Steric and electronic effects in capsule-confined green fluorescent protein chromophores

Article abstract of DOI:10.1021/ja1094606

The turn-on of emission in fluorescent protein chromophores sequestered in an "octaacid" capsule is controlled by stereoelectronic effects described by a linear free energy relationship. The stereochemical effects are governed by both the positions and volumes of the aryl substituents, while the electronic effects, including ortho effects, can be treated with Hammett σ parameters. The use of substituent volumes rather than A values reflects packing of the molecule within the confines of the capsule.

Full text of DOI:10.1021/ja1094606

COMMUNICATION
pubs.acs.org/JACS
Steric and Electronic Effects in Capsule-Confined Green Fluorescent
Protein Chromophores
†
‡
‡
,‡
Anthony Baldridge, Shampa R. Samanta, Nithyanandhan Jayaraj, V. Ramamurthy,* and
,
†
Laren M. Tolbert*
†
School of Chemistry and Biochemistry, Georgia Institute of Technology, 901 Atlantic Drive, Atlanta, Georgia 30332-0400,
United States
‡
Department of Chemistry, University of Miami, Coral Gables, Florida 33146, United States
S Supporting Information
b
ABSTRACT: The turn-on of emission in fluorescent
protein chromophores sequestered in an “octaacid” capsule
is controlled by stereoelectronic effects described by a linear
free energy relationship. The stereochemical effects are
governed by both the positions and volumes of the aryl
substituents, while the electronic effects, including ortho
effects, can be treated with Hammett σ parameters. The use
of substituent volumes rather than A values reflects packing
of the molecule within the confines of the capsule.
Figure 1. Torsional modes in excited states of BMIs.
1
guest/host ratio, no unbound guest could be observed by H
he way in which confinement affects fluorescence presents a
NMR spectroscopy, and additional fluorescence studies confirmed
7
T
remarkable set of challenges in supramolecular photochem-
istry. We have been investigating the sensitivity of the chromo-
phores derived from fluorescent proteins (FPs) to their environ-
ment through studies that mimic the effect of the sequestering
that the fluorescence was at or near saturation. Notably,
CH CH - as an ortho substituent did not lead to a stable complex.
3
2
We next considered the electronic and steric effects on the
excited-state emission using steady-state fluorescence measure-
1
-
5
β-barrel. Such chromophores are known to exhibit fast internal
ments. In each case, 10 M solutions of the BMI in benzene and
-
5
conversion through two torsional modes, φ and τ, that are
in a 4 ꢀ 10 M solution of OA in 10 mM borate buffer were
prepared and then excited at 349 nm. To eliminate the contribu-
tion from Raman scattering, the emission spectra of the solvents
alone were recorded and subtracted from the emission spectra of
the sample solutions. This correction was <5%.
inhibited in the protein (see Figure 1). This facile internal
4
conversion, which represents a 10 -fold decrease in fluorescence
intensity, makes this chromophore very unusual and provides a
unique opportunity to investigate details of excited-state decay
in confined environments. For instance, we have discovered
that sequestration of these benzylidenemethylimidazolidi-
Although excited-state decay rates can be obtained by single-
photon counting, for such a large number of samples, we appealed to
X
nones (BMIs) within a deep hydrophobic cavitand, the so-called
the fluorescence quantum yields Φ as a function of the substituent X
f
2
X
X
X
X
“
octaacid” (OA), mimics the β-barrel and turns on the fluores-
(R and R ). Since Φ = k /k , where k is the rate constant for
1 2 f f dt f
X X X
cence in a way that depends on the substitution at the ortho
fluorescence, k = k þ k is the total rate constant for decay, and
dt
f
nr
3
X
nr rel f f
X
X
H
position. In order to separate the influence of electronic and
k isthenonradiativedecayrateconstant,weobtainΦ =Φ /Φ =
X H X H H
steric effects on this phenomenon, we now report how the
fluorescence turn-on is affected by a number of substituted
chromophores that differ in both their steric volumes and
electronic effects.
(k /k )/(kdt/kdt), whereΦ refers to the quantum yield for H/CH
f f f 3
(entry 1 in Table 1). Making the somewhat less than rigorous
assumption that the oscillator strengths are not greatly affected
X
H
by weakly perturbing substituents, we conclude that k
≈ k
, which
f
f
X
H
X
The BMI chromophores were synthesized using previously
gives Φrel ≈ kdt/kdt. Strictly speaking, we are interested in the ratio
4
H
X
X
described methods (see R and R in Table 1) and then exposed
of nonradiative decay rates, knr/knr, but for relatively low Φ we
f
1
2
to buffered solutions of OA in D O. In all cases, complexation of
the chromophores was validated by the upfield H NMR chemical
can assume that total decay is dominated by internal conversion. The
quantum yield was obtained by integrating the area under the
emission curve from 360 to 650 nm.
Substituent effects are classically a function of electronic
and steric effects, as represented by a linear free energy
2
5
1
shifts of the guests relative to the values in CD CN solvent,
3
corresponding to placement of protons within the shielding region
of the aromatic cavity [see the Supporting Information (SI)]. As
3,6
seen in our earlier studies, the alkyl derivatives experienced
strong shielding of both aryl-ring and N-alkyl groups within the
cavity (see the SI). At the concentrations used here, with a 1:4
Received: October 20, 2010
Published: December 21, 2010
r 2010 American Chemical Society
712
dx.doi.org/10.1021/ja1094606
|J. Am. Chem. Soc. 2011, 133, 712–715

Journal of the American Chemical Society
COMMUNICATION
Table 1. Linear Free Energy and Fluorescence Parameters for
Various BMIs Included in the OA Cavitand
3
a
ΔV (Å )
10-
3
Φ
OA
b
log(Φrel
X
)
c
fit
d
R
1
/R
2
σ
R
1
R
2
H/Me
o-F/Me
0.00
0.29
0.34
0.06
0.0
4.6
4.6
4.6
0.0
0.0
3.1
69
0.0
1.3
0.0
0.6
m-F/Me
p-F/Me
0.0
130
2.2
78
1.6
1.0
0.0
-0.2
1.4
0.1
o-Cl/Me
m-Cl/Me
o-Br/Me
m-Br/Me
p-Br/Me
0.50 13.8
0.37 13.8
0.55 18.2
0.39 18.2
0.23 18.2
0.0
1.1
0.0
19
0.8
1.0
0.0
140
13
1.7
1.3
0.0
0.6
1.0
0.0
15
0.7
0.5
o-Me/Me -0.13 18.3
m-Me/Me -0.07 18.3
p-Me/Me -0.17 18.3
o-OMe/Me -0.37 27.5
0.0
1.2
1.0
1.5
2.2
3.7
-0.4
-0.4
-0.3
-0.2
0.1
0.0
0.0
-0.4
-0.3
-0.3
0.1
Figure 2. Fit of eq 2 to the experimental data.
0.0
0.0
where β is the proportionality constant for the steric constant
of substituent 2. Steric constants have been represented in a
m-OMe/Me
0.12 27.5
0.0
8
variety of ways, including the Charton ν constants and A
p-OMe/Me -0.27 27.5
0.0
0.93
440
75
-0.5
2.2
-0.5
2.0
9
values. However, these generally capture only the steric hindrance
o-CF
3
/Me
/Me
0.81 32.0
0.43 32.0
0.0
of a substituent and not its total volume, which may be
m-CF
3
0.0
1.4
1.1
1
0
inappropriate for the effect within a confined space. For instance,
a methoxy group has a larger volume but a smaller A value than a
methyl group because the nearest-neighbor atom is smaller.
Instead, we deemed it more appropriate to represent χ and χ
by the density functional theory (DFT)-calculated changes in
molecular volume, ΔV, upon replacement of hydrogen with the
H/Pr
0.00
0.00
37.0
18.5
37.0
55.2
73.8
37.0
37.0
37.0
37.0
37.0
37.0
37.0
37.0
37.0
37.0
37.0
37.0
37.0
37.0
15
0.7
0.4
o-Me/Et -0.13 18.3
o-Me/Pr -0.13 18.3
o-Me/Bu -0.13 18.3
o-Me/Pn -0.13 18.3
m-Me/Pr -0.07 18.3
p-Me/Pr -0.17 18.3
3.0
9.7
4.8
2.0
1.8
2.1
220
7.2
160
20
0.0
0.2
0.5
0.4
1
2
0.2
0.6
-0.2
-0.2
-0.2
1.8
0.8
-0.2
-0.2
1.2
substituent on benzene in the case of R and the methyl group in
1
toluene by the appropriate alkyl group in the case of R2,
respectively (Table 1). Thus ΔV for CH in the case of R is
m-F/Pr
p-F/Pr
0.34
0.06
4.6
4.6
3
1
3
3
1
8.3 Å and in the case of R , 0.0 Å .
0.4
0.2
2
Use of the Hammett equation for excited-state processes has a
somewhat checkered history, and a number of substituent
constants have been proposed for photochemical processes.
As is generally the case for LFERs, we appeal to the success of the
method at reproducing the data. Moreover, while LFERs are
well-developed for meta and para substituents, the use of σ values
for ortho substituents is fraught with problems, almost all of
o-Cl/Pr
m-Cl/Pr
o-Br/Pr
m-Br/Pr
p-Br/Pr
0.50 13.8
0.37 13.8
0.55 18.2
0.39 18.2
0.23 18.2
1.7
1.5
0.8
1.2
1
1
180
7.5
29
1.8
1.7
0.4
1.2
1.0
0.6
o-OMe/Pr -0.37 27.5
m-OMe/Pr 0.12 27.5
p-OMe/Pr -0.27 27.5
8.8
11
0.5
0.1
1
2
0.5
0.3
which have to do with dissecting electronic versus steric effects.
Thus, we adopted the ortho σ values developed by Tribble and
1.2
150
170
-0.4
1.7
-0.4
2.4
13
o-CF
3
/Pr
/Pr
0.81 32.0
0.43 32.0
Traynham, which are based upon the apparently sterically
independent chemical shifts of phenols.
m-CF
3
1.7
1.2
14
a
c
b
Using benzene solvent as a proxy for the aromatic OA interior,
Determined by DFT calculations. Emission quantum yield in OA.
d
we performed a linear regression on eq 2 for each of the ortho,
meta, and para substituents, obtaining the correlation (open
circles) shown in Figure 2. In both benzene and OA, the CF3
group did not correlate well and was neglected from the analysis.
We now consider the interplay of the steric and electronic factors.
Relative emission quantum yield in OA with respect to H/Me. Fit to
X
log(Φrel) in eq 2 for OA.
relationship (LFER) (eq 1):
!
P
P
2
2
H
Using the sums of the correlation factors, (σ F ) , (R χ ) ,
k
k
P
i
i i
i
i 1i
X
nr
2
logðΦ Þ ꢁ log
¼ σ F þ R χ
ð1Þ
and (β χ ) , we can estimate the percent contribution to the
rel
X
3
3
i
i 2i
nr
variance from each factor (see Table 2). First, we consider the
substituent effects in benzene. We note that the electronic effect
of the substituent, σ, dominates the decay, and the effects of the
ortho and para substituents are similar, consistent with the usual
ortho/para formalism. Second, the effects of the substituents are
almost entirely electronic. The one exception, of course, is the
ortho substituent, where 12% of the variance is due to the steric
effect. Again, this is not surprising given the difficulty in dissect-
ing steric from electronic effects in ortho substituents. Third, the
Here σ is the Hammett substituent constant, F is the reaction
constant, χ is a steric constant, and R is the related propor-
tionality constant. In the case of two substituents, eq 1 can be
modified as
!
H
k
X
nr
logðΦ Þ ꢁ log
¼ σ F þ R χ þ β χ
ð2Þ
rel
X
3
3
1
3
2
k
nr
7
13
dx.doi.org/10.1021/ja1094606 |J. Am. Chem. Soc. 2011, 133, 712–715

Journal of the American Chemical Society
COMMUNICATION
Table 2. Relative Contributions of Electronic and Volumetric Factors
F
R (R
1
)
2
β (R )
substituent
value
% contrib.
value
% contrib.
value
% contrib.
in benzene
0.010 ( 0.003
0.001 ( 0.001
-0.001 ( 0.001
ortho
meta
para
1.3 ( 0.1
2.1 ( 0.2
1.8 ( 0.1
88
99
99
12
-0.001 ( 0.002
0.001 ( 0.002
<1%
<1%
<1%
<1%
<1%
-0.001 ( 0.001
in OA
ortho
meta
para
2.2 ( 0.3
3.2 ( 0.7
2.6 ( 0.3
82
87
93
0.018 ( 0.007
-0.009 ( 0.011
0.005 ( 0.003
10
12
5
0.013 ( 0.005
0.005 ( 0.008
0.005 ( 0.004
8
2
2
3
impact. Similarly, the ethyl group, at 36.8 Å , prevents encapsula-
3
tion by OA, while the CF group, at 32.0 Å , barely allows it when
3
present in the ortho position.
Finally, we consider the nature of the Hammett F value in the
capsule. With some variation for the ortho, meta, and para substituents,
which presumably results from the use of a different set of σ values in
each case, we note that F(ΟΑ) > F(PhH) in all cases. Thus, internal
conversion within OA must require more charge transfer back to the
ground state, resulting in the observed higher values for F within OA
than in benzene solution. This may reflect a larger amount of twist
within the capsule than in benzene. Thus, the role of the host molecule
in rigidifying the chromophore is clearly demonstrated.
In other guest molecules, including biomolecules, we have
observed rigidification of FP chromophores that is dependent
15
upon structure in analogous ways. We believe that such topological
control of fluorescence represents a diverse method for creating
molecular probes.
X
Figure 3. Effect of the N-alkyl substituent on log(Φrel).
N-alkyl substituent (R ) has almost no effect, in agreement with
’ ASSOCIATED CONTENT
2
little steric perturbation of the transition state. Finally, the F value
is positive with magnitude >1, indicating substantial negative
charge transfer to the aryl ring in the transition state for decay.
Since the excited state (represented by a diradical in Figure 1)
is known to have significant negative charge transfer to the
imidazolidinone ring, it is not surprising that internal conversion
would reflect that reverse charge transfer.
S
Supporting Information. Experimental information, syn-
b
thetic details, characterization, and additional spectroscopic infor-
mation. This material is available free of charge via the Internet at
http://pubs.acs.org.
’
AUTHOR INFORMATION
We now consider the effect of encapsulation. Unlike the
results for benzene, the remote N-alkyl substituent has a sig-
nificant effect (see Table 2), approaching that of the ortho
substituent and decreasing in the order ortho > meta > para, as
Corresponding Authors
murthy1@miami.edu; tolbert@chemistry.gatech.edu
3
suggested by our previous results. Curiously, as the volume
’ ACKNOWLEDGMENT
increases in the series R = Me, Et, Pr, Bu, and Pn (pentyl) while
R1 = Me, we see a linear increase and then a decrease in the
fluorescence enhancement (see Figure 3), presumably reflecting
2
We thank the National Science Foundation (CHE-0809179
and CHE-0848017) for financial support. A.B. acknowledges a
fellowship from the Center for Organic Photonics and Electro-
nics (COPE) at Georgia Tech.
the effect of locking and then overcrowding the capsule. Par-
enthetically, we note that the Charton ν constants and the A
8
values for Pr, Bu, and Pn are nearly identical, validating our use
’
REFERENCES
of molecular volume. For this reason, we restrict our considera-
tion to Me and Pr as the N-alkyl groups but maintain the linearity
with volume as the contribution to the LFE.
(
1) Enoki, S.; Saeki, K.; Maki, K.; Kuwajima, K. Biochemistry 2004,
3, 14238–14248.
2) (a) Gibb, C. L. D.; Gibb, B. C. J. Am. Chem. Soc. 2004, 126,
1408–11409. (b) Jayaraj, N.; Zhao, Y.; Parthasarathy, A.; Porel, M.; Liu,
4
(
Comparing the steric effects for ortho, meta, and para, we see a
decreasing steric effect for the phenyl substituent and a more
dramatic decrease for the N-alkyl substituent. The “locking”
1
R. S. H.; Ramamurthy, V. Langmuir 2009, 25, 10575–10586.
(3) Baldridge, A.; Samanta, S. R.; Jayaraj, N.; Ramamurthy, V.;
3
effect we suggested earlier is seen more clearly, with the ortho
Tolbert, L. M. J. Am. Chem. Soc. 2010, 132, 1498–1499.
(4) Lerestif, J. M.; Perrocheau, J.; Tonnard, F.; Bazureau, J. P.;
Hamelin, J. Tetrahedron 1995, 51, 6757–6774.
and meta substituents having a major steric effect and the para
substituent, which cannot prevent aryl rotation, having less
7
14
dx.doi.org/10.1021/ja1094606 |J. Am. Chem. Soc. 2011, 133, 712–715

Journal of the American Chemical Society
COMMUNICATION
(
5) Unfortunately, use of the more hydrophilic p-OH group (R
Me) precluded complexation.
6) Parthasarathy, A.; Kannumalle, L. S.; Ramamurthy, V. Org. Lett.
007, 9, 5059–5062.
7) Further details of the complexation behavior, including associa-
tion constants, will be reported in a forthcoming paper.
8) (a) Charton, M. J. Am. Chem. Soc. 1975, 97, 1552–1556. (b)
Charton, M. J. Org. Chem. 1976, 41, 2217–2220.
9) Eliel, E. L.; Wilen, S. H. Stereochemistry of Organic Compounds;
Wiley: New York, 1993; p 696.
2
=
(
2
(
(
(
(
10) Turro, N. J. Acc. Chem. Res. 2000, 33, 637–646.
(11) Cordes, T.; Schadendorf, T.; Priewisch, B.; Rueck-Braun, K.;
Zinth, W. J. Phys. Chem. A 2008, 112, 581–588.
12) For a recent discussion, see: Gustafson, J. L.; Sigman, M. S.;
Miller, S. J. Org. Lett. 2010, 12, 2794–2797.
13) Tribble, M. T.; Traynham, J. G. J. Am. Chem. Soc. 1969, 91,
79–388.
14) Porel, M.; Jayaraj, N.; Kannumalle, L. S.; Maddipatla, M. V. S.
N.; Parthasarathy, A.; Ramamurthy, V. Langmuir 2009, 25, 3473–3481.
15) Lee, J.-S.; Baldridge, A.; Feng, S.; SiQiang, Y.; Kim, Y. K.; Tolbert,
L. M.; Chang, Y.-T. ACS Combi. Chem. DOI: 10.1021/co100012k.
(
(
3
(
(
’
NOTE ADDED AFTER ASAP PUBLICATION
The Supporting Information PDF file published ASAP December
1, 2010, was an early version with data that did not match the data
2
in the published Communication. The final version Supporting
Information was published December 28, 2010.
7
15
dx.doi.org/10.1021/ja1094606 |J. Am. Chem. Soc. 2011, 133, 712–715