Novel retinylidene iminium salts for defining opsin shifts: Synthesis and intrinsic chromophoric properties

Article abstract of DOI:10.1039/b600121a

Retinal Schiff bases serve as chromophores in many photoactive proteins that carry out functions such as signalling and light-induced ion translocation. The retinal Schiff base can be found as neutral or protonated, as all-trans, 11-cis or 13-cis isomers

Full text of DOI:10.1039/b600121a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Novel retinylidene iminium salts for defining opsin shifts: synthesis
and intrinsic chromophoric properties
a
b
b
a
b
˚
Michael Axman Petersen, Iben B. Nielsen, Michael B. Kristensen, Anders Kadziola, Lutz Lammich,
Lars H. Andersen*b and Mogens Brøndsted Nielsen*a
Received 5th January 2006, Accepted 17th February 2006
First published as an Advance Article on the web 13th March 2006
DOI: 10.1039/b600121a
Retinal Schiff bases serve as chromophores in many photoactive proteins that carry out functions such
as signalling and light-induced ion translocation. The retinal Schiff base can be found as neutral or
protonated, as all-trans, 11-cis or 13-cis isomers and can adopt different conformations in the protein
binding pocket. Here we present the synthesis and characterisation of isomeric retinylidene iminium
salts as mimics blocked towards isomerisation at the C11 position and conformationally restrained. The
intrinsic chromophoric properties are elucidated by gas phase absorption studies. These studies reveal a
small blue-shift in the S0→S1 absorption for the 11-locked derivative as compared to the unlocked one.
The gas phase absorption spectra of all the cationic mimics so far investigated show almost no
absorption in the blue region. This observation stresses the importance of protein interactions for
colour tuning, which allows the human eye to perceive blue light.
Bacteriorhodopsin shows a maximum at 570 nm.7 Moreover,
the cascade of conformational changes occurring after light-
induced isomerisation of both bacteriorhodopsin and rhodopsin
provides a series of conformers each of which experiences a unique
absorption maximum.5,10
We have recently determined experimentally the gas phase
absorption spectra of all-trans retinal derivatives 1 and 2 (Fig. 1)
in the protonated or methylated Schiff base form, respectively
(cationic mimics).11 We used a technique where photo-absorption
was registered by detection of photo-fragments. Briefly, the
Introduction
Many photoactive proteins that perform visual signalling and
light-induced ion translocation contain the retinal chromo-
phore.1–5 Thus, all vertebrate visual pigments are integral-
membrane proteins containing a protonated Schiff base of 11-
cis-retinal at the seventh a-helix of opsin. Vertebrates have two
kinds of photoreceptor cells, called cones and rods. Cones function
in bright light and are responsible for colour vision, whereas
rod cells function in dim light but do not perceive colour. The
primary event in visual excitation is isomerisation of the 11-
cis-retinal to all-trans-retinal, which leads to significant changes
in the geometry and to significant strain in the protein. After
a series of conformational changes, the imine is deprotonated,
and all-trans-retinal is eventually liberated through a hydrolysis
step.5 Bacteriorhodopsin, present in the purple membrane of
Halobacterium salinarum,6 contains a protonated Schiff base of
all-trans-retinal. This protein is responsible for transmembrane
proton-pumping,7 a process that is fuelled by light-induced 13-
trans→13-cis isomerisation.
Our ability to distinguish colours relies on a fascinating tuning
of the chromophore absorption maximum via subtle protein
interactions.1,2,5 Absorption data are usually compared to those
of protonated retinal Schiff bases in methanol solution where
the absorption maximum is at about 440 nm for both cis- and
trans-configurations of the chromophore.8 The shift in absorption
maximum from this value as experienced by a protein is termed
‘the opsin shift’.9 Rhodopsin present in rods exhibits an absorption
maximum at ca. 500 nm, i.e. an opsin shift of 60 nm, whereas
proteins present in human cone pigments absorb with a maxi-
mum somewhere in the region between ∼420 and ∼560 nm.2,5
Fig. 1 Compounds 1 and 2 are mimics of the protonated retinal Schiff
base present in proteins (R1 = lysine side chain and R2 = H). The 11-locked
derivative 3 presents a new target molecule for the investigation of intrinsic
chromophoric properties.
aDepartment of Chemistry, University of Copenhagen, Universitetsparken 5,
DK-2100, Copenhagen Ø, Denmark. E-mail: mbn@kiku.dk
bDepartment of Physics and Astronomy, University of Aarhus, DK-8000,
Aarhus C, Denmark. E-mail: lha@phys.au.dk
1546 | Org. Biomol. Chem., 2006, 4, 1546–1554
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
chromophore molecules were transferred to the gas phase by
electrospray ionisation and injected into an electrostatic ion
storage ring where they were irradiated by a tuneable, pulsed
laser.12 On account of the increased energy (typically 2 eV) after
absorption of a single photon, the microcanonical temperature
of the chromophore ions was increased by several hundreds of
Kelvin and spontaneous Arrhenius-type fragmentation happened
over a time scale of about a millisecond.13 The storage-ring
technique enabled us to keep excited chromophores and wait
for delayed decays to occur. Upon spontaneous fragmentation,
neutral photo-fragments were not kept stored by the storage ring,
but were detected on a dedicated particle detector, the signals
of which made up the absorption spectrum, after normalisation
to the number of photons in the laser pulse and the number
of ions in the storage ring. The method allows investigation
of charged molecules that cannot be studied in the gas phase
by simple evaporation. Moreover, absorption data of isolated
chromophores in the gas phase are highly desirable for a number
of reasons: Firstly, they provide the best reference for quantum
chemical calculations. Secondly, for protonated chromophores like
protonated Schiff base retinal, there is always a negative counter
ion in the solutions which may not easily be controlled. Thirdly, a
series of studies have indeed suggested that the conditions inside
many photo-active proteins like the Green Fluorescent Protein are
almost vacuum like.14
Scheme 1 Reagents and conditions: (a) (EtO)2POCH2CN, NaH, THF, rt.
salt 3 together with the (13Z)-3 isomer in a ratio of 9 : 1 (Scheme 2).
These iminium salt isomers were impossible to separate. Reduction
of nitrile (13Z)-5 gave aldehyde (13Z)-6 that, however, under
chromatographic work-up partly isomerised to the all-E isomer 6.
Further isomerisation occurred during reaction with dimethylam-
monium tetrafluoroborate to provide a mixture of (13Z)-3 and 3 in
favour of the all-E salt. The (13Z)-3 : 3 ratio of 1 : 8 is close to that
obtained (1 : 9) when starting from the all-E aldehyde 6. Reduction
of (9Z)-5 gave almost isomerically pure aldehyde (9Z)-6. However,
treatment with dimethylammonium tetrafluoroborate gave a 1 :
1 mixture of the expected iminium salt (9Z)-3 together with the
all-E isomer 3 (including traces of 13Z isomers). Almost the
same isomer distribution was obtained when starting the synthetic
sequence with the nitrile (9Z,13Z)-5 instead. Here isomerisation
prevented us from obtaining the targeted product (9Z,13Z)-3.
From the iminium salt product distributions, it is very clear
that 9Z→9E and, in particular, 13Z→13E isomerisations occur
readily. In consequence, the isolated products were in favour of the
all-E isomer 3 (or at least ∼50%), whatever aldehyde was chosen
as precursor.
To shed further light on the intrinsic chromophoric properties of
cationic Schiff bases of retinal and any possible differences between
isomers as well as conformers, we decided to lock the C11 double
bond towards isomerisation. Since the absorption behaviour of
compounds 1 and 2 is very similar,11 the methylated iminium
derivative 3 was chosen as the next target compound rather than a
protonated one. A locked chromophore may naturally have slightly
different vibrational properties which may be displayed in the
electronic absorption spectrum.
Results and discussion
Synthesis
The synthesis of isomeric 11-locked retinylidene iminium salts
starts from ketones 4 (all-E) and (9Z)-4 (Scheme 1).15 The
conversion of these ketones into immediate precursors for the
iminium salts follows a synthetic protocol previously reported
by Albeck et al.,15 but in contrast to this protocol we isolated
all intermediate isomers by careful chromatographic separation
when possible in order to subject each isomer to individual
transformations. For simplicity, we shall only specify the positions
of Z-configuration in the compound labeling. Isomer designations
were performed by a combination of 1H NMR spectroscopy and
X-ray crystallographic analyses (vide infra).
First, the ketone 4 was subjected to an Emmons–Horner
reaction with diethyl cyanomethylphosphonate to afford the two
nitriles 5 and (13Z)-5 (Scheme 1). These two isomers were
separated by careful column chromatographic work-up. Similarly,
the ketone (9Z)-4 was transformed into nitriles (9Z)-5 and
(9Z,13Z)-5.
1H NMR spectroscopy
The 1H NMR chemical shifts of the locked retinal isomers
are collected in Table 1; the atom numbering is depicted in
Fig. 2. Significant downfield shifts are observed for the protons
at C12 (H–C12) and C15 (H–C15) in the 13Z-isomers [(13Z)-6 and
(9Z,13Z)-6] relative to those in the 13E-isomers [6 and (9Z)-6].
Reduction of 5 with diisobutyl aluminiumhydride (DIBALH)
provided the aldehyde 6 that was subsequently treated with
dimethylammonium tetrafluoroborate to provide the new iminium
Fig. 2 Atom numbering; cf. Table 1.
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 1546–1554 | 1547
©

View Article Online
Table 1 1H NMR chemical shifts (dH/ppm) of locked retinal derivatives in CDCl3
Compd.
H–C22 H–C21 H–C20 H–C19 H–C18
H–C16/17 H–C15 H–C14 H–C12 H–C10 H–C8
H–C7
H–C4 H–C3 H–C2
6
2.88
2.86
2.47
1.86
1.83
1.84
1.83
2.49
2.38
2.47
2.37
2.07
2.03
2.08
2.00
1.71
1.72
1.71
1.71
1.03
1.02
1.03
1.02
10.06
10.05
10.17
10.14
5.81
5.77
5.70
5.67
6.25
6.23
7.18
7.10
5.95
5.85
5.98
5.89
6.11
6.55
6.12
6.56
6.28
6.33
6.27
6.33
2.02
2.03
2.02
2.00
1.62
1.60
1.62
1.60
1.47
1.45
1.47
1.45
(9Z)-6
(13Z)-6
(9Z,13Z)-6 2.46
Scheme 2 Reagents and conditions: (a) DIBALH, pentane, −78 ◦C; (b) SiO2, Et2O, 0 ◦C; (c) BF4NH2Me2, 4 A sieves, EtOH, 0 ◦C.
˚
In contrast, the protons at C14 and C22 are significantly upfield-
shifted in the 13Z-isomers. The 9Z-isomers experience upfield
shifts for the protons at C10, C19, and C20 relative to those in the
9E-isomers, while downfield shifts are observed for the protons at
C7 and C8.
X-Ray crystallographic analysis
Single crystals of the nitriles (9Z)-5 and (9Z,13Z)-5 were grown
from heptane by slow evaporation at ca. 5 ◦C. The solvent
was allowed to completely evaporate since the crystals dissolve
at room temperature. The X-ray crystal structures confirm the
isomer designations and are revealed in Figs. 3 and 4. Both
structures adopt the s-cis conformation at the “C8–C11” bond
(defined according to crystal structure atom numbering). The
Fig. 3 X-ray crystal structure of (9Z)-5. Drawing made by ORTEP-II.16
The atom ellipsoids are drawn at the 50% probability level.
conjugated section is non-planar for both molecules and can be
separated in three intersecting planes defined by atoms C9, C8, C11
1548 | Org. Biomol. Chem., 2006, 4, 1546–1554
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
Computational study
The locked derivatives 3, (9Z)-3, (13Z)-3 and (9Z,13Z)-3 were
subjected to semiempirical (PM3) geometry optimisations (in
differently preselected s-cis and s-trans conformations), employing
the Gaussion 03 program package.17 The PM3 method was chosen
for geometry optimisation since it was previously found to describe
the structures of retinal Schiff bases well, whereas it largely
underestimates rotational barriers.18 Frequencies were calculated
to verify that the calculated structures are local minima on the
potential energy surface and to correct for zero-point kinetic
energies. Density Functional Theory (DFT) single-point energies
of the optimised structures were calculated at the B3LYP/6–31 +
G(d) and B3LYP/6–311++G(2d,p) levels. Optimised structures
and relative energies (at 0 K) are revealed in Fig. 5. The 6,7
s-cis conformer of the all-E isomer 10 was found to be 0.3–
0.5 kcal mol−1 more stable than the s-trans conformer.19 The
conjugated section is almost planar in both conformers, not
including, however, the cyclohexene double bond (C5–C6–C7–
C8 torsional angle of ca. 42◦). The conjugated sections of the 9Z
and 13Z isomers in 10,11 s-trans or s-cis conformations deviate
considerably from planarity, and these isomers are higher in energy
than the all-E isomer. It is worth pointing out that the two
Fig.
4
X-ray crystal structure of (9Z,13Z)-5. Drawing made by
ORTEP-II.16 The atom ellipsoids are drawn at the 50% probability level.
(plane 1), C11, C12, C13, C15, C16 (plane 2) and C16, C17, C19,
C22, C23 (plane 3). For (9Z)-5, planes 1 and 2 intersect at an angle
of 48.6◦, planes 2 and 3 at an angle of 46.5◦, and planes 1 and 3 at
an angle of 78.6◦. For (9Z,13Z)-5, the corresponding angles are
44.4◦, 51.7◦ and 82.9◦.
Fig. 5 PM3-optimised structures and relative energies calculated at the a) B3LYP/6–31 + G(d) and b) B3LYP/6–311++G(2d,p) level.
This journal is The Royal Society of Chemistry 2006 Org. Biomol. Chem., 2006, 4, 1546–1554 | 1549
©

View Article Online
optimised conformations of the 9Z,13Z isomer are considerably
energetic in view of the fact that this isomer could not be isolated
in our attempted synthesis. The 10,11 s-cis conformation of both
(9Z)- and (13Z)-3 is more stable than the corresponding s-trans
conformation, while the opposite is true for (9Z,13Z)-3 where the
10,11 s-cis conformation enforces unfavourable steric interactions.
These interactions are less severe for the nitrile (9Z,13Z)-5 for
which the 10,11 s-cis conformation is in fact more stable (by 2.1–
2.2 kcal mol−1) than the s-trans conformation. Indeed, the 10,11
s-cis conformation is adopted by (9Z,13Z)-5 in the solid state
according to the X-ray crystallographic analysis (vide supra).
Absorption spectroscopy
Fig. 6 Absorption spectra of 2 (2.8 × 10−5 M; all-E/13Z 7 : 1) and 3
(5.2 × 10−5 M; all-E/13Z 9 : 1) in CH2Cl2.
The longest wavelength absorption maxima for the nitrile and
aldehyde isomers as well as for the iminium salts in different
solvents are listed in Tables 2 and 3. Moreover, the gas phase
absorption maxima (vertical transitions) previously measured for
1 and slightly modified data for 2 are included in the table,11
together with the new gas phase data obtained for the locked
cationic mimic 3 (including isomer impurities).
Despite the preference for non-planar structures of (9Z)-3 and
(13Z)-3 as predicted from the calculational study, these isomers
experience absorption maxima in solution close to that of the all-
E isomer 3, for which the linear chain is almost planar according
to the calculations. Thus, all isomers absorb in the region kmax
489–494 nm in CH2Cl2 as inferred from a careful comparison of
the absorption spectra of the different isomer mixtures isolated.
This region is only slightly blue-shifted relative to the maximum
of iminium salt 2 (Fig. 6). A slightly more significant blue-shift is
observed when comparing the related bacteriorhodopsins. Thus,
Albeck et al.15 incubated aldehyde 6 with bacteriorhodopsin and
obtained pigments that absorbed at kmax 556 nm, while the natural
bacteriorhodopsin absorbs at kmax 570 nm.
Fig. 7 Absorption spectra of a) 2 (all-E/13Z 7 : 1) and b) 3 (all-E/13Z
9 : 1) in the gas phase.
In the gas phase (i.e., without an environment containing
negative counter-ions), compound 3 exhibits two main absorption
peaks at kmax = 546 and 594 nm (Fig. 7). The peak at the
longest wavelength is attributed to a vertical S0→S1 transition
without significant vibrational excitation in the S1 state. The energy
difference between the peaks corresponds to 1480 cm−1 which
indicates that the structure at 546 nm may be related to C–C
vibrations in the excited S1 state. Thus, the calculated IR spectrum
shows a strong absorption in this neighbourhood at 1410 cm−1
(value obtained after scaling with the PM3 scaling factor 0.9740).
It should be noted that the transition to the second excited singlet
state (S2) is expected at a wavelength slightly below 400 nm.20 In
the gas phase, compound 2 exhibits two main absorption bands
which peak at kmax = 540 and 608 nm (Fig. 7). It appears that many
vibrations may be excited in the S0→S1 transition; according to
the spectrum, vibrational energies near 1400 and 3400 cm−1 may
in particular be noticed.
Table 2 Longest wavelength absorption maxima (kmax/nm) of nitriles and
aldehydes in solution
Compd.
Hexane
Compd.
Hexane
CH2Cl2
5
344
343
343
348
6
358 (356a)
359
378
379
368
373
382
(9Z)-5
(13Z)-5
(9Z,13Z)-5
(9Z)-6
(13Z)-6
(9Z,13Z)-6
all-trans-retinal
352 (350a)
355
369
The gas phase maximum (594 nm) of the locked chromophore 3
is slightly blue-shifted relative to that observed for chromophore 2
(608 nm), measured under exactly the same conditions. The small
a Ref. [15].
Table 3 Longest wavelength absorption maxima (kmax/nm) of retinylidene iminium compounds in solution and gas phase
CH2Cl2
MeOH
Dioxane
DMSO
447
Gas phase
1
2
498
494
489–494
440
384, 450sh
540sh, 610a
540, 608b
546, 594
448
442–446
3, (9Z)-3, (13Z)-3
a Ref. [11]. b Slightly modified relative to those values published in ref. [11]; sh = shoulder.
1550 | Org. Biomol. Chem., 2006, 4, 1546–1554
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
blue-shift (14 nm) may be ascribed to a more limited conjugation
in the locked chromophore. The intrinsic absorption maxima of
compounds 2 and 3 are both red-shifted by 38 nm relative to
those of the natural and artificial bacteriorhodopsins, respectively,
which they are mimicking. Notably, all absorptions measured in
solution (even in dichloromethane) are at higher energy than that
of bacteriorhodopsin, and one would therefore have to be very
careful in describing the influence of protein interactions by direct
comparison with solvent values. The protein induces an increase,
and not a decrease, in the energy of the transition relative to that
of the isolated chromophore; this conclusion can only be drawn
unambiguously from the gas phase investigation.
in Hz. Samples were prepared using CDCl3 (Cambridge Isotope
˚
Labs) dried over molecular sieves (4 A) and neutralised with
basic Al2O3. Fast atom bombardment (FAB) mass spectra were
obtained on a Jeol JMS-HX 110 Tandem Mass spectrometer in
the positive ion mode using either 3-nitrobenzyl alcohol (NBA)
or glycerol as matrix. Microanalyses were performed at the
Microanalytical Laboratory at the Department of Chemistry,
University of Copenhagen. Solution absorption spectra were
recorded on a Shimadzu UV 1601PC instrument. Gas phase
absorption spectra were obtained by applying the ELectrostatic
Ion Storage ring in Aarhus (ELISA) as described elsewhere.12
Finally, we would like to emphasise that the cationic mimics
experience gas phase absorption spectra where the absorption
in the blue region is very low. The large absorption blue-shift
experienced in particular by blue cone pigments (kmax 425 nm)
is explained by a polar protein environment in proximity to
the protonated iminium nitrogen as well as a closely situated
negatively charged counter-ion.2,21 These influences lead to a
higher stabilisation of the ground-state relative to the excited state
and hence a blue-shift. Similarly, a polar environment accounts
for the high-energy absorption experienced in methanol solution
(Table 3).
Compounds 5 and (13Z)-5
NaH (60% in mineral oil, 130 mg, 3.25 mmol) was washed
with dry n-pentane and taken up in dry DMF (60 mL) under
N2. Diethyl cyanomethylphosphonate (600 mg, 3.3 mmol) was
added, and the solution was stirred for 15 min at rt followed
by addition of isomerically pure 4 (800 mg, 2.8 mmol). After
stirring at rt for 3 h, the reaction was quenched with sat. aqueous
NH4Cl (50 mL) followed by extraction with diethyl ether (3 ×
50 mL). The combined organic phases were dried (MgSO4) and
evaporated in vacuo. Purification with repeated (3 ×) flash column
chromatography (SiO2, EtOAc–heptane 15 : 85 v/v) gave the two
isomers 5 (360 mg, 42%) and (13Z)-5 (227 mg, 26%) as yellow oils
which solidified upon freezing. Compound 5: Rf = 0.53 (EtOAc–
heptane 3 : 17); dH (300 MHz, CDCl3, 25 ◦C, 7.26 ppm) 6.25 (1 H,
d, J = 16.1 Hz, CH), 6.18 (1 H, s, CH), 6.08 (1 H, d, J = 16.1 Hz,
CH), 5.90 (1 H, s, CH), 5.00 (1 H, s, CHCN), 2.62 (2 H, t, J =
6.2 Hz, CH2), 2.42 (2 H, t, J = 6.2 Hz, CH2), 2.01–1.99 (5 H, m,
CH3, CH2), 1.81 (2 H, m, J = 6.2 Hz, CH2), 1.70 (3 H, s, CH3),
1.63–1.57 (2 H, m, CH2), 1.48–1.44 (2 H, m, CH2), 1.02 (6 H, s,
CH3); dC (75 MHz, CDCl3, 25 ◦C, 77.0 ppm) 158.5, 147.4, 139.1,
138.1, 137.5, 131.0, 129.8, 128.9, 127.1, 118.3, 91.4, 39.5, 34.2,
33.0, 30.2, 28.9, 28.1, 22.2, 21.7, 19.2, 14.7; m/z (FAB+, NBA) 308
(MH+). Compound (13Z)-5: Rf = 0.47 (EtOAc–heptane 3 : 17);
Conclusions
Although isomerically pure samples were not obtained on ac-
count of ready isomerisation, solution studies have revealed that
isomeric 11-locked retinal Schiff bases exhibit almost identical
absorption maxima, i.e., structural deviations from planarity have
a negligible influence. Along the same line, only a small blue-shift
in absorption maximum is observed in the gas phase relative to
a conformationally flexible, unlocked derivative. In our previous
studies on unlocked retinal Schiff base derivatives,11 the presence of
two absorptions in the gas phase spectrum was somewhat puzzling.
We have now confirmed that both these absorptions are indeed real
and they are most likely the result of vibrational excitation in the
S1 state. Moreover, we can conclude that in the absence of protein
interactions, the human eye would not detect light of wavelength
shorter than 450 nm. In other words, we would not be able to
perceive blue light.
◦
dH (300 MHz, CDCl3, 25 C, 7.26 ppm) 6.67 (1 H, s, CH), 6.26
(1 H, d, J = 16.1 Hz, CH), 6.10 (1 H, d, J = 16.1 Hz, CH), 5.95
(1 H, s, CH), 4.90 (1 H, s, CHCN), 2.44–2.38 (4 H, m, CH2), 2.08
(1 H, s, CH3), 2.01 (2 H, t, J = 6.2 Hz, CH2), 1.78 (2 H, m, CH2),
1.70 (3 H, s, CH3), 1.61 (2 H, m, CH2), 1.46 (2 H, m, CH2), 1.02
(6 H, s, CH3); dC (75 MHz, CDCl3, 25 ◦C, 77.0 ppm) 158.0, 147.8,
139.6, 138.2, 137.5, 131.1, 129.8, 129.0, 124.9, 117.6, 90.1, 39.5,
34.2, 33.0, 30.6, 30.2, 28.9, 22.5, 21.7, 19.2, 14.7; m/z (HR-FAB+,
NBA) 308.2351 (MH+, C22H30N requires 308.2378).
Experimental
General methods
Chemicals were purchased from Aldrich and used as received.
All reactions were carried out in the dark under an atmosphere
of Ar or N2. Thin-layer chromatography (TLC) was carried out
using aluminium sheets pre-coated with silica gel 60F (Merck
5554). The plates were inspected under UV light. Dry column
chromatography was carried out using silica gel 60 (Merck 15111,
0.015–0.40 mm). Flash column chromatography was carried out
using silica gel 60 (Merck 9385, 0.040–0.063 mm). Melting points
were measured on a Reichert melting point apparatus equipped
with a microscope. 1H NMR (300 MHz) and 13C NMR (75 MHz)
spectra were recorded on a Varian instrument, using the residual
solvent as the internal standard. All chemical shifts are quoted
on the d scale (ppm), and all coupling constants (J) are expressed
Compounds (9Z)-5 and (9Z,13Z)-5
Compounds (9Z)-5 (286 mg, 33%) and (9Z,13Z)-5 (450 mg, 52%)
were obtained as yellow solids in a similar way (with stirring
for 18 h) from isomerically pure (9Z)-4 (801 mg). Samples of
both compounds were recrystallised from heptane. Compound
(9Z,13Z)-5: m.p. 84.5–85.5 ◦C. Rf = 0.48 (EtOAc–heptane 3 : 17);
◦
dH (300 MHz, CDCl3, 25 C, 7.26 ppm) 6.58 (1 H, s, CH), 6.55
(1 H, d, J = 16.4 Hz, CH), 6.34 (1 H, d, J = 16.4 Hz, CH), 5.90
(1 H, s, CH), 4.90 (1 H, s, CHCN), 2.39 (4 H, m, CH2), 2.09–2.01
(5 H, m, CH2, CH3), 1.82–1.71 (5 H, m, CH2, CH3), 1.61 (2 H,
m, CH2), 1.47 (2 H, m, CH2), 1.04 (6 H, s, CH3); dC (75 MHz,
CDCl3, 25 ◦C, 77.0 ppm) 157.8, 147.8, 138.4, 137.7, 131.1
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 1546–1554 | 1551
©

View Article Online
(2 C-double intensity), 130.1, 129.5, 124.8, 117.4, 90.3, 39.6, 34.2,
33.1, 30.8, 30.4, 29.0, 22.5, 22.0, 21.7, 19.2; m/z (HR-FAB+,
glycerol) 308.2382 (MH+, C22H30N requires 308.2378). Compound
(9Z)-5: m.p. 104–104.5 ◦C (Found: C, 85.81; H, 9.64; N, 4.52.
C22H29N (307.47) requires C, 85.94; H, 9.91; N, 4.56%). Rf = 0.41
(EtOAc–heptane 3 : 17); dH (300 MHz, CDCl3, 25 ◦C, 7.26 ppm)
6.51 (1 H, d, J = 16.2 Hz, CH), 6.31 (1 H, d, J = 16.2 Hz, CH),
6.17 (1 H, s, CH), 5.82 (1 H, s, CH), 4.96 (1 H, s, CHCN), 2.62
(2 H, t, J = 5.9 Hz, CH2), 2.33 (2 H, t, J = 5.9 Hz, CH2), 2.06–1.99
(5 H, m, CH2, CH3), 1.80 (2 H, m, J = 6.4 Hz, CH2), 1.71 (3 H, s,
CH3),1.62 (2 H, m, CH2), 1.47 (2 H, m, CH2), 1.02 (6 H, s, CH3);
dC (75 MHz, CDCl3, 25 ◦C, 77.0 ppm) 158.5, 147.1, 138.3, 137.9,
131.4, 130.7, 130.0, 129.2, 127.0, 118.3, 91.5, 39.5, 34.2, 33.0, 30.3,
28.9, 28.2, 22.1, 22.0, 21.7, 19.1; m/z (FAB+, glycerol) 308.
CHO), 7.18 (1 H, s, CH), 6.27 (1 H, d, J = 15.8 Hz, CH), 6.12
(1 H, d, J = 15.8 Hz, CH), 5.98 (1 H, s, CH), 5.70 (1 H, d, J =
8.2 Hz, CH), 2.47 (4 H, t, J = 6.2 Hz, CH2), 2.08 (3 H, s, CH3),
2.02 (2 H, t, J = 6.2 Hz, CH2), 1.84 (2 H, m, J = 6.2 Hz, CH2), 1.71
(3 H, s, CH3), 1.62 (2 H, m, 2H, CH2), 1.47 (2 H, m, CH2), 1.03
(6 H, s, CH3); dC (75 MHz, CDCl3, 25 ◦C, 77.0 ppm) 189.6, 156.7,
148.6, 139.4, 138.2, 137.6, 131.7, 129.9, 129.1, 123.5, 122.7, 39.5,
34.3, 33.0, 32.1, 30.9, 28.9, 22.6, 21.7, 19.2, 14.8; m/z (HR-FAB+,
NBA) 311.2367 (MH+, C22H31O requires 311.2375).
Compound (9Z,13Z)-6
An isomeric mixture of (9Z,13Z)-9 (66%) and (13Z)-9 (19%)
was obtained from (9Z,13Z)-5 (together with unreacted starting
material (15%)). Further purification was not possible due to
isomerisation. Compound (9Z,13Z)-6: Rf = 0.45 (EtOAc–heptane
3 : 17)-isomerisation on plate; dH (300 MHz, CDCl3, 25 ◦C,
7.26 ppm) 10.14 (1 H, d, J = 8.3 Hz, CHO), 7.10 (1 H, s, CH), 6.56
(1 H, d, J = 15.8 Hz, CH), 6.33 (1 H, d, J = 15.8 Hz, CH), 5.89
(1 H, s, CH), 5.67 (1 H, d, J = 8.3 Hz, CH), 2.46 (2 H, t, J = 6.2 Hz,
CH2), 2.37 (2 H, t, J = 6.2 Hz, CH2), 2.00 (5 H, m, CH2, CH3),
1.83 (2 H, m, J = 6.2 Hz, CH2), 1.71 (3 H, s, CH3), 1.60 (2 H, m,
CH2), 1.45 (2 H, m, CH2), 1.02 (6 H, s, CH3); dC (75 MHz, CDCl3,
Compound 6
Isomerically pure 5 (234 mg, 0.76 mmol) was dissolved in dry
n-pentane (50 mL) and cooled to −78 ◦C. DIBALH (1 M in
hexanes, 2.4 mL, 2.4 mmol) was added, and the reaction mixture
was allowed to stir at −78 ◦C for 212 h. Then diethyl ether (40 mL)
and silica (5 g) were added, and after stirring overnight at 5 ◦C
the reaction mixture was filtered through Celite and evaporated in
vacuo. Purification by dry column chromatography (SiO2, EtOAc–
heptane 5% ) gave 6 as a yellow oil (233 mg, 99%) with a trace of
the 13Z isomer. Rf = 0.27 (EtOAc–heptane 3 : 17); dH (300 MHz,
◦
25 C, 77.0 ppm) 189.6, 156.7, 148.3, 138.3, 137.8, 131.3, 131.0,
130.2, 129.9, 123.6, 122.3, 39.5, 34.2, 33.0, 32.2, 31.0, 29.0, 22.5,
22.0, 21.7, 19.1; m/z (HR-FAB+, NBA) 311.2368 (MH+, C22H31O
requires 311.2375).
◦
CDCl3, 25 C, 7.26 ppm) 10.06 (1 H, d, J = 8.2 Hz, CH), 6.28
(1 H, d, J = 15.4 Hz, CH), 6.25 (1 H, s, CH), 6.11 (1 H, d, J =
15.4 Hz, CH), 5.95 (1 H, s, CH), 5.81 (1 H, d, J = 8.2 Hz, CH),
2.88 (2 H, t, J = 6.2 Hz, CH2), 2.49 (2 H, t, J = 6.2 Hz, CH2), 2.07
(3 H, s, CH3), 2.02 (2 H, m, CH2), 1.86 (2 H, m, J = 6.2 Hz, CH2),
1.71 (3 H, s, CH3), 1.62 (2 H, m, CH2), 1.47 (2 H, m, CH2), 1.03
(6 H, s, CH3); dC (75 MHz, CDCl3, 25 ◦C, 77.0 ppm) 190.7, 157.2,
149.1, 139.5, 138.2, 137.6, 131.5, 130.2, 129.9, 129.1, 125.0, 39.5,
34.3, 33.0, 30.6, 28.9, 25.0, 22.4, 21.7, 19.2, 14.8; m/z (HR-FAB+,
NBA) 311.2355 (MH+, C22H31O requires 311.2375).
Compounds 3 and (13Z)-3
Aldehyde 6 (49 mg, 0.16 mmol) was dissolved in absolute ethanol
at 0 ◦C (3 mL), whereupon dimethylammonium tetrafluoroborate
˚
(27 mg, 0.20 mmol) and 4 A molecular sieves wer◦e added. The
reaction mixture was allowed to stir carefully at 5 C overnight,
and then it was filtered and evaporated in vacuo. The deep red
solid residue was dissolved in dry CH2Cl2, filtered and evaporated
in vacuo followed by washing with dry n-pentane, which gave a
mixture of the two isomers 3 and (13Z)-3 (64 mg, 95%) in a 9 :
1 ratio. Compound 3: dH (300 MHz, CDCl3, 25 ◦C, 7.26 ppm)
8.53 (1 H, d, J = 11.6 Hz, CHN), 6.52 (1 H, s, CH), 6.43 (1 H,
d, J = 16.0 Hz, CH), 6.16 (1 H, d, J = 16.0 Hz, CH), 6.14 (1 H,
d, J = 11.6 Hz, CHCHN), 6.05 (1 H, s, CH), 3.66 (3 H, s, CH3),
3.42 (3 H, s, CH3), 2.84 (2 H, t, J = 6.2 Hz; CH2), 2.60 (2 H,
t, J = 6.2 Hz; CH2), 2.14 (3 H, s, CH3), 2.03 (2 H, m, CH2),
1.88 (2 H, m, J = 6.1 Hz, CH2), 1.71 (3 H, s, CH3), 1.61 (2 H,
m, CH2), 1.47 (2 H, m, CH2), 1.03 (6 H, s, CH3); dC (75 MHz,
CDCl3, 25 ◦C, 77.0 ppm) 169.4, 163.0, 158.1, 144.7, 138.0, 137.5,
131.9, 131.5, 131.4, 130.8, 112.9, 48.8, 40.0, 39.5, 34.3, 33.2, 31.1,
28.9, 25.8, 22.2, 21.8, 19.1, 15.3; NOESY experiments confirmed
structure; m/z (HR-FAB+, NBA): 338.2842 ([M-BF4]+, C24H36N
requires 338.2848). Compound (13Z)-3: dH (300 MHz, CDCl3,
25 ◦C, 7.26 ppm) – only a few characteristic signals can be assigned
due to signal overlap: 8.67 (1 H, d, J = 11.6 Hz, CHN), 7.07
(1 H, s, CH), 3.38 (3 H, s, CH3).
Compound (9Z)-6
Compound (9Z)-6 was obtained as a thick, orange oil (256 mg,
99%) in a similar way from (9Z)-5 (255 mg). Trace of isomer
impurities. Rf = 0.36 (EtOAc–heptane 3 : 17); dH (300 MHz,
CDCl3, 25 ◦C, 7.26 ppm) 10.05 (1 H, d, J = 8.3 Hz, CHO),
6.55 (1 H, d, J = 16.0 Hz, CH), 6.33 (1 H, d, J = 16.0 Hz, CH),
6.23 (1 H, s, CH), 5.85 (1 H, s, CH), 5.77 (1 H, d, J = 8.3 Hz,
CH), 2.86 (2 H, t, J = 5.9 Hz, CH2), 2.38 (2 H, t, J = 5.9 Hz,
CH2), 2.03 (5 H, m, CH2, CH3), 1.83 (2 H, m, J = 6.2 Hz, CH2),
1.72 (3 H, s, CH3), 1.60 (2 H, m, CH2), 1.45 (2 H, m, CH2), 1.02
(6 H, s, CH3); dC (75 MHz, CDCl3, 25 ◦C, 77.0 ppm) 190.7, 157.0,
148.6, 138.5, 137.9 131.5, 130.9, 130.1, 129.8, 129.6, 125.1, 39.5,
34.2, 33.0, 30.7, 29.0, 25.1, 22.4, 22.0, 21.8, 19.2; m/z (HR-FAB+,
NBA) 311.2362 (MH+, C22H31O requires 311.2375).
Compound (13Z)-6
An isomeric mixture of compound (13Z)-6 and 6 was obtained as
a yellow oil (132 mg, 85%) in a similar way from (9Z)-5 (154 mg).
Dry column chromatography led to further isomerisation to
6. Compound (13Z)-6: Rf = 0.32 (EtOAc–heptane 3 : 17); dH
Other isomer mixtures of 3
Aldehydes (9Z)-6, (13Z)-6 and (13Z)-6/6 were converted in a
similar way to isomer mixtures of 3. An unequivocal assignment
of 1H NMR resonances of isomers other than 3 was not possible.
◦
(300 MHz, CDCl3, 25 C, 7.26 ppm) 10.17 (1 H, d, J = 8.2 Hz,
1552 | Org. Biomol. Chem., 2006, 4, 1546–1554
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
Table 4 X-Ray diffraction parameters and statistics of compounds (9Z)-5 and (9Z,13Z)-5
(9Z)-5
(9Z,13Z)-5
Empirical formula
C22H29N
307.46
122(2)
C22H29N
307.46
122(2)
Mr
T/K
˚
k/A
0.71073
0.71073
Crystal system
Space group
Orthorhombic
P212121
9.8840(16)
90.0
11.7110(11)
90.0
16.091(2)
90.0
Orthorhombic
P212121
9.8070(8)
90.0
11.7490(12)
90.0
16.027(2)
90.0
˚
a/A
a/◦
˚
b/A
b/◦
˚
c/A
c /◦
3
˚
V/A
1862.6(4)
4
1.096
1846.7(4)
4
1.106
Z
qcalc/g cm−3
lMoKa/mm−1
0.062
0.063
F(000)
672
672
Crystal size/mm3
0.72 × 0.36 × 0.18
2.15–32.98
50833
0.59 × 0.49 × 0.37
2.15–35.01
52673
h range/◦
Reflections collected
Unique data
6990 [Rint = 0.0674]
8043 [Rint = 0.0449]
Obsd data [I > 2r(I)]
6132
7116
GOF on F2
1.094
1.117
R indices (all data)
Larg.diff.peak/hole/e A
R1 = 0.055, wR2 = 0.113
R1 = 0.055, wR2 = 0.109
−3
˚
0.31/−0.19
0.36/−0.14
Crystal structure determination of compounds (9Z)-5 and
(9Z,13Z)-5†
7 D. Oesterhelt, Angew. Chem., Int. Ed. Engl., 1976, 15, 17–24;
W. Stoeckenius, Acc. Chem. Res., 1980, 13, 337–344; U. Haupts, J.
Tittor and D. Oesterhelt, Ann. Rev. Biophys. Biomol. Struct., 1999, 28,
367–369; J. K. Lanyi, Nature, 1995, 375, 461–463.
8 R. S. Becker and K. Freedman, J. Am. Chem. Soc., 1985, 107, 1477–
1485; K. A. Freedman and R. S. Becker, J. Am. Chem. Soc., 1986, 108,
1245–1251.
9 N. K. Nakanishi, V. Balogh-Nair, M. Arnaboldi, K. Tsujimoto and
B. Honig, J. Am. Chem. Soc., 1980, 102, 7945–7947.
10 R. R. Birge, T. M. Cooper, A. F. Lawrence, M. B. Masthay, C. Vasilakis,
C.-F. Zhang and R. Zidovetzki, J. Am. Chem. Soc., 1989, 111, 4063–
4074.
Intensity data for (9Z)-5 and (9Z,13Z)-5 were collected at 122 K
on a Bruker-Nonius KappaCCD diffractometer equipped with an
Oxford Cryostream unit. Data were reduced with EvalCCD,22 the
structures solved by direct methods using SIR9723 and refined by
least-squares against F2 with SHELXL9724 as incorporated in the
maXus program.25 For both structures all non-hydrogen atoms
were anisotropically refined. All hydrogen atoms were found in
subsequent difference Fourier maps and refined using a riding
model. Experimental parameters and statistics are summarized in
Table 4.
11 L. H. Andersen, I. B. Nielsen, M. B. Kristensen, M. O. A. El Ghazaly,
˚
S. Haacke, M. B. Nielsen and M. A. Petersen, J. Am. Chem. Soc., 2005,
127, 12347–12350.
12 S. P. Møller, Nucl. Instrum. Methods Phys. Res., Sect. A, 1997, 394,
281–286; J. U. Andersen, P. Hvelplund, S. B. Nielsen, S. Tomita,
H. Wahlgreen, S. P. Møller, U. V. Pedersen, J. S. Forster and T. J. D.
Jørgensen, Rev. Sci. Instrum., 2002, 73, 1284–1287; L. H. Andersen,
O. Heber and D. Zajfman, J. Phys. B.: At., Mol. Opt. Phys., 2004,
37, R57–R88; S. B. Nielsen, J. U. Andersen, P. Hvelplund, B. Liu and
S. Tomita, J. Phys. B.: At., Mol. Opt. Phys., 2004, 37, R25–R56.
13 L. H. Andersen, H. Bluhme, S. Boye´, T. J. D. Jørgensen, H. Krogh,
I. B. Nielsen, S. Brøndsted Nielsen and A. Svendsen, Phys. Chem.
Chem. Phys., 2004, 6, 2617–2627.
Acknowledgements
The Lundbeck Foundation and the Velux Foundation are grate-
fully acknowledged for supporting this project. Moreover, we
acknowledge financial support from the Danish Natural Science
Research Council (grants # 2111-04-0018 and 21-03-0330).
14 L. H. Andersen, S. B. Nielsen and S. U. Pedersen, Eur. Phys. J. D, 2002,
20, 597–600; I. B. Nielsen, S. Boye´-Peronne, M. O. A. El Ghazaly,
M. B. Kristensen, S. B. Nielsen and L. H. Andersen, Biophys. J., 2005,
89, 2597–2604; S. Boye´, S. B. Nielsen, H. Krogh, I. B. Nielsen, U. V.
Pedersen, A. F. Bell, X. He, P. J. Tonge and L. H. Andersen, Phys.
Chem. Chem. Phys., 2005, 5, 3021–3026.
15 A. Albeck, N. Friedman, M. Sheves and M. Ottolenghi, J. Am. Chem.
Soc., 1986, 108, 4614–4618.
16 C. K. Johnson, 1976, ORTEP-II, A Fortran Thermal-Ellipsoid Plot
Program. Report ORNL-5138, Oak Ridge National Laboratory, Oak
Ridge, Tennessee, USA.
17 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A.
Robb, J. R. Cheeseman, J. A. Montgomery, Jr., T. Vreven, K. N.
Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone,
B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson,
H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X.
Li, J. E. Knox, H. P. Hratchian, J. B. Cross, C. Adamo, J. Jaramillo,
References
1 G. Wald, Science, 1968, 162, 230–239.
2 G. G. Kochendoerfer, S. W. Lin, T. Sakmar and R. A. Mathies, Trends
Biochem. Sci., 1999, 24, 300–305.
3 B. K. Fung and L. Stryer, Proc. Natl. Acad. Sci., 1980, 77, 2500–2504.
4 K. Palczewski, T. Kumasaka, T. Hori, C. A. Behnke, H. Motoshima,
B. A. Fox, I. Le Trong, D. C. Teller, T. Okada, R. E. Stenkamp,
M. Yamamoto and M. Miyano, Science, 2000, 289, 739–745.
5 L. Stryer, Biochemistry, 2000, (Freeman, New York).
6 D. Oesterhelt and W. Stoeckenius, Nature New Biol., 1971, 233, 149–
152.
† CCDC reference numbers 294256 and 294257. For crystallographic data
in CIF or other electronic format see DOI: 10.1039/b600121a
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 1546–1554 | 1553
©

View Article Online
R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi,
C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth,
P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D.
Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K.
Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S.
Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz,
I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y.
Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W.
Chen, M. W. Wong, C. Gonzalez and J. A. Pople, Gaussian 03, Revision
B.03, Gaussian, Inc., Pittsburgh PA, 2003.
conformation: R. van der Steen, P. L. Biesheuvel, R. A. Mathies and
J. Lugtenburg, J. Am. Chem. Soc., 1986, 108, 6410–6411.
20 I. B. Nielsen, L. Lammich and L. H. Andersen, Phys. Rev. Lett., 2006,
96, 018304.
21 For studies of the influence exerted by a charge in proximity to the
retinal chromophore, see for example: ref. 9; M. G. Motto, M. Sheves,
K. Tsujimoto, V. Balogh-Nair and K. Nakanishi, J. Am. Chem. Soc.,
1992, 114, 7947–7949; A. Albeck, N. Livnah, H. Gottlieb and M.
Sheves, J. Am. Chem. Soc., 1992, 114, 2400–2411.
22 A. J. M. Duisenberg, L. M. J. Kroon-Batenburg and A. M. M. Schreurs,
J. Appl. Crystallogr., 2003, 36, 220–229.
23 A. Altomare, M. C. Burla, M. Camalli, G. L. Cascarano, C. Giaco-
vazzo, A. Guagliardi, A. G. G. Moliterni and R. Spagna, J. Appl.
Crystallogr., 1999, 32, 115–119.
24 G. M. Sheldrick, 1997, SHELXL97. Program for the Refinement of
Crystal Structures, University of Go¨ttingen, Germany.
25 S. Mackey, C. J. Gilmore, C. Edwards, N. Stewart and K. Shankland,
1999, maXus. Computer Program for the Solution and Refinement
of Crystal Structures, Bruker Nonius, The Netherlands, Macscience,
Japan & The University of Glasgow.
18 H. Zhoua, E. Tajkhorshida, T. Frauenheim, S. Suhai and M. Elstner,
Chem. Phys., 2002, 277, 91–103.
19 The 6-s-trans conformation is adopted, however, for many retinylidene
iminium salts in the solid state; see for example: S. Harbison, S. Smith, J.
Pardoen, J. Courtin, J. Lugtenburg, J. Herzfeld, R. Mathies and R. G.
Griffin, Biochemistry, 1985, 24, 6955–6962; G. S. Harbison, P. P. J.
Mulder, H. Pardoen, J. Lugtenburg, J. Herzfeld and R. G. Griffin,
J. Am. Chem. Soc., 1985, 107, 4809–4816; G. R. Elia, R. F. Childs,
J. F. Britten, D. S. C. Yang and B. D. Santarsiero, Can. J. Chem.,
1996, 74, 591–601. Moreover, bacteriorhodopsin requires the 6-s-trans
1554 | Org. Biomol. Chem., 2006, 4, 1546–1554
This journal is
The Royal Society of Chemistry 2006
©