Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

Journal of Molecular Structure xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Journal of Molecular Structure


journal homepage: www.elsevier.com/locate/molstruc

Binding phenomena and fluorescence quenching. II: Photophysics


of aromatic residues and dependence of fluorescence spectra on protein
conformation q
Patrik R. Callis ⇑
Department of Chemistry and Biochemistry, Montana State University, Bozeman, MT 59717, United States

h i g h l i g h t s g r a p h i c a l a b s t r a c t

 Survey of useful tryptophan models


for testing outside the protein
environment.
 Survey of natural quenchers and non-
quenchers of tryptophan found in
proteins.
 Introduction to principles behind why
excited Trp is easily quenched by
electron transfer.
 Introduction to why Trp is so
sensitive to electrostatic environment
whereas similar sensor molecules are
not.

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

Article history: The three amino acids with aromatic ring side chains—phenylalanine (Phe), tyrosine (Tyr), and especially
Received 15 April 2014 tryptophan (Trp) have played a long and productive role in helping unlock the secrets of protein behavior
Accepted 15 April 2014 by optical spectroscopy (absorption, fluorescence, circular dichroism, etc.) In principle, an appropriately
Available online xxxx
placed Trp will undergo fluorescence wavelength and/or intensity changes upon whatever functional
process a protein performs. Although perceived to be enigmatic and not well understood, Trp is arguably
Keywords: now better understood than many of the extrinsic probes currently in use. Basic principles of intrinsic
Protein
tryptophan fluorescence quenching and wavelength shifts in proteins are presented, with strong
Tryptophan
Fluorescence
emphasis on the importance of electrostatics. The condensed description of findings from recent
Quenching experiments and simulations of tryptophan fluorescence and intrinsic quenching in proteins is designed
Electron transfer to help authors in planning and interpreting experimental results of ligand binding studies.
Molecular dynamics simulations Ó 2014 Elsevier B.V. All rights reserved.

Introduction chains—phenylalanine (Phe), tyrosine (Tyr), and especially trypto-


phan (Trp) have played a long and productive role in helping
Proteins are the working machines of life forms. As such, there unlock the secrets of protein behavior by optical spectroscopy
is intense interest in understanding the mechanistic details of pro- (absorption, fluorescence, circular dichroism, etc.) [1–10]. This sec-
tein function. The three amino acids with aromatic ring side tion seeks to provide an in-depth, up-to-date resource about prop-
erties of these intrinsic aromatic residues that will familiarize
those planning to use fluorescence as a tool for proteins—especially
q
This article is part of a special issue titled ‘‘Fluorescence studies of biomolecular
in the context of ligand binding.
association processes. Towards a detailed understanding of spectroscopic, thermo-
dynamic and structural aspects’’. Trp dominates the following presentation for the same reason
⇑ Tel.: +1 1 406 994 5414. that Trp is the most widely exploited of the aromatic amino acids.
E-mail address: pcallis@montana.edu The larger p-system makes it the strongest and the most

http://dx.doi.org/10.1016/j.molstruc.2014.04.051
0022-2860/Ó 2014 Elsevier B.V. All rights reserved.

Please cite this article in press as: P.R. Callis, J. Mol. Struct. (2014), http://dx.doi.org/10.1016/j.molstruc.2014.04.051
2 P.R. Callis / Journal of Molecular Structure xxx (2014) xxx–xxx

red-shifted absorber and emitter, and both its emission wave- that both the radiative and non-radiative rates for 3MI in solution
length and quantum yield are quite sensitive to the local electro- increase considerably as solvent polarity decreases. As a result, the
static environment (electric field direction) in proteins. Normally, quantum yield remains relatively independent of polarity while
this means it is possible to exclusively excite Trp, thereby eliminat- the lifetime decreases from 9 ns in water to 4 ns in cyclohexane,
ing complexities due to FRET from Tyr and Phe. In principle, an with intermediate values in ether or alcohol. The radiative rate
appropriately placed Trp will undergo fluorescence wavelength increases at shorter wavelengths (non-polar solvents) for two rea-
and/or intensity changes upon whatever functional process a pro- sons: (1) all radiative rates are proportional to the cube of the fre-
tein performs. quency of the light emitted, and (2) the transition dipole (the
A primary theme of this paper is that understanding of fluores- square of which determines the radiative rate) increases with
cence spectral shifts and quenching due to electron transfer decreasing polarity. The latter is because the excited state wave-
depends on having a basic (but not extensive) grasp of electrostatic function has less charge transfer character in non-polar solvents.
principles. To aid the reader, an Appendix is provided that gives These together explain the otherwise curious phenomenon that
rapid access to useful numerical values in relation to spectral shifts sf for the buried, high quantum yield Trp48 of azurin is only 4 ns
and energy gaps in terms of the extremely high electric fields and whereas Trp in water-exposed positions may have an equally high
potentials found in proteins. quantum yield with sf of 9 ns.
In contrast, the mechanism by which the internal non-radiative
rates (internal conversion and intersystem crossing) for 3MI
Tryptophan
increase as the 1La absorption is blue shifted does not seem to be
known.
Trp has a reputation for being enigmatic and not well under-
stood. This was, indeed, once the case, but it is arguably now better
N-acetyltryptophanamide (NATA)
understood than many of the extrinsic probes currently in use
NATA in water is a good model for a solvent-exposed Trp in a
[1,10–57]. The just-cited references are a somewhat biased repre-
non-terminus protein site. This is because the two nearby amides
sentative selection from over 10,000 papers in the last 60 years
are chemically very similar to the two nearby backbone amides
addressing tryptophan or indole fluorescence. Other papers cited
that all protein Trps experience. The quantum yield is 0.14, about
in this review are also significant.
half that of 3MI, virtually proving that electron transfer quenching
The perceived complexity of Trp comes from the nature of the
to backbone amides is a viable mechanism, as was recognized
pyrrole-benzene ring system which causes the two lowest excited
quite early [46,61–63]. NATA in alcohols and ethers is a good
singlet states, 1Lb and 1La, to be accidently degenerate, unlike both
model for more buried Trps. In these less polar environments it
Phe and Tyr, which have widely separated 1Lb and 1La states [58].
is found that electron transfer to the amides is suppressed because
For many years there was belief that either state might be the
the CT-1La energy gap fluctuations are smaller, not because the
emitting state, depending on the protein environment. An ad hoc
energy gap is smaller [64]. An interesting and revealing study
practice of assigning fluorescence with vibronic structure to 1Lb
has been performed by Alston et al. [65,66], demonstrating that
has turned out to be merely the consequence that the large perma-
the local amino acid sequence affects Trp fluorescence in small
nent dipole change for the 1La transition (not to be confused with
peptides and denatured proteins.
transition dipole) leads to sufficiently large inhomogeneous broad-
ening in most protein environments that the natural vibronic
Tryptophan (free amino acid)
structure is obscured. Studies in cold vacuum have verified the pre-
The free amino acid at any pH is not a good model for Trp in a
dicted vibronic structure for the 1La transition [29], and several
protein because there is no comparable bonding environment hav-
studies have since shown that the more intensely absorbing, sol-
ing the charged ammonium and carboxylate groups tethered in
vent-sensitive 1La state is the emitting state in the vast majority
such close proximity to the indole ring.
of proteins [27,29,48,59,60]. For only two mutant proteins showing
exceptionally blue shifted fluorescence has evidence of 1Lb emis-
Trp-containing dipeptides
sion been observed [48]. Even the extremely blue shifted fluores-
These can be used to model cases in which Trp in a protein is in
cence of Trp48 of azurin, whose fluorescence wavelength
the N- or C-terminal position [37,40].
maximum (kmax) is 306 nm, has been shown to emit entirely from
1
La [48,60].
5-Fluoro3MI, 5-fluoroNATA, and 5-fluorotryptophan
Whereas the previously listed models are useful for mimicking
Model reference molecules Trp in various environments in proteins, their 5-fluorotryptophan
(5-FW) derivatives provide valuable information for why they gen-
Studies that depend on the quenching of a fluorophore in a erally do not mimic Trp in proteins. 5-FW has found considerable
polymer will always be strengthened by collecting data under use as a modified Trp that can be incorporated into proteins in
identical conditions for model systems in the absence of the pro- place of Trp. Broos and coworkers [67,68] found that when incor-
tein environment. We first identify and describe those most used poration was near 100%, most Trps that showed multi-exponential
for Trp. fluorescence decay curves became effectively mono-exponential
when replaced by 5FW. This striking behavior was subsequently
3-Methylindole (3MI) explained by the slightly higher (0.2 eV) ionization potential of
3MI has only a methyl at the 3-position where Trp connects to 5FW compared to the model 3MI because the highest (energy)
the protein backbone. This is a useful reference because the occupied molecular orbital (HOMO) is stabilized by 0.2 eV [32].
absence of the backbone section eliminates intrinsic quenching That such a seemingly small decrease in electron donating power
by electron transfer, while other spectroscopic properties are not can produce such a dramatic effect on the sensitively of the local
significantly affected. This is a better model than indole because electrostatic environment strongly supports the discrete state
the 3-methyl substitution shifts the 1La state to lower energies, hypothesis underlying non-exponential fluorescence decay of Trp
allowing it to cross the 1Lb state and become the fluorescing state, in proteins. This modification has virtually no effect on the sensi-
even in hydrocarbon environments, as noted above. The little- tivity of wavelength to environment, so its use is valuable for
cited, but unique and important study by Meech et al. [11] revealed removing unwanted heterogeneity caused by sensitivity to

Please cite this article in press as: P.R. Callis, J. Mol. Struct. (2014), http://dx.doi.org/10.1016/j.molstruc.2014.04.051
P.R. Callis / Journal of Molecular Structure xxx (2014) xxx–xxx 3

electron-transfer based quenching while retaining nearly the same electron transfer-based quenchers of 3MI (and therefore of Trp,
wavelength sensitivity [43]. The quantum yield of 5-FW indicates in principle) commonly found in proteins. Note that the proton-
that 5-F substitution, for unknown reasons, does not inhibit ation state is quite important. The second order collisional quench-
quenching by proton transfer [69]. ing rate constant, kq, is also shown in Table 2 to indicate the
relative effectiveness of the quenchers [16]. Acrylamide, a common
Trps in proteins are usually ‘‘pre-quenched’’ collisional quencher used to sense whether Trp is solvent exposed
is included for reference [16].
Virtually all reported quantum yields of single Trps in proteins Proton transfer-based quenching. Table 2 lists groups that are
fall in the range 0.35 down to 0.01 or less [1,6]. Recall that 0.35 is found to quench indole and 3MI in water by a not so well under-
essentially the quantum yield exhibited by 3MI in any solvent [11]. stood mechanism involving proton transfer to the 4, 7, or 2 posi-
The interesting physical basis for this remarkable—and very use- tions of the excited indole ring [16,53]. Robb, Olivucci and
ful—variability will be discussed in Section 2.4. For now, we note coworkers [74,75] have proposed a mechanism for this type of
that the implications of this ‘‘pre-quenched’’ state of many of the quenching.
candidate Trps for sensing ligand binding must be kept in mind. Proton-transfer based quenching is distinguished from electron
For an intrinsically quenched Trp with Uf = 0.15, for example, transfer-based quenching by its pronounced deuterium isotope
ligand binding may actually increase the fluorescence by partially effect [16]. Intramolecular proton transfer from the ammonium
or wholly removing the quenching mechanism, or may cause group of free Trp zwitterion at neutral pH reduces Uf to 0.14 rela-
pseudo-quenching, by enhancing the existing intrinsic quenching tive to the value near 0.35 found for 3MI and Trp at pH 10 [14,76].
mechanism, or by introducing a new one. These may come about,
for example, by local structure conformational change, electro- Intrinsic quenching in proteins
static effects, changing the nature of water exposure, etc. This does
not compromise the use of Trp as a sensor, but merely means that While considerable progress has been made towards under-
one cannot safely assume that the ligand is actually quenching, standing Trp lifetimes and quantum yields in proteins, the subject
even if it is capable of doing so. remains complex for two reasons:
One important example is the monitoring of Trp fluorescence
changes in myosin upon addition of ATP for fundamental studies (a) There are ten groups in proteins that in principle can quench
of muscle contraction. Because ATP itself does not quench Trp fluo- Trp fluorescence.
rescence, an early study by Morales and coworkers concluded that (b) Potential quenchers of a Trp in a particular protein may or
Trp fluorescence quantum yield could be affected by electrostatic may not actually quench; the extent of quenching depends
effects alone [70]. Similarly, there are numerous accounts of Trp critically on the precise electrostatic environment for both
fluorescence changes upon binding one of the most important the Trp and the quencher—affecting whether electron trans-
ligands in biology, Ca2+, which is not a quencher [71]. fer is uphill or downhill in free energy.
For example, quenching of Trp214 fluorescence of human
serum albumin by peptides observed by Gonzalez-Beja and The fluorescence quantum yield varies by 30-fold for single
coworkers [72] cannot be assigned with certainty to quenching Trps in different proteins. This has turned out to be because of
by the peptide binding, given the relative weakness of amides as extreme variation among proteins regarding the ability of the local
quenchers and the likelihood that structural changes accompany- electrostatic environment to assist or inhibit electron transfer to
ing binding could alter the intrinsic quenching by the local protein the nearest backbone amide groups [31,77]. Such sensitivity to
backbone near the Trp. quenching by electron transfer is the primary reason that Trp
fluorescence is widely used for monitoring so many protein
Quenchers of Trp models in water processes, including ligand binding. Only recently, however, was
the mechanism of this sensitivity rationalized [31,77]. Below,
Before tackling the more involved subject of intrinsic quenching we: (a) address why excited Trp is able to transfer electrons easily,
of Trp fluorescence in proteins, we consider the more straightfor- (b) why electron transfer quenches fluorescence, (c) what types of
ward information that exists for collisional quenching of 3MI in environments favor or disfavor electron transfer, and (d) why Trp
aqueous solution by groups that are potential quenchers in is so sensitive to electrostatic environment whereas similar sensor
proteins. molecules are not.
There are two primary intrinsic quenching mechanisms at work
in the case of Trp: (1) quenching by electron transfer from the
excited indole ring to the nearby backbone amides or electron- Table 1
accepting side chains; (2) quenching by proton transfer to elec- Electron transfer quenchers.
tron-rich atoms of the indole ring. The protons come from acidic Quencher kq  107 M 1
s 1

side chains such as tyrosine, and the ammonium groups of lysine


Disulfide bonds >1000
and N-termini. Barkley and coworkers have made a careful study Acrylamide 720
of isotope effects on these quenchers that clearly distinguishes Histidine cation (pH 5.3) 240
between the two mechanisms by the higher deuterium isotope Cysteine anion (pH 10.4) 190
effect for proton transfer quenching [16]. Intrinsic quenching of Neutral cysteine (pH = 7.6) 140
Neutral acetic acid 43
Trp fluorescence by FRET, however, is typically not possible
Neutral glutamic acid (pH 4.5) 43a
because no other amino acid has an absorption spectrum that over- Neutral aspartic acid (pH 4.5) 43a
laps the Trp emission (except in rare cases of tyrosine anion). Neutral C-terminal carboxyl 43a
Electron transfer quenchers. In water, there is compelling exper- N-acetylasparagine 8.8b
imental evidence that the following groups can act as collisional N-acetylglutamine 6.5b
Protein backbone amides 0.1 to 500
quenchers of the pertinent model chromophore, 3MI, by an elec- Neutral histidine (pH > 6) 3.7
tron transfer mechanism [15,16,73]. Barkley and coworkers [16]
a
have carefully distinguished this mechanism from a separate Behavior is expected to be very similar to neutral acetic acid.
b
N-acetylation is to remove proton transfer quenching by N terminal NH+3.
mechanism due to proton transfer. Table 1 lists the important

Please cite this article in press as: P.R. Callis, J. Mol. Struct. (2014), http://dx.doi.org/10.1016/j.molstruc.2014.04.051
4 P.R. Callis / Journal of Molecular Structure xxx (2014) xxx–xxx

Table 2
Proton transfer quenchers.

Quencher kq  107 M 1
s 1

Neutral tyrosine side chain 98


N-terminal ammonium groups 26
Protonated lysine side chain 10

(a) Trp fluorescence is easily quenched because it can donate an


electron with relative ease. This is directly related to the fact
that Trp is one of the best reducing agents in biology. Both
facts stem from the remarkably low ionization energy of
the ground state (7.4 eV) compared to many aromatic sys-
tems, e.g., benzene (9.2 eV). This simply means an electron
occupying the HOMO of the Trp side chain lies high in
energy relative to that of most other molecules. Even in
the ground state, Trp acts as a reducing agent in some
well-known biological redox reactions. The lowest energy
excited state of Trp is 4 eV higher than the ground state,
meaning that an electron can be transferred with even more
ease—if the electrostatic environment is favorable.
(b) Why does electron transfer quench fluorescence? The
answer is that following electron transfer, neither the
indole ring radical cation nor the radical anion of the accep-
tor entity are able to fluoresce (they are ground state radi- Fig. 1. Quantum mechanical excitation energies for the fluorescing state (1La (red)
cals), and CT fluorescence that would reverse the electron and CT state (black), computed from molecular dynamics simulations for A. Staph.
nuclease; B. Dsba. (For interpretation of the references to colour in this figure
transfer is extremely weak. Furthermore, molecular dynam-
legend, the reader is referred to the web version of this article.)
ics combined with quantum mechanical computation
(MD + QM) simulations suggest that solvent relaxation
about the CT state stabilizes the CT state to an energy well the CT state and the fluorescing states must be in resonance,
below that of the 1La fluorescing state within 100 fs! There i.e., must have the same energy at the instant of transfer.
is therefore little time for the electron to return, and the This requires that positive charges are close to the acceptor,
electron remains trapped until a dark electron transfer back and/or negative charges are near the donor.
to the indole ring returns the system to the ground state.
This was first understood semi-quantitatively through
Fig. 1 shows MD + QM trajectories of computed excitation ener- MD + QM simulations [31,77]. Fig. 1 compares representative
gies to the 1La and lowest CT states for two proteins bathed in MD + QM results for Staphylococcus nuclease (STNase), whose sin-
explicit water at room temperature. The energies are given in thou- gle Trp is very fluorescent, and Dsba, whose single Trp is weakly
sands of cm 1, where 33000 cm 1 is 300 nm. Throughout the first fluorescent. The large CT-1La energy gap found for STNase is consis-
50 ps, the water and protein is equilibrated to the charge distribu- tent with a high quantum yield, because the CT state has little
tion of the 1La state. At the arbitrary time of 50 ps, the charges on probability to have the same energy as 1La. Likewise, the small
the Trp were changed to those of the CT state. The subsequent energy gap seen for Dsba offers much higher opportunity for elec-
rapid stabilization of the CT state in each case by 16,000 cm 1 tron transfer. Fig. 2 is a cartoon of why the energy gaps differ so
(2 eV or 192 kJ/mol) is what appears to be the primary act of much between these two cases, and exemplifies the general under-
quenching. lying cause of the variability. Fig. 2A shows the specific case of
positive and negative groups, e.g., lysine and glutamate, stabilizing
(c) What causes the variability? From early times the root basis electron transfer to a backbone amide. Such an arrangement will
for most of the variability was believed to be electron trans- lead to a weakly fluorescing Trp, in the absence of other nearby
fer from the excited indole ring to the closest backbone charges. The opposite arrangement shown in Fig. 2B obviously
amides [46,62,63]. A study of cyclic hexapeptides, designed inhibits electron transfer to the amide, and leads to a highly fluo-
to contain no quencher candidates except the backbone rescent Trp. This is an example of what could be called an internal
[78,79], incisively demonstrated that the backbone amides resonance Stark effect, similar to the external resonance Stark
are effective quenchers, and also showed strong variation effect described by Boxer and coworkers [80]. What is left out of
depending on sequence. The precise mechanism, however, the cartoon is the considerable contribution from water molecules,
by which some Trps in proteins exhibited the full 0.35 value especially with regard to H-bonding to the amides.
for quantum yield while others had a quantum yield <0.01 in
the absence of obvious strong quenching side chains, was (d) Why Trp is so sensitive to electrostatic environment was
unknown until 10 years ago. Even with X-ray structures, it alluded to in Section 2.1 on 5-fluorotryptophan [32]. This
was difficult to understand why many Trps in proteins were has to do with the size of the CT-1La energy gap being such
not quenched at all while others were strongly quenched, as to give a quenching rate close to kr + knr in most proteins.
given that all Trps are close to at least two backbone amides. According to Eq. (1), the quantum yield would be near 0.15.
As noted above, the answer to this question lies in the ability If the quenching rate is substantially increased or
of the enormously strong electrostatic fields (potential dif- decreased, the quantum yield will span the 0–0.3 range.
ferences) typical of proteins to assist or inhibit the electron Fluorophores that have much larger or much smaller energy
transfer process to an amide. Electron transfer—as for all gaps will have very low or very high quantum yields that
processes in nature—requires that energy be conserved; are less sensitive to small changes in the energy gap.

Please cite this article in press as: P.R. Callis, J. Mol. Struct. (2014), http://dx.doi.org/10.1016/j.molstruc.2014.04.051
P.R. Callis / Journal of Molecular Structure xxx (2014) xxx–xxx 5

Fig. 2. How the location of environment charges (circled) determines the quenching by electron transfer to a backbone amide. The uncircled charges represent the dipole
created by electron transfer from the indole ring to an amide. In (A), charges stabilize the dipole of the CT state with electron transferred towards the positive and away from
the negative. The CT state is in resonance with the fluorescing (1La) state, electron transfer is fast, and fluorescence is weak. In (B), charges are arranged oppositely from case A,
greatly destabilizing the CT state. There can be little opportunity for resonance; quenching is therefore slight, and fluorescence is strong. Water and other polar groups are
equally important in contributing to the overall electrostatic influence on electron transfer in both cases. (For interpretation of the references to colour in this figure legend,
the reader is referred to the web version of this article.)

The general picture portrayed in Fig. 2 is valid, provided electro- Quenching by some ligands in proteins
statics of H-bonding from water and other groups is included. It is
supported by considerable circumstantial computational evidence Trp fluorescence quenching by ligands have provided useful
that demonstrates the crucial importance of the local electric field insights. A few are mentioned here, including quenching by heme,
direction and magnitude in stabilizing the electron occupation of Cu2+, Eu3+, and Yb3+.
the empty p orbital localized mostly on the amide carbonyl car- Tryptophan fluorescence is highly quenched in heme-contain-
bon [17,31–34,41,54,77,81–85]. This striking variability is not ing proteins. Hochstrasser and Negus found Trp fluorescence life-
due to differences in the local polarity because the quantum yield times on the order of 10 s of ps for several myoglobins, in which
is nearly independent of solvent polarity for indole chromophores the separation between Trp and heme is 20 Å [90]. Their analysis
[11]. Electron capture into the same p orbital has been strongly for Sperm Whale and Tuna myoglobins assuming quenching was
implicated for fragmentation mechanisms in mass spectroscopy solely by FRET gave results in agreement with experiment.
of peptides [86]. Recently, Parson, Anderson and co-workers were Trp 48 of azurin has one of the most hydrophobic environments
able to predict fluorescence quantum yields for Trp in several known, leading to extremely blue-shifted fluorescence and maxi-
folded hairpin peptides using a similar, but independent MD + QM mally high fluorescence quantum yield in the absence of Cu II. In
method [54]. its active form, a Cu II ion is bound in a site 10 Å from the closest
An example of how electrostatic effects can cause unexpected ring atom of Trp48, and quenches the fluorescence by 20-fold. A
results comes from an extensive study of the conserved, weakly thorough study of azurins from different species by Petrich et al.
fluorescing Trps 68 and 156 of human gamma D and gamma S tentatively assigned the quenching mechanism as electron transfer
crystallin mutants, in which Chen et al. [33] replaced Cys32 whose from the excited Trp48, reducing the Cu II to Cu I.
S is only 4 Å distant from Trp 68 ring atoms. No increase in quan- Horrocks et al. showed definitively that electron transfer occurs
tum yield was found, however, apparently because of negative over 10 Å in cod parvalbumin when the Ca2+ binding sites are occu-
groups near the cysteine, which suppress its electron accepting pied by Eu3+ or Yb3+ [91,92]. In the lanthanide series, Eu3+ and Yb3+
power. In other cases, there is evidence of moderate quenching have by far the most positive reduction potentials, and are the only
by Cys [67,87]. lanthanides that quench Trp fluorescence when bound to cod par-
In another example, histidine cation was documented as a valbumin. Electron transfer was shown as the only viable mecha-
quencher in proteins, typically by showing a decrease in Uf upon nism for Yb3+. FRET is ruled out because Yb3+ has no absorption
lowering pH [88,89]. For the Q105H mutant of T4 lysozyme, how- bands that overlap Trp emission.
ever, Van Gilst and Hudson [52] found that lowering the pH
increased Uf of the nearby Trp138. We have argued that this seem- Nonquenchers in proteins
ingly anomalous behavior is caused by the electrostatic effect of
the His cation positive charge near the indole ring, which reduces Chen and Barkley found no detectable quenching of 3-methyl-
the already substantial quenching by the backbone amide, but not indole fluorescence in water at neutral pH by N-acetyl derivatives
enough to send the electron to the His [31]. This behavior illus- of glycine, alanine, valine, leucine, phenylalanine, proline,
trates the difficulties of predicting Trp fluorescence quenching hydroxyproline, serine, threonine, methionine, and arginine. The
pathways in proteins. lower limit for detection was kq < 1  107 M 1 s 1. Early quenching
Proton transfer quenching in proteins. There is much less evi- studies reported that most of these as free amino acid zwitterions
dence for proton transfer quenching compared to that from elec- quenched tryptophan fluorescence at pH 5–6 [93,94]. The use of N-
tron transfer in proteins. There is often association of Trp with acetyl derivatives, however, subsequently revealed that in many
Lys or Arg in what are called cation-p complexes, but Trps in such cases the unacetylated amino acids quenched by proton transfer
complexes are almost always highly fluorescent. Apparently, in from the N-terminal ammonium group [16].
such complexes, the ammonium group of the Lys is either con- For emphasis and clarity, additional comments are made con-
formationally hindered from donating a proton, or a more aqueous cerning glutamate, aspartate, and phenylalanine, which are some-
environment is required. The high Uf values in these complexes are times stated to be quenchers, but for which there is little or no
easily understood in terms of electrostatics, as indicated in Figs. 1 evidence that these can act as quenchers in solution or in proteins.
and 2. Close proximity of neutral tyrosines to Trps is common in Glutamate and aspartate, being negatively charged, almost cer-
proteins, but does not seem to correlate with quenching. tainly cannot quench by accepting an electron. All charged groups,

Please cite this article in press as: P.R. Callis, J. Mol. Struct. (2014), http://dx.doi.org/10.1016/j.molstruc.2014.04.051
6 P.R. Callis / Journal of Molecular Structure xxx (2014) xxx–xxx

however, may appear to be quenchers in the protein environment A serious caveat, however, is that the strong effect of nearby
because of electrostatic effects. For example, a glutamate near the protein charges will be considerably offset by dielectric compensa-
indole ring of Trp will greatly assist electron transfer to a backbone tion from large numbers of water molecules out to 15 Å, which
amide. Confusion may also arise because of the variability of pKa may be partially oriented by charges near the chomophore. Only
values of glutamic and aspartic acids in proteins, which have been when these very substantial contributions are included will mod-
observed as high as 7.5, i.e., 3 units above the normal values. Thus, eling predictions be accurate [30]. A spectacular example is the
when somewhat buried, glutamate and aspartate may quench at case of Trp140 in STNase, which is flanked by two lysine ammo-
pH7, but only because they are protonated in that environment. nium groups over the benzene ring. The nearby positive charges
A conserved phenylalanine (Phe8) juxtaposed to Trp48 in the nominally would contribute a 90 nm Coulombic red shift, but
large class of homeo domains was implicated as the reason for mutations of these lysines to neutral sidechains is experimentally
the very low quantum yield of this Trp in non-polar environment found to result in only tiny (0.5 nm) shifts of the fluorescence
[95–97]. This appears to be a case of correlation instead of cause. kmax [87]. This effect is well reproduced by MD + QM modeling
Nominally, the HOMO of benzene is too low in energy and the low- wherein polarization of distant water contributes compensating
est unoccupied MO (LUMO) is too high to facilitate quenching of blue shifts that in each case leave the resultant shift nearly
Trp. The picture is clouded by the fact that benzene quenches unchanged [35].
Trp fluorescence in concentrated solutions. Subramanian et al., Water–protein relaxation. Shifts of absorption spectra are much
however, found that the Trp quantum yield remained very low in smaller than those of fluorescence because the ground state per-
the F8A mutant of Bicoid-homeo domain, somewhat dispelling manent dipole is much smaller than the 1La permanent dipole,
the notion that Phe is the quencher. There is MD + QM evidence although both point in roughly the same direction (with the pyr-
[98] indicating that the ubiquitous strong quenching of the con- role ring being more positive than the benzene ring). Immediately
served Trp 48 in homeo domains is more likely due to the back- upon excitation the system is in a non-equilibrium state, and the
bone amide, facilitated by a large helix dipole, whose field surroundings begin to respond in a way to return to electrostatic
strongly directs electrons from the Trp ring to the amide. Similarly, equilibrium. The largest component of this relaxation typically is
the low quantum yield of conserved Trps 68 and 156 of cD and cS from water, because of its fast rotational and translational mobility
crystallins from many species all have fluorescence quantum yields [105]. For solvent exposed Trps in proteins, the longest kmax is
near 0.01, yet extensive mutation of nearby plausible quenchers 352 nm, considerable shorter than that of 3MI in water
fails to increase the quantum yield [33,34]. Similar to the homeo (374 nm) and NATA (362 nm) [14]. This is apparently because in
domains, the conformation and surroundings result in efficient proteins, one side of the indole ring is shielded from solvent.
quenching by the nearby backbone amide, helped considerably In recent years, the rate of this relaxation has been thoroughly
by firmly trapped water molecules that stabilize the CT state. studied with ultrafast time resolved fluorescence upconversion
methods [38,39,41,49,50,106,107]. The timescale of the relaxation
Tryptophan and tyrosine are quenchers of dyes and flavins is almost always observed to be biphasic, with the bulk of relaxa-
tion being completed within 2 ps, and a smaller component
Because of the high-lying HOMO energy, ground state Trp can relaxing over 10 ps ? >1 ns [43]. There is ongoing discussion con-
also act as a quencher of many fluorophores, e.g., flavins and dyes cerning whether protein, water, or both are responsible for the
[99–101]. When the HOMO of the dye lies substantially below that slower times [35,107,108].
of Trp, the vacancy in the HOMO of the excited dye may be filled by Trp–Trp FRET. When multiple Trps are present in a protein, there
an electron from the ground state Trp HOMO. (To a somewhat les- is a high probability of Trp–Trp FRET, given that differences in elec-
ser extent, tyrosine quenches dyes and flavins by the same mecha- trostatic environment will favor net unidirectional FRET to Trps
nism.) The final state of the Trp ring immediately following with more red-shifted fluorescence. An example is bacteriophage
electron transfer is the same as when excited Trp is quenched by T4 lysozyme, which has three Trps approximately at the corners
an electron acceptor (the indole ring is a ground state radical cat- of an equilateral triangle with 15 Å sides. Observing phosphores-
ion and the dye is a ground state radical anion). Solvation behavior cence and optically detected magnetic resonance at 4.2 K of the
following electron transfer is much the same as when excited Trp wild type and all single and double Trp ? Tyr mutants, Ghosh
donates an electron [82]. Because the HOMO of 5FW is lower than et al. [109] demonstrated that most of the excitation to W126 is
that of Trp, 5FW is not expected to quench dyes as efficiently as transferred to W158.
Trp.

Wavelength variation and Trp–Trp FRET Tyrosine and phenylalanine

Although focus has been on quenching in the above sections, Fluorescence from tyrosine (Tyr) and phenylalanine (Phe) is
there are cases for which binding induces quantifiable changes in also a valuable resource for investigating protein function, but
fluorescence or absorption spectra with little or no quenching. This has seen far less use than that of Trp. This is because these absorb
becomes especially useful when isosbestic points are associated at shorter wavelengths, have weaker absorptivity, have less wave-
with binding [102–104]. length sensitivity to environment, and transfer much of their exci-
As with electron transfer-induced quenching, electrostatics tation to Trp if present nearby. In addition, both are far more
plays a critical role in determining fluorescence emission shape abundant than Trp in most proteins.
and wavelength maximum (kmax) [30,84,105]. Numerous quantum Some useful miscellaneous facts are that the low quantum
mechanical studies on indoles have been summarized as unani- yields of Tyr and Phe in proteins is almost certainly due to quench-
mously agreeing that the transition from the ground to the 1La ing by electron transfer to the local backbone amides, just as for
state moves substantial electron density from the pyrrole ring to Trp [110]. An important fact associated with Tyr is that in certain
the benzene ring [26]. Therefore, positive charges near the benzene environments, or at high pH, tyrosinate may be present. This spe-
ring will contribute a shift of the absorption and fluorescence to cies absorbs and emits near where Trp does, and can have a broad
longer wavelengths (red shift) while positive charges near the pyr- red shifted emission similar to Trp.
role ring will contribute a blue shift. Just the opposite will hold for Nevertheless, for proteins with no Trp and a limited number of
negative charges. Phe and/or Tyr, circumstances exist for which Tyr and Phe fluores-

Please cite this article in press as: P.R. Callis, J. Mol. Struct. (2014), http://dx.doi.org/10.1016/j.molstruc.2014.04.051
P.R. Callis / Journal of Molecular Structure xxx (2014) xxx–xxx 7

cence can be exploited to obtain useful information. Some exam- Wavelength shift
ples have been reviewed recently [2]. Although Phe fluorescence
is very difficult to detect in the presence of Tyr and Trp, it becomes For back-of-envelope calculations, one may use the shift of elec-
quite useful for those proteins not containing Tyr and Trp [2]. One tron density for excitation to 1La as 0.25 electron from the pyrrole
important protein meeting this criterion is calmodulin, for which to benzene ring. Therefore if the average potential of the benzene
Shea and co-workers [111] have made useful studies of Ca2+ bind- ring is 1 V more positive than the pyrrole ring, the energy differ-
ing. These studies found a >50% decrease in the combined fluores- ence will be about 0.25 eV = 2000 cm 1, which is a 20 nm
cence of the 5 Phe residues upon binding one Ca2+. Given that Ca2+ red shift for a wave length of 320 nm. If the average potential
is not a quencher, this is a case in which the structural change and/ of the benzene ring is 1 V more negative than the pyrrole ring,
or electrostatic effect has enhanced the normal quenching by elec- the energy difference will be about +0.25 eV = +2000 cm 1, which
tron transfer to amides. is a 20 nm blue shift at 320 nm.

References
Conclusions
[1] M.R. Eftink, Methods Biochem. Anal. 35 (1991) 127–205.
When using tryptophan (Trp) as the fluorophore in a protein, [2] J.B.A. Ross, W.R. Laws, M.A. Shea, in: V. Uversky, E.A. Permyakov (Eds.),
Protein Structures: Methods in Protein Structure and Structure Analysis, Nova
one should keep in mind the dozen or so potential intrinsic
Science Publishers, Inc., New York, 2007, pp. 1–18.
quenchers, and the consequences that binding of an external [3] A.P. Demchenko, Ultraviolet Spectroscopy of Proteins, Springer-Verlag, New
quenching ligand may have on any pre-existing intrinsic quench- York, 1986.
[4] G. Weber, Biochem. J. 75 (1960) 335–345.
ing. Binding of a ligand near a Trp may enhance or reduce intrinsic
[5] S.V. Konev, Fluorescence and Phosphorescence of Proteins and Nucleic Acids,
quenching by a direct electrostatic effect or indirectly by changing Plenum, New York, 1967.
the local structure, water exposure, etc. This is mentioned primar- [6] J.R. Lakowicz, Principles of Fluorescence Spectroscopy, Springer, New York,
ily to help investigators avoid unwarranted conclusions regarding 2006.
[7] C.A. Royer, Chem. Rev. 106 (2006) 1769–1784.
the mechanism for the fluorescence change. For example, Ca2+ [8] E.A. Burstein, S.M. Abornev, Y.K. Reshetnyak, Biophys. J. 81 (2001) 1699–1709.
and ATP binding cause useful Trp fluorescence changes, but neither [9] Y.K. Reshetnyak, E.A. Burstein, Biophys. J. 81 (2001) 1710–1734.
can quench Trp fluorescence. The binding constant determined by [10] Y.K. Reshetnyak, Y. Koshevnik, E.A. Burstein, Biophys. J. 81 (2001) 1735–1758.
[11] S.R. Meech, A. Lee, D. Phillips, Chem. Phys. 80 (1983) 317–328.
the fluorescence change should not depend on the exact mecha- [12] J.M. Beechem, L. Brand, Ann. Rev. Biochem. 54 (1985) 43–71.
nism by which the fluorescence changes, however. [13] M.R. Eftink, L.A. Selvidge, P.R. Callis, A.A. Rehms, J. Phys. Chem. 94 (1990)
The effect of temperature upon quenching will always contain a 3469–3479.
[14] M.R. Eftink, Y. Jia, D. Hu, C.A. Ghiron, J. Phys. Chem. 99 (1995) 5713–5723.
large component from the temperature dependence of fluores- [15] Y. Chen, B. Liu, H.-T. Yu, M.D. Barkley, J. Am. Chem. Soc. 118 (1996) 9271–
cence from the indole ring itself. The quantum yield of fluorescence 9278.
from 3-methylindole in water drops rapidly with increasing tem- [16] Y. Chen, M.D. Barkley, Biochemistry 37 (1998) 9976–9982.
[17] C.P. Pan, P.L. Muino, M.D. Barkley, P.R. Callis, J. Phys. Chem. B 115 (2011)
perature by a mechanism that is still not understood. Care must
3245–3253.
be taken, therefore, when interpreting the meaning of the temper- [18] A.A. Rehms, P.R. Callis, Chem. Phys. Lett. 140 (1987) 83–89.
ature dependence of quenching by ligand binding. [19] D.M. Sammeth, S. Yan, L.H. Spangler, P.R. Callis, J. Phys. Chem. 94 (1990)
7340–7342.
[20] P.R. Callis, J. Chem. Phys. 95 (1991) 4230–4240.
[21] D.M. Sammeth, S.S. Siewert, L.H. Spangler, P.R. Callis, Chem. Phys. Lett. 193
Acknowledgements (1992) 532–538.
[22] L.S. Slater, P.R. Callis, J. Phys. Chem. 99 (1995) 8572–8581.
The author is extremely indebted to Drs. Mickey Schurr, Bruce [23] B.J. Fender, D.M. Sammeth, P.R. Callis, Chem. Phys. Lett. 239 (1995) 31–37.
[24] B.J. Fender, P.R. Callis, Chem. Phys. Lett. 262 (1996) 343–348.
Hudson, Dan Harris, Andy Albrecht, Tom Scott, Mary Barkley, Bob [25] D.K. Hahn, P.R. Callis, J. Phys. Chem. 101 (1997) 2686–2691.
Woody, Lennie Brand, Lee Spangler, Jaap Broos, Sandy Ross, Jay [26] P.R. Callis, Methods Enzymol. 278 (1997) 113–150.
Knutson, Franck Prendergast, and Dongping Zhong for helpful [27] K.W. Short, P.R. Callis, J. Chem. Phys. 108 (1998) 10189–10196.
[28] B.J. Fender, K.W. Short, D.K. Hahn, P.R. Callis, Int. J. Quantum Chem. (1999).
interactions.
[29] K.W. Short, P.R. Callis, J. Chem. Phys. 113 (2000) 5235–5244.
[30] J.T. Vivian, P.R. Callis, Biophys. J. 80 (2001) 2093–2109.
[31] P.R. Callis, T. Liu, J. Phys. Chem. B 108 (2004) 4248–4259.
Notes on electrostatics [32] T.Q. Liu, P.R. Callis, B.H. Hesp, M. de Groot, W.J. Buma, J. Broos, J. Am. Chem.
Soc. 127 (2005) 4104–4113.
[33] J.J. Chen, S.L. Flaugh, P.R. Callis, J. King, Biochemistry 45 (2006) 11552–11563.
Potential has the units of volts, i.e., joules/coulomb. An electron [34] J. Chen, P.R. Callis, J. King, Biochemistry 48 (2009) 3708–3716.
has lower energy in a more positive potential (closer to positive [35] J.N. Scott, P.R. Callis, J. Phys. Chem. B 117 (2013) 9598–9605.
[36] P.R. Callis, J.R. Tusell, Fluorescence spectroscopy and microscopy: methods
charge or farther from negative charge). From Coulomb’s Law,
and protocols, in: Y. Engelborghs, A.J.W.G. Visser (Eds.), Book Series Methods
one can show that at a point 1 Å distant from a proton, the poten- in Molecular Biology, 1076, Springer, Clifton, N.J., 2014, pp. 171–214.
tial is +14.40 V. Therefore the energy of an electron 1 Å from a pro- [37] R.F. Chen, J.R. Knutson, H. Ziffer, D. Porter, Biochemistry 30 (1991) 5184–
ton is 14.40 eV or 332 kcal/mol or 1389 kJ/mole. The electric 5195.
[38] X.H. Shen, J.R. Knutson, J. Phys. Chem. B 105 (2001) 6260–6265.
fields in proteins are on the order of a few times 107 V/cm, i.e., a [39] J.H. Xu, D. Toptygin, K.J. Graver, R.A. Albertini, R.S. Savtchenko, N.D. Meadow,
few tenths of V/Å [112,113]. The permanent dipole change during S. Roseman, P.R. Callis, L. Brand, J.R. Knutson, J. Am. Chem. Soc. 128 (2006)
excitation (not to be confused with transition dipole) to the 1La state 1214–1221.
[40] J.H. Xu, J.R. Knutson, J. Phys. Chem. B 113 (2009) 12084–12089.
of Trp is 5–10 Debyes, meaning that a field of 107 V/cm can cause a [41] J.H. Xu, J.J. Chen, D. Toptygin, O. Tcherkasskaya, P. Callis, J. King, L. Brand, J.R.
10 nm shift in the fluorescence. Knutson, J. Am. Chem. Soc. 131 (2009) 16751–16757.
[42] A. Biesso, J.H. Xu, P.L. Muiño, P.R. Callis, J.R. Knutson, J. Am. Chem. Soc. ASAP
(2014). dx.doi.org/10.1021/ja406126a.
[43] D. Toptygin, A.M. Gronenborn, L. Brand, J. Phys. Chem. B 110 (2006) 26292–
Electron transfer-induced quenching
26302.
[44] D. Toptygin, T.B. Woolf, L. Brand, J. Phys. Chem. B 114 (2010) 11323–11337.
For electron transfer, if the amide C atom of Trp is more positive [45] A. Ababou, E. Bombarda, Protein Sci. 10 (2001) 2102–2113.
than the indole ring by 1 V, the energy to create the CT state will be [46] J.W. Petrich, M.C. Chang, D.B. McDonald, G.R. Fleming, J. Am. Chem. Soc. 105
(1983) 3824–3832.
1 eV = 8000 cm 1 = 23 kcal/mol = 96.5 kJ/mol less than if there [47] J.W. Petrich, J.W. Longworth, G.R. Fleming, Biochemistry 26 (1987) 2711–
were no potential difference. 2734.

Please cite this article in press as: P.R. Callis, J. Mol. Struct. (2014), http://dx.doi.org/10.1016/j.molstruc.2014.04.051
8 P.R. Callis / Journal of Molecular Structure xxx (2014) xxx–xxx

[48] K.T. Jensen, G. Strambini, G. Gonnelli, J. Broos, J.B. Jackson, Biophys. J (2008). [81] L.C. Kurz, B. Fite, J. Jean, J. Park, T. Erpelding, P. Callis, Biochemistry 44 (2005)
[49] J. Peon, S.K. Pal, A.H. Zewail, Proc. Natl. Acad. Sci. USA 99 (2002) 10964– 1394–1413.
10969. [82] P.R. Callis, T.Q. Liu, Chem. Phys. 326 (2006) 230–239.
[50] S.K. Pal, J. Peon, A.H. Zewail, Proc. Natl. Acad. Sci. USA 99 (2002) 1763–1768. [83] P.R. Callis, A. Petrenko, P.L. Muino, J.R. Tusell, J. Phys. Chem. B 111 (2007)
[51] D.L. Harris, B.S. Hudson, Chem. Phys. 158 (1991) 353–382. 10335–10339.
[52] M. Van Gilst, B.S. Hudson, Biophys. Chem. 63 (1996) 17–25. [84] P.R. Callis, in: C.D. Geddes (Ed.), Reviews in Fluorescence 2007, vol. 4,
[53] H. Shizuka, M. Serizawa, H. Kobayashi, K. Kameta, H. Sugiyama, T. Matsuura, I. Springer, NY, 2009, pp. 199–248.
Saito, J. Am. Chem. Soc. 110 (1988) 1726–1732. [85] P.R. Callis, Methods Enzymol. (2011) 1–38.
[54] A.W. McMillan, B.L. Kier, I. Shu, A. Byrne, N.H. Andersen, W.W. Parson, J. Phys. [86] I. Swierszcz, P. Skurski, J. Simons, J. Phys. Chem. A 116 (2012) 1828–1837.
Chem. B 117 (2013) 1790–1809. [87] W.H. Qiu, Y.T. Kao, L.Y. Zhang, Y. Yang, L.J. Wang, W.E. Stites, D.P. Zhong, A.H.
[55] D.W. Pierce, S.G. Boxer, Biophys. J. 68 (1995) 1583–1591. Zewail, Proc. Natl. Acad. Sci. USA 103 (2006) 13979–13984.
[56] D.R. Demmer, G.W. Leach, S.C. Wallace, J. Phys. Chem. 98 (1994) 12834– [88] R. Vos, Y. Engelborghs, Photochem. Photobiol. 60 (1994) 24–32.
12843. [89] K. Willaert, R. Loewenthal, J. Sancho, M. Froeyen, A. Fersht, Y. Engelborghs,
[57] T.E.S. Dahms, K.J. Willis, A.G. Szabo, J. Am. Chem. Soc. 117 (1995) 2321–2326. Biochemistry 31 (1992) 711–716.
[58] P.R. Callis, Int. J. Quantum. Chem. S18 (1984) 579–588. [90] R.M. Hochstrasser, D.K. Negus, Proc. Natl. Acad. Sci. USA 81 (1984) 4399–
[59] K.W. Short, P.R. Callis, Chem. Phys. 283 (2002) 269–278. 4403.
[60] J. Broos, K. Tveen-Jensen, E. de Waal, B.H. Hesp, J.B. Jackson, G.W. Canters, P.R. [91] W.D. Horrocks, J.P. Bolender, W.D. Smith, R.M. Supkowski, J. Am. Chem. Soc.
Callis, Angew. Chem. Int. Ed. Engl. 46 (2007) 5137–5139. 119 (1997) 5972–5973.
[61] R.W. Cowgill, Arch. Biochm. Biophys. 100 (1963) 36–44. [92] R.M. Supkowski, J.P. Bolender, W.D. Smith, L.E.L. Reynolds, W.D. Horrocks,
[62] R.W. Cowgill, Biochim. Biophys. Acta 200 (1970) 18–25. Coord. Chem. Rev. 185 (1999) 307–319.
[63] J. Feitelson, Israel J. Chem. 8 (1970) 241–252. [93] R.F. Steiner, E.P. Kirby, J. Phys. Chem. 73 (1969) 4130–4135.
[64] P.L. Muino, P.R. Callis, J. Phys. Chem. B 113 (2009) 2572–2577. [94] M. Van Gilst, C. Tang, A. Roth, B. Hudson, J. Fluorescence 4 (1994) 203–207.
[65] R.W. Alston, M. Lasagna, G.R. Grimsley, J.M. Scholtz, G.D. Reinhart, C.N. Pace, [95] V. Nanda, S.M. Liang, L. Brand, Biochem. Biophys. Res. Commun. 279 (2000)
Biophys. J. 94 (2008) 2280–2287. 770–778.
[66] R.W. Alston, M. Lasagna, G.R. Grimsley, J.M. Scholtz, G.D. Reinhart, C.N. Pace, [96] V. Nanda, L. Brand, Proteins 40 (2000) 112–125.
Biophys. J. 94 (2008) 2288–2296. [97] V. Subramaniam, T.M. Jovin, R.V. Rivera-Pomar, J. Biol. Chem. 276 (2001)
[67] J. Broos, F. Maddalena, B.H. Hesp, J. Am. Chem. Soc. 126 (2004) 22–23. 21506–21511.
[68] J. Broos, Fluorescence spectroscopy and microscopy: methods and protocols, [98] T. Liu, P.R. Callis, Biophys. J. 88 (2005) 161A–162A.
in: Y. Engelborghs, A.J.W.G. Visser (Eds.), Book Series Methods in Molecular [99] A.C. Vaiana, H. Neuweiler, A. Schulz, J. Wolfrum, M. Sauer, J.C. Smith, J. Am.
Biology, 1076, Springer, Clifton, N.J., 2014, pp. 359–370. Chem. Soc. 125 (2003) 14564–14572.
[69] C.-Y. Wong, M.R. Eftink, Biochemistry 37 (1998) 8938–8946. [100] S. Doose, H. Neuweiler, M. Sauer, Chem. Phys. Chem. 6 (2005) 2277–2285.
[70] D.B. Bivin, S. Kubota, R. Pearlstein, M.F. Morales, Proc. Natl. Acad. Sci. USA 90 [101] S. Doose, H. Neuweiler, M. Sauer, Chem. Phys. Chem. 10 (2009) 1389–1398.
(1993) 6791–6795. [102] A.S.R. Koti, N. Periasamy, J. Chem. Phys. 115 (2001) 7094–7099.
[71] J.M. Francois, C. Gerday, F.G. Prendergast, J.D. Potter, J. Muscle Res. Cell. M. 14 [103] A.P. Demchenko, J. Fluorescence 20 (2010) 1099–1128.
(1993) 585–593. [104] A.P. Demchenko, in: A.P. Demchenko (Ed.), Springer Series on Fluorescence,
[72] M. Gonzalez-Bejar, E. Alarcon, H. Poblete, J.C. Scaiano, J. Perez-Prieto, vol. 8, Springer, New York, 2010, pp. 3–24.
Biomacromolecules 11 (2010) 2255–2260. [105] P.L. Muino, P.R. Callis, J. Chem. Phys. 100 (1994) 4093–4109.
[73] R.W. Ricci, J.M. Nesta, J. Phys. Chem. 80 (1976) 974–980. [106] L.Y. Zhang, Y. Yang, Y.T. Kao, L.J. Wang, D.P. Zhong, J. Am. Chem. Soc. 131
[74] L. Blancafort, D. Gonzalez, M. Olivucci, M.A. Robb, J. Am. Chem. Soc. 124 (2009) 10677–10691.
(2002) 6398–6406. [107] D.P. Zhong, S.K. Pal, A.H. Zewail, Chem. Phys., Lett. 503 (2011) 1–11.
[75] A. Sinicropi, R. Pogni, R. Basosi, M.A. Robb, G. Gramlich, W.M. Nau, M. [108] B. Halle, L. Nilsson, J. Phys. Chem. 113 (2009) 8210–8213.
Olivucci, Angew. Chem. Int. Ed. Engl. 40 (2001) 4185–4189. [109] S. Ghosh, L.H. Zang, A.H. Maki, J. Chem. Phys. 88 (1988) 2769–2775.
[76] G.R. Fleming, J.M. Morris, R.J. Robbins, G.J. Woolfe, P.J. Thistlethwaite, G.W. [110] M. Noronha, J.C. Lima, P. Lamosa, H. Santos, C. Maycock, R. Ventura, A.L.
Robinson, Proc. Natl. Acad. Sci. USA 75 (1978) 4652–4656. Macanita, J. Phys. Chem. A 108 (2004) 2155–2166.
[77] P.R. Callis, J.T. Vivian, Chem. Phys. Lett. 369 (2003) 409–414. [111] W.S. VanScyoc, M.A. Shea, Protein Sci. 10 (2001) 1758–1768.
[78] C.P. Pan, P.R. Callis, M.D. Barkley, J. Phys. Chem. B 110 (2006) 7009–7016. [112] P.R. Callis, B.K. Burgess, J. Phys. Chem. 101 (1997) 9429–9432.
[79] P.D. Adams, Y. Chen, K. Ma, M.G. Zagorski, F.D. Sonnichsen, M.L. McLaughlin, [113] C. Dedonder-Lardeux, C. Jouvet, S. Perun, A.L. Sobolewski, Phys. Chem. Chem.
M.D. Barkley, J. Am. Chem. Soc. 124 (2002) 9278–9286. Phys. 5 (2003) 5118–5126.
[80] T.P. Treynor, S.G. Boxer, J. Phys. Chem. B 108 (2004) 13513–13522.

Please cite this article in press as: P.R. Callis, J. Mol. Struct. (2014), http://dx.doi.org/10.1016/j.molstruc.2014.04.051

You might also like