A novel fluorescence sensing system using a photochromism-based assay (P-CHROBA) technique for the detection of target proteins

Article abstract of DOI:10.1039/b501877n

In the post-genomic era a number of biological technologies, including protein-detecting microarrays which can detect molecular interactions based on changes in fluorescence intensity, have been developed to investigate complicated protein functions and n

Full text of DOI:10.1039/b501877n

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/materials | Journal of Materials Chemistry
A novel fluorescence sensing system using a photochromism-based assay
(P-CHROBA) technique for the detection of target proteins
Kin-ya Tomizaki and Hisakazu Mihara*
Received 4th February 2005, Accepted 17th March 2005
First published as an Advance Article on the web 6th April 2005
DOI: 10.1039/b501877n
In the post-genomic era a number of biological technologies, including protein-detecting
microarrays which can detect molecular interactions based on changes in fluorescence intensity,
have been developed to investigate complicated protein functions and networks. However, the
ability of such techniques to obtain reproducible and quantitative results can be compromised
due to the need for the immobilization of capture agents and the labelling analytes with
chromophores. In the present study, first we report the design and synthesis of photochromic
spiropyran-containing peptides and then demonstrate a unique fluorescence sensing system
comprising a photochromism-based assay (P-CHROBA) technique to distinguish between target
proteins. The spiropyran moiety in the peptides exhibited characteristic physicochemical
properties in the SP-to-MC isomerization (thermocoloration) and the MC-to-SP
photoisomerization (photobleaching) depending upon changes in micro-environments such as
the dielectric constants of solvents and steric hindrances generated by molecular interactions.
We attempted to detect protein–peptide interactions using reproducible MC-to-SP
photoisomerization properties by monitoring the fluorescence decay of the MC form in the
peptide. This can reduce background fluorescence signals caused by emission from excess reagents
and avoid the laborious introduction of probing molecules to analytes and the immobilization of
capture agents onto solid surfaces. The protein fingerprints (PFPs) based on the
photoisomerization properties could successfully distinguish between six different model proteins,
and the combination of the P-CHROBA and PFP technique would be a powerful tool for
profiling target proteins with reproducible and reliable results.
However, methods for the production of highly potent anti-
Introduction
bodies must be developed because a number of antibodies
The complete genome sequencing of various organisms has
corresponding to the target proteins is required. We have
allowed us to obtain a large amount of valuable information
previously designed and synthesized fluorophore-containing
peptide libraries with secondary structures such as the a-helix,20
b-strand,21 and b-loop.22 We have demonstrated that they can
for understanding complex cellular events. Consequently, a
number of biological technologies including oligonucleotide
arrays and protein-detecting microarrays have been emerging
be used to distinguish between proteins of interest using protein
with increasing frequency. The oligonucleotide arrays, which
fingerprints (PFPs) which are based on changes in the
provide important information on changes in mRNA expres-
fluorescence intensity characteristic of complexes formed by
sion levels in response to a variety of physiological stimuli, are
the protein–peptide interactions. Our PFP technique does not
now commercially available and are a promising tool for
require the antibodies and/or the analytes to be labelled with a
analyzing gene expression.1 Protein-detecting microarrays,
chromophore because the dyes have already been site-
which can detect antigen–antibody, protein–protein, protein–
ligand, protein–saccharide, and protein–small molecule inter-
actions, will become a powerful biological research tool in the
near future.2 However, the protein-detecting microarrays
need to meet several requirements: ease of manipulation; a
competitive price; highly sensitive labelling reagents, and for
an efficient detection method to be developed for practical
protein-detecting chips. From this point of view, dozens of
exciting reviews on protein-detecting microarrays have been
reported in past few years.3–17 Today, antibodies are the most
promising capture agents for detecting target proteins and
for antibody chips, where antibodies are immobilized onto
glass plates, and have just been released commercially.18,19
specifically incorporated into the peptide chains by a synthetic
chemical method. Although a protein-detection system is
currently available, it requires improvements such as
a
reduction in the relatively high background signals resulting
from modification of the capture agents with chromophores.
One means of improving the PFP technique is to alter the
wavelengths of the fluorescence maxima of the protein–peptide
interactions. Another way is to use measurements which are
independent of the background fluorescence signals from excess
reagents. Thus, desirable labelling groups of this kind could be
exploited to probe the surfaces of proteins of interest and to
make the emerging PFP technique more broadly applicable.
Spiropyrans, well known photochromic compounds, have
been vigorously studied in the field of molecular switches and
memories.23 Each spiropyran derivative has an equilibrium
*hmihara@bio.titech.ac.jp
2732 | J. Mater. Chem., 2005, 15, 2732–2740
This journal is ß The Royal Society of Chemistry 2005

View Article Online
interactions. In order to extend the applicability of this tech-
nique, we report herein the first example of a photochromism-
based assay (P-CHROBA) technique for profiling proteins of
interest in which, the MC-to-SP photoisomerization rate
constants that are characteristic of changes in the micro-
environments surrounding a spiropyran moiety in peptides are
utilized as a factor to distinguish between proteins in complexes.
Fig. 1 Photochromic properties of spiropyran derivatives in common
organic solvents. The spiropyran (SP) form can be converted to the
merocyanine (MC) form by UV light irradiation with a photoisome-
rization rate constant, kSP-to-MC, also the MC form can be isomerized
by darkness, or by visible light irradiation with a rate constant,
Results and discussion
Design and synthesis of spiropyran-containing peptides
kMC-to-SP
.
The rationale behind the design of a novel protein-detecting
system using spiropyran-containing peptides is based on the
photophysical properties of a spiropyran moiety varying
between a colorless ‘‘closed’’ SP-form and a highly colored
‘‘open’’ MC-form, these transitions being controllable by UV
and visible light irradiation (and sometimes incubation in the
dark). The amino acid sequences of peptides were chosen from
the substrates of four different kinases [protein kinase A
(PKA),30 c-Src kinase,31 c-Abl tyrosine kinase,31 and protein
kinase C (PKC)32] having a relatively similar charge balance
but different arrangement of amino acid components (Fig. 2).
The spiropyran derivative was attached to the side chain
of lysine (for K-PKA-SP (1), K-cSrc-SP (6), K-cAbl-SP (7),
and K-PKC-SP (8)), ornithine (for O-PKA-SP (2)),
L-a,c-diaminobutyric acid (Dab) (for Db-PKA-SP (3)), and
L-a,b-diaminopropionic acid (Dap) (for Dp-PKA-SP (4)), all
of which were placed at the N-terminus of the interaction site
through a glycine spacer. An S-protected cysteine residue
was attached to the N-terminus of the sequence for future
conjugation with another functional group. All peptides were
capped with an acetyl (Ac) group at their N-termini, except
Nt-PKA-SP (5) in which the N-terminus was modified with the
spiropyran derivative.
constant between that of the fluorescent merocyanine (MC)
form (colored in pink) and the non-fluorescent colorless
spiropyran (SP) form characteristic of the dielectric constant
of solvents.24 The chemical structure of spiropyrans in
common organic solvents can be controlled by UV and visible
light irradiation or UV light and incubation in the dark
(Fig. 1).25 Using such unique photochromic properties,
spiropyrans have been employed not only in materials
chemistry but also in bio-organic chemistry, especially as an
analytical tool for the transport of amino acid derivatives
across bilayers or membranes,26 or as capture agents for amino
acids on the surface of gold nanoparticles27 indicating that the
photomerocyanine form of the spiropyran derivatives is able to
recognize appropriately charged molecules via electrostatic
interactions. The MC-to-SP isomerization properties of
spiropyran derivatives are also affected by the sterically-
hindered environments present in organogels.23b,28
Recently, we have successfully demonstrated a chromism-
based assay (CHROBA) technique for the in situ detection of
the protein kinase A (PKA)-catalyzed phosphorylation, based
on changes in the thermocoloration from the colorless
SP-form to the fluorescent MC-form, depending upon the
phosphorylation of a serine side chain in the spiropyran-
containing peptide substrate.29 This technique offers a system
that reduces the influence of background signals from
reagents, does not require the immobilization of the peptide
substrates, isotope-labeled ATP and specific anti-phos-
phoamino acid antibodies, and thus facilitates production
and manipulation. However, the method based on the chromic
properties of spiropyrans appears to be an early stage in the
evolution of a promising technology for detecting molecular
The eight spiropyran-containing peptides were prepared via
three different synthetic routes (Scheme 1). In routes 1 and 3,
the spiropyran moiety is incorporated into the side chain or
N-terminus of the full length peptide on the resin after
completion of the peptide chain elongation by solid phase
synthesis. In route 2, spiropyran-tethered Fmoc-amino acids
are prepared in advance and then employed in the common
solid phase peptide synthesis. First, the known spiropyran
derivative, 2-(39,39-dimethyl-6-nitro-39H-spiro[chromene-2,29-
indole]-19-yl)ethanol (9) was prepared from commercially
Fig. 2 Amino acid sequences of spiropyran-containing peptides comprosed of three parts: a modification site, a spiropyran moiety, and an
interaction site. ‘SP’ denotes the spiropyran moiety.
This journal is ß The Royal Society of Chemistry 2005
J. Mater. Chem., 2005, 15, 2732–2740 | 2733

View Article Online
available 2,3,3-trimethyl-3H-indole via three steps resulting
in y50% overall yield.25 Compound 9 was treated with
p-nitrophenyl chloroformate, giving an activated spiropyran
(10) with 56% yield. In route 1 for preparation of K-PKA-SP
(1), K-cSrc-SP (6), K-cAbl-SP (7), and K-PKC-SP (8), after
completion of the peptide chain elongation by solid phase
synthesis using Fmoc chemistry,33 the 4-methyltrityl (Mtt)
group protecting the lysine side chain was removed with
trifluoroacetic acid (TFA)/triisopropylsilane (TIS)/CH2Cl2
(1/5/94, v/v/v). Compound 10 was reacted with the generated
amino group in the peptides. In route 2 for preparation of
O-PKA-SP (2), Db-PKA-SP (3), and Dp-PKA-SP (4), the
Fmoc-L-a-amino acids tethering a spiropyran moiety to the
v-amino groups were synthesized by coupling 10 with an
v-free amino group in Fmoc-L-a-amino acids, giving com-
pounds 11, 12, and 13 with 40, 68, and 61% yield, respectively.
The Fmoc-L-a-amino acids tethering the spiropyran moiety
were used in the usual solid phase peptide synthesis by the
Fmoc strategy.33 In route 3 for Nt-PKA-SP (5), the spiropyran
derivative was connected to the N-terminus of the sequence on
the resin. The fully-protected peptide-bound resins were
treated with TFA/m-cresol/ethanedithiol (EDT)/thioanisole
(TA) (40/1/3/3, v/v/v/v) at room temperature and the crude
peptides obtained were purified by reversed phase high
performance liquid chromatography (RP-HPLC), followed
by lyophilization, giving pure compounds as a yellow powder
with 28% yield for K-PKA-SP (1), 3.6% for O-PKA-SP (2),
3.8% for Db-PKA-SP (3), 4.0% for Dp-PKA-SP (4), 7.3% for
Nt-PKA-SP (5), 19% for K-cSrc-SP (6), 25% for K-cAbl-SP
(7), and 18% for K-PKC-SP (8). Molecular ion peaks for all
peptides were observed by matrix-assisted laser desorption/
ionization time-of-flight mass spectrometry (MALDI-TOFMS).
Yields of some peptides, especially O-PKA-SP (2), Db-PKA-SP
(3), and Dp-PKA-SP (4) synthesized via route 2 were signifi-
cantly lower than for others, probably due to incomplete
introduction of spiropyran-tethered Fmoc-amino acids and/or
side reactions of the spiropyran moiety during the peptide chain
elongation. Therefore, route 1 would be a better way to synthe-
size spiropyran-containing peptides by the solid phase technique.
SP-to-MC thermocoloration properties of spiropyran-containing
peptides
First, we investigated photochromic properties of spiropyran-
containing peptides in neutral aqueous media, then employed
them to profile proteins of interest using the MC-to-SP photo-
isomerization rate constants as a probing factor.
When the yellow powder, colored by a co-existing proto-
nated MC-form, was dissolved in a neutral aqueous solution
the yellow solution immediately turned pink due to deproto-
nation of the phenolic proton of the MC moiety, resulting in
the fluorescent zwitterion MC-form. Fig. 3A shows the
absorption and fluorescence spectra of K-PKA-SP (1) in
aqueous buffered solution acquired immediately after pre-
paration of the sample solution. An intense visible absorption
band at 510 nm and fluorescence emission at 600 nm
(lex 5 510 nm) were observed in the absorption and
fluorescence spectra, respectively, indicating a mixture of the
SP and MC forms in the spiropyran moiety. The absorption
Scheme 1 Synthesis of spiropyran-containing peptides via three
different routes.
2734 | J. Mater. Chem., 2005, 15, 2732–2740
This journal is ß The Royal Society of Chemistry 2005

View Article Online
with an elapsed time of 0 h of thermocoloration in the dark
and corresponding to the SP-form of the spiropyran moiety in
the peptide. When the incubation was performed continuously
the peak area of the SP-form after 19.0 min retention time
decreased and a new peak at 14.9 min appeared corresponding
to the pink sample solution. Each spiropyran-containing
peptide showed characteristic increases in HPLC peak area
with progression of the relaxation which depended on micro-
environments of the spiropyran moiety in the peptide.
Prolonged incubation (t . 200 h at 4 uC) caused decomposi-
tion of the spiropyran moiety in the peptides. Therefore, to
obtain kinetic parameters the HPLC data were acquired within
a 200 h time frame.
Fig. 3 Absorption (solid lines) and fluorescence (dashed lines, lex
5
510 nm) spectra of K-PKA-SP (1) taken immediately after dissolving
the yellow powder in 20 mM Tris HCl buffer (pH 7.4) (A) and after
completion of photoisomerization by irradiation with 510 nm light for
30 min at 25 uC (B).
The kinetic parameters and equilibrium constants are
summarized in Table 1. The SP-to-MC isomerization (thermo-
coloration) rate constants (kSP-to-MC) vary between 2.3 and 6.7,
and the MC-to-SP photoisomerization (photobleaching) rate
constants (kMC-to-SP) vary between 0.51 and 4.4, resulting in
the equilibrium constants (Keq) being in the range 0.52–6.1.
These findings strongly support preferable stabilization of the
zwitterion MC-forms in the peptides by solvation with bulk
water molecules, as explained above. Nt-PKA-SP (5), in
particular, showed the greatest Keq constant at 6.1 (53.1/0.51),
probably due to the loss of mobility of the spiropyran moiety
at the N-terminus of the peptide being different from that at
side chains in the peptides. On the other hand, K-cSrc-SP (6)
showed a reversed kinetic property in that the kMC-to-SP was
greater than the kSP-to-MC, resulting in a smaller Keq constant,
at 0.52, implying destabilization of the MC-form in the
peptide. Although the net charge of the K-cSrc-SP (6) peptide
sequence is +2, similar to those of the X-PKA-SP series (X 5 K,
O, Db, Dp, and Nt, (1–5)), mutual repulsion of negative
charges between the MC-form and the glutamic acid side chain
in the sequence might affect the stabilization of the MC-form
in K-cSrc-SP (6). These results suggest that the physicochemial
properties in the SP-to-MC thermocoloration of spiropyran
derivatives in aqueous media are strongly correlated with the
micro-environments surrounding the spiropyran moieties in
the peptides. Solution of the three dimensional structures of
and fluorescence spectra of the SP-form in the K-PKA-SP (1)
solution taken after irradiation with 510 nm light for 30 min
at 25 uC are shown in Fig. 3B. Both the intense absorption
band at 510 nm and the fluorescence emission at 600 nm
disappeared on completion of the MC-to-SP photoisomeriza-
tion. When the solution containing predominantly the SP-form
was incubated in the dark at room temperature for several
hours it turned pink again, indicating that the chemical
structure of the spiropyran moiety in K-PKA-SP (1) is
reversibly controlled by visible light for the MC-to-SP
photoisomerization (photobleaching) and by incubation in
the dark for the SP-to-MC isomerization (thermocoloration).
Other peptides also exhibited a quite similar photochromism
involving the photobleaching and thermocoloration in neutral
aqueous media (data not shown).
In order to obtain the kinetic parameters of spiropyran-
containing peptides in neutral aqueous solution, relaxation
(thermocoloration) processes from an SP-form to an equili-
brium state (a mixture of SP- and MC-forms) in all peptides
were followed by RP-HPLC to follow the progression of ring-
opening reactions in the dark. Fig. 4 shows HPLC profiles of
K-PKA-SP (1) incubated at 4 uC in the dark after exposure to
ambient laboratory light to give the SP-form. A single peak
was observed after 19 min in the chromatogram associated
Table 1 Physicochemical parameters in the photochromic properties
of spiropyran-containing peptidesa
103kSP-to-MC
/
103kMC-to-SP
/
Net
Keq charge
Peptide
h21
h21
K-PKA-SP (1)
O-PKA-SP (2)
Db-PKA-SP (3) 6.7
Dp-PKA-SP (4) 4.8
Nt-PKA-SP (5) 3.1
K-cSrc-SP (6)
K-cAbl-SP (7)
K-PKC-SP (8)
a
3.5
5.6
1.3
2.2
2.9
1.6
0.51
4.4
1.6
2.6
2.7
2.5
2.3
3.0
6.1
+/+ (+2)
+/+ (+2)
+/+ (+2)
+/+ (+2)
+/+ (+2)
2.3
4.6
3.9
0.52 2/+/+/+ (+2)
2.9
1.5
+/+/+ (+3)
+/+/+/+ (+4)
Physicochemical parameters were obtained from HPLC peak areas
at retention times corresponding to the MC- and SP-forms in
K-PKA-SP (1) (14.9 min/19.0 min for MC-/SP-forms); O-PKA-SP
(2) (14.7/18.6); Db-PKA-SP (3) (14.6/18.4); Dp-PKA-SP (4) (14.7/
18.4); Nt-PKA-SP (5) (14.7/18.8); K-cSrc-SP (6) (14.9/18.6); K-cAbl-
SP (7) (14.5/18.4); K-PKC-SP (8) (13.6/17.4) under incubation
conditions in 20 mM Tris HCl buffer (pH 7.4), 150 mM NaCl at
4 uC in the dark. The analytical procedure is described in the
experimental section.
Fig. 4 HPLC profiles of K-PKA-SP (1) during the relaxation process
from the SP- (grey line) to MC- (black line) forms acquired 0, 24, 96,
and 198 h after starting incubation at 4 uC in the dark.
This journal is ß The Royal Society of Chemistry 2005
J. Mater. Chem., 2005, 15, 2732–2740 | 2735

View Article Online
spiropyran-containing peptides by NMR would help in the
understanding of the molecular details of such properties of
spiropyran derivatives in aqueous media.
methanol in the buffer (v/v), n-hexanesulfonic acid sodium
salt (n-HexSO3Na), cetyl trimethylammonium bromide
(CTAB), b-cyclodextrin (b-CD), and c-cyclodextrin (c-CD)
(Fig. 6A). Neither negatively nor positively charged surfac-
tants such as n-HexSO3Na and CTAB affected the MC-to-SP
photoisomerization rate constants compared to those without
additives. The addition of b-CD and c-CD lowered the rate
constants by 3.1% and 1.3%, respectively (negative effects).
MC-to-SP photoisomerization properties of spiropyran-
containing peptides
Next we examined the photophysical properties in the MC-to-
SP photoisomerization (photobleaching) of the spiropyran-
containing peptides. The fluorescence intensity of K-PKA-SP
(1) at 600 nm was recorded as a function of irradiation time
with 510 nm light in aqueous solution, giving a fluorescence
decay curve that could be fitted with the following first-order
equation, F 5 F‘ + (F0 2 F‘)exp(2kMC-to-SPt), where F is the
fluorescence intensity at each time step, F0 is the fluorescence
intensity at t 5 0 h, F‘ is the fluorescence intensity of the SP-
form, kMC-to-SP is the MC-to-SP photoisomerization rate
constant, and t is the elapsed time under visible light
irradiation (Fig. 5A). The photobleaching properties of MC-
form in the K-PKA-SP (1) were investigated further. Fig. 5B
shows the relative MC-to-SP photoisomerization rate con-
stants as a function of the absorbance of the fluorescent MC-
form at 510 nm in the sample solutions. In the case of
absorbance ranging from 0.007 to 0.034, the peptide gave quite
similar relative rate constants, but with an increase in
absorbance to above 0.07, the relative rate constants tended
to be sharply lowered, suggesting that analyses using the
absorbance-independent region obtained above would be more
suitable for highly reproducible results than analyses using the
conventional fluorescence sensing systems. Furthermore, these
findings also imply that protein–peptide interactions with a
broad range of affinity constants would be detectable because
the chemical structure of spiropyrans, controllable by UV
light or irradiation or incubation in the dark, was independent
of concentrations of the capturing peptides below 0.07 in
absorbance.
Fig. 6 (A) The positive/negative effects (Dk/k0, Dk 5 k 2 k0) in the
MC-to-SP photoisomerization of K-PKA-SP (1) in the absence (k0)/
presence of n-HexSO3Na, CTAB, b-CD, c-CD, and MeOH ([K-PKA-
SP (1) ] 5 2.2 mM, [n-HexSO3Na], [CTAB], [b-CD], or [c-CD] 5
1.0 mM, [MeOH] 5 20% in 20 mM Tris HCl buffer (pH 7.4), 150 mM
In order to understand the MC-to-SP photobleaching in
K-PKA-SP (1) in more depth, we measured photoisomeriza-
tion rate constants under various conditions including 20%
NaCl at 4 uC. (B) The relationships between the MC-to-SP
photoisomerization rate constants and absorption maxima.
Fig. 5 (A) Fluorescence decay curve of K-PKA-SP (1) monitored at 600 nm obtained by irradiation of the sample solution with 510 nm light to
photoisomerize from the MC- to SP-forms. The decay curve could be fitted with a simple first-order equation. [Peptide] 5 2.2 mM and [PKA] 5
0.76 mg mL21 in 20 mM Tris HCl buffer (pH 7.4), 150 mM NaCl at 4 uC. (B) Concentration dependence of the MC-to-SP photoisomerization rate
constants. Absorbance at 510 nm shows the concentration of the MC-form in the peptide. Samples of K-PKA-SP (1) ([peptide] 5 0.14–28 mM) were
irradiated with a 510 nm light in 20 mM Tris HCl buffer (pH 7.4), 150 mM NaCl at 4 uC. The standard photoisomerization rate constant is marked
with an asterisk.
2736 | J. Mater. Chem., 2005, 15, 2732–2740
This journal is ß The Royal Society of Chemistry 2005

View Article Online
Meanwhile, the MC-to-SP photoisomerization in 20%
methanol/buffer exhibited an increased rate constant by 8.2%
(positive effects). Fig. 6B shows the relationships between the
positive/negative effects in the photoisomerization and absorp-
tion bands of the MC-forms with various additives. The
addition of CTAB, b-CD, c-CD, and methanol shifted the
absorption maxima of the MC-form to lower wavenumbers,
indicating that the spiropyran moiety had been placed in
reduced dielectric environments. In particular, b-CD and
c-CD provided not only reduced dielectric environments,
but also sterically-hindered surroundings for the spiropyran
moiety in the peptides by forming an inclusion complex.34
These results indicated that the MC-to-SP photoisomerization
was accelerated in solutions with lower dielectric constants
(positive effects) and decelerated by sterically-hindered
micro-environments (negative effects), motivating us to use
the positive/negative effects in the MC-to-SP photoisomeriza-
tion rate constants as a factor for the detection of target
proteins.
Protein-peptide binding assay
We chose six different proteins: PKA, a-amylase, b-galacto-
sidase, lysozyme, hexokinase, and S-100. Fig. 7A–F show the
positive/negative effects in the photoisomerization of eight
different spiropyran-containing peptides by the addition of the
six different model proteins. The eight peptides gave charac-
teristic patterns against model proteins: ‘‘+/2/+/2/+/+/+/2’’
for PKA, ‘‘2/2/2/2/2/+/0/2’’ for a-amylase, ‘‘2/2/2/2/+/+/
+/2’’ for b-galactosidase, ‘‘2/2/2/2/2/+/+/2’’ for lysozyme,
‘‘2/+/+/+/+/2/+/2’’ for hexokinase, and ‘‘2/2/+/2/+/+/+/2’’
for S-100, with highly reproducible results. Protein fingerprints
(a colored pattern for each protein) were created based on the
positive/negative signals in the photoisomerization described
above, in which the accelerated rate constants (positive effects)
were expressed in red, the decelerated ones (negative effects) in
green, and no effect in black (Fig. 7G). The resulting colored
bar-code obtained for each protein was sufficiently unique to
distinguish each protein from the others, suggesting that
protein fingerprints based on the photoisomerization rate
constants obtained by this P-CHROBA technique would be
useful for profiling proteins. The P-CHROBA technique using
the photoisomerization rate constants as a distinguishing
factor has the advantage that background fluorescence signals
from reagents do not affect the photobleaching of the MC-
form during measurements because only the MC-form in the
peptides is responsive to 510 nm light. Even the small number
of spiropyran-containing peptides employed in the present
study successfully distinguished between the target proteins.
Increasing the size of a peptide library might improve the
reliability for profiling numerous proteins.
Fig. 7 (A)–(F) Data for the positive/negative effects (Dk/k0, Dk 5 k 2
k0) in the MC-to-SP photoisomerization rate constants on addition
of the six different target proteins. Peptides ([peptide] 5 2.0–2.2 mM)
were irradiated with 510 nm light in the presence and absence
of proteins ([protein] 5 0.76 mg mL21) to give the MC-to-SP
photoisomerization rate constants, k and k0 in 20 mM Tris HCl
buffer (pH 7.4), 150 mM NaCl at 4 uC, respectively. Experimental
errors were estimated to be mostly within ¡2.0%. (G) The protein
fingerprints created on the basis of the positive/negative effects
obtained in Fig. 7A–F. The positive and negative effects are red and
green, respectively. (H) Changes in the MC-to-SP photoisomerization
rates of K-PKA-SP (1) depending upon concentrations of PKA (closed
circles) and a-amylase (open circles). [K-PKA-SP (1) ] 5 2.2 mM,
[protein] 5 0, 0.095, 0.19, 0.38, and 0.76 mg mL21 in 20 mM Tris HCl
buffer (pH 7.4), 150 mM NaCl at 4 uC.
K-PKA-SP (1) exhibited the positive and negative effects in
photochromism on addition of 0.76 mg mL21 of PKA and
a-amylase, respectively (Fig. 7A and B), indicating that such
photobleaching rate shifts were associated with the properties
of the binding between proteins and peptides. We further
investigated relationships between the protein concentration
and the MC-to-SP photoisomerization rate constants to extend
the applicability of the P-CHROBA technique. Fig. 7H shows
This journal is ß The Royal Society of Chemistry 2005
J. Mater. Chem., 2005, 15, 2732–2740 | 2737

View Article Online
the dependence of the accelerated/decelerated photoisomeriza-
tion rate constants of K-PKA-SP (1) on the concentrations of
PKA and a-amylase. The protein concentrations varying from
0 to 0.76 mg mL21 gave a linear correlation between Dk/k0 and
the protein concentrations. With increasing PKA concentra-
tion, the photoisomerization accelerated linearly, conversely
the addition of a-amylase reduced the Dk/k0 values, indicating
that the P-CHROBA technique, using the working curves
based on the Dk/k0 values, is applicable to the quantification of
protein concentrations. However, the method showed rela-
tively high detection limits, indicating that it still requires
further development to become a practical technique. Although
analytical methods utilizing fluorescent probes are generally
quite sensitive, it is relatively difficult to provide reproducible
results and reduce background signals from excess reagents.
The present technique can solve such difficulties because it is
based on measurement of the MC-to-SP photoisomerization
rate constants. These values can be stabilized if fluorescence
decay curves are acquired under the standard conditions, and
the method can avoid the need for immobilization of capture
agents onto solid surfaces, facilitating removal of excess
reagents. Furthermore, combination of the P-CHROBA and
PFP techniques would be one of the most versatile analytical
tools for proteomic studies.
quartz cell (10 mm path length) at various temperatures.
Fluorescence spectra were aquired on a Hitachi F-2500
fluorescence spectrophotometer equipped with a magnetic
stirrer and a thermoregulator using a quartz cell (5.0 mm in
width 6 10 mm in path length) at various temperatures.
All proteins (PKA, a-amylase, b-galactosidase, lysozyme,
hexokinase, and S-100) were purchased from Sigma-
Aldrich, Japan. All solvents and reagents (except amino acid
derivatives) were purchased from Wako Pure Chemical
Industries (Osaka, Japan) and used as received. Fmoc-
amino acid derivatives including Fmoc-Orn(Boc)-OH,
Fmoc-Dab-OH, and Fmoc-Dap-OH were purchased
from Watanabe Chemical Industries (Hiroshima, Japan).
Acetonitrile (HPLC grade) was used for the HPLC analysis
and purification.
Chromatography
Adsorption column chromatography was performed using
flash silica gel (70–230 mesh, Merck). Analytical HPLC
and purification of the peptides were performed on
a
Hitachi L7000 or a Shimadzu LC2010C system equipped
with a Wakosil 5C18 or a YMC-pack ODS-A (4.6 6 150 mm)
with a linear gradient of acetonitrile/0.1% TFA at a flow
rate of 1.0 mL min21 for analysis, and a YMC ODS A323
(10 6 250 mm) at a flow rate of 3.0 mL min21 for preparative
purification.
Conclusions
In the present study we have described the synthesis and
characterization of photochromic spiropyran-containing pep-
tides, and the feasibility of a unique fluorescence sensing
system based on the photochromic properties of the peptides
(P-CHROBA). The spiropyran moiety in the peptides
exhibited characteristic physicochemical properties in SP-to-
MC isomerization (thermocoloration) and MC-to-SP photo-
isomerization (photobleaching) which depend on changes in
the micro-environment such as the dielectric constant of the
solvent and steric hindrances generated by interactions. The
MC-to-SP photobleaching of the MC-form in the peptides was
faster than the SP-to-MC thermocoloration in neutral
aqueous media. Thus, we successfully demonstrated the use
of photobleaching-process-shifts resulting from protein–
peptide interactions, which can leas to reduced background
fluorescence signals from excess reagents, avoid laborious
introduction of probing molecules to analytes, and omit the
need for immobilization of capture agents onto solid surfaces.
The PFP technique using peptides also had the advantages that
short peptides are (i) easy to prepare and (ii) responsive
enough to distinguish between target proteins by a colored
bar-code based on changes in fluorescence decay curves, even
though they bind relatively poorly to proteins. The combina-
tion of the P-CHROBA and PFP techniques would provide a
powerful tool for profiling target proteins with reproducible
and reliable results.
Mass spectrometry
Mass spectra were obtained by matrix-assisted laser
desorption/ionization time-of-flight mass spectrometry
(MALDI-TOFMS, KOMPACT MALDI III, Shimadzu,
Japan) with 3,5-dimethoxy-4-hydroxycinnamic acid (SA) as
the matrix.
Non-commercial compounds
2-(39,39-Dimethyl-6-nitro-39H-spiro[chromene-2,29-indole]-19-
yl)ethanol (9) was prepared as described in the literature.25
p-Nitrophenyl 2-(39,39-dimethyl-6-nitro-39H-spiro[chromene-
2,29-indole]-19-yl)ethylcarbonate (10)
A solution of p-nitrophenyl chloroformate (572 mg, 2.84 mmol)
in CH2Cl2 (5.0 mL) was added to a mixture of 9 (1.00 g,
2.84 mmol) and diisopropylethylamine (DIEA, 2.0 mL) in
CH2Cl2 (20 mL) dropwise for 1 min at 0 uC and the resulting
reaction mixture was stirred at 0 uC continuously. After 1 h,
the addition of the solution of p-nitrophenyl chloroformate
(572 mg, 2.84 mmol) in CH2Cl2 (5.0 mL) was repeated and the
mixture was stirred at 0 uC. After 1 h, another solution of
p-nitrophenyl chloroformate (572 mg, 2.84 mmol) in CH2Cl2
(5.0 mL) was added and the resulting mixture was stirred at
0 uC. After 1 h, the reaction mixture was concentrated and
purified by column chromatography [silica, ethyl acetate/
petroleum ether (1 : 2)]. The residue was decanted with diethyl
ether/petroleum ether twice, followed by vacuum filtration,
giving a faintly pink solid (827 mg, 56%): mp 99–101 uC (dec).
Analytically calculated for C27H23N3O8: C, 62.67; H, 4.48; N,
8.12. Found: C, 61.53; H, 4.56; N, 7.98.
Experimental
General
Absorption spectra were acquired on a Shimadzu UV-2550
spectrophotometer equipped with a thermoregulator using a
2738 | J. Mater. Chem., 2005, 15, 2732–2740
This journal is ß The Royal Society of Chemistry 2005

View Article Online
L-a-(9-Fluorenylmethoxycarbonyl)-v-[2-(39,39-dimethyl-6-nitro-
39H-spiro[chromene-2,29-indole]-19-yl)ethoxycarbonyl]ornithine
(Fmoc-Orn(SP)-OH, 11)
were protected as follows: acetamidomethyl (Acm) for Cys,
t-butyloxycarbonyl (Boc) or 4-methyltrityl (Mtt) for Lys,
2,2,4,6,7-pentamethyldihydrobenzofurane-5-sulfonyl (Pbf) for
Arg, t-butyl ester (OtBu) for Glu, and t-Bu ether (tBu) for
Tyr, Thr, and Ser. Initially the Ac–Cys(Acm)–Gly–Lys(Mtt)–
Gly–[various sequences]–Gly resin was prepared in 30 mmol
quantities. The peptide-bound resin was dried in vacuo and
divided into two portions. Half of the peptide-bound resin
(15 mmol) was treated with TFA/TIS/CH2Cl2 (1/5/94, v/v/v)
at room temperature to remove the Mtt protecting group
of the Lys residue. The resulting peptide-bound resin
was washed with 1% DIEA/N-methyl-2-pyrrolidone (NMP)
(three times), then neat NMP (three times) and treated with
10 (3 eq) and DIEA (6 eq) at room temperature in the dark.
After 6 h, DIEA (6 eq) was added again and the reaction
mixture was shaken at room temperature in the dark. After
12 h the reaction mixture was filtered and washed with
NMP then CHCl3. All of the protecting groups except
Acm were removed by the treatment of TFA/m-cresol/EDT/
TA (40/1/3/3, v/v/v/v) for 1 h at room temperature. The
crude peptide obtained was purified by RP-HPLC and
characterized by MALDI-TOFMS (SA), giving a fluffy
yellow powder: K-PKA-SP (1, 6.7 mg, 28%): observed,
1608.5 [(M + H)+]; calculated 1607.8. K-cSrc-SP (6, 5.6 mg,
19%): observed, 1929.0 [(M + Na)+], 1906.1 [(M + H)+];
calculated, 1905.2. K-cAbl-SP (7, 6.7 mg, 25%): observed,
1834.1 [(M + Na)+], 1813.3 [(M + H)+]; calculated 1811.1.
K-PKC-SP (8, 4.6 mg, 18%): observed, 1723.4 [(M + H)+];
calculated 1722.0.
Fmoc-Orn(Boc)-OH (175 mg, 0.386 mmol) was treated with
TFA (3.0 mL) at 0 uC for 30 min. The reaction mixture
was concentrated, precipitated from diethyl ether/petroleum
ether, and dried in vacuo, giving a colorless solid. The solid
was reacted with 10 (100 mg, 0.193 mmol) in the presence of
DIEA (134 mL, 0.772 mmol) in N,N-dimethylformamide
(DMF, 5.0 mL) at 0 uC for 30 min then at room temperature
for 1 h. Ethyl acetate was added to the reaction mixture, and
the organic phase was washed with 10% citric acid and water
and dried over MgSO4. Silica gel column chromatography
[CHCl3/methanol/acetic acid (90/10/10)] followed by precipita-
tion from diethyl ether/petroleum ether gave a faintly yellowish
solid (57 mg, 40%): mp 114–116 uC (dec); MALDI-TOFMS
(SA) observed, 733.4 [M + H]+; calculated, 732.8. Calculated
for C41H40N4O9 + 3H2O: C, 61.18; H, 6.02; N, 6.96. Found: C,
61.35; H, 5.51; N, 6.81.
L-a-(9-Fluorenylmethoxycarbonyl)-v-[2-(39,39-dimethyl-6-nitro-
39H-spiro[chromene-2,29-indole]-19-yl)ethoxycarbonyl]-
diaminobutyric acid (Fmoc-Dab(SP)-OH, 12)
Following the general method described in the synthesis of
compound 11, Fmoc-Dab-OH (79.0 mg, 0.232 mmol) was
reacted with 10 (100 mg, 0.193 mmol) in the presence of
DIEA (81 mL, 0.464 mmol) in DMF (10 mL) at 0 uC for
30 min, then at room temperature for 5 h. Chromatographic
purification followed by precipitation from diethyl ether/
petroleum ether gave a faintly orange solid (95 mg, 68%):
mp 114–116 uC (dec); MALDI-TOFMS (SA) observed, 718.4
General procedure of preparation of the spiropyran-containing
peptides (Route 2)
[M + H]+; calculated, 718.8. Calculated for C40H38N4O9
+
H2O: C, 65.21; H, 5.47; N, 7.60. Found: C, 65.74; H, 5.40;
The fully-protected peptide-bound resins were prepared and
deprotected/deresinated by the general method described
above. The crude peptide obtained was purified by RP-
HPLC and characterized by MALDI-TOFMS (SA), affording
a fluffy yellow powder: O-PKA-SP (2, 1.8 mg, 3.6%):
observed, 1596.8 [(M + H)+]; calculated, 1593.8. Db-PKA-SP
(3, 1.8 mg, 3.8%): observed, 1582.3 [(M + H)+]; calculated
1579.8. Dp-PKA-SP (4, 1.9 mg, 4.0%): observed, 1566.6
[(M + H)+]; calculated 1565.8.
N, 7.56.
L-a-(9-Fluorenylmethoxycarbonyl)-v-[2-(39,39-dimethyl-6-nitro-
39H-spiro[chromene-2,29-indole]-19-yl)ethoxycarbonyl]-
diaminopropionic acid (Fmoc-Dap(SP)-OH, 13)
Following the general method described in the synthesis of
compound 11, Fmoc-Dap-OH (80.0 mg, 0.232 mmol) was
reacted with 10 (100 mg, 0.193 mmol) in the presence of
DIEA (81 mL, 0.464 mmol) in DMF (3.0 mL) at 0 uC for
30 min, then at room temperature for 3 h. Chromatographic
purification followed by precipitation from diethyl ether/
petroleum ether gave a faintly yellowish solid (83 mg,
61%): mp 118–121 uC (dec); MALDI-TOFMS (SA) observed,
704.6 [M + H]+; calculated, 704.7. Calculated for C39H36N4O9 +
2H2O: C, 63.24; H, 5.44; N, 7.56. Found: C, 63.27; H, 5.22;
N, 7.44.
General procedure of preparation of the spiropyran-containing
peptides (Route 3)
The fully-protected Fmoc–Cys(Acm)–Gly–Ala–Gly–Leu–
Arg(Pbf)–Arg(Pbf)–Ser(tBu)–Leu–Gly–resin was prepared by
the general method described above. The Fmoc protecting
group at the N-terminus was removed with 20% piperidine/
NMP and washed with NMP 10 (47 mg, 90 mmol). It was
reacted with the amino group in the peptide on resin in the
presence of DIEA (31 mL, 0.18 mmol) in NMP. After 2 h,
DIEA (31 mL, 0.18 mmol) was added and the resulting mixture
was shaken overnight. The peptide-bound resin was treated
with TFA/m-cresol/EDT/TA (40/1/3/3, v/v/v/v) for 1 h at
room temperature. The crude peptide obtained was purified
by RP-HPLC and characterized by MALDI-TOFMS (SA),
giving a fluffy yellow powder: Nt-PKA-SP (5, 3.3 mg, 7.3%)
observed, 1509.9 [(M + H)+]; calculated 1508.7.
General procedure of preparation of the spiropyran-containing
peptides (Route 1)
Peptides were synthesized by means of Fmoc chemistry on
Rink amide MBHA resin with 2-(1H-benzotriazole-1-yl)-
1,1,3,3-tetramethyluronium hexafluorophosphate (HBTU)
and 1-hydroxybenzotriazole monohydrate (HOBt) as
coupling reagents.33,35 The side chains of the amino acids
This journal is ß The Royal Society of Chemistry 2005
J. Mater. Chem., 2005, 15, 2732–2740 | 2739

View Article Online
P. O. Brown and R. W. Davis, Proc. Natl. Acad. Sci. USA, 1997,
94, 13057–13062.
2 (a) Protein Microarray Technology, ed. D. Kambhampati, Wiley-
VCH, Weinheim, 2003; (b) Protein Arrays: Methods and Protocols;
Methods in Molecular Biology, ed. E. T. Fung, Humana Press,
New Jersey, 2004, vol. 264.
3 G. Walter, K. Bu¨ssow, D. Cahill, A. Lueking and H. Lehrach,
Curr. Opin. Microbiol., 2000, 3, 298–302.
4 H. Zhu and M. Snyder, Curr. Opin. Chem. Biol., 2001, 5, 40–45.
5 T. Kodadek, Chem. Biol., 2001, 8, 105–115.
6 Y.-S. Lee and M. Mrksich, Trends Biotechnol., 2002, 20, 12 suppl.,
S14–S18.
Thermally equilibrium studies of the spiropyran-containing
peptides
The peptide solution (300 mM) in a 20 mM Tris HCl buffer
containing 150 mM NaCl (pH 7.4) was irradiated in ambient
laboratory light at 4 uC for 30 min to convert it to the SP-form.
An aliquot (10 mL) was removed from the peptide solution and
evaluated by RP-HPLC (with a linear gradient of 0–100%
acetonitrile/0.1% TFA at a flow rate of 1.0 mL min21 for
30 min detected at 300 nm) as a HPLC chromatogram at
t 5 0 h. The solution was incubated at 4 uC in the dark and
evaluated by RP-HPLC at each time step. The increasing
and decreasing HPLC peak areas of the MC- and SP-forms,
respectively, were plotted and fitted with the following
7 B. Schweitzer and S. F. Kingsmore, Curr. Opin. Biotechnol., 2002,
13, 14–19.
8 M. Yanagida, J. Chromatogr. B, 2002, 771, 89–106.
9 M. F. Templin, D. Stoll, M. Schrenk, P. C. Traub, C. F. Vo¨hringer
and T. O. Joos, Trends Biotechnol., 2002, 20, 160–166.
10 M. F. Lopez and M. G. Pluskal, J. Chromatogr. B, 2003, 787,
19–27.
11 H. Zhu and M. Snyder, Curr. Opin. Chem. Biol., 2003, 7, 55–63.
12 J. Glo¨kler and P. Angenendt, J. Chromatogr. B, 2003, 797,
229–240.
13 A. Bradbury, N. Velappan, V. Verzillo, M. Ovecka, L. Chasteen,
D. Sblattero, R. Marzari, J. Lou, R. Siegel and P. Pavlik, Trends
Biotechnol., 2003, 21, 312–317.
14 D. S. Wilson and S. Nock, Angew. Chem. Int. Ed., 2003, 42,
494–500.
15 D. N. Howbrook, A. M. van der Valk, M. C. O’Shaughnessy,
D. K. Sarker, S. C. Baker and A. W. Lloyd, Drug Discovery Today,
2003, 8, 642–651.
16 S. Hanash, Proteomics, 2003, 3, 2075–2076, special issue on Protein
Microarrays.
17 K.-Y. Tomizaki, K. Usu and and H. Mihara, ChemBioChem, 6,
DOI: 10.1002/cbic.200400232.
18 http://www.invitrogen.com.
19 http://www.protometrix.com.
20 K. Usui, M. Takahashi, K. Nokihara and H. Mihara, Mol.
Diversity, 2004, 8, 209–218.
21 K. Usui, T. Ojima, M. Takahashi, K. Nokihara and H. Mihara,
Biopolymers, 2004, 76, 129–139.
22 M. Takahashi, K. Nokihara and H. Mihara, Chem. Biol., 2003, 10,
53–60.
23 (a) G. Berkovic, V. Krongauz and V. Weiss, Chem. Rev., 2000, 100,
1741–1753; (b) N. Tamai and H. Miyasaka, Chem. Rev., 2000, 100,
1875–1890.
24 J. B. Flannery, Jr., J. Am. Chem. Soc., 1968, 90, 5660–5671.
25 (a) F. M. Raymo and S. Giordani, J. Am. Chem. Soc., 2001, 123,
4651–4652; (b) F. M. Raymo and S. Giordani, J. Am. Chem. Soc.,
2002, 124, 2004–2007.
equation to obtain the thermodynamic parameters, A 5 A‘
+
(A0 2 A‘)exp(2t21t), where A is the peak area of the MC-/SP-
form in a HPLC chromatogram at each time step, A0 is the
peak area of the MC/SP-form at t 5 0 h, A‘ is the peak area of
the MC-/SP-form in an equilibrium state, t is the relaxation
constant (t21 5 kSP-to-MC + kMC-to-SP), and t is the time elapsed.
The equilibrium constants (Keq) were calculated from the equa-
tion, Keq 5 [MC]/[SP] 5 A‘(MC)c/A‘(SP)c, where A‘(MC)c
and A‘(SP)c are corrected peak areas in an equilibrium state
[calculated from the following relation, A‘(MC) 2 A0(MC) 5
A0(SP) 2 A‘(SP)]. Finally, both kSP-to-MC and kMC-to-SP
were calculated by the following equations, t21 5 kSP-to-MC
+
kMC-to-SP and Keq 5 kSP-to-MC/kMC-to-SP
.
Kinetic studies of the MC-to-SP photoisomerization
A sample solution of a spiropyran-containing peptide with or
without a protein of interest in 20 mM Tris HCl buffer
containing 150 mM NaCl (pH 7.4) was incubated overnight at
4 uC in the dark. The sample was transferred to a quartz cell
(5.0 6 10 mm) placed at the thermoregulator in the Hitachi
F-2500 fluorescence spectrophotometer, and stirred with a
magnetic bar at a desired temperature for 30 min before
measurement was started. The fluorescence decay of MC
species at 600 nm (excitation at 510 nm) was detected after a
time of 1800 s. The fluorescence decay curve obtained was
26 J. Sunamoto, K. Iwamoto, Y. Mohri and T. Kominato, J. Am.
Chem. Soc., 1982, 104, 5502–5504.
27 B. I. Ipe, S. Mashima and K. G. Thomas, J. Am. Chem. Soc., 2003,
125, 7174–7175.
28 A. Shumburo and M. C. Biewer, Chem. Mater., 2002, 14,
3745–3750.
29 K.-Y. Tomizaki, J. Xu and H. Mihara, Bioorg. Med. Chem. Lett.,
2005, 15, 1731–1735.
fitted with the first-order kinetic equation, F 5 F‘ + (F0
2
F‘)exp(2kMC-to-SPt) and analyzed by KaleidaGraph software
(Synergy Software), where F is the fluorescence intensity at
each time step, F‘ is the fluorescence intensity at infinity, F0 is
the initial fluorescence intensity, kMC-to-SP is the MC-to-SP
photoisomerization rate constant, and t is the elapsed time.
30 S. Cox and S. S. Taylor, Biochemistry, 1995, 34, 16203–16209.
31 B. T. Houseman, J. H. Huh, S. J. Kron and M. Mrksich, Nat.
Biotechnol., 2002, 20, 270–274.
32 L. E. Heasley and G. L. Johnson, J. Biol. Chem., 1989, 264,
8646–8652.
33 Fmoc Solid Phase Peptide Synthesis: A Practical Approach, ed.
W. C. Chen and P. D. White, Oxford University Press, New York,
2000.
34 F. Hamada, K. Hoshi, Y. Higuchi, K. Murai, Y. Akagami and
A. Ueno, J. Chem. Soc., Perkin Trans. 2, 1996, 2567–2570.
35 R. Knorr, A. Trzeciak, W. Bannwarth and D. Gillessen,
Tetrahedron Lett., 1989, 30, 1927–1930.
Kin-ya Tomizaki and Hisakazu Mihara*
Department of Bioengineering and the COE21 Program, Graduate
School of Bioscience and Biotechnology, Tokyo Institute of Technology,
B-40, 4259 Nagatsuta, Midori, Yokohama 226-8501, Japan.
E-mail: hmihara@bio.titech.ac.jp; Fax: +81-45-924-5833
References
1 (a) R. J. Lipshutz, S. P. Fodor, T. R. Gingeras and D. J. Lockhart,
Nat. Genet., 1999, 21, 20–24; (b) D. A. Lashkari, J. L. DeRisi,
J. H. McCusker, A. F. Namath, C. Gentile, S. Y. Hwang,
2740 | J. Mater. Chem., 2005, 15, 2732–2740
This journal is ß The Royal Society of Chemistry 2005