out any direct involvement of the adeninegroup in the electron transfer repair mechanism of DNAphotolyase but suggest that the adenine moiety mediates therepair by anchoring the thymine dimer through H-bonds andthus modulates the electron jump by a superexchange mecha-nism.47 Very recently, Acocella et al. have used a quantummechanical time-dependent approach to show that the adeninemoiety in photolyase provides the necessary electrostaticinteractions to promote the electron transfer and sterically keepsthe influence of the surrounding medium under control.48 Thus,in our previous study, we felt it was quite intriguing toinvestigate the conformational changes of FAD as well as themodulation in its photophysical properties and intramolecularelectron transfer behavior in the presence of the biomimeticbinding pockets of CD hosts. Our results revealed that thebinding of �-CD to FAD leads to a change in its conformationfrom the “closed” to the “open” form.22 This conformationalchange increases the distance between the electron donor(adenine) and the electron acceptor (isoalloxazine) moieties, andthus inhibits the electron transfer process, which in turn leadsto a significant fluorescence enhancement as well as an increasein the fluorescence lifetime of FAD. Comparative studies withthe parent molecule, RF, showed no characteristic changes, thusindicating that the �-CD host binds preferably to the adeninemoiety of FAD rather than to its isoalloxazine ring.

In the present work, we have investigated the interaction ofRFandFADwithanotherversatilemacrocyclichost,cucurbit[7]uril(CB7, Scheme 1c), using absorption, steady-state fluorescence,and time-resolved fluorescence measurements. Although thedimensions of the host cavities are quite similar for both �-CDand CB7, they have large structural and functional dissimilari-ties, which lead to the diverse behavior of the two macrocyclichosts. While CB7 is a cyclic polymer of seven glycoluril units,�-CD is composed of seven glucopyranose units. Unlike �-CD,which is shaped like a torus and whose ends are laced withhydroxyl groups, CB7 is a symmetrical pumpkin shaped mole-cule having carbonyl laced portals. This variation in thecomposition of the peripheral functional groups leads tosignificant differences in the nature of interactions of �-CD andCB7 hosts with various guest molecules. Quite understandably,due to the presence of the carbonyl groups at the portals ofCB7, the dipole-dipole and/or ion-dipole interactions play amajor role in the stable complex formation of CB7 with suitableguest molecules. In the present study, interactions between theCB7 host and the biologically important molecules, RF andFAD, have been investigated, considering the present host-guest systems as an alternative and simple model for the enzyme-substrate interactions in biological systems.

2. Experimental Section

Riboflavin (RF) and cucurbit[7]uril (CB7) were obtained fromAldrich and used as such. FAD was obtained from Sigma andused after purification by ion exchange chromatography for theremoval of the possible degradation products (FMN and RF)in the sample. Thus, the FAD sample from Sigma (purity 96%)was eluted at an ion strength gradient between 1 and 10 mMphosphate buffer (pH 7.0) using a DEAE Sephacil column(Sigma). The concentrations of RF and FAD were calculatedfrom their molar extinction coefficients (ε450, RF ) 12200M-1cm-1, ε450, FAD ) 11300 M-1cm-1)49 and were maintained

in the range 4-10 µM in all of the experimental solutions. Theinteraction of RF or FAD with CB7 was studied by addingdifferent weighed amounts of CB7 to the respective fluorophoresolutions. All studies with FAD at pH 7 were performed in 10mM phosphate buffer in Nanopure water (Millipore Elix3/A10water purification system; conductivity of <0.1 µS cm-1) atambient temperature.

Absorption spectra were recorded with a Shimadzu UV-visspectrophotometer (model UV-160A), and steady-state fluores-cence spectra were recorded with a Hitachi spectrofluorimeter(F-4500). The time-resolved fluorescence measurements werecarried out with a time-correlated single photon counting(TCSPC) spectrometer (Horiba Jobin Yvon IBH, U.K.). A 374nm diode laser (∼100 ps, 1 MHz repetition rate) was used asthe excitation source, and an MCP-PMT based detection modulewas used for the measurement of the fluorescence decays.Except for anisotropy measurements, all other fluorescencedecays were collected at magic angle (54.7°) with respect tothe vertically polarized excitation light, to avoid the effect ofrotational depolarization of the dye on the measured fluorescencelifetimes.50,51 The DAS-6 software from IBH was used for thedeconvolution analysis of the observed decays, consideringeither monoexponential or biexponential decay functions. Thequality of the fits and consequently the mono- and biexponentialnatures of the decays were judged by the reduced chi-square(�2) values and the distribution of the weighted residuals amongthe data channels. For a good fit, the �2 value was close to unityand the weighted residuals were distributed randomly amongthe data channels.50,51 For fluorescence anisotropy measurements,the polarized fluorescence decays for the parallel (I|(t)) andperpendicular (I⊥(t)) emission polarizations with respect to thevertical excitation polarization were first collected. Using theseI|(t) and I⊥(t) decays, the anisotropy decay function, r(t), wasconstructed as

where G is a correction factor for the polarization bias of thedetection set up.50,51 The G factor was obtained independentlyby measuring the two perpendicularly polarized fluorescencedecays and using horizontally polarized light for sampleexcitation. Computational studies were performed with theGaussian 92 package.52

3. Results and Discussion

3.1. Absorption, Steady-State Fluorescence, and Time-Resolved Fluorescence Changes of RF in the Presence ofCB7. Figure 1 shows the absorption spectra of RF in water inthe absence and presence of the macrocyclic host, CB7. Thespectra in the presence of CB7 are presented after subtractionof the background absorption by the respective CB7 concentra-tions to eliminate the contribution of the overlapping absorptionof the CB7 host in the observed wavelength region. Theabsorption bands, with a maximum around 450 and 370 nm,correspond to the S0 to S1 and S0 to S2 transitions of RF,respectively.30 With increasing concentration of CB7, there isa decrease in the absorbance of the 450 nm band and a clearred shift of the 370 nm band. The changes in the absorptionspectra definitely indicate some interaction of RF with the CB7host. An apparent isosbestic point is observed around 413 nmwhich indicates the conversion of the free RF to CB7-RFcomplex. The red shift of the 370 nm band is indicative of

r(t) )I|(t) - GI⊥(t)

I|(t) + 2GI⊥(t)(1)

Differential Interaction of RF and FAD with CB7 J. Phys. Chem. B, Vol. 114, No. 33, 2010 10719

H-bonding interaction between RF and CB7.32 The marked redshift of the 370 nm band in contrast to the 450 nm band is dueto the stronger H-bonding of RF in the S2 state than in the S1

state due to the greater dipole moment of RF in the formerstate.32 The absorption spectra of RF in water and methanol,depicting the red shift of the 370 nm band in water, are shownin the inset of Figure 1, for comparison.

The fluorescence spectrum of RF is broad with a maximumaround 530 nm. On addition of CB7, the fluorescence intensitygradually decreases and is accompanied by a small (∼2 nm)red shift of the emission spectra (Figure 2). It should bementioned here that these changes in the absorption andfluorescence spectral features are characteristic of the interactionof the CB7 host, since no interaction was observed betweenRF and the cyclodextrin (CD) hosts in our previous study.22 Itis also interesting to note that the specific interaction betweenRF and CB7 leads to a reduction in the fluorescence intensityrather than to the commonly expected fluorescence enhance-ment. In general, during host-guest interactions, the guestmolecule is either encapsulated within the host cavity orexternally bound to the host by noncovalent forces, dependingon the relative dimensions of the guest and host cavity.53 For afluorescent guest molecule, such complexation generally results

in fluorescence enhancement, since the vibrational and rotationalmotions of the guest are restricted and consequently thenonradiative decay rates of the excited guest are reduced. Froma comparison of the relative sizes of the CB7 host (portaldiameter 5.4 Å) and the isoalloxazine group (4.7 Å, cf. Scheme1) of RF, and also considering the steric hindrance provided bythe ribityl side chain, either a partial inclusion of RF within thehost cavity or an exclusion complex formation can be envisioned(see the geometry optimization studies discussed later). In thelatter case, the fluorescence enhancement may not be as highas that for an inclusion complex formation;53 nevertheless, thereduction in the fluorescence intensity is quite surprising.

For a better understanding of the fluorescence quenchingprocess, the fluorescence lifetimes of RF have been measuredin the absence and presence of CB7. A decrease in thefluorescence lifetime is observed with increasing CB7 concen-trations (Figure 3 and Table 1). However, the lifetime changesare substantially less in comparison to the significant decreasein the steady-state fluorescence intensities. This indicates thatthe fluorescence quenching primarily results from a ground stateinteraction of RF with the CB7 host. A typical Stern-Volmerplot (I0/I vs quencher concentration) for the steady-statefluorescence quenching of RF shows a clear negative deviationfrom linearity (inset I, Figure 2). Such negative deviations canbe observed if the fluorophore population is not completelyaccessible to the quencher, as commonly seen in biomolecularenvironments and in heterogeneous media, or if the ground stateinteraction between the fluorophore and the quencher leads tothe formation of a less fluorescent complex (rather than anonfluorescent complex).50 Since in the present case theinteracting species are simply the free RF and CB7, the partialaccess of the fluorophores toward the quenchers is an unlikelyproposition. Thus, the formation of a relatively less fluorescentCB7-RF complex is the more likely mechanism for the negativedeviation observed in the Stern-Volmer plot. Representing thehost-guest interaction as

Figure 1. Representative absorption spectra of RF (6 µM) in 10 mMphosphate buffer at pH 7 with different CB7 concentrations, aftercorrection for the background absorption by CB7. [CB7]/mM: (1) 0.0,(2) 0.05, (3) 0.10, (4) 0.20, (5) 0.65, (6) 1.0. The inset shows theabsorption spectra of RF (9.5 µM) in (a) methanol and (b) water.

Figure 2. Steady-state fluorescence spectra of RF (6 µM) in 10 mMphosphate buffer (pH 7) at different CB7 concentrations. [CB7]/mM:(1) 0.0, (2) 0.02, (3) 0.05, (4) 0.10, (5) 0.29, (6) 0.65, (7) 1.0. Theexcitation wavelength was 420 nm. Inset I shows the typicalStern-Volmer plot, and inset II shows the binding isotherm forCB7•FAD with the fitted curve according to eq 5.

Figure 3. Fluorescence decay traces for RF (6 µM) monitored at 530nm in the absence (1) and in the presence (2) of CB7 (1 mM). L is thelamp profile. The inset shows the fluorescence anisotropy decays, r(t),for RF (6 µM) in 10 mM phosphate buffer (pH 7) in the absence (a)and in the presence (b) of CB7 (1 mM). The excitation wavelengthwas 374 nm.

TABLE 1: Fluorescence Decay Parameters for RF in 10mM Phosphate Buffer (pH 7) Monitored at 530 nm in theAbsence and Presence of CB7

CB7 (mM) 0 0.05 0.20 0.65 1.00τ1 (ns) 4.70 4.58 4.45 4.35 4.32�2 1.08 1.12 1.16 1.14 1.09

10720 J. Phys. Chem. B, Vol. 114, No. 33, 2010 Dutta Choudhury et al.

and considering [RF]0 and [CB7]0 as the total concentrationsof RF and CB7, respectively, the concentration of the free(uncomplexed) RF in equilibrium with CB7•RF complex canbe expressed as eq 3.16,54,55

Since the guest exchange rate constants, i.e., the conversion ofthe uncomplexed guest to the complexed one and vice versa,are very small for macrocyclic hosts (occurs in microseconds),the exchange of RF during its excited-state lifetime (∼4.6-4.2ns) can be neglected.55 Thus, the observed fluorescence intensitycan be understood as a composite of the fluorescence intensitycontributions from the complexed and uncomplexed forms ofRF and should be expressed as eq 4:56

where IRF0 is the initial fluorescence intensity in the absence of

CB7 and ICB7•RF∞ corresponds to the fluorescence intensity if all

of the RF molecules in the solution were complexed with CB7.The expression for the changes in the fluorescence intensity (∆If

λ)in the presence of CB7 can therefore be obtained by therearrangement of eq 4 and is given as16,55

where [RF]eq can be obtained from eq 3. In the present case,since the host-guest interaction leads to fluorescence quenching,the fluorescence titration curve was obtained by plotting -∆If

λ

vs [CB7]0 (instead of ∆Ifλ vs [CB7]0), as shown in inset II of

Figure 2. It is to be mentioned here that a plot of I0/I vs CB7concentration should also follow a trend similar to the -∆If

λ vs[CB7]0 plot, as indeed observed in the present case (cf. Inset Iof Figure 2). These observations thus support the formation ofa ground state host-guest complex between CB7 and RF. Byfitting the fluorescence titration curve shown in inset II of Figure2 using eqs 3 and 5, the equilibrium constant value, K, for theinteraction of RF with CB7 is estimated to be 6700 ((50) M-1.

Evidence for complex formation in the CB7-RF system hasalso been obtained from time-resolved anisotropy measurements.Since the anisotropy decay rate depends upon the effective sizeof the fluorescent molecule and the microviscosity of its localenvironment, it is expected that the rotational diffusion rate ofRF should decrease (rotational correlation time should increase)upon complexation with the CB7 host due to an increase in theeffective size of the fluorescent moiety (dye + host).56 Thefluorescence anisotropy decay for free RF in aqueous solutionis found to be single exponential in nature with a rotationalcorrelation time of ∼130 ps. In the presence of CB7, theanisotropy decay is considerably slower and the rotationalcorrelation time is found to be ∼400 ps, thus supporting theformation of a supramolecular complex between RF and CB7.

The anisotropy decay traces for RF in the absence and presenceof CB7 are shown in the inset of Figure 3.

To understand the mechanism of the fluorescence quenchingin the present systems, the radiative and nonradiative decay rateconstants (kf and knr, respectively) for RF were estimatedfollowing eqs 6 and 7, in the absence and presence of CB7(∼1 mM CB7, saturation concentration for the bindinginteraction).50

Here, φf is the quantum yield and τf is the fluorescence lifetimeof the dye, RF. It is found that the nonradiative rate constantfor RF increases from 1.58 × 108 to 1.98 × 108 s-1, whereasthe radiative rate constant decreases from 5.54 × 107 to 3.38 ×107 s-1 upon complex formation. The increase in the nonradia-tive rate can be attributed to the H-bonding interaction betweenthe guest and the host, as also evidenced by the red shift of the370 nm absorption band of RF in the presence of CB7. However,the decrease in the radiative rate seems to be more significantin the present case and needs to be examined more critically.

As mentioned earlier, fluorescence quenching due to host-guest interaction is rather unusual. However, there are someliterature reports of such a phenomenon, although the reasonsfor the fluorescence quenching are not always very clear.57-61

Schuette et al. have explained the reduction in the fluorescenceof acridine dye in the presence of �-CD in terms of increasedintersystem crossing of acridine in the relatively aprotic mediuminside the �-CD cavity.57 Liu et al. have attributed thefluorescence quenching of rhodamine B in the presence of �-CDto the preferential complexation of �-CD with the nonfluorescentlactone form of rhodamine B and a conversion of somefluorescent zwitterionic form of the dye to the nonfluorescentlactone form in the presence of �-CD.58 Very recently, Zhou etal. have reported the fluorescence quenching of RF in thepresence of CB7, although no explanation was provided for theobserved effect.59 In the interaction of CB7 with lumichrome(7,8-dimethylalloxazine), which is an important member of theflavin family, Miskolczy et al. have observed a CB7 inducedtautomerization of lumichrome from the alloxazine to theisoalloxazine form due to proton transfer from N(1) to the N(10)position, leading to the characteristic changes in the fluorescencespectra.62 Such a tautomerization is not possible for RF (7,8-dimethy-10-ribitylisolalloxazine) due to the ribityl substitutionat the N(10) position. Moreover, such tautomerization leads todrastic spectral changes which are not observed in the presentcase. However, for RF, the isoalloxazine ring can undergolactam-lactim tautomerization leading to three possible con-formations (Scheme 1b), even though these forms are notdistinguishable from absorption and fluorescence spectral char-acterstics.63 It is generally believed that the flavins are mainlypresent in the lactam form (A) and there is hardly anycontribution from the two hydroxy-imino (lactim) forms (B andB′); i.e., the prototropic equilibria in Scheme 1b are apparentlyshifted almost completely toward the lactam form.63 In thepresence of CB7, however, conversion of the lactam form tothe two lactim forms may be facilitated if the CB7 host hasbetter binding affinities with the lactim forms compared to thelactam form. Moreover, since there are two lactim forms of RF

RF + CB7 y\zK

CB7 • RF (2)

[RF]eq ) {K[RF]0 - K[CB7]0 - 1 +

√(K[RF]0 + K[CB7]0 + 1)2 - 4K2[RF]0[CB7]0}/2K (3)

If ) IRF0

[RF]eq

[RF]0+ ICB7•RF

∞ [CB7 • RF]eq

[RF]0(4)

∆Ifλ ) (1 -

[RF]eq

[RF]0)(ICB7•RF

∞ - IRF0 ) (5)

kf )φf

τf(6)

knr )1τf

- kf (7)

Differential Interaction of RF and FAD with CB7 J. Phys. Chem. B, Vol. 114, No. 33, 2010 10721

compared to the one lactam form in the equilibria, the effectiveinteraction of CB7 for the lactim forms will be higher than thelactam form, resulting in an enhanced conversion from lactamto the lactim forms. Therefore, considering the two lactim formsof RF to have lower fluorescence quantum yields than that ofthe lactam form, binding of RF to CB7 will obviously lead tofluorescence quenching as well as a reduction in the effectiveradiative rate constant. The situation may be explicitly repre-sented as in Scheme 2. Following this scheme, if KB, KB′ > KA,it follows that K2 > K1 and K2′ > K1′ (see Note S1 in theSupporting Information). In analyzing the fluorescence titrationcurve, however, it is quite justified to consider an effectivebinding constant K for all of the tautomeric forms, as given byeq 2 earlier, which can be more explicitly written as eq 8.

It should be stressed here that the complexation of RF withCB7 is not likely to convert the lactam form completely to thelactim forms; rather, a good fraction of the complexes shouldbe present in the CB7•RFlactam form. It is important to mentionfurther that the value of the equilibrium constant, K ∼6700 M-1,as estimated from the analysis of the fluorescence titration curve,can at the most be considered as the effective binding constantfor all of the different tautomeric forms of RF (Note S1,Supporting Information). Considering that KB, KB′ > KA, theobserved binding constant is expected to be greater than KA,which is the true binding constant of the lactam form with CB7.

We feel that the lactam to lactim conversion of RF in thepresence of CB7, which arises due to the higher binding affinityof CB7 with the lactim forms, is the reason for the reduction inthe fluorescence intensity of RF upon interaction with CB7.However, the question that arises is why the lactim forms shouldhave a better binding affinity for the CB7 host than the lactamform. It is known that the dipole-dipole interaction plays amajor role in the binding of various guest molecules with theCB7 host. Thus, if the dipole moments of the lactim forms arelarger than that of the lactam form, a preferential interaction ofthe lactim forms with CB7 can be anticipated. To havequalitative support for this idea, we have theoretically estimatedthe ground state dipole moments for the different tautomericforms of RF by PM3 calculations with molecular mechanics(MM) correction using the Gaussian 92 package.52 These valuesare listed in Table 2. It is seen that both of the lactim formshave higher dipole moments than that of the lactam form.Geometry optimization studies were also carried out to have

some insight on the mode of interaction of RF with CB7 andto have an estimate of the relative stabilities of the host-guestcomplexes. The ∆Hf values for the complex formation asestimated from the optimized parameters are listed in Table 2,and the optimized geometries are presented in Figure 4. The∆Hf values clearly show that the CB7 complexes with the lactimforms, which have higher dipole moments, are more stable thanthe CB7 complex with the lactam form of RF. These resultsthus support our proposition for the CB7 induced lactam tolactim tautomerization of RF. The geometry optimized structuresfurther indicate that, in the present systems, exclusion typecomplexes (rather than inclusion complexes) are formed for boththe lactam (A) and lactim (B and B′) forms of RF with CB7. Itis proposed that the exclusion complexes are possibly stabilizedby the H-bonding interactions of the carbonyl portals of CB7,with the -NH group in the lactam form and the -OH groupsin the lactim forms of RF. Such H-bonding interaction is infact supported by the observed red shift in the 370 nm absorptionband in the presence of the CB7 host. It has been suggested inthe literature that in biological systems the isoalloxazine moietiesare capable of undergoing tautomeric conversion into the enolicforms as a result of the H-bonding interactions with proteins.63,64

An analogous situation thus seemed to prevail in the presenceof the macrocyclic host, CB7.

3.2. Absorption, Steady-State Fluorescence, and Time-Resolved Fluorescence Changes of FAD in the Presence ofCB7. The absorption spectra of FAD in the absence andpresence of CB7 are presented in Figure 5 after correction forthe background absorption by the respective CB7 concentrations.Similar to RF, two absorption bands are observed with maximaaround 450 and 370 nm, corresponding to the S0 to S1 and S0

to S2 transitions of the isoalloxazine chromophore, respectively.The changes in the absorption spectra with increasing CB7concentrations are qualitatively similar to those of RF. A redshift of the 370 nm band is observed. However, unlike RF, noclear isosbestic point is observed in the present case due to alarger decrease in absorbance for both of the bands. The changesin the absorption spectra indicate that the isoalloxazine ring ofFAD interacts with the CB7 host. Like RF, the fluorescencespectrum of FAD is broad with maximum emission around 530nm (Figure 6). However, in this case with increasing CB7concentration, there is an enhancement in the fluorescenceintensity, unlike the fluorescence quenching observed in the caseof RF (cf. Figure 2).

In addition to the isoalloxazine ring, the FAD molecule hasanother potential site for interaction with CB7, namely, theadenine moiety. The dimension of the adenine moiety (∼4.8Å, cf. Scheme 1) is also quite compatible for a possible inclusionwithin the CB7 cavity. In fact, in our earlier study on theinteraction of FAD with the �-CD host, we observed that themajor host-guest interaction takes place via the inclusion ofthe adenine moiety of FAD into the �-CD cavity. Since thesize of the host cavity for CB7 is quite similar to that of �-CD,in the CB7-FAD system also, a sizable interaction is expectedvia the inclusion of the adenine moiety into the CB7 cavity.This interaction, however, is not likely to cause any effect onthe absorption and fluorescence spectra of FAD, since the basicchromophore unit in FAD is its isoalloxazine ring. Consideringthe interaction of CB7 with isoalloxazine, it is expected thatthe fluorescence intensity should reduce rather than increase,as observed in the case of RF, due to the proposed lactam-lactimtautomerization induced by CB7. Thus, the observed enhance-ment in the fluorescence intensity for the CB7-FAD systemsuggests that, in addition to the above tautomerization process,

SCHEME 2: The Binding Equilibrium of CB7 with theDifferent Tautomeric Forms of RF

TABLE 2: Dipole Moments (µ) of the Different TautomericForms of RF and the Optimized Energies (∆Hf) for TheirComplexes with CB7 as Obtained by Computational Studies

tautomer lactam (A) lactim (B) lactim (B′)

µ (D) 7.6 8.7 9.6-∆Hf (kcal mol-1) 8.8 9.7 10.2

RFlactam + CB7 y\zK

CB7 • RFlactam/lactims (8)

10722 J. Phys. Chem. B, Vol. 114, No. 33, 2010 Dutta Choudhury et al.

there must be some other mechanism which leads to a significantincrease in the fluorescence intensity of FAD on its interactionwith CB7.

As mentioned earlier, an enhancement in the fluorescenceintensity of FAD was observed in our previous study on itsinteraction with �-CD, in spite of a very weak binding in thehost-guest system. In this case, the fluorescence enhancementwas explained in terms of the conformational change of FADfrom the “closed” to the “open” form on its binding (via theinclusion of the adenine moiety) with the �-CD host. A similarsituation is also very likely for the CB7-FAD system, althoughin the present case FAD possibly interacts with CB7 involvingboth the adenine and the isoalloxazine moieties. It is reported

that, in aqueous solution, a large fraction of the FAD moleculesin the ground state exists in the “closed” conformation, wherethe isoalloxazine ring and the adenine moiety are stacked oneach other and this stacking-unstacking process is dynamic innature and occurs in the time scale of about 19 ns.38,65 Theproximity of the isoalloxazine and the adenine moieties in the“closed” conformer results in the efficient fluorescence quench-ing due to an ultrafast photoinduced intramolecular electrontransfer from the adenine to the isoalloxazine moiety.35,37,38,40,65

In the presence of CB7, the steric repulsion between the CB7complexed adenine and/or isoalloxazine moieties will lead to alarge change in the conformational dynamics of FAD, causingthe stacking-unstacking equilibrium to shift largely toward the“open” conformation. This conformational change will thussignificantly reduce the photoinduced electron transfer from theadenine moiety to the isoalloxazine ring due to an increase inthe distance between the electron donor and acceptor groups,and accordingly will reduce the inherent fluorescence quenchingobserved in the free FAD molecule. Recent studies by Radosz-kowicz et al. based on fluorescence measurements and MDsimulations have shown that the fluorescence quenching due tointramolecular electron transfer is possible only when thedistance between the adenine and the isoalloxazine moieties is<5.5 Å.66 Thus, the change over from the “closed” to the “open”conformer and the concomitant reduction in the intramolecularelectron transfer process can lead to a substantial increase inthe fluorescence intensity of FAD in the presence of CB7. Itshould be remembered that, parallel to the fluorescence en-hancement caused by the conformational changes, the lactam-lactim tautomerization of the isoalloxazine ring on binding toCB7 will cause a decrease in the fluorescence intensity. Thus,there is an interplay between two opposing effects and theobserved results indicate that the fluorescence enhancement dueto the “closed” to “open” conformational changes dominatesover the quenching effect arising due to the tautomerizationprocess. It may be mentioned here that the estimation of thebinding constant for the CB7-FAD interaction by fitting the∆If

λ vs [CB7]0 plot (inset of Figure 6) using eqs 2-5 was notvery convincing because the plot does not reach its saturationlimit even at a very high concentration of CB7, unlike in thecase of the CB7-RF system. In the present case, the situationis further aggravated due to the presence of two binding sitesfor FAD (adenine and isoalloxazine moieties) which might havedissimilar binding interaction with CB7. Moreover, in thepresent case, the observed fluorescence enhancement does notcorrespond to a single effect but is the result of two opposingeffects, i.e., enhancement due to the conformational change ofFAD and quenching due to the CB7-isoalloxazine interaction.Since the two effects might not vary in the same proportionwith the gradual increase in the CB7 concentration, the ∆If

λ vs[CB7]0 plot may not follow the usual binding isotherm.

Figure 4. Geometry optimized structures for the complexes of CB7 with the different tautomeric forms of RF: (a) CB7•RFlactam; (b) CB7•RFlactim,B;(c) CB7•RFlactim,B′.

Figure 5. Representative absorption spectra of FAD (5 µM) in 10mM phosphate buffer at pH 7 with different CB7 concentrations, aftercorrection for the background absorption by CB7. [CB7]/mM: (1) 0.0,(2) 0.10, (3) 0.33, (4) 0.45, (5) 0.79, (6) 1.0.

Figure 6. Steady-state fluorescence spectra of FAD (5 µM) in 10 mMphosphate buffer (pH 7) at different CB7 concentrations. [CB7]/mM:(1) 0.0, (2) 0.20, (3) 0.45, (4) 0.76, (5) 1.0, (6) 1.5. The excitationwavelength was 420 nm. The inset shows the increase in thefluorescence intensity of FAD with respect to CB7 concentration. Asexplained in the text, no reasonable binding isotherm could beconstructed for any satisfactory quantitative analysis.

Differential Interaction of RF and FAD with CB7 J. Phys. Chem. B, Vol. 114, No. 33, 2010 10723

Therefore, the determination of the binding constant for theCB7-FAD interaction has not been attempted in the presentstudy. However, to have some insight on the complex formationprocess, time-resolved fluorescence anisotropy measurementswere carried out for FAD in the absence and presence of CB7.The fluorescence anisotropy decays were found to be single-exponential in nature in both cases. As anticipated, the anisot-ropy decay for FAD was found to be significantly slower inthe presence of CB7 (inset of Figure 7) with a rotationalcorrelation time (τr) of 890 ps in comparison with the τr valueof 200 ps as observed for the free FAD. This observation clearlysupports the complexation of FAD with the CB7 host.

To have a better idea about the binding mechanism and theinhibition of the intramolecular electron transfer process in FADon binding to CB7, the fluorescence decays of FAD in aqueoussolution were also investigated using TCSPC measurements inthe presence of different CB7 concentrations. Some representa-tive decay traces in the absence and presence of CB7 are shownin Figure 7. The fluorescence decay of FAD in aqueous solutionis found to be biexponential in nature, with a major (∼97%)long lifetime component of ∼2.4 ns and a small (∼3%) shortlifetime component of ∼0.3 ns. In the presence of CB7, thefluorescence decay kinetics of FAD becomes significantlyslower. In the CB7-FAD systems, the decays are seen to fitbetter with a biexponential function having one of the decaytime constants in the range of ∼2.4 ns and the other in the rangeof ∼4.3 ns. Table 3 lists the fluorescence decay parameters ofFAD in aqueous solution in the presence of different CB7concentrations. As indicated in Table 3, with increasing CB7concentration, the contribution of the longer lifetime component(∼4.3 ns) increases gradually with a corresponding decrease inthe contribution of the ∼2.4 ns component.

Previous studies on time-resolved fluorescence measurementsusing nanosecond time resolution have reported a significantlyshorter and nonexponential excited state decay of FAD ascompared to its analogue flavin mononucleotide (FMN).38 Thesefindings are consistent with the quenching of FAD fluorescencedue to intramolecular electron transfer in its “closed” conformer.While the decay analysis of FMN revealed a fluorescencelifetime of 4.7 ns, FAD yielded two lifetime components of2.8 and 0.3 ns.38 Recent studies with higher time resolution haveshown that FAD displays a 5-10 ps component that ischaracteristic of ultrafast fluorescence quenching and a 2.6 ns

contribution resulting from less efficient quenching.22,34,40 It isalso known that, in a slightly less polar solvent, like formamide,the π-stacked complex formation is prevented, leading pre-dominantly to the “open” conformer. In fact, a recent transientabsorption study has reported comparable excited state lifetimesof FAD and FMN.37 We have also observed similar long decaytimes of ∼5.2 ns for both FAD and RF in the relatively lesspolar and highly viscous solvent, glycerol.22 Thus, in essence,it is indicated that the short 5-10 ps decay components arisedue to the efficient excited state quenching by intramolecularelectron transfer in the stacked conformation of FAD and thelong ∼2.4 ns component (in water) probably corresponds tothe average lifetime of several plausible intermediate conforma-tions of FAD excluding the stacked one. Since the timeresolution of the time-correlated single photon counting (TC-SPC) instrument is not sufficient to resolve the ultrafast decaycomponent, apparently the shorter component of some of theconformations is displayed as the small 0.3 ns component (∼3%)in the measured decays of the free FAD. In the presence ofCB7, the contribution of this component is reduced further andcannot be detected at all in the TCSPC measurement. Theexistence of several partially folded conformations of free FADwithin its excited state lifetime have in fact been confirmed byrecent MD simulation studies.40 It is also implied from the abovediscussion that the “open” conformer of FAD should have alifetime similar to that of FMN, i.e., ∼4.7 ns. Thus, theobservation of the long ∼4.3 ns component in the fluorescencedecay of FAD in the presence of CB7, the gradual increase inthe contribution of this component, and the concomitant decreasein the 2.4 ns component with the increasing host concentrationsuggests that the conformation of FAD gradually changes fromthe “closed” to the “open” form upon binding with the CB7host. Therefore, similar to our previous results with the �-CDhost, in the presence of CB7 too, a dynamic equilibrium isestablished between the free and CB7 bound FAD moleculesin the ground state. Since the formation and dissociation of thehost-guest complexes occur only in the microsecond time scale,within the excited state lifetime of FAD (few ns), there is hardlyany change in the equilibrium between the free and bound dye.Accordingly, the free FAD molecules, following their indepen-dent excitation, decay predominantly with ∼2.4 ns decay time,as also observed for FAD in aqueous solution in the absence ofCB7. Similarly, the bound FAD molecules having an “open”conformation decay predominantly with a lifetime in the rangeof ∼4.3 ns, following their independent excitation. As theconcentration of the host molecule increases in the solution,the population of the FAD molecules bound to CB7 alsoincreases, leading to a gradual increase in the long lifetime

Figure 7. Fluorescence decay traces for FAD (5 µM) in 10 mMphosphate buffer (pH 7) monitored at 530 nm in the presence of CB7.[CB7]/mM: (1) 0.0, (2) 0.20, (3) 0.76, (4) 1.5. L is the lamp profile.The inset shows the fluorescence anisotropy decays, r(t), for FAD (5µM) in 10 mM phosphate buffer (pH 7) in the absence (a) and in thepresence (b) of CB7 (1.5 mM). The excitation wavelength was 374nm.

TABLE 3: Fluorescence Decay Parameters for FAD in 10mM Phosphate Buffer (pH 7) Monitored at 530 nm in theAbsencea and Presence of CB7 (A1 and A2 Correspond to theRelative Contributionsb of the Two Lifetimes, τ1 and τ2)

CB7 (mM) A1 (%) τ1 (ns) A2 (%) τ2 (ns) �2

0 97a 2.40 1.210.20 76 2.31 24 4.30 1.060.35 68 2.33 32 4.30 1.050.78 52 2.40 48 4.33 1.011.5 39 2.37 61 4.37 1.09

a A small contribution (3%) of a 0.3 ns component is alsoobserved in the absence of CB7. This short component is, however,not detectable in the presence of CB7. b The fluorescence decayswere fitted according to I(t) ) ∑i ai exp(-t/τi) for single andbiexponential decays, and the relative contributions were estimatedas Ai ) aiτi/∑i aiτi.

10724 J. Phys. Chem. B, Vol. 114, No. 33, 2010 Dutta Choudhury et al.

component. It may be noted, however, that the lifetime of theCB7 complexed FAD (∼4.3 ns) is somewhat lower than thevalue otherwise expected for the “open” conformer of FAD(∼4.7 ns). In the case of the �-CD-FAD system, the fluores-cence lifetime for the complexed FAD was in fact found to be∼4.8 ns, very similar to that of the open conformer of FAD.22

Thus, the slightly lower lifetime in the present case can be dueto the influence of the lactam-lactim tautomerization of theisoalloxazine ring, which should have a reducing effect on thefluorescence lifetime, as observed in the interaction of RF withthe CB7 host.

To further confirm that the fluorescence enhancement andthe lifetime changes of FAD in the presence of CB7 are due tothe conformational change of FAD from the “closed” to the“open” form, absorption as well as steady-state and time-resolved fluorescence studies have been carried out for theCB7-FAD system at pH 2.8. This acidic pH range has beenchosen because it is reported in the literature that the pKa valuefor the adenine moiety in FAD is 3.6 and below this pH, wherethe adenine moiety is protonated, FAD changes its conformationfrom the “closed” to the “open” form.35 It is believed that thesolvation of the protonated, hydrophilic adenine ring disruptsthe π-π interaction between the isoalloxazine and the adeninemoieties of FAD, thereby leading to the conformationalchange.35

The absorption spectra (after correction for backgroundabsorption by the CB7 host) and the fluorescence spectra ofFAD in the absence and presence of CB7 at pH 2.8 are presentedin Figure 8. The changes in the absorption spectra (inset ofFigure 8) on addition of CB7 were similar to those observed atneutral pH (cf. Figure 5), although the red shift was moreprominent and no significant decrease in absorbance for the 370nm band was observed. In this case, an apparent isosbestic pointwas also indicated. The fluorescence spectrum of FAD at pH2.8 was also similar to that at neutral pH, but the fluorescenceintensity was much higher due to the presence of the “open”form of FAD and a consequent reduction in the intramolecularelectron transfer quenching at the acidic pH. On addition ofCB7, interestingly, quenching of fluorescence was observed incontrast to the fluorescence enhancement observed at neutralpH. This observation clearly supports our proposition that thefluorescence enhancement of FAD in the presence of CB7 atneutral pH is due to the conformational change of FAD from

the “closed” to the “open” form as a result of its binding withCB7. At pH 2.8, since FAD is already predominantly presentin the “open” form, such an effect is not expected. However,the interaction of CB7 with the isoalloxazine moiety leads tothe fluorescence quenching due to the ensuing lactam-lactimtautomerization of this moiety in the presence of CB7. Nonethe-less, it may be noted here that the quenching effect for theCB7-FAD system at acidic pH is considerably less than thatobserved in the interaction of RF with CB7 (cf. Figure 2). Thisis possibly due to the fact that at pH 2.8, which is just 0.8 unitslower than the pKa of the adenine moiety, not all of the FADmolecules will be protonated at the adenine moiety and thus beconverted to the “open” conformation. A significant FADpopulation will still remain in the neutral form for which the“closed” conformation is quite possible in the free state butwhich undergoes a conformational change to the “open” formin the presence of CB7. This latter effect will thus cause someenhancement in the fluorescence, which will act against thequenching effect due to the tautomerization process. However,considering the fact that the CB7 host also has a pKa value ofabout 2.2 and that the FAD molecule becomes unstable at verylow pH, the pH of the solution in the present measurement wasnot lowered any further. Using the Hendersen-Hasselbalchequation, the fraction of FAD molecules present in the neutralform at pH 2.8 is estimated to be 13.6%.67 Further, consideringthat about 80% of FAD in solution remains in the “closed”conformation,38 the population of the “closed” FAD conformersin the experimental condition (at pH 2.8) is estimated to be about11%.

The fluorescence decay trace of FAD at pH 2.8 waseffectively single exponential in nature with a decay time of3.59 ns. The lower lifetime value (3.59 ns instead of ∼4.7 ns,the latter being the lifetime of FMN) indicates that FAD is notpresent entirely in the “open” form but there are also contribu-tions from other conformations, i.e., the “closed” conformer aswell as several other intermediate conformations. Though the“closed” conformers hardly contribute to the observed decaymeasured by the TCSPC instrument, the other intermediateconformers may have discrete but quite close lifetimes in thesimilar range of the “open” conformer. Such close lifetimecomponents cannot be individually resolved from the observeddecay trace, and hence, the decay effectively appears to be singleexponential in nature with an average lifetime value of 3.59ns. On addition of CB7, the decay traces gradually becomeslower, as depicted in Figure 9. The decay time constants atdifferent CB7 concentrations obtained by fitting the decay tracesto a single exponential function are listed in Table 4. Thisincrease in the fluorescence decay time of FAD in the presenceof CB7 at pH 2.8 can be understood as the combination of twoeffects, namely, the slight quenching in the lifetime due to thelactam-lactim tautomerization of the isoalloxazine moiety anda significant enhancement in the lifetime due to the “closed” to“open” conformational change of the fraction of FAD moleculesthat remain unprotonated and in the “closed” conformation atpH 2.8 in the absence of CB7. These two contrasting effectslead to an overall increase in the lifetime of the CB7-FADsystem at pH 2.8. It may be mentioned here that, due to thelarge size of the CB7-FAD system and the large changes inthe conformation of FAD on interaction with CB7, geometryoptimization studies at the PM3 level were not attempted inthis case. That the binding of CB7 with FAD does take placeat the acidic pH is, however, confirmed by time-resolvedanisotropy measurements (inset of Figure 9). The rotationalcorrelation time of FAD is found to increase from ∼200 ps in

Figure 8. Steady-state fluorescence spectra of FAD (5 µM) at pH 2.8with different CB7 concentrations. [CB7]/mM: (1) 0.0, (2) 0.05, (3)0.10, (4) 0.32, (5) 0.60, (6) 1.0. The excitation wavelength was 420nm. The inset shows the absorption spectra of FAD (5 µM) at pH 2.8with different CB7 concentrations, after correction for the backgroundabsorption by CB7. [CB7]/mM: (1) 0.0, (2) 0.05, (3) 0.32, (4) 0.60,(5) 1.0.

Differential Interaction of RF and FAD with CB7 J. Phys. Chem. B, Vol. 114, No. 33, 2010 10725

the absence of CB7 to ∼800 ps in the presence of CB7, thussupporting the complex formation between the host CB7 andthe guest FAD at pH 2.8.

4. Conclusions

The observed results indicate that CB7 forms host-guestcomplexes with both RF and FAD. The unusual fluorescencequenching of RF in the presence of CB7 is proposed to arisedue to the tautomerization of the isoalloxazine ring of RF, fromthe lactam form to the less fluorescent lactim forms upon bindingwith the host. This CB7 induced tautomerization is possiblebecause, in comparison to the lactam form, the lactim formshave larger dipole moments and hence larger dipole-dipoleinteraction with the host, which leads to more stable complexformation with CB7. Geometry optimization studies indicatethe formation of an exclusion complex, rather than an inclusioncomplex between RF and CB7. The binding of RF to CB7 isalso supported by the retardation in the fluorescence anisotropydecay rate of RF in the presence of CB7. In contrast to thefluorescence quenching observed for RF, enhancement in thefluorescence intensity is observed for FAD in the presence ofCB7. This enhancement is proposed to arise from a change inthe conformation of FAD from the “closed” to the “open” formdue to the binding of CB7 with the adenine and/or theisoalloxazine moiety of FAD. Thus, in this case, the host-guestinteraction inhibits the inherent fluorescence quenching of theFAD molecule by blocking the intramolecular electron transferfrom adenine to the isoalloxazine ring, which is prevalent inthe “closed” conformation of FAD. The lactam-lactim tau-tomerization process also participates in the interaction of CB7with the isoalloxazine ring of FAD, but the fluorescenceenhancement due to the conformational change predominates,and masks the quenching effect of the tautomerization process.The conformational change of FAD is supported by theobservation of a long lifetime component in the fluorescencedecay of the CB7-FAD system, which is characteristic of the

“open” form of FAD. Further, in acidic pH conditions, whenFAD exists mainly in the “open” form due to protonation ofthe adenine moiety, the conformational change no longerremains the dominant process in the presence of CB7 and thetautomerization effect can be detected, leading to observablefluorescence quenching as in the case of RF. Thus, the resultsindicate that CB7 can provide a biomimicking environment,thereby changing the tautomerization equilibrium of the isoal-loxazine moiety and modulating the conformation of FAD fromthe “closed” to the more biologically predominant “open” form.

Acknowledgment. We acknowledge Dr. G. B. Maralihalliand Ms. R. Agarwal, MBD, BARC, for helping in the purifica-tion of FAD and Dr. H. P. Upadhyaya, RPCD, BARC, forhelping with the computational studies. We thank Dr. T.Mukherjee, Director, Chemistry Group, BARC, and Dr. S. K.Sarkar, Head, RPCD, BARC, for their encouragement andsupport in the course of this study.

Supporting Information Available: Note S1 showingequations for the calculation of binding constants. This materialis available free of charge via the Internet at http://pubs.acs.org.

References and Notes

(1) Uekama, K.; Hirayama, F.; Irie, T. Chem. ReV. 1998, 98, 2045.(2) Nepogodiev, S. A.; Stoddart, J. F. Chem. ReV. 1998, 98, 1959.(3) Ghosh, S.; Isaacs, L. J. Am. Chem. Soc. 2010, 132, 4445.(4) Basilio, N.; Garcıa-Rıo, L.; Moreira, J. A.; Pessego, M. J. Org.

Chem. 2010, 75, 848.(5) Saleh, N.; Koner, A. L.; Nau, W. M. Angew. Chem., Int. Ed. 2008,

47, 5398.(6) Mohanty, J.; Bhasikuttan, A. C.; Dutta Choudhury, S.; Pal, H. J.

Phys. Chem. B 2008, 112, 10782.(7) Hennig, A.; Bakirci, H.; Nau, W. M. Nat. Methods 2007, 4, 629.(8) Bakirci, H.; Koner, A. L.; Dickman, M. H.; Kortz, U.; Nau, W. M.

Angew. Chem., Int. Ed. 2006, 45, 7400.(9) Szejtli, J. Chem. ReV. 1998, 98, 1743.

(10) Lagona, J.; Mukhopadhyay, P.; Chakrabarti, S.; Isaacs, L. Angew.Chem., Int. Ed. 2005, 44, 4844.

(11) Ikeda, A.; Shinkai, S. Chem. ReV. 1997, 97, 1713.(12) Schneider, H.-J.; Hacket, F.; Rudiger, V. Chem. ReV. 1998, 98,

1755.(13) Li, S.; Purdy, W. C. Chem. ReV. 1992, 92, 1457.(14) Lee, J. W.; Samal, S.; Salvapalam, N.; Kim, H.-J.; Kim, K. Acc.

Chem. Res. 2003, 36, 621.(15) Liu, Y.; Han, B.-H.; Chen, Y.-T. J. Phys. Chem. B 2002, 106, 4678.(16) Mohanty, J.; Bhasikuttan, A. C.; Nau, W. M.; Pal, H. J. Phys. Chem.

B 2006, 110, 5132.(17) Shaikh, M.; Mohanty, J.; Bhasikuttan, A. C.; Uzunova, V. D.; Nau,

W. M.; Pal, H. Chem. Commun. 2008, 3681.(18) Wenz, G. Angew. Chem., Int. Ed. 1994, 33, 803.(19) Zadmard, R.; Schrader, T. J. Am. Chem. Soc. 2005, 127, 904.(20) Mohanty, J.; Dutta Choudhury, S.; Upadhyaya, H. P.; Bhasikuttan,

A. C.; Pal, H. Chem.sEur. J. 2009, 15, 5215.(21) Dutta Choudhury, S.; Mohanty, J.; Pal, H.; Bhasikuttan, A. C. J. Am.

Chem. Soc. 2010, 132, 1395.(22) Kandoth, N.; Dutta Choudhury, S.; Mohanty, J.; Bhasikuttan, A. C.;

Pal, H. J. Phys. Chem. B 2010, 114, 2617.(23) Penzer, G. R.; Radda, G. K. Q. ReV. Chem. Soc. 1967, 1, 43.(24) Dong, C.; Flecks, S.; Unversucht, S.; Haupt, C.; van Pee, K.-H.;

Naismith, J. H. Science 2005, 309, 2216.(25) Fitzpatrick, P. F. Acc. Chem. Res. 2001, 34, 299.(26) Chosrowjan, H.; Taniguchi, S.; Mataga, N.; Phongsak, T.; Sucha-

ritakul, J.; Chaiyen, P.; Tanaka, F. J. Phys. Chem. B 2009, 113, 8439.(27) van den Berg, P. A. W.; Mulrooney, S. B.; Gobets, B.; van Stokkum,

I. H. M.; van Hoek, A.; Williams, C. H. Jr.; Visser, A. J. W. G. ProteinSci. 2001, 10, 2037.

(28) Mataga, N.; Chosrowjan, H.; Taniguchi, S. J. Phys. Chem. B 2002,106, 8917.

(29) Zhong, D.; Zewail, A. H. Proc. Natl. Acad. Sci. U.S.A. 2001, 98,11867.

(30) Heelis, P. F. Chem. Soc. ReV. 1982, 11, 15.(31) Sun, M.; Moore, T. A.; Song, P.-S. J. Am. Chem. Soc. 1972, 94,

1730.(32) Yagi, K.; Ohishi, N.; Nishimoto, K.; Choi, J. D.; Soong, P.-S.

Biochemistry 1980, 19, 1553.

Figure 9. Fluorescence decay traces for FAD (5 µM) at pH 2.8monitored at 530 nm in the presence of CB7. [CB7]/mM: (1) 0.0, (2)0.32, (3) 1.0. L is the lamp profile. The inset shows the fluorescenceanisotropy decays, r(t), for FAD (5 µM) at pH 2.8 in the absence (a)and in the presence (b) of CB7 (1.0 mM). The smooth lines show thebest fit curves. The excitation wavelength was 374 nm.

TABLE 4: Fluorescence Decay Parameters for FAD at pH2.8 Monitored at 530 nm in the Absence and Presence ofCB7

CB7 (mM) 0 0.05 0.20 0.50 1.0τ1 (ns) 3.59 3.67 3.75 3.86 3.96�2 1.17 1.19 1.19 1.18 1.20

10726 J. Phys. Chem. B, Vol. 114, No. 33, 2010 Dutta Choudhury et al.

(33) Weber, G. Biochem. J. 1949, 47, 115.(34) Chosrowjan, H.; Taniguchi, S.; Mataga, N.; Tanaka, F.; Visser,

A. J. W. G. Chem. Phys. Lett. 2003, 378, 354.(35) Li, G.; Glusac, K. D. J. Phys. Chem. A 2008, 112, 4573.(36) Li, G.; Glusac, K. D. J. Phys. Chem. B 2009, 113, 9059.(37) Stanley, R. J.; MacFarlane, A. W., IV. J. Phys. Chem. A 2000,

104, 6899.(38) Visser, A. J. W. G. Photochem. Photobiol. 1984, 40, 703.(39) Wahl, P.; Auchet, J. C.; Visser, A. J. W. G.; Muller, F. FEBS Lett.

1974, 44, 67.(40) van den Berg, P. A. W.; Feenstra, K. A.; Mark, A. E.; Berendsen,

H. J. C.; Visser, A. J. W. G. J. Phys. Chem. B 2002, 106, 8858.(41) Miles, D. W.; Urry, D. W. Biochemistry 1968, 7, 2791.(42) Kainosho, M.; Kyogoku, Y. Biochemistry 1972, 11, 741.(43) Copeland, R. A.; Spiro, T. G. J. Phys. Chem. 1986, 90, 6648.(44) Dym, O.; Eisenberg, D. Protein Sci. 2001, 10, 1712.(45) Pan, J.; Byrdin, M.; Aubert, C.; Eker, A. P. M.; Brettel, K.; Vos,

M. H. J. Phys. Chem. B 2004, 108, 10160.(46) Antony, J.; Medvedev, D. M.; Stuchebrukhov, A. A. J. Am. Chem.

Soc. 2000, 122, 1057.(47) Kao, Y.-T.; Saxena, C.; Wang, L.; Sancar, A.; Zhong, D. Proc.

Natl. Acad. Sci. U.S.A. 2005, 102, 16128.(48) Acocella, A.; Jones, G. A.; Zerbetto, F. J. Phys. Chem. B 2010,

114, 4101.(49) Whitby, L. G. Biochem. J. 1953, 54, 437.(50) Lakowicz, J. R. Principles of Fluorescence Spectroscopy, 3rd ed.;

Springer: New York, 2006.(51) O’Connor, D. V.; Phillips, D. Time Correlated Single Photon

Counting; Academic Press: New York, 1984.

(52) Frisch, M. J.; et al. Gaussian 92, revision E.1; Gaussian Inc.:Pittsburgh, PA, 1992.

(53) Dutta Choudhury, S.; Mohanty, J.; Upadhyaya, H. P.; Bhasikuttan,A. C.; Pal, H. J. Phys. Chem. B 2009, 113, 1891.

(54) Kandoth, N.; Dutta Choudhury, S.; Mukherjee, T.; Pal, H. Photo-chem. Photobiol. Sci. 2009, 8, 82.

(55) Nau, W. M.; Zhang, X. J. Am. Chem. Soc. 1999, 121, 8022.(56) Singh, M. K.; Pal, H.; Koti, A. S. R.; Sapre, A. V. J. Phys. Chem.

A 2004, 108, 1465.(57) Schuette, J. M.; Ndou, T. T.; Warner, I. M. J. Phys. Chem. 1992,

96, 5309.(58) Liu, Y.; You, C.-C. J. Phys. Org. Chem. 2001, 14, 11.(59) Zhou, Y.; Sun, J.; Yu, H.; Wu, L.; Wang, L. Supramol. Chem.

2009, 21, 495.(60) Karunanithi, P.; Ramamurthy, P.; Ramakrishnan, V. T. J. Inclusion

Phenom. Macrocyclic Chem. 1999, 34, 105.(61) Sarkar, B.; Das, U.; Bhattacharyya, S.; Bose, S. K. Bull. Chem.

Soc. Jpn. 1995, 68, 1807.(62) Miskolczy, Z.; Biczok, L.; Gorner, H. J. Photochem. Photobiol., A

2009, 207, 47.(63) Berezovskii, V. M.; Tul’chinskaya, L. S.; Polyakova, N. A. Russ.

Chem. ReV. 1972, 41, 574.(64) Nishimoto, K. Bull. Chem. Soc. Jpn. 1967, 40, 2493.(65) Barrio, J. R.; Tolman, G. L.; Nelson, J. L.; Spencer, R. D.; Weber,

G. Proc. Natl. Acad. Sci. U.S.A. 1973, 70, 941.(66) Radoszkowicz, L.; Huppert, D.; Nachliel, E.; Gutman, M. J. Phys.

Chem. A 2010, 114, 1017.(67) Atkins, P. W. Physical Chemistry, 6th ed.; Oxford University Press:

Oxford, U.K., 1998.

JP1041662

Differential Interaction of RF and FAD with CB7 J. Phys. Chem. B, Vol. 114, No. 33, 2010 10727