Chemistry and Physics of 183 (2014) 1–8

Contents lists available at ScienceDirect

Chemistry and Physics of Lipids

jou rnal homepage: www.elsevier.com/locate/chemphyslip

Location, dynamics and relaxation of a nile red-based

phase-sensitive fluorescent membrane probe

a,1 a,1 a b

Roopali Saxena , Sandeep Shrivastava , Sourav Haldar , Andrey S. Klymchenko , a,∗

Amitabha Chattopadhyay

a

CSIR-Centre for Cellular and Molecular Biology, Uppal Road, Hyderabad 500 007, India

b

Laboratoire de Biophotonique et Pharmacologie, UMR 7213 CNRS, Université de Strasbourg, Faculté de Pharmacie, 74, Route du Rhin,

67401 Illkirch Cedex, France

a

r t i c l e i n f o a b s t r a c t

Article history: Fluorescent membrane probes offer the advantage of high sensitivity, suitable time resolution, and mul-

Received 18 March 2014

tiplicity of measurable parameters, and provide useful information on model and cell membranes. In

Received in revised form 22 April 2014

this paper, we have explored the location, dynamics, and solvent relaxation characteristics of a novel

Accepted 27 April 2014

Nile Red-based phase-sensitive probe (NR12S). Unlike Nile Red, NR12S enjoys unique orientation and

Available online 4 May 2014

location in the membrane, and is localized exclusively in the outer leaflet of the membrane bilayer. By

analysis of membrane depth using the parallax approach, we show that the fluorescent group in NR12S is

Keywords:

localized at the membrane interface, a region characterized by slow solvent relaxation. Our results show

Fluorescent membrane probe

REES that NR12S exhibits REES (red edge excitation shift), consistent with its interfacial localization. More

interestingly, REES of NR12S displays sensitivity to the membrane phase. In addition, fluorescence emis-

Nile Red

NR12S sion maximum, anisotropy, and lifetime of NR12S are dependent on the membrane phase. We envision

Membrane penetration depth that NR12S may prove to be a useful probe in future studies of complex natural membranes.

Liquid-ordered phase © 2014 Elsevier Ireland Ltd. All rights reserved.

1. Introduction ability to monitor molecules by a variety of physicochemical

approaches at increasing spatiotemporal resolution (Eggeling et al.,

Biological membranes are complex two-dimensional, non- 2009). In particular, application of spectroscopic and microscopic

covalent assemblies of a diverse variety of lipids and proteins. They techniques using fluorescent lipid analogs represents a convenient

impart an identity to the cell and its organelles, and represent an approach for monitoring membrane lipid organization and dynam-

ideal milieu for the proper function of a diverse set of membrane ics. Fluorescence-based approaches are preferred due to their high

proteins. The eukaryotic cell is composed of diverse lipids (van Meer sensitivity, suitable time resolution, and multiplicity of measur-

and de Kroon, 2011) and tracking lipids in a crowded cellular milieu able parameters. Lipids covalently linked to extrinsic fluorophores

poses considerable challenge. In this scenario, use of lipid probes with suitable fluorescence properties are generally used for such

assumes significance (Chattopadhyay, 2002). Various types of lipid studies. The advantage of using fluorescently labeled lipids is the

probes have proved to be useful in membrane biology due to their choice available for the fluorescent label. Probes with appropriate

characteristics can therefore be designed for specific applications.

A major criterion of a fluorescent membrane probe is its sensitiv-

ity to environmental factors. Nile Red, an uncharged phenoxazone

Abbreviations: 2-AS, 2-(9-anthroyloxy)stearic acid; 12-AS, 12-(9-anthroyloxy)- dye (see Fig. 1), is such a probe whose fluorescence properties

stearic acid; 5-PC, 1-palmitoyl-2-(5-doxyl)stearoyl-sn-glycero-3-phosphocholine;

are altered by the polarity of its immediate environment due to

12-PC, 1-palmitoyl-2-(12-doxyl)stearoyl-sn-glycero-3-phosphocholine; DMPC,

a large change in its dipole moment upon excitation (Greenspan

1,2-dimyristoyl-sn-glycero-3-phosphocholine; DOPC, 1,2-dioleoyl-sn-glycero-3-

and Fowler, 1985; Golini et al., 1998). This large change in dipole

phosphocholine; DPPC, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine; ET(30),

[2,6-diphenyl-4-(2,4,6-triphenyl-N-pyridino)phenoxide]; LUV, large unilamellar moment has been attributed to charge separation between the

␣

vesicle; Nile Red, 9-diethylamino-5H-benzo[ ]-5-one; POPC, 1- diethylamino group which acts as the electron donor and the

palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine; REES, red edge excitation shift;

quinoid part of the molecule which serves as the electron accep-

Tempo-PC, 1,2-dioleoyl-sn-glycero-3-phosphotempocholine.

∗ tor. Nile Red has been used as a fluorescent probe for monitoring

Corresponding author. Tel.: +91 40 2719 2578; fax: +91 40 2716 0311.

hydrophobic surfaces in proteins (Sackett et al., 1990), and as a

E-mail address: [email protected] (A. Chattopadhyay).

1

These authors contributed equally to this work. lipid stain in membranes (Gao et al., 2006). It has also been used for

http://dx.doi.org/10.1016/j.chemphyslip.2014.04.007

0009-3084/© 2014 Elsevier Ireland Ltd. All rights reserved.

2 R. Saxena et al. / Chemistry and Physics of Lipids 183 (2014) 1–8

O and 12-PC were obtained from Avanti Polar Lipids (Alabaster, AL).

S 2-AS and 12-AS were obtained from Molecular Probes (Eugene, OR).

O–

Lipids were checked for purity by thin layer chromatography on

O N+ O

silica gel precoated plates (Sigma) in chloroform//

(65:35:5, v/v/v) and were found to give only one spot in all

O

cases with a phosphate-sensitive spray and on subsequent char-

ring (Dittmer and Lester, 1964). Concentrations of phospholipids

N

were determined by phosphate assay subsequent to total oxidation

O by perchloric acid (McClare, 1971). DMPC was used as an inter-

nal standard to assess lipid digestion. NR12S was synthesized as

described previously (Kucherak et al., 2010). The concentration of

N a stock solution of NR12S prepared in DMSO was estimated using

−1 −1

its molar extinction coefficient (ε) of 45,000 M cm at 550 nm

in . All other chemicals used were of the highest purity

available. used were of spectroscopic grade. Purity of sol-

vents was further confirmed by the ET(30) procedure (Mukherjee

Fig. 1. Chemical structure of NR12S. The part of NR12S responsible for its membrane et al., 1994). Water was purified through a Millipore (Bedford, MA)

anchoring property is shown in red and the fluorophore (Nile Red) is shown in blue. Milli-Q system and used throughout.

(For interpretation of the references to color in this figure legend, the reader is

referred to the web version of this article.)

2.2. ET(30) procedure

monitoring organization and heterogeneity induced by cholesterol The ET(30) dye is a solvatochromic dye which undergoes one of

in model (Krishnamoorthy and Ira, 2001; Mukherjee et al., 2007b) the largest known solvent-induced shifts in absorption maximum.

and natural (Mukherjee et al., 2007a) membranes. The extremely large solvent-induced shift has been used to intro-

The chemical structure and design of a membrane probe are duce an empirical parameter of solvent polarity, called the ET(30)

crucial in terms of its usefulness as a reporter molecule. It is desir- value. The ET(30) value for a solvent is defined as the transition

able that the probe should be able to intercalate with membrane energy of the dissolved ET(30) dye measured in kcal/mol according

lipids with the fluorophore part suitably embedded in the mem- to the following equation:

brane. Hydrophobic molecules (such as Nile Red) partition into −3

ET = hcN¯ A = 2.859 × 10 ¯ (1)

the membrane, but do not get oriented in a specific conforma-

tion due to lack of anchoring. As a result, these probes essentially where h is Planck’s constant, c is the velocity of light, ¯ is the

−1

provide a weighted average information, depending on the num- wavenumber of the photon in cm which produces the electronic

ber of locations they occupy, and their fluorescence properties in transition, and NA is Avogadro’s number. Due to extremely large

these locations. The information obtained from such probes there- solvatochromism of this dye, the ET(30) values serve as sensitive

fore lacks specificity. This can be avoided by covalently linking the indicators of solvent polarity and are sensitive to any impurity

probe with a fatty acyl chain (which helps in alignment of the probe present in trace amounts. ET(30) values have been previously deter-

in the membrane) and an anchoring group which is often charged mined for a large number of solvents. A few grains of the ET(30) dye

(Shynkar et al., 2007). This strategy was recently utilized to gener- were dissolved in a given solvent, and its absorption maximum was

ate NR12S (see Fig. 1 for chemical structure), a fluorescent probe monitored. From this absorption maximum, the ET(30) value was

based on Nile Red, which has unique orientation and location in calculated using Eq. (1). The ET(30) values so obtained were com-

the membrane (Kucherak et al., 2010). Besides being environment- pared with the literature values (Reichardt, 1988). The ET(30) values

sensitive, NR12S has the advantage of being localized exclusively obtained showed a maximum deviation of <0.5% from the reported

in the outer leaflet of the membrane (Chiantia et al., 2012; Darwich values.

et al., 2012). Nonetheless, the exact location of the fluorescent

group in NR12S in membranes is not known. In this paper, we 2.3. Sample preparation

determined the exact depth of the NR12S fluorophore using the

parallax approach (Chattopadhyay and London, 1987). Our results All experiments were performed using large unilamellar vesi-

show that the fluorophore in NR12S is localized at the membrane cles (LUVs) of 100 nm diameter of POPC, DPPC, or POPC/40 mol%

interface, a region characterized by unique motional and dielectric cholesterol. While POPC and POPC with 40 mol% cholesterol sam-

characteristics different from both the bulk aqueous phase, and the ples contained 1 mol% NR12S, DPPC samples contained 0.5 mol%

more isotropic hydrocarbon-like deeper regions of the membrane NR12S. In general, 640 nmol of total lipid and 6.4 (or 3.2 in case of

(Haldar et al., 2011). This specific region of the membrane exhibits DPPC) nmol of NR12S were mixed well and dried under a stream of

◦

slow rates of solvent relaxation (Chattopadhyay and Mukherjee, nitrogen while being warmed gently (∼40 C). After further dry-

1993; Jurkiewicz et al., 2006; Das et al., 2008), a feature that has ing under a high vacuum for at least 3 h, the lipid mixture was

been associated with red edge excitation shift (REES). Consistent hydrated (swelled) by addition of 1.5 ml of buffer A (10 mM sodium

with this, we report here that NR12S exhibits REES and the mag- phosphate, 150 mM sodium chloride, pH 7.4), and each sample was

nitude of REES displays sensitivity to membrane phase. Our results vortexed for 3 min to uniformly disperse the lipids and form homo-

are useful for future studies of NR12S in natural membranes with geneous multilamellar vesicles. The buffer was always maintained

complex behavior. at a temperature well above the phase transition temperature of

the phospholipid used as the vesicles were made. Lipids were

◦ ◦

2. Materials and methods therefore swelled at a temperature of 40 C for POPC and 60 C

for DPPC samples. LUVs of 100 nm diameter were prepared by the

2.1. Materials extrusion technique using an Avestin Liposofast Extruder (Ottawa,

Ontario, Canada) as previously described (MacDonald et al., 1991;

DMPC, cholesterol and ET(30) dye were obtained from Sigma Mukherjee and Chattopadhyay, 2005). Briefly, the multilamellar

Chemical Co. (St. Louis, MO). DOPC, DPPC, POPC, Tempo-PC, 5-PC vesicles were freeze–thawed five times using liquid nitrogen to

R. Saxena et al. / Chemistry and Physics of Lipids 183 (2014) 1–8 3

ensure solute equilibration between trapped and bulk solutions with bandpass of 5 nm were used. Fluorescence was measured at

◦

and then extruded through polycarbonate filters (pore diameter room temperature (∼23 C) and averaged over two 5 s readings.

of 100 nm) mounted in an extruder fitted with Hamilton syringes Intensities were found to be stable over time. In all cases, the

(Hamilton Company, Reno, NV). The samples were subjected to intensity from background samples without NR12S was subtracted.

11 passes through the polycarbonate filters to give the final LUV Membrane penetration depths were calculated using Eq. (5) (see

suspension. Background samples were prepared in the same way Section 3).

except that NR12S was not added to them. The optical density of

the samples measured at 530 nm was less than 0.15 in all cases,

2.6. Time-resolved fluorescence measurements

which rules out any possibility of inner filter effect or scattering

artifacts. Samples were incubated in dark for 12 h at room tempera-

Fluorescence lifetimes were calculated from time-resolved flu-

◦

ture (∼23 C) for equilibration prior to fluorescence measurements.

orescence intensity decays using IBH 5000F NanoLED equipment

All experiments were performed with at least three sets of sam-

(Horiba Jobin Yvon, Edison, NJ) with DataStation software in the

◦

ples at room temperature (∼23 C). For experiments using solvents,

time-correlated single photon counting (TCSPC) mode. A pulsed

1 nmol of NR12S in DMSO was mixed with 2 ml of a given solvent

light-emitting diode (LED) (NanoLED-01) was used as an excita-

used in the study.

tion source. This LED generates optical pulses at 490 nm with pulse

duration of 1.2 ns and is run at 1 MHz repetition rate. The LED pro-

2.4. Depth measurements using the parallax method file (instrument response function) was measured at the excitation

wavelength using Ludox (colloidal silica) as the scatterer. To opti-

The actual spin (nitroxide) contents of the spin-labeled phos- mize the signal-to-noise ratio, 10,000 photon counts were collected

pholipids (Tempo-, 5- and 12-PC) were assayed using fluorescence in the peak channel. All experiments were performed using emis-

quenching of anthroyloxy-labeled fatty acids (2- and 12-AS) as sion slits with bandpass of 4 or 8 nm. The sample and the scatterer

described previously (Abrams and London, 1993). For depth mea- were alternated after every 5% acquisition to ensure compensation

surements, liposomes were made by the ethanol injection method for shape and timing drifts occurring during the period of data col-

(Kremer et al., 1977). These samples were made by codrying lection. This arrangement also prevents any prolonged exposure of

160 nmol of DOPC containing 10 mol% spin-labeled phospholipid the sample to the excitation beam, thereby avoiding any possible

(Tempo-, 5- or 12-PC) and 1 mol% NR12S under a steady stream of photodamage of the fluorophore. Data were stored and analyzed

◦

∼

nitrogen with gentle warming ( 35 C), followed by further dry- using DAS 6.2 software (Horiba Jobin Yvon, Edison, NJ). Fluores-

ing under a high vacuum for at least 3 h. The dried lipid film was cence intensity decay curves so obtained were deconvoluted with

dissolved in ethanol to give a final concentration of 40 mM. The the instrument response function and analyzed as a sum of expo-

ethanolic lipid solution was then injected into 1.5 ml of buffer A, nential terms

 

while vortexing to give a final concentration of 0.11 mM DOPC in  −t

F t = ˛

buffer. The lipid composition of these samples was 90% DOPC and ( ) i exp (3)

i i

10% spin-labeled PC (Tempo-, 5- or 12-PC). Duplicate samples were

prepared in each case except for samples lacking the quencher where F(t) is the fluorescence intensity at time t and ˛i is a pre-

(Tempo-, 5- or 12-PC), for which triplicates were prepared. Back- exponential factor representing the fractional contribution to the

␶

ground samples lacking NR12S were prepared in all cases, and their time-resolved decay of the component with a lifetime i such that

fluorescence intensity was subtracted from the respective sample i˛i = 1. The decay parameters were recovered using a nonlinear

fluorescence intensity. Samples were kept in dark for 12 h before least squares iterative fitting procedure based on the Marquardt

fluorescence measurements. algorithm (Bevington, 1969). The program also includes statistical

and plotting subroutine packages (O’Connor and Phillips, 1984).

2.5. Steady state fluorescence measurements The goodness of fit of a given set of observed data and the cho-

2

sen function was evaluated by the  ratio, the weighted residuals

Steady state fluorescence measurements were performed with (Lampert et al., 1983), and the autocorrelation function of the

a Hitachi F-4010 spectrofluorometer (Tokyo, Japan) using 1 cm path weighted residuals (Grinvald and Steinberg, 1974). A fit was con-

length quartz cuvettes. Excitation and emission slits with bandpass sidered acceptable when plots of the weighted residuals and the

of 5 nm were used for all measurements. Background intensities of autocorrelation function showed random deviation about zero with

2



samples in which NR12S was omitted were subtracted from each a minimum value not more than 1.5. Intensity-averaged mean



sample spectrum to cancel out any contribution due to the sol- lifetimes for biexponential decays of fluorescence were calcu-

vent Raman peak and other scattering artifacts. The spectral shifts lated from the decay times and pre-exponential factors using the

obtained with different sets of samples were identical in most cases. following equation (Lakowicz, 2006):

In other cases, the values were within ±1 nm of those reported. 2 2

˛1 + ˛2

Fluorescence anisotropy measurements were performed at  = 1 2 (4)

◦ ˛

 + ˛ 

room temperature (∼23 C) using a Hitachi Glan-Thompson polar- 1 1 2 2

ization accessory. Anisotropy values were calculated from the

equation (Lakowicz, 2006): 3. Results

IVV − GIVH

r = (2) 3.1. Membrane penetration depth of NR12S

IVV + 2GIVH

where IVV and IVH are the fluorescence intensities (after appropri- Knowledge of the exact location of the fluorescent group in a

ate background subtraction) with the excitation polarizer oriented membrane probe is crucial since it allows to correlate the fluo-

vertically and the emission polarizer vertically and horizontally ori- rescence parameters of the probe with a defined location in the

ented, respectively. G is the ratio of the efficiencies of the detection membrane. This is important since some probes (such as probes

system for vertically and horizontally polarized light and is equal labeled with the 7-nitrobenz-2-oxa-1,3-diazol-4-yl (NBD) group)

to IHV/IHH. tend to loop up in the membrane (Chattopadhyay, 1990; Haldar

For depth measurements, samples were excited at 530 nm, and and Chattopadhyay, 2013). Knowledge of the precise depth of a

emission was collected at 600 nm. Excitation and emission slits membrane-embedded group or molecule often helps define the

4 R. Saxena et al. / Chemistry and Physics of Lipids 183 (2014) 1–8

30 S)

T 1.0 of

UNI

20 RY 0.8 (Å) the center

ayer

FLUORESCENCE 0.6

bil

from

10 (ARBITRA

the 0.4 Distance 0 ------

0.2 NORMALIZED

Center of the bilayer INTENSITY

560 580 600 620 640

Fig. 2. A schematic representation of one-half of the membrane bilayer showing the

localization of NR12S in phosphatidylcholine membranes. The horizontal line at the

EMISSION WAVELENGTH (nm)

bottom indicates the center of the bilayer.

Fig. 3. Effect of membrane phase on fluorescence emission spectra of NR12S. Rep-

resentative fluorescence emission spectra of NR12S in membranes of POPC ( ),

conformation and topology of membrane probes and proteins

DPPC (—) and POPC/cholesterol (—) are shown. Spectra are intensity-normalized at

(Chattopadhyay, 1992; London and Ladokhin, 2002). Interestingly,

the respective emission maximum. Measurements were carried out at room tem-

◦

properties such as polarity, fluidity, segmental motion, ability

perature (∼23 C). The excitation wavelength used was 530 nm for POPC and DPPC

to form hydrogen bonds, extent of solvent penetration, and the membranes, and 525 nm for membranes containing POPC/40 mol% cholesterol. The

ratio of NR12S to total lipid was 1:100 (mol/mol) in case of membranes of POPC

resultant environmental heterogeneity are known to vary in a

and POPC/cholesterol, and 1:200 (mol/mol) for DPPC membranes. The total lipid

depth-dependent manner in the membrane (Chattopadhyay and

concentration was 0.43 mM in all cases. See Section 2 for other details.

Mukherjee, 1999a; Chattopadhyay, 2003; Haldar et al., 2012). The

membrane penetration depth of the NR12S fluorophore was calcu-

lated by the parallax method (Chattopadhyay and London, 1987) conformation in the gel phase, they are fluid and disordered in

using the equation: the liquid-disordered phase. The liquid-ordered phase represents

 

an interesting phase, and is characterized by acyl chains that are

2

(−1/C) ln(F1/F2) − L

21 extended and ordered (such as in the gel phase), but display high zcF = Lc + (5)

1 2L

21 lateral mobility similar to the liquid-disordered phase (Mouritsen,

2010). The liquid-ordered phase exists above a threshold level of

where zcF is the distance of the fluorophore from the center of

cholesterol for binary lipid mixtures (Mouritsen, 2010). The fluores-

the bilayer, Lc1 is the distance of the center of the bilayer from

cence emission spectra of NR12S in these membranes are shown

the shallow quencher (Tempo-PC in this case), L21 is the dif- 2

in Fig. 3. The emission maximum of NR12S in POPC (fluid) and

ference in depth between the two quenchers (i.e., the vertical

DPPC (gel) vesicles was found to be 594 and 592 nm, respectively,

distance between the shallow and deep quenchers), and C is the

when excited at 530 nm. In POPC vesicles with 40 mol% choles-

two dimensional quencher concentration in the plane of the mem-

2 terol (the liquid-ordered phase), the emission maximum displayed

brane (molecules/A˚ ). Here, F1/F2 is the ratio of F1/Fo and F2/Fo, in

a blue shift toward lower wavelength and was found to be 583 nm

which F1 and F2 are the fluorescence intensities in the presence

when excited at 525 nm. Such blue shift in emission maximum

of the shallow quencher (Tempo-PC) and the deep quencher (5-

of NR12S in membranes of liquid-ordered phase has previously

PC), respectively, both at the same quencher concentration C; Fo

been reported, and assigned to decreased membrane hydration

is the fluorescence intensity in the absence of any quencher. All

and defined vertical orientation of the fluorophore in this phase

the bilayer parameters used were the same as described previously

(Kucherak et al., 2010).

(Chattopadhyay and London, 1987). Our results show that the aver-

age depth of penetration of the fluorescent group of NR12S is ∼18 A˚

3.3. REES in membranes of varying phase

from the center of the bilayer (see Fig. 2). This suggests that the

fluorescent group in NR12S is localized at the interfacial region of

REES is defined as the shift in the wavelength of maximum flu-

the membrane. This region of the membrane typically exhibits rel-

orescence emission toward higher wavelengths, caused by a shift

atively slow solvent relaxation and therefore is suitable for REES

in the excitation wavelength toward the red edge of the absorp-

measurements (see below) (Haldar et al., 2011; Chattopadhyay,

tion band. This effect assumes relevance for polar fluorophores

2003; Chattopadhyay and Haldar, 2014). It should be noted that this

in motionally restricted environment where the dipolar relax-

is a time-averaged value of depth and does not take into account

ation time for the solvent shell around a fluorophore becomes

any distribution in probe depth.

comparable to or longer than its fluorescence lifetime (Haldar

et al., 2011; Chattopadhyay, 2003; Chattopadhyay and Haldar,

3.2. Fluorescence characteristics of NR12S in membranes of

2014; Mukherjee and Chattopadhyay, 1995; Raghuraman et al.,

varying phase

2005; Demchenko, 2008). An attractive aspect of REES is that

it allows to monitor the mobility parameters of the environ-

The fluorescence properties of NR12S have been reported to be

ment itself (represented by the relaxing solvent molecules) using

sensitive to membrane phase (Kucherak et al., 2010). Membrane

phase represents a crucial determinant of membrane physical

properties (Van Meer et al., 2008). We therefore monitored the

2

fluorescence properties of NR12S in LUVs made of POPC, DPPC We have used the term maximum of fluorescence emission in a somewhat wider

sense here. In every case, we have monitored the wavelength corresponding to

and POPC with 40 mol% cholesterol, as these vesicles repre-

maximum fluorescence intensity, as well as the center of mass of the fluorescence

sent liquid-disordered (fluid), gel (ordered), and liquid-ordered

emission, in the symmetric part of the spectrum. In most cases, both these methods

phase membranes, respectively (Brown and London, 1998). While

yielded the same wavelength. In cases where minor discrepancies were found, the

the lipid acyl chains are ordered and extended in all trans center of mass of emission has been reported as the fluorescence maximum.

R. Saxena et al. / Chemistry and Physics of Lipids 183 (2014) 1–8 5

(A) 605 0.22 PY ) RO

nm 600

OT 0.20

595 ANIS

XIMUM ( 0.18

MA 590

ON ON 0.16 585 MISSI

E 0.14 FLUORESCENCE 580 POPC POPC + Chol DPPC

525 540 555 570 585

Fig. 5. Fluorescence anisotropy of NR12S in membranes of varying phase. The exci-

EXCITATION WAVELENGTH (nm)

tation wavelength used was 530 nm and emission was monitored at 596 nm in all

cases. Data shown are means ± S.E. of at least three independent measurements. All

other conditions are as in Fig. 3. See Section 2 for other details.

(B) 20

Chattopadhyay and Haldar, 2014). Importantly, although REES has

been earlier reported in liquid-disordered and gel phase mem-

branes (Chattopadhyay and Mukherjee, 1999b; Raghuraman et al.,

16

2007), our present results of REES of NR12S in POPC/cholesterol

membranes constitute one of the early reports of REES for probes

in liquid-ordered phase membranes. 12

ES (nm)

3.4. Fluorescence anisotropy and lifetime in membranes of RE

varying phase

8

Fluorescence anisotropy is extensively used to monitor the

rotational diffusion rate of membrane embedded probes, which

4

is sensitive to the packing of lipid fatty acyl chains (Lakowicz,

POPC POPC + Chol

2006; Jameson and Ross, 2010). This is due to the fact that fluo-

rescence anisotropy depends on the extent to which the probe is

Fig. 4. (A) Effect of changing excitation wavelength on the wavelength of maximum

able to reorient after excitation, and probe reorientation is depend-

emission for NR12S in membranes of POPC (᭹), and POPC/40 mol% cholesterol ().

ent on local lipid packing. Fig. 5 shows the steady state anisotropy

The lines joining the data points are provided merely as viewing guides. (B) Com-

parison of the magnitude of REES of NR12S in these membranes. The magnitude of NR12S in membranes of varying phase. As apparent from the

of REES corresponds to the total shift in emission maximum when the excitation figure, the anisotropy is lowest in the liquid-disordered phase

wavelength is changed from 525 to 580 nm. All other conditions are as in Fig. 3. See

(POPC). This is due to the relatively loose packing in the liquid-

Section 2 for other details.

disordered phase. Interestingly, the anisotropy of NR12S in the

liquid-ordered (POPC/cholesterol) phase, although higher than that

in liquid-disordered phase, appears to be much lower than the cor-

the fluorophore merely as a reporter group. REES has proved

responding value in the gel (DPPC) phase. This is surprising since,

to be a useful tool to monitor probe environment in mem-

as mentioned above, the acyl chains in the liquid-ordered phase

branes and membrane-mimetic environments (Mukherjee and

are reported to be extended and ordered, similar to the packing

Chattopadhyay, 2005; Chattopadhyay and Mukherjee, 1999a;

arrangement in the gel phase (but see later).

Shrivastava et al., 2009; Rawat and Chattopadhyay, 1999;

Fluorescence lifetime serves as a reliable indicator of the

Mukherjee et al., 2004; Raghuraman et al., 2004; Kelkar and

local environment in which a given fluorophore is localized

Chattopadhyay, 2004).

(Prendergast, 1991). A typical decay profile of NR12S in liquid-

The shifts in the maxima of fluorescence emission of NR12S in

ordered phase vesicles of POPC/cholesterol with its biexponential

vesicles of POPC and POPC/cholesterol as a function of excitation

fitting and the statistical parameters used to check the goodness

wavelength are shown in Fig. 4A. As the excitation wavelength is

of fit is shown in Fig. 6. The lifetimes of NR12S in membranes

changed from 525 to 580 nm, the emission maximum of NR12S

of varying phase are shown in Table 1. All fluorescence decays

exhibits a shift toward longer wavelengths in both cases. The emis-

could be fitted well to a biexponential function. We chose to use

sion maximum is shifted from 594 to 603 nm in liquid-disordered

the intensity-averaged mean fluorescence lifetime as an important

phase POPC vesicles and from 583 to 602 nm in liquid-ordered

POPC/cholesterol membranes. These shifts correspond to REES of

∼

9 and 19 nm for liquid-disordered and liquid-ordered phase vesi- Table 1

a

Representative fluorescence lifetimes of NR12S in membranes of varying phase.

cles, respectively, and are shown in Fig. 4B. Such dependence of

emission spectrum on excitation wavelength is characteristic of Condition ˛1 1 (ns) ˛2 2 (ns)  (ns)

the red edge effect. Observation of REES in both cases suggests that

POPC 0.12 1.05 0.88 3.96 3.86

the fluorescent moiety of NR12S experiences motionally restricted

DPPC 0.30 0.76 0.70 3.72 3.48

environment in these membranes. This is consistent with the inter- POPC/cholesterol 0.11 1.39 0.89 4.36 4.25

facial localization of the fluorescent group of NR12S in membranes a

The excitation wavelength was 490 nm and emission was monitored at 596 nm.

(see Fig. 2), since the membrane interface exhibits characteristic All other conditions are as in Fig. 3. Mean fluorescence lifetimes were calculated

slow solvent relaxation (Haldar et al., 2011; Chattopadhyay, 2003; using Eq. (4). See Section 2 for other details.

6 R. Saxena et al. / Chemistry and Physics of Lipids 183 (2014) 1–8

Fig. 6. Time-resolved fluorescence intensity decay of NR12S in POPC membranes containing 40 mol% cholesterol. The excitation wavelength used was 490 nm corresponding

to pulsed diode light source and emission was monitored at 596 nm. The sharp peak on the left corresponds to the profile of the pulsed light emitting diode (LED). The

relatively broad peak on the right is the decay profile, fitted to a biexponential function. The two lower plots show the weighted residuals and the autocorrelation function

of the weighted residuals. All other conditions are as in Fig. 3. See Section 2 for other details.

parameter for describing the behavior of membrane embedded is perturbed in gel phase DPPC vesicles due to extreme tight packing

NR12S because it is independent of the method of analysis and the of the lipid fatty acyl chains. As a result, the probe is pushed toward

number of exponentials used to fit the time-resolved fluorescence the membrane surface, thereby experiencing more polar environ-

decay. The mean fluorescence lifetimes of NR12S in membranes of ment which results in a shorter lifetime. This is in accordance with

varying phase were calculated from data shown in Table 1 using the observed red shifted emission in gel phase compared to the

Eq. (4), and are shown in Table 1 and Fig. 7A. The mean fluores- liquid ordered phase (Fig. 3) (Kucherak et al., 2010). Fluorescence

cence lifetime of NR12S was found to be minimum in gel phase lifetime of NR12S in POPC vesicles was found to be intermediate

DPPC membranes with a value of ∼3.5 ns. The mean fluorescence between gel and liquid-ordered phases (Fig. 7A).

lifetimes of NR12S in liquid-disordered (POPC) and liquid-ordered In order to ensure that the anisotropy values measured for

(POPC/cholesterol) phase membranes were longer, with values of NR12S (Fig. 5) are not influenced by lifetime-induced artifacts, the

∼

3.9 and 4.2 ns, respectively. In order to explore the effect of sol- apparent (average) rotational correlation times were calculated

vent polarity on the fluorescence lifetime of NR12S, we measured using Perrin’s equation (Lakowicz, 2006):

lifetime of NR12S in solvents of varying polarity. The fluorescence  r

 =

c (6) lifetimes of NR12S in different solvents are shown in Fig. S1. Our r

◦ − r

results show that fluorescence lifetime of NR12S is relatively high

in solvents of low polarity, and low in solvents of high polarity. where ro is the limiting (fundamental) anisotropy of the fluorescent

This suggests that fluorescence lifetime of NR12S is sensitive to group in NR12S (in the absence of any other depolarizing pro-

polarity of the surrounding medium. The longer mean fluorescence cesses such as rotational diffusion), r is the steady state anisotropy

lifetime of NR12S in POPC-cholesterol vesicles could therefore be (Fig. 5), and  is the mean fluorescence lifetime from Table 1.

due to the relatively tight packing of the lipid fatty acyl chains in the Although Perrin’s equation is not strictly applicable to this system,

liquid-ordered phase, which results in less water penetration. Inter- it is assumed that this equation will apply to a first approximation,

estingly, the relatively low fluorescence lifetime of NR12S in gel especially because we used mean fluorescence lifetimes for the

phase DPPC vesicles, characterized by compact packing of the lipid analysis of multiple component lifetimes. The values of the appar-

fatty acyl chains, merits comment. It is possible that the interfacial ent rotational correlation times, calculated using Eq. (6) with a ro

location of NR12S observed in liquid-disordered phase membranes value of 0.34 (Ferrer and del Monte, 2005), are shown in Fig. 7B.

R. Saxena et al. / Chemistry and Physics of Lipids 183 (2014) 1–8 7

(A) fluorescence, and exclusive localization in the outer leaflet

4.5

(Chiantia et al., 2012; Darwich et al., 2012). Although Nile Red is

frequently used as a lipid stain and in exploring membrane organi-

zation (Gao et al., 2006; Krishnamoorthy and Ira, 2001; Mukherjee

4.0

et al., 2007a,b), NR12S enjoys certain advantages over Nile Red. LIFETIME

Unlike Nile Red, NR12S specifically labels the outer leaflet of plasma

membranes in cells due to its very slow flip-flop rate. In this work,

3.5 we analyzed the membrane penetration depth of the fluorescent

(ns)

group in NR12S using the parallax method. We show here that the

fluorescent group in NR12S is localized at the membrane interfa-

3.0 cial region, characterized by an average depth of penetration of

FLUORESCENCE ∼

18 A˚ from the center of the bilayer (see Fig. 2). The interfacial

region of the membrane is uniquely characterized by relatively

slow solvent relaxation and offers an appropriate environment for

MEAN 2.5

POPC POPC + Chol DPPC REES measurements (Haldar et al., 2011; Chattopadhyay, 2003;

Chattopadhyay and Haldar, 2014). As a consequence, our results

show that NR12S displays phase-sensitive REES in membranes

(B)

(Fig. 4). These results constitute the first report on membrane pen-

etration depth and solvent relaxation characteristics of this novel

5 probe in membranes. In addition, we show that the fluorescence

(ns)

emission maximum, anisotropy, and lifetime of NR12S are depend-

ent on the phase of the membrane.

TIONAL

TIME The phase-dependent solvent relaxation properties of NR12S

4 could be potentially useful since cellular membranes display com-

ROTA

plex solvent relaxation patterns, depending on the membrane TION

A

phase. Taken together, our results show that NR12S can distin-

RENT

guish different membrane phases through a variety of fluorescence

PA

3 parameters such as emission maximum, REES, anisotropy and life- AP

CORREL

time. We conclude that NR12S appears to be a promising probe

to explore membrane organization in model and biological mem-

POPC POPC + Chol DPPC branes.

Fig. 7. (A) Mean fluorescence lifetimes of NR12S in membranes of varying phase.

Mean fluorescence lifetimes were calculated using Eq. (4). The excitation wave- Conflict of interest

length used was 490 nm and emission was monitored at 596 nm. Data shown are

±

means S.E. of at least three independent measurements. All other conditions are

The authors declare that there are no conflict of interest.

as in Fig. 3. See Section 2 for other details. (B) Apparent rotational correlation times

of NR12S in membranes of varying phase. Apparent rotational correlation times

were calculated from fluorescence anisotropy values of NR12S from Fig. 5 and mean

fluorescence lifetimes from panel (A) of this figure using Eq. (6). See text for other Transparency document

details.

The Transparency document associated with this article can be

found in the online version.

As expected, the apparent rotational correlation time was found

to be minimum in the liquid-disordered (POPC) phase. Interest-

ingly, the apparent rotational correlation times are almost same

Acknowledgments

in the liquid-ordered (POPC/cholesterol) and gel (DPPC) phases.

This is consistent with similar compact packing of lipid acyl chains

This work was supported by the Council of Scientific and Indus-

in both these phases. The apparent disagreement of these results

trial Research (CSIR, India), and Centre National de la Recherche

with anisotropies shown in Fig. 5 reveals that the anisotropy val-

Scientifique (CNRS, France). R.S. and S.H. thank the Council of Sci-

ues were indeed influenced by fluorescence lifetimes. The results

entific and Industrial Research for the award of Senior Research

from anisotropy measurements therefore should be interpreted

Fellowships. A.C. is an Adjunct Professor at the Special Centre for

with caution.

Molecular Medicine of Jawaharlal Nehru University (New Delhi,

India) and Indian Institute of Science Education and Research

4. Discussion

(Mohali, India), and Honorary Professor at the Jawaharlal Nehru

Centre for Advanced Scientific Research (Bangalore, India). A.C.

It is becoming increasingly evident that the eukaryotic cellular

gratefully acknowledges J.C. Bose Fellowship (Department of Sci-

environment is characterized by multiple membranes with vary-

ence and Technology, Govt. of India). We thank Arunima Chaudhuri

ing phase, differing in physical dimensions such as thickness (Van

and G. Aditya Kumar for help during the preparation of the

Meer et al., 2008; Sharpe et al., 2010). These membranes provide

manuscript, and members of A.C.’s research group for critically

the varied and dynamic backdrop for carrying out cellular signaling

reading the manuscript.

modulated by a range of slow solvent relaxation dynamics. In this

work, we have monitored the organization, dynamics and solvent

relaxation characteristics of NR12S in membranes of varying phase Appendix A. Supplementary data

utilizing fluorescence-based approaches including REES and the

parallax approach for depth analysis. NR12S is a recently devel- Supplementary data associated with this article can be found,

oped membrane probe based on Nile Red with several special in the online version, at http://dx.doi.org/10.1016/j.chemphyslip.

features. These include environmental sensitivity, phase-sensitive 2014.04.007.

8 R. Saxena et al. / Chemistry and Physics of Lipids 183 (2014) 1–8

References Kremer, J.M.H., van der Esker, M.W., Pathmamanoharan, C., Wiersema, P.H., 1977.

Vesicles of variable diameter prepared by a modified injection method. Bio-

chemistry 16, 3932–3935.

Abrams, F.S., London, E., 1993. Extension of the parallax analysis of membrane pen-

Krishnamoorthy, G., Ira, 2001. Fluorescence lifetime distribution in characterizing

etration depth to the polar region of model membranes: use of fluorescence

membrane heterogeneity. J. Fluoresc. 11, 247–253.

quenching by a spin-label attached to the phospholipid polar headgroup. Bio-

Kucherak, O.A., Oncul, S., Darwich, Z., Yushchenko, D.A., Arntz, Y., Didier, P., Mély,

chemistry 32, 10826–10831.

Y., Klymchenko, A.S., 2010. Switchable nile red-based probe for cholesterol

Bevington, P.R., 1969. Data Reduction and Error Analysis for the Physical Sciences.

and lipid order at the outer leaflet of biomembranes. J. Am. Chem. Soc. 132,

McGraw-Hill, New York.

4907–4916.

Brown, D.A., London, E., 1998. Structure and origin of ordered lipid domains in

Lakowicz, J.R., 2006. Principles of Fluorescence Spectroscopy, 3rd ed. Springer, New

biological membranes. J. Membr. Biol. 164, 103–114.

York.

Chattopadhyay, A., 1990. Chemistry and biology of N-(7-nitrobenz-2-oxa-1,3-diazol

Lampert, R.A., Chewter, L.A., Phillips, D., O’Connor, D.V., Roberts, A.J., Meech, S.R.,

4-yl)-labeled lipids: fluorescence probes of biological and model membranes.

1983. Standards for nanosecond fluorescence decay time measurements. Anal.

Chem. Phys. Lipids 53, 1–15.

Chem. 55, 68–73.

Chattopadhyay, A., 1992. In: Gaber, B.P., Easwaran, K.R.K. (Eds.), Biomembrane Struc-

London, E., Ladokhin, A.S., 2002. In: Benos, D., Simon, S. (Eds.), Current Topics in

ture and Function: The State of the Art. Adenine Press, Schenectady, NY, pp.

153–163. Membranes. Elsevier, New York, pp. 89–115.

MacDonald, R.C., MacDonald, R.I., Menco, B.Ph.M., Takeshita, K., Subbarao, N.K., Hu,

Chattopadhyay, A., 2002. Lipid probes in membrane biology. Chem. Phys Lipids 116,

1–188. L.R., 1991. Small-volume extrusion apparatus for preparation of large, unilamel-

lar vesicles. Biochim. Biophys. Acta 1061, 297–303.

Chattopadhyay, A., 2003. Exploring membrane organization and dynamics by the

McClare, C.W.F., 1971. An accurate and convenient organic phosphorus assay. Anal.

wavelength-selective fluorescence approach. Chem. Phys. Lipids 122, 3–17.

Biochem. 39, 527–530.

Chattopadhyay, A., Haldar, S., 2014. Dynamic insight into protein structure utilizing

Mouritsen, O.G., 2010. The liquid-ordered state comes of age. Biochim. Biophys. Acta

red edge excitation shift. Acc. Chem. Res. 47, 12–19.

1798, 1286–1288.

Chattopadhyay, A., London, E., 1987. Parallax method for direct measurement of

Mukherjee, S., Chattopadhyay, A., 1995. Wavelength-selective fluorescence as a

membrane penetration depth utilizing fluorescence quenching by spin-labeled

novel tool to study organization and dynamics in complex biological systems. J.

phospholipids. Biochemistry 26, 39–45.

Fluoresc. 5, 237–246.

Chattopadhyay, A., Mukherjee, S., 1993. Fluorophore environments in

Mukherjee, S., Chattopadhyay, A., 2005. Influence of ester and ether linkage in

membrane-bound probes: a red edge excitation shift study. Biochemistry

phospholipids on the environment and dynamics of the membrane interface:

32, 3804–3811.

a wavelength-selective fluorescence approach. Langmuir 21, 287–293.

Chattopadhyay, A., Mukherjee, S., 1999a. Depth-dependent solvent relaxation in

Mukherjee, S., Chattopadhyay, A., Samanta, A., Soujanya, T., 1994. Dipole moment

membranes: wavelength-selective fluorescence as a membrane dipstick. Lang-

change of NBD group upon excitation studied using solvatochromic and quan-

muir 15, 2142–2148.

tum chemical approaches: implications in membrane research. J. Phys Chem.

Chattopadhyay, A., Mukherjee, S., 1999b. Red edge excitation shift of a deeply

98, 2809–2812.

embedded membrane probe: implications in water penetration in the bilayer.

Mukherjee, S., Kombrabail, M., Krishnamoorthy, G., Chattopadhyay, A., 2007a.

J. Phys. Chem. B 103, 8180–8185.

Dynamics and heterogeneity of bovine hippocampal membranes: role of choles-

Chiantia, S., Klymchenko, A.S., London, E., 2012. A novel leaflet-selective fluorescence

terol and proteins. Biochim. Biophys. Acta 1768, 2130–2144.

labeling technique reveals differences between inner and outer leaflets at high

Mukherjee, S., Raghuraman, H., Chattopadhyay, A., 2007b. Membrane localization

bilayer curvature. Biochim. Biophys. Acta 1818, 1284–1290.

and dynamics of Nile Red: effect of cholesterol. Biochim. Biophys. Acta 1768,

Darwich, Z., Klymchenko, A.S., Kucherak, O.A., Richert, L., Mély, Y., 2012. Detection of

59–66.

apoptosis through the lipid order of the outer plasma membrane leaflet. Biochim.

Mukherjee, S., Raghuraman, H., Dasgupta, S., Chattopadhyay, A., 2004. Organiza-

Biophys. Acta 1818, 3048–3054.

tion and dynamics of N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)-labeled lipids: a

Das, R., Klymchenko, A.S., Duportail, G., Mély, Y., 2008. Excited state proton transfer

fluorescence approach. Chem. Phys. Lipids 127, 91–101.

and solvent relaxation of a 3-hydroxyflavone probe in lipid bilayers. J. Phys.

O’Connor, D.V., Phillips, D., 1984. Time-Correlated Single Photon Counting. Academic

Chem. B 112, 11929–11935.

Press, London, pp. 180–189.

Demchenko, A.P., 2008. Site-selective red-edge effects. Methods Enzymol. 450,

59–78. Prendergast, F.G., 1991. Time-resolved fluorescence techniques: methods and appli-

cations in biology. Curr. Opin. Struct. Biol. 1, 1054–1059.

Dittmer, J.C., Lester, R.L., 1964. A simple, specific spray for the detection of phospho-

Raghuraman, H., Kelkar, D.A., Chattopadhyay, A., 2005. Novel insights into protein

lipids on thin-layer chromatograms. J. Lipid. Res. 5, 126–127.

structure and dynamics utilizing the red edge excitation shift approach. In: Ged-

Eggeling, C., Ringemann, C., Medda, R., Schwarzmann, G., Sandhoff, K., Polyakova,

des, C.D., Lakowicz, J.R. (Eds.), Reviews in Fluorescence. Springer, New York, pp.

S., Belov, V.N., Hein, B., von Middendorff, C., Schönle, A., Hell, S.W., 2009. Direct

199–222.

observation of the nanoscale dynamics of membrane lipids in a living cell. Nature

Raghuraman, H., Pradhan, S.K., Chattopadhyay, A., 2004. Effect of urea on the organi-

457, 1159–1163.

zation and dynamics of Triton X-100 micelles: a fluorescence approach. J. Phys.

Ferrer, M.L., del Monte, F., 2005. Enhanced emission of nile red fluorescent nanopar-

Chem. B 108, 2489–2496.

ticles embedded in hybrid sol–gel glasses. J. Phys. Chem. B 109, 80–86.

Raghuraman, H., Shrivastava, S., Chattopadhyay, A., 2007. Monitoring the looping

Gao, F., Mei, E., Lim, M., Hochstrasser, R.M., 2006. Probing lipid vesicles by bimolec-

up of acyl chain labeled NBD lipids in membranes as a function of membrane

ular association and dissociation trajectories of single molecules. J. Am. Chem.

phase state. Biochim. Biophys. Acta 1768, 1258–1267.

Soc. 128, 4814–4822.

Rawat, S.S., Chattopadhyay, A., 1999. Structural transition in the micellar assembly:

Golini, C.M., Williams, B.W., Foresman, J.B., 1998. Further solvatochromic, ther-

a fluorescence study. J. Fluoresc. 9, 233–244.

mochromic, and theoretical studies on Nile Red. J. Fluoresc. 8, 395–404.

Reichardt, C., 1988. Solvents and Solvent Effects in Organic Chemistry. VCH Publish-

Greenspan, P., Fowler, S.D., 1985. Spectrofluorometric studies of the lipid probe, nile

ers, Weinheim, Germany.

red. J. Lipid Res. 26, 781–789.

Sackett, D.L., Knutson, J.R., Wolff, J., 1990. Hydrophobic surfaces of tubulin probed

Grinvald, A., Steinberg, I.Z., 1974. On the analysis of fluorescence decay kinetics by

by time-resolved and steady-state fluorescence of Nile Red. J. Biol. Chem. 265,

the method of least-squares. Anal. Biochem. 59, 583–598.

14899–14906.

Haldar, S., Chattopadhyay, A., 2013. In: Mely, Y., Duportail, G. (Eds.), Fluorescent

Sharpe, H.J., Stevens, T.J., Munro, S.A., 2010. Comprehensive comparison of

Methods to Study Biological Membranes. Springer, Heidelberg, pp. 37–50.

transmembrane domains reveals organelle-specific properties. Cell 142,

Haldar, S., Chaudhuri, A., Chattopadhyay, A., 2011. Organization and dynamics of

158–169.

membrane probes and proteins utilizing the red edge excitation shift. J. Phys.

Shrivastava, S., Haldar, S., Gimpl, G., Chattopadhyay, A., 2009. Orientation and

Chem. B 115, 5693–5706.

dynamics of a novel fluorescent cholesterol analogue in membranes of varying

Haldar, S., Kombrabail, M., Krishnamoorthy, G., Chattopadhyay, A., 2012.

phase. J. Phys. Chem. B 113, 4475–4481.

Depth-dependent heterogeneity in membranes by fluorescence lifetime distri-

Shynkar, V.V., Klymchenko, A.S., Kunzelmann, C., Duportail, G., Muller, C.D.,

bution analysis. J. Phys. Chem. Lett. 3, 2676–2681.

Demchenko, A.P., Freyssinet, J.M., Mély, Y., 2007. Fluorescent biomembrane

Jameson, D.M., Ross, J.A., 2010. Fluorescence polarization/anisotropy in diagnostics

probe for ratiometric detection of apoptosis. J. Am. Chem. Soc. 129, 2187–

and imaging. Chem. Rev. 110, 2685–2708.

2193.

Jurkiewicz, P., Olzynska, A., Langner, M., Hof, M., 2006. Headgroup hydration and

van Meer, G., de Kroon, A.I.P.M., 2011. Lipid map of the mammalian cell. J. Cell Sci.

mobility of DOTAP/DOPC bilayers: a fluorescence solvent relaxation study. Lang-

124, 5–8.

muir 22, 8741–8749.

Van Meer, G., Voelker, D.R., Feigenson, G.W., 2008. Membrane lipids: where they are

Kelkar, D.A., Chattopadhyay, A., 2004. Depth-dependent solvent relaxation in

and how they behave. Nat. Rev. Mol. Cell Biol. 9, 112–124.

reverse micelles: a fluorescence approach. J. Phys. Chem. B 108, 12151–12158.