Novel bacteriorhodopsin analogues based on azo chromophores

Article abstract of DOI:10.1021/ja954286x

Bacteriorhodopsin analogues BR-I, BR-II and BR-III containing azo chromophores 4-[[4'-(N,N-dimethylamino)phenyl-1']azo]benzaldehyde, 3-[4-[[4'-(N,N-dimethylamino)phenyl-1']azo]phenyl-1]prop-2-enal and 3-[4-[[4'-(N,N-dimethylamino)phenyl-1']azo]phenyl-1]-2

Full text of DOI:10.1021/ja954286x

J. Am. Chem. Soc. 1996, 118, 6185-6191
6185
Novel Bacteriorhodopsin Analogues Based on Azo
Chromophores
Anil K. Singh,* Joydip Das, and Nirmalya Majumdar
Contribution from the Department of Chemistry, Indian Institute of Technology,
Powai, Bombay 400 076, India
ReceiVed December 28, 1995X
Abstract: Bacteriorhodopsin analogues BR-I, BR-II and BR-III containing azo chromophores 4-[[4′-(N,N-
dimethylamino)phenyl-1′]azo]benzaldehyde, 3-[4-[[4′-(N,N-dimethylamino)phenyl-1′]azo]phenyl-1]prop-2-enal and
3-[4-[[4′-(N,N-dimethylamino)phenyl-1′]azo]phenyl-1]-2-methylprop-2-enal, respectively, were prepared and character-
ized by UV-vis spectroscopy, opsin shift, competitive binding with retinal, fluorescence spectroscopy, light-induced
pH change, and flash photolysis. BR-I, BR-II, and BR-III had UV-vis absorption maxima at 458, 597, and 485
nm and opsin shifts of -329, 3091, and 43 cm-1, respectively. Competitive binding studies showed that the azo
chromophores could not be easily displaced by retinal. Quenching of protein fluorescence by the azo chromophores
indicated intimate interactions occurring between the respective azo chromophores and the protein bound residues.
The proteins were also found to show functional characteristics (light-induced pH change and flash photolysis profiles)
different from those of native bacteriorhodopsin. The results are discussed in terms of the nature of interaction
between the azo chromophore and the surrounding protein.
Introduction
Experimental Section
Materials and Methods. all-trans-Retinal (Sigma), trans-cinna-
maldehyde, trans-R-methylcinnamaldehyde, and N,N-dimethylaniline
(Aldrich) were used as received. n-Butylamine (Aldrich) was stored
over potassium hydroxide pellets and freshly distilled under nitrogen
before use. All other chemicals were from Sisco Research Laboratories
(India). Purple membrane fragments were isolated and purified from
H. halobium (R1M1) using the procedure described by Oesterhelt and
Stoeckenius.4 Bacterioopsin was obtained by bleaching the purple
membrane according to the procedure described earlier.5 The apomem-
brane was pelleted and washed three times with water to remove excess
hydroxylamine and retinal oxime. The apomembranes were suspended
in Tris‚HCl buffer (1.0 × 10-2 M). The pH of the buffer was adjusted
by addition of dilute HCl or dilute NaOH solution as required. All
pH measurements were done on a Radiometer PHM84 research pH
meter equipped with GK 2401 C combination electrodes. UV-vis
measurements were recorded on a Hitachi U-2000 spectrophotometer.
IR spectra were recorded on a Nicolet Impact 400 FTIR spectrometer.
NMR spectra were recorded on a Varian VXR-300MHz spectrometer.
Centrifugations were done on a Beckman L-8-55M preparative ultra-
centrifuge using a 45 Ti rotor. Flash photolysis profiles were obtained
on an Applied Photophysics flash kinetic spectrometer (model 858),
the output of which was fed to a Philips PM3350 digital storage
oscilloscope which recorded the flash photolysis traces. The excitation
flash was provided by a U-tube flash lamp (40 J, 900 V, <100 µs)
with a range from 400 to 700 nm mounted adjacent to the sample holder,
and the detection system consisted of a tungsten-halogen lamp (12
V, 20 W) and a photomultiplier tube. Fluorescence spectra were
recorded on a Spex Spectrofluorolog spectrofluorimeter with a 450 W
xenon arc lamp as the light source. Data were analyzed on a DM-1B
data processor interfaced with the spectrofluorimeter. Molecular
mechanics calculations were carried out using the PCMODEL molecular
modeling software (4th ed., copyright 1990 Serena Software, Box 3076,
Bloomington, IN 47402-3076) on an 80486 IBMPC.
Bacteriorhodopsin (BR), the membrane bound protein in the
purple membrane (PM) of Halobacterium halobium contains
the all-trans/13-cis-retinylidene Schiff base chromophore and
functions as a light-driven proton pump.1 In recent years much
effort has been directed toward understanding the unique spectral
and photochemical properties of BR. In this context, a large
number of synthetic chromophore analogues have been incor-
porated into the binding site of BR, and the observation of the
resulting opsin shifts and/or photochemical properties has been
analyzed in terms of chromophore-protein interactions.2,3 More
recently, research on the development of BR variants has also
gained momentum due to the possibility of producing proteins
of enhanced performance for applications in molecular electron-
ics.3 BR analogues having chromophores which alter the
protein’s absorption characteristics while maintaining the pho-
tochemical activity and those with chromophores which slow
the decay of the M intermediate have been developed.3
We report herein the preparation and characterization of BR
analogues based on azobenzene chromophores which show
unique spectral and photochemical properties.
* To whom correspondence should be addressed. Fax: 91-22-578 3480.
E-mail: retinal@ether.chem.iitb.ernet.in.
X Abstract published in AdVance ACS Abstracts, June 1, 1996.
(1) (a) Oesterhelt, D.; Stoeckenius, W. Nature, New Biol. 1971, 233,
149. (b) Stoeckenius, W.; Bogomolni, R. A. Annu. ReV. Biochem. 1982,
52, 587.
(2) (a) Honig, B.; Dinur, U.; Nakanishi, K.; Balogh-Nair, V.; Gawinow-
icz, M. A.; Arnaboldi, M.; Motto, M. G. J. Am. Chem. Soc. 1979, 101,
7084. (b) Arnaboldi, M.; Motto, M. G.; Tsujimoto, K.; Balogh-Nair, V.;
Nakanishi, K. J. Am. Chem. Soc. 1979, 101, 7082. (c) Derguini, F.; Dunn,
D.; Eisenstein, L.; Nakanishi, K.; Odashima, K.; Rao, J. V.; Shastry, L.;
Termini, J. Pure Appl. Chem. 1986, 58, 719. (d) Lugtenburg, J.; Muradin-
Szweykowska, M.; Heeremans, C.; Pardoen, J. A.; Harbison, G. S.; Herzfeld,
J.; Griffin, R. G.; Smith, S. O.; Mathies, R. A. J. Am. Chem. Soc. 1986,
108, 3104. (e) Asato, A. E.; Li, X. Y.; Mead, D.; Patterson, G. M. L.; Liu,
R. S. H. J. Am. Chem. Soc. 1990, 112, 7398.
(3) (a) Birge, R. R. Annu. ReV. Phys. Chem. 1990, 41, 683 and references
cited therein. (b) Brauchle, C.; Hampp, N.; Oesterhelt, D. AdV. Mater. 1991,
3, 420. (c) Oesterhelt, D.; Brauchle, C.; Hampp, N. Q. ReV. Biophys. 1991,
24, 425 and references cited therein.
Synthesis of Azo Chromophores I, II, and III. The synthetic
procedure of the compounds 4-[[4′-(N,N-dimethylamino)phenyl-1′]azo]-
benzaldehyde (I), 3-[4-[[4′-(N,N-dimethylamino)phenyl-1′]azo]phenyl-
1]prop-2-enal (II), and 3-[4-[[4′-(N,N-dimethylamino)phenyl-1′]-
azo]phenyl-1]-2-methylprop-2-enal (III) is shown in Scheme 1. 4-Ami-
(4) Oesterhelt, D.; Stoeckenius, W. Methods Enzymol. 1974, 31, 667.
(5) Oesterhelt, D.; Schuhmann, L.; Gruber, H. FEBS Lett. 1974, 44, 262.
S0002-7863(95)04286-7 CCC: $12.00 © 1996 American Chemical Society

6186 J. Am. Chem. Soc., Vol. 118, No. 26, 1996
Singh et al.
8.7 Hz, aromatic), 7.90 (4H, m, aromatic), 9.61 (1H, s, CHO). UV-
vis (MeOH, λmax, nm): 251 (ꢀ ) 7422), 310 (ꢀ ) 8698), 450 (ꢀ )
17 460). IR (CHCl3, cm-1) 3010, 1676, 1603, 1521, 1478, 1425, 1366.
Formation and Protonation of the n-Butylamine Schiff Bases of
Compounds I-III. Azo chromophore I (2.8 × 10-3 M) was taken in
0.8 mL of dry methanol and 0.2 mL of n-butylamine added to it. To
this was added a small amount of anhydrous sodium sulfate (5 mg)
and a bead of Linde 4 Å molecular sieves. The mixture was kept under
nitrogen at 4 °C overnight. The dark orange-red solution was filtered
through a small Buchner filter directly into a Schlenk type storage vessel
in which it was carefully evaporated to dryness using high vacuum
(10-3 Torr). The solid was redissolved in 1.0 mL of dry methanol and
stored under nitrogen at -20 °C. Protonation of Schiff base (1.0 ×
10-5 M) was carried out by addition of dry methanolic HCl under
nitrogen at 25 °C. Schiff bases of chromophores II and III were
similarly prepared and handled.
Scheme 1
Formation of Azo-Protein Complexes BR-I, BR-II, and BR-
III and Competitive Binding Studies. The azo chromophores I, II,
and III in ethanol were incubated with the apomembrane (4.0 × 10-5
M) in an aqueous medium, pH 5 (1.0 × 10-2 M Tris‚HCl buffer) at
ambient temperature in dim red light. The azo chromophore concentra-
tions were 9.0 × 10-4 M in all cases, and the final ethanol concentration
in the protein suspension was less than 3% v/v. Formation of the
respective azo-protein complexes was monitored by UV-vis absorp-
tion spectra over a period of hours. After 24 h, the azo-protein
complexes formed were washed free of excess chromophore by
centrifugation (50000g, 10 min). At the end of 24 h, each of the
incubation mixtures was supplemented with all-trans-retinal and the
increase of absorbance at 560 nm was measured. The percent
displacement was determined by measuring the ratio of the absorbance
at the UV-vis maximum for the respective BR analogue at 60 min
(the time taken for maximum displacement by retinal) and expressed
as a percentage.
Fluorescence Studies. BR-I, BR-II, and BR-III were washed
three times with 1.0 × 10-2 M Tris‚HCl (pH 5) buffer and centrifuged
(25000g, 30 min) to remove excess chromophore. The pellet was
resuspended in 1.0 × 10-2 M Tris‚HCl (pH 5). Bacterioopsin (BOP)
and BR were also suspended in Tris buffer (pH 5). The optical density
of all three solutions was adjusted to 0.5 (A280 ) 0.5) before recording
the fluorescent emission for 280 nm excitation.
Light-Induced pH Change Studies. Suspensions of native BR and
the modified proteins BR-I, BR-II, and BR-III in basal salt of high
optical density (A ) 0.8 at the wavelength of maximum absorbance in
the visible region) were prepared with the pH adjusted to 5 by the
addition of HCl. No buffers were used as they inhibit pH changes.
These were sonicated for 3 min using a Branson 450 sonifer at 4 °C.
This suspension (4 mL) was taken in a tube with an outer jacket
circulating water to maintain the temperature at 25 °C. The samples
were illuminated with a Philips tungsten-halogen lamp (230 V, 500
W). The distance between the sample and the light source was 12 cm.
pH changes were monitored with the pH meter. The amounts of protons
displaced as a function of time were monitored by recording the change
in pH upon turning the illumination source on and upon switching it
off.
Flash Photolysis Studies. Basal salt suspensions of purple mem-
brane fragments (A570 ) 0.8) and the BR analogues BR-I, BR-II,
and BR-III (A ) 0.8 at the wavelength of maximum absorbance in
the visible region) in Tris‚HCl buffer (pH 5) were taken in a quartz
cell for recording the flash photolysis decay profiles. The flash
photolysis profiles of the transient species were monitored between
400 and 420 nm.
nobenzaldehyde was synthesized following literature procedure.6 Nitra-
tion of trans-cinnamaldehyde and trans-R-methylcinnamaldehyde was
done as reported elsewhere.7 The reduction of the nitro compounds to
the corresponding amines was done following known procedures.8
Coupling of N,N-dimethylaniline with the corresponding amines was
done using a slight modification of known procedures.9 This was done
to avoid self-condensation of the compounds having both the amino
and aldehyde groups into polymeric Schiff bases. In a typical
procedure, concentrated HCl (0.4 mL) was dissolved in 20 mL of water
and cooled to 0 °C with stirring. A solution of amino compound (0.5
mmol) in acetone was added dropwise to the cooled solution over a
period of 30 min with constant stirring. A precooled solution of NaNO2
(37.0 mg, 0.53 mmol) in 2 mL of water was added dropwise to the
stirred suspension. A pinch of urea was added to the vigorously stirred
solution to destroy any excess nitrous acid remaining in solution. To
this solution was added dropwise a solution of N,N-dimethylaniline
(63.0 µL, 0.5 mmol) in 3 mL of glacial acetic acid with stirring. An
opaque yellow-red suspension resulted. This solution was vigorously
stirred for another 10 min, after which a saturated solution of sodium
acetate was added. Bright orange-yellow (for compound I) or orange-
red (for compounds II and III) flakes of the azo compounds appeared.
The crystalline solids were filtered and dissolved in acetone. The
compounds were purified by column chromatography over a silica gel
column and characterized by NMR, UV-vis, and IR spectroscopy. The
characterization data for the three azo compounds are as follows (s )
singlet, d ) doublet, m ) multiplet; ꢀ values are in L mol-1 cm-1).
4-[[4′-(N,N-Dimethylamino)phenyl-1′]azo]benzaldehyde (I). mp:
164-165 °C (uncorrected). Yield: 92%. NMR (CDCl3, δ, ppm): 3.12
(6H, s, NMe2), 6.75 (2H, d, J ) 9.0 Hz, aromatic), 7.94 (6H, m,
aromatic), 10.05 (1H, s, CHO). UV-vis (MeOH, λmax, nm): 229 (ꢀ
) 14 500), 284 (ꢀ ) 18 200), 449 (ꢀ ) 22 500). IR (CHCl3, cm-1):
3010, 1690, 1590, 1515, 1425, 1360.
3-[4-[[4′-(N,N-Dimethylamino)phenyl-1′]azo]phenyl-1]prop-2-
enal (II). mp: 204-205 °C (uncorrected). Yield: 97%. NMR
(CDCl3, δ, ppm) 3.11 (6H, s, NMe2), 6.76 (3H, m, aromatic + olefinic),
7.53 (1H, d, J ) 16.0 Hz, Ar-CH, olefinic), 7.68 (2H, d, J ) 8.4 Hz,
aromatic), 7.84 (4H, m, aromatic), 9.72 (1H, d, J ) 7.7 Hz, CHO).
UV-vis (MeOH, λmax, nm): 253 (ꢀ ) 5382), 313 (ꢀ ) 10 928), 461
(ꢀ ) 18 836). IR (CHCl3, cm-1): 3010, 1678, 1602, 1522, 1475, 1426,
1365.
3-[4-[[4′-(N,N-Dimethylamino)phenyl-1′]azo]phenyl-1]-2-methyl-
prop-2-enal (III). mp: 150-151 °C (uncorrected). Yield: 94%. NMR
(CDCl3, δ, ppm): 2.14 (3H, s, Me), 3.11 (6H, s, NMe2), 6.77 (2H, d,
J ) 9.0 Hz, aromatic), 7.30 (1H, s, Ar-CH, olefinic), 7.66 (2H, d, J )
Results and Discussion
pH Dependent UV-Vis Characteristics of Chromophores
I, II, and III. As evident from Table 1, the azo chromophores
I-III show pH dependent UV-vis behavior. This is explained
in Scheme 2. At high pH, the chromophore remains unproto-
nated. As the pH decreases, the availability of protons increases,
and at sufficiently low pH, one of the nitrogen atoms of the
azo group (-NdN-) becomes protonated, giving rise to large
red-shifted bands (e.g., 511, 517, and 511 nm for chromophores
(6) Beard, H. G.; Hodgson, H. H. J. J. Chem. Soc. 1944, 4.
(7) Diehl, L.; Einhorn, A. Ber. Dtsch. Chem. Ges. 1885, 18, 2335.
(8) Hirata, Y.; Nakata, H.; Yamada, K.; Okuhara, K.; Naito, T.
Tetrahedron 1961, 14, 25.
(9) Vogel’s Textbook of Practical Organic Chemistry; Longman Group
Ltd.: Harlow, U.K., 1984; p 718.

Bacteriorhodopsin Analogues Based on Azo Chromophores
J. Am. Chem. Soc., Vol. 118, No. 26, 1996 6187
Table 1. UV-Vis Characteristics of Azo Chromophores I, II, and
III in Methanol and at pH 3, 5, 7, and 9
Table 2. UV-Vis Characteristics of Schiff Bases of I, II, and III
with n-Butylamine and Their Protonated Salts
UV-Vis λmax of chromophore (nm)a
UV-vis λmax (nm)a
c
d
medium
I
II
III
chromophore
RCHO
SBb
PSB1
PSB2
pH 3b
511
405
412
416
449
517
388
387
385
461
511
395
393
392
450
I
II
III
449
461
450
436
446
434
465
504
484
520
541
542
pH 5b
pH 7b
pH 9b
a In methanol; [chromophore] ) 1.0 × 10-5 M in each case. b SB
) Schiff base. [SB] ) 1.0 × 10-5 M. c PSB1 ) protonated Schiff
base formed on protonation with 1:1 dry methanolic HCl (1.0 × 10-5
M HCl). d PSB2 ) protonated Schiff base formed on protonation with
1:2 dry methanolic HCl (2.0 × 10-5 M).
methanol
a [Chromophore] ) 1.0 × 10-5 M. b 1.0 × 10-2 M Tris‚HCl buffer.
Scheme 2
Figure 1. UV-vis characteristics of the n-butylamine Schiff base of
azo chromophore I: [Schiff base] ) 1.0 × 10-5 M, (s) 436 nm (in
the absence of acid), (- - -) 465 nm (1:1 dry HCl, [HCl] ) 1.0 × 10-5
M), (‚‚‚) 520 nm (1:2 dry HCl, [HCl] ) 2.0 × 10-5 M).
I, II, and III, respectively). The more-red-shifted band (517
nm) of protonated chromophore II is due to the extended
conjugation available in II which is not so in the case of I.
However, for protonated chromophore III, the extended con-
jugation is prevented from playing a greater role in the
stabilization of the positive charge due to the presence of a
methyl group which can cause out-of-plane twisting of the bonds
and disruption of extended conjugation. Thus, protonated I and
protonated III show similar UV-vis absorption maxima at 511
nm despite dissimilarities in their chemical structures.
Scheme 3
The protonation behavior of chromophores I-III throws light
on the role of the availability or nonavailability of protons in a
given environment and its importance in the wavelength-
regulating property of the azo chromophore with changing pH.
This may have some implications on the availability of protons
at the active site environment of retinal proteins wherein the
relative abundance of amino acid residues with ionizable H
atoms in the vicinity of the retinylidene chromophore plays an
important role in protonation/deprotonation and hence in
wavelength regulation. Wavelength regulation could be effected
by the partial (H-bonding type) or full protonation of the azo
chromophore by the protein bound ionizable residues, resulting
in species of the type shown in Scheme 2. This could be the
cause for the appearance of the new red-shifted bands obtained
by the interaction of the azo chromophores I, II, and III with
bacterioopsin (BOP), resulting in the formation of BR-I, BR-
II, and BR-III, respectively.
UV-Vis Characteristics of the n-Butylamine Schiff Bases
of Chromophores I, II, and III and Their Protonation
Behavior. The Schiff bases of azo chromophores I-III with
n-butylamine were prepared, and their protonation behavior was
studied to compare their UV-vis characteristics with those of
the protein complexes BR-I, BR-II, and BR-III and to
determine the respective opsin shifts.10 The Schiff bases of azo
chromophores I-III show interesting UV-vis characteristics
on protonation with dry methanolic HCl (Table 2). The
absorption bands of the Schiff bases are found to be sensitive
to the availability of protons in the medium. The Schiff base
IV (Scheme 3) shows an absorption maximum at 436 nm and
another intense band in the UV region (Figure 1). On addition
of 1 equiv of methanolic HCl, the visible (436 nm) band shifts
to 465 nm whereas a hypsochromic shift becomes apparent for
the band in the UV region. This UV band resolves into two
bands with absorption maxima at 254 and 295 nm. Addition
of a larger amount of methanolic HCl (2 equiv) causes the
visible band to further shift toward the red to 520 nm. Three
(10) Nakanishi, K.; Balogh-Nair, V.; Arnaboldi, M.; Tsujimoto, K.;
Honig, B. J. Am. Chem. Soc. 1980, 102, 7945.

6188 J. Am. Chem. Soc., Vol. 118, No. 26, 1996
Singh et al.
Table 3. UV-Vis Absorption Properties of Bacteriorhodopsin and Analogues Based on Chromophores I-III
λ
max (nm)
chromophore
RCHOa
RCHOb
SBa
PSBa
BRb
OSc (cm-1
)
I
II
III
retinal
449
461
450
381
405
388
395
385
436
446
434
357
465
504
484
440
458
597
485
560
-329
3091
43
4870
a In methanol. SB ) Schiff base of the respective chromophore with n-butylamine (1.0 × 10-5 M). PSB ) protonated Schiff base. Protonation
effected by 1:1 dry methanolic HCl. b In Tris‚HCl buffer (1.0 × 10-2 M, pH 5). c Opsin shift.
bands at 254, 295, and 356 nm (minor band) in the UV region
and isobestic points at 345 and 373 nm are also observed (Figure
1).
Protonation appears to take place in two stages; the -CdN-
group is protonated first followed by the -NdN- group. This
is consistent with the fact that while the pKa of -CdN- is
∼7.5,11 that of -NdN- is ∼3.0.12 The stepwise protonation
of the Schiff base of I is represented in Scheme 3. The
nonprotonated Schiff base IV absorbs at 436 nm (Figure 1).
On protonation with 1 equiv of acid, the imine nitrogen becomes
protonated and the UV-vis absorption maximum shifts to 465
Figure 2. Formation of the azo-protein complex (progress with time)
([I] ) 9.0 × 10-4 M, [BOP] ) 4.0 × 10-4 M in 1.0 × 10-2 M Tris‚HCl
nm. The resonance-stabilized cation is shown by structures IVa
and IVb. On addition of a further 1 equiv of acid, one of the
azo nitrogens becomes protonated and the UV-vis absorption
maximum shifts to 520 nm. The resultant dication is represented
by the resonating structures IVc and IVd. Of these, structure
Vd is expected to be the predominant contributing structure due
to less repulsion between the positive charges. Moreover,
aldehyde I, which does not contain the -CdN- group, on
treatment with 1 equiv of dry methanolic HCl, shows a peak at
511 nm due to protonation of the -NdN- group. Therefore,
the peak at 465 nm for the protonated Schiff base (IVa, IVb)
could only be due to protonation of the imine (-CdN-)
nitrogen atom.
The protonation behavior of the n-butylamine Schiff bases
of II and III can be similarly rationalized. The UV-vis
absorption properties of the n-butylamine Schiff bases of I, II,
and III as well as their regulated, stepwise protonation with a
distinct UV-vis absorption maximum for each protonated
species afford a new approach to wavelength regulation in
organic chromophores in general and azo-modified BR in
particular.
Interaction of Azo Chromophores I, II, and III with BOP.
Incubation of the ethanolic solutions of I, II, and III (9.0 ×
10-4 M) with BOP (4.0 × 10-5 M) in a pH 5 buffer (Tris‚HCl,
1.0 × 10-2 M) in the dark at 25 °C resulted in the formation of
protein-chromophore complexes BR-I, BR-II, and BR-III
derived from the azo chromophores I, II, and III, respectively.
These were detected by following the change in UV-vis
absorption (Table 3) with time. As typical examples, formation
of the protein complexes BR-I and BR-II are shown in
Figures 2 and 3, respectively. Chromophores I-III show pH
dependent UV-vis absorption bands. At pH 5 (the pH at which
binding studies were performed) chromophore I shows strong
absorption bands in the range of 300-400 nm (λmax ) 289 and
405 nm). The band below 300 nm overlaps with the apoprotein
absorption. After addition of compound I to the apoprotein at
pH 5, the peak at 405 nm initially blue shifts to 395 nm, but
over a period of time the band red shifts to 458 nm Via a peak
at 418 nm with an isobestic point at 408 nm (Figure 2). Binding
of chromophore II with the apoprotein first results in a pigment
buffer): (a) before addition of azo chromophore I; (b) at 0 min, 395
nm; (c) after 30 min, 418 nm; (d) after 75 min, 458 nm.
Figure 3. Formation of the 597 nm BR analogue from interaction of
azo chromophore II ([II] ) 9.0 × 10-4 M, [BOP] ) 4.0 × 10-5 M in
1.0 × 10-2 M Tris‚HCl buffer, pH 5, 25 °C). Spectra were recorded
at different time intervals: (a) 0 min; (b) 5 min; (c) 10 min; (d) 20
min; (e) 30 min; (f) 40 min; (g) 50 min.
with UV-vis absorption at 487 nm which slowly shifts to 597
nm with isobestic points at 422 and 515 nm. Thus, analogue
BR-II first forms a metastable pigment with an absorption
maximum at 487 nm which becomes stabilized to a pigment
with an absorption maximum at 597 nm over a period of time
(Figure 3). Such stepwise binding of retinal analogues with
BOP has been reported earlier.2e Chromophore III also binds
with BOP, similar to chromophore I, with an initial UV-vis
absorption maximum at 465 nm and a final absorption maximum
at 485 nm obtained over a period of time, but without any clear
cut isobestic point. It appears that, for all the cases, metastable
pigments are formed before the final stable pigment is generated.
An estimation of half-lives and the rate of formation (k) of
BR-I, BR-II, and BR-III (Table 4) indicates that chro-
mophores I-III interact readily with BOP. Chromophores I,
II, and III on reaction with BOP give new protein chromophores
with absorption bands at 458, 597, and 485 nm, respectively.
The rate constants (k) for the formation of BR-I, BR-II, and
BR-III are 1.06 × 10-1, 6.5 × 10-2, and <6.9 × 10-2 min-1
,
respectively, assuming first-order kinetics. Further, the largest
bathochromic shifts are observed at pH 5. Studies at several
other pH values (3, 7, and 9) did not result in significantly-
red-shifted pigments though minor shifts of up to 10 nm were
recorded for all the chromophores (I-III) incubated with BOP
for up to 24 h at 25 °C. That such interaction occurs primarily
at pH 5 is due to the pH sensitive nature of the chromophore,
the particular protein conformation at that pH, and the interac-
(11) Schaffer, A. M.; Yamaoka, T.; Becker, R. S. Photochem. Photobiol.
1975, 21, 297.
(12) CRC Handbook of Chemistry and Physics; Weast, R. C., Ed.; CRC
Press Inc.: Boca Raton, FL, 1985; p D-159.

Bacteriorhodopsin Analogues Based on Azo Chromophores
J. Am. Chem. Soc., Vol. 118, No. 26, 1996 6189
Table 4. Rate Constants for Formation of BR, BR-I, BR-II, and
BR-III and Competitive Binding of Retinal with the Modified BR
Pigments
Table 5. Fluorescent Emission Data for BOP, BR, and
Azo-Protein Complexes BR-I, BR-II, and BR-III (Excitation λ
) 280 nm, 25 °C, 1.0 × 10-2 M Tris‚HCl, pH 5)
percentb,c
displacement
of the azo
λmax
K
SV (M-1) ×
λmax
K
SV (M-1) ×
rate constanta
for formation
of BR,
fluorophorea (nm)
106
fluorophorea (nm)
106
UV-vis
maximum
(nm)
BOP
BR
BR-I
339
334
341
BR-II
BR-III
335
345
1.18c
0.71c
t1/2
chromophore
4.63b
5.13c
protein
k (min-1) × 10-1 (min) by retinal (%)
BR-I
BR-II
BR-III
BR
458
597
485
560
1.06
0.65
<0.69
1.50
6.5
10.6
<1.0
4.6
26.3
15.4
80.0
a [Fluorophore] ) 1.5 × 10-6 M. b Obtained by measuring the
quenching of protein fluorescence by retinal within bacteriorhodopsin
in comparison to bacterioopsin. c Obtained by measuring the quenching
of protein fluorescence by the chromophores I, II, and III within the
chromophore-protein complex in comparison to bacterioopsin.
a First-order rate constants at 25 °C. In 1.0 × 10-2 M Tris‚HCl, pH
5. b On addition of excess retinal in the dark. c Monitored for up to 24
h at 25 °C. Results obtained during the first 60 min do not change
significantly at longer equilibration times.
the protein bound amino acid residues from approaching this
region of the chromophore, leading to a lack of some vital
interactions which enhance conjugation. Theoretical molecular
mechanics calculations (PCMODEL) carried out on cinnama-
ldehyde and R-methylcinnamaldehyde show that while for
R-methylcinnamaldehyde the olefinic side chain attached to the
aromatic nucleus (as in compound III) is twisted by an angle
of 64° with respect to the plane of the neighboring phenyl ring,
for cinnamaldehyde, the chain (as in compound II) and the
phenyl ring are found to lie on the same plane. These results
are in agreement with theoretical and experimental studies
carried out earlier on â-ionone and retinal which indicate a
distorted s-cis conformation about the C6-C7 single bond with
the torsional angle in the range 30-70°.10,13 The protein-
chromophore complex of I shows a still lower UV-vis
absorption maximum (458 nm) as well as a negative opsin shift
(-329 cm-1). It is evident from the structure of I that the degree
of stabilization attained for II and III is not possible in the case
of I due to the absence of olefinic bonds in I, i.e., the lack of
longitudinal proximity of the chromophore to the apoprotein
charges.
Fluorescence Behavior of BR-I, BR-II, and BR-III.
The azo chromophores I-III are found to quench the fluorescent
emission of the apoprotein of BR, BOP, when combined with
it as azo-protein complexes BR-I, BR-II, and BR-III (Table
5). The fluorescent emission of BOP becomes quenched when
it combines with retinal to form BR. Similar quenching is
observed when BOP combines with azo chromophores I-III.
The azo chromophores I-III also quench the fluorescent
emission of tryptophan (Trp) in aqueous solutions.
It is known that the fluorescent emission of BR is quenched
relative to BOP due to the interaction between the Trp residues
lining the retinal pocket and the retinylidene chromophore. A
similar interaction is envisaged to explain the observed quench-
ing of the protein fluorescence by azo chromophores I-III in
the azo-protein complexes BR-I, BR-II, and BR-III in
comparison to BOP.
Light-Induced pH Change Activity of BR-I, BR-II, and
BR-III. The light-induced pH change activity of the modified
proteins BR-I, BR-II, and BR-III were studied and compared
with that of native BR (Table 6). The studies were conducted
at pH 5 as the modified BR analogues were formed and remain
stable at this pH. It was found that while native BR is able to
sustain repeated light-on/light-off cycles, only BR-II shows
similar behavior. BR-I undergoes only one such cycle while
BR-III does not exhibit any light-induced pH change activity.
This is because, while in the case of native BR, the protein
is able to complete the photocycle (BR570 f J f K f L f M
f ... f BR570), such may not be the case with the modified
proteins BR-I and BR-III. BR-II is functional and sustains
Scheme 4
tions between the chromophores and the charge-carrying surface
residues of the protein which are available to the chromophores.
The stability of the BR analogues, BR-I, BR-II, and BR-
III was checked by addition of retinal to the respective BR
analogue solutions and by monitoring the growth of UV-vis
absorbance at 560 nm with time. For BR-I, there was 26.3%
displacement of the chromophore by retinal, for BR-II, 15.4%
displacement of the chromophore, while for BR-III, there was
80% displacement of the chromophore by retinal (Table 4). The
competitive binding of retinal with BR-I, BR-II, and BR-
III was monitored over a period of 24 h. Maximum displace-
ment of the respective azo chromophores from the BR analogues
by retinal occurred during the first 60 min. Insignificant
displacement (not exceeding 5%) occurred during the remaining
period of incubation; i.e., the results obtained during the first
60 min do not change significantly at longer equilibration times.
Chromophores I-III showed considerable variation in their
absorption properties and opsin shift data on interaction with
BOP (Table 3). While the analogue BR-I showed a negative
opsin shift of -329 cm-1, analogue BR-II showed a large
positive opsin shift of 3091 cm-1 and analogue BR-III showed
only a small positive opsin shift of 43 cm-1. The difference in
the behavior of the three chromophores within the protein
environment can be explained by considering the resonance
structures Va and Vb (Scheme 4). The stabilization afforded
by the extended conjugation and delocalization of the positive
charge on the imine N atom of the protein complex of
chromophore II results in the large red-shifted UV-vis maxi-
mum (597 nm as compared to 560 nm for native BR) and opsin
shift (3091 cm-1). The much lower UV-vis maximum (485
nm) and opsin shift (43 cm-1) for the protein complex of III
are due to the presence of the methyl group in the chain which
offers steric hindrance to the proper recognition of chromophore
III by the BOP binding site. This methyl group may also cause
the out-of-plane twisting of the olefinic bond in the side chain
of III within the restricted protein environment and consequent
disruption of the extended conjugation and loss of stereoelec-
tronic stability for the protein-chromophore complex. It is
possible for the steric bulk of the R-methyl group to prevent
(13) Honig, B.; Hudson, B.; Sykes, B. D.; Karplus, M. Proc. Natl. Acad.
Sci. U.S.A. 1971, 68, 1289.

6190 J. Am. Chem. Soc., Vol. 118, No. 26, 1996
Singh et al.
Table 6. Functional Characteristics of Chromophore-Substituted
Bacteriorhodopsins
isomerization cannot be ruled out in the case of BR-II. BR-
III does not show any detectable transient absorption which
may be due to the steric hindrance of the methyl group which
restricts the isomerization of III within BR-III.
transientb
absorption
lifetime, τ (ms)
lighta-induced
pH change
protein
chromophore
Whether trans-azobenzene is indeed isomerized to cis-
azobenzene was tested by steady state photolysis of chro-
mophores I and II. Photolysis of the two chromophores I and
II in methanol and in aqueous buffer (1.0 × 10-2 M Tris‚HCl,
pH 5) at their respective UV-vis absorption maxima using a
monochromated source (450 W xenon arc lamp with a Spex
1681B Miniate-2 monochromator) does not result in any change
in their UV-vis absorption maxima or absorbance. Thus, large
spectral changes in the absorption spectra of the chromophores
I and II are not observed under steady state photolysis
conditions. However, this does not imply absence of cis-trans
isomerization. It is known that extremely fast Z to E thermal
recovery in protic solvents occurs when (E)-azobenzenes are
photolyzed. Therefore, large spectral changes are not de-
tected.14,15 The chromophores studied by us have both electron-
donating and electron-accepting groups; i.e., they belong to the
class of push-pull substituted azobenzenes. Steady state
photolysis of these chromophores in methanol and in aqueous
buffer does not result in observable spectral shifts probably due
to hydrogen bond interactions between methanol/aqueous buffer
and the chromophores I and II.
BR-I
BR-II
BR-III
BR
I
II
III
retinal
+
+
-
+
1.5
1.0
7.0
a 1.0 × 10-2 M KCl-HCl, pH 5. b Excitation flash 400-700 nm,
<100 µs.
Moreover, in contrast to the usual time scale for Z to E
thermal recovery of azobenzenes (e.g., 100-500 µs for 4-(N,N-
dimethylamino)-4′-nitro-1,1′-azodibenzene in various organic
solvents at 30 °C),16 the considerably different values for the
azo-modified BR analogues indicates the influence of the
surrounding protein environment on the photoisomerization of
the chromophores I-III.
Nature of the Protein-Chromophore Interaction in BR-
I, and BR-II, and BR-III in Terms of Longitudinal
Restrictions of the Protein Binding Site and the External
Point Charge Model. It is evident from the above discussion
that there is strong interaction between the azo chromophores
I, II, and III and the apoprotein. This is manifested by the
altered physicochemical characteristics of BR-I, BR-II, and
BR-III in comparison to native BR. An understanding of the
factors involved for this altered behavior is necessary to explain
the interactions occurring within these proteins. This may also
throw new light on chromophore-protein interactions in native
BR.
The chromophore binding site of BR imposes longitudinal
restrictions on the retinylidene chromophore, which results in
the highly regiospecific photoisomerization process.17 For all-
trans-retinal, the distance between the center of the trimethyl-
cyclohexene ring and C-15 is 12.3 Å, whereas the 13-cis-retinal
it is 11.2 Å. Similar distances calculated for azo chromophores
I-III in the fully extended geometry are 11.0 Å for I and 13.8
Å for both II and III. Thus, the distances for I-III fall within
the binding zone of BOP.
The role of point charges in causing the large bathochromic
shift in native BR is much discussed.10 A negative point charge
is located at a distance of 3.5 Å above C-5 of the retinylidene
chromophore of BR. A second negative point charge is located
at a distance of 3.0 Å below the protonated Schiff base nitrogen
and acts as the counterion. On the basis of the recently proposed
Figure 4. Flash photolysis traces for (a) native BR and (b) analogue
BR-II with a maximum at 410 nm (x axis 1.0 ms per division; Y axis
0.5 V, 4.0% T per division) at 25 °C.
repeated light-on/light-off cycles due to the greater similarity
between the chromophore of native BR and the chromophore
of BR-II. This is not the case with BR-I and BR-III. Azo
chromophore I can undergo isomerization in the presence of
light inside the protein which may then change conformation,
exposing fresh surface residues with ionizable protons. These
protons are expelled by the protein, causing the observed pH
changes. However, the protein may not be able to provide the
energy to reverse the isomerization of azo chromophore I, and
so the photocycle is not completed. Therefore, the pH change
initially observed is not repeated. In the case of BR-III the
protein environment is probably too restrictive for the molecular
geometry of chromophore III to allow its photoisomerization
and resultant light-induced pH change effects. However, more
studies are required to fully understand this light-induced pH
change activity of the analogues.
Flash Photolysis Studies. Modified BR containing azo
chromophores I-III were flash photolyzed, and formation of
transient species was recorded. Figure 4 shows the transient
absorption for analogue BR-II and native BR. The transient
lifetime (τ) is given by the time at 1/eImax where Imax is the
highest intensity of the signal. BR-I and BR-II produced
transient species with lifetimes of 1.5 and 1.0 ms, respectively.
BR-III, however, did not exhibit formation of such transient
species. Under similar photolytic conditions, native BR pro-
duced a transient species with a lifetime of 7.0 ms at 410 nm.
Thus, compared to the transient lifetime for native BR (7.0 ms),
the modified BR analogues exhibit transient species with shorter
lifetimes (Table 6). Formation of transient species in the case
of BR-I and BR-II may be due to the photoisomerization of
the protein bound azo chromophore (-NdN-). Formation of
such transient species even for a chromophore which is devoid
of a -CdC- bond, e.g., BR-I, reinforces this assumption.
However, contribution to the transient profile from -CdC-
(14) Lovrien, R.; Pesheck, P.; Tisel, W. J. Am. Chem. Soc. 1974, 96,
244.
(15) Sanchez, A. M.; deRossi, R. H. J. Org. Chem. 1995, 60, 2974.
(16) Asano, T.; Okada, T.; Shinkai, S.; Shigematsu, K.; Kusano, Y.;
Manabe, O. J. Am. Chem. Soc. 1981, 103, 5161.
(17) Matsumoto, H.; Liu, R. S. H.; Simmons, C. J.; Seff, K. J. Am. Chem.
Soc. 1980, 102, 4259.

Bacteriorhodopsin Analogues Based on Azo Chromophores
J. Am. Chem. Soc., Vol. 118, No. 26, 1996 6191
three-dimensional model of BR, the charge above C-5 is due
to Asp 96 and the counterion to the Schiff base is Asp 212.18
The electronic and steric properties of the azo chromophores I,
II, and III differ greatly from those of retinal, and the effect of
the two protein bound point charges is expected to be very
altered. While the trimethylcyclohexene ring of retinal is
replaced by the phenyl ring in chromophores I-III, the presence
of the -NdN- group provides additional sites for protona-
tion/H bond interaction with charge-carrying protein bound
residues. These interactions may place additional restrictions
on the isomerization of the azo chromophore. The negative
opsin shift of BR-I could have a partial contribution from the
repulsive interaction between Arg 82 and the positive charge
on the N,N-dimethylamino group of chromophore I, though in
native BR the trimethylcyclohexene group is situated away from
Arg 82. The effect of the positive charge on the N,N-
dimethylamino group is offset by the availability of extended
conjugation in the chromophores of BR-II and BR-III,
resulting in net positive opsin shifts.
to BOP also points toward intimate interactions occurring
between the respective azo chromophores and the protein bound
residues.
Thus, careful evaluation of the available information on the
azo-substituted BR-I, BR-II, and BR-III reveals several
interesting facets of chromophore-protein interaction in BR.
Conclusion
The formation of BR-I, BR-II, and BR-III by the
interaction of chromophores I, II, and III, respectively, with
BOP indicates that the apoprotein can, under appropriate
conditions, accommodate molecules in its binding site which
have molecular geometries markedly different from those of
retinal. It is also evident that the chromophores are effective
pH sensitive probes for the inner hydrophobic domain of the
protein.
The current investigations have defined a new direction for
preparing novel BR-based pigments and also provide new
insight into BOP-chromophore interactions.
The residues Trp 86, Trp 138, Trp 182, and Trp 189 are
present in the retinal pocket of the protein, and the quenching
of fluorescent emission in BR-I, BR-II, and BR-III relative
Acknowledgment. The authors thank the reviewers for their
valuable suggestions and the funding agencies (DST, DAE,
CSIR, Government of India) for financial support.
(18) Henderson, R.; Baldwin, J. M.; Ceska, T. A.; Zemlin, F.; Beckmann,
E.; Downing, K. H. J. Mol. Biol. 1990, 213, 899.
JA954286X