Chemoenzymatic synthesis of 11-cis-retinal photoaffinity analog by use of squid retinochrome

Full text of DOI:10.1021/ja970956i

5758
J. Am. Chem. Soc. 1997, 119, 5758-5759
Scheme 1
Chemoenzymatic Synthesis of 11-cis-Retinal
Photoaffinity Analog by Use of Squid Retinochrome
Babak Borhan, Roxanne Kunz, Andrew Y. Wang, and
Koji Nakanishi*
Department of Chemistry, Columbia UniVersity
New York, New York 10027
Nina Bojkova and Kazuo Yoshihara
Suntory Institute for Bioorganic Research
Osaka, Japan
ReceiVed March 26, 1997
We expect the photoactive 11-cis-3-diazo-4-oxoretinal (1) to
be crucial in clarifying the visual transduction process, but its
synthesis has not been successful. However, as described below,
the preparation has been achieved by incubation of trans isomer
2 with squid retinochrome which smoothly performed the critical
trans f 11-cis isomerization to yield 1.
The visual transduction process is initiated by isomerization
of the 11-cis-retinal chromophore in rhodopsin to all-trans,
cleavage of the chromophore/opsin bond and terminates in all-
trans-retinal and opsin.1 The cis f trans isomerization triggers
a chain of conformational changes in the opsin which induces
an enzymatic cascade leading to vision.1a,2 Scheme 1 depicts
the intermediates which have been identified based on low-
temperature spectroscopy.3 However, the movement of the
chromophore relative to the receptor opsin along the isomer-
ization pathway remains unknown.
Our objective is to clarify this crucial aspect on a molecular
structural basis using photoaffinity labeling as the main tool.
Photolysis of a pigment incorporating the nonisomerizable 11-
cis-locked retinal analog 3 (Scheme 1) resulted in clear-cut
cross-linking to Leu-266, thus revealing that the ionone C-3 is
in close contact with helix F of rhodopsin in the dark.4 On the
other hand, studies using a photoreactive 11-cis-retinal analog
in which the 11-ene is not locked showed that the C-3 region
cross-linked to both helices C and F;5 recent studies with spin
labels6 also showed that movements of these two helices were
involved in the light activation process. It is thus possible that
the cis f trans isomerization results in a “flip-over motion of
the ring” from the proximity of helix F to helix C as well as
movements of helices F and C and that this induces the
conformational changes responsible for the enzymatic cascade.
This scheme has been further corroborated by Sakmar and co-
workers who have demonstrated that the relative movements
of helices C and F is required for activation of the G-protein-
coupled receptor rhodopsin.7
We deemed it necessary to follow the isomerization pathway
in a temperature-resolved manner by using, in contrast to locked
analog 3, the rhodopsin analog incorporating unlocked retinal
1. After irradiation at 500 nm (-140 °C, “batho”), the
chromophore is cross-linked (254 nm) and the amino acid(s)
are sequenced. Similiar experiments will be performed at -40
°C (“lumi”), -15 °C (“meta-I”), etc. Since the photoaffinity
label should traverse the same path of 11-cis-retinal isomeriza-
tion, such sequential cross-linking experiments should allow one
to trace the relative movements of the chromophore and the
receptor opsin.8
Synthesis of 11-cis-retinal analogs, however, face problems
due to the chromophore instability. The problem is magnified
for analogs with functional groups which have to be introduced
after formation of the 11-cis-ene. Most schemes geared toward
introducing the 11-cis geometry close to the end of the synthesis
have either failed or resulted in low yields and formation of
complex isomeric mixtures.9 In contrast to 11-cis analogs, it
is known that synthesis of the more stable all-trans analogs are
more straightforward. It would therefore be advantageous to
synthesize the corresponding all-trans isomers and then intro-
duce the 11-cis geometry at the end of the synthesis. Herein,
we report a method for enzymatically introducing the 11-cis
moiety at the last stage of the synthesis from an all-trans retinal,
a reaction that could be performed with the necessary tritiated
analog and could be quite general.
(1) (a) Rando, R. R. Chem. Biol. 1996, 3. (b) Rando, R. R. Angew. Chem.,
Int. Ed. Engl. 1990, 29. (c) Nakanishi, K.; Crouch, R. Isr. J. Chem. 1995,
35, 253-272. (d) Birge, R. R.; Hubbard, L. M. Biophys. J. 1981, 34, 517.
(2) (a) Longstaff, C.; Calhoon, R. D.; Rando, R. R. Proc. Natl. Acad.
Sci. U. S. A. 1986, 83, 4209. (b) Longstaff, C.; Rando, R. R. Biochemistry
1986, 24, 8137. (c) Hofmann, K. P. Photobiochem. Photobiophys. 1986,
13, 309. (d) Wheeler, G. L.; Bitensky, M. W. Proc. Natl. Acad. Sci. U.S.A.
1977, 74, 4238-4242. (e) Fung, B. K.; Hurley, J. B.; Stryer, L. Proc. Natl.
Acad. Sci. U.S.A. 1981, 78, 152-156. (f) DeGrip, W. J.; Gray, D.; Gillespie,
J.; Bovee, P. H.; Van den Berg, E. M.; Lugtenburg, J.; Rothschild, K. J.
Photochem. Photobiol. 1988, 48, 497-504. (g) Konig, B.; Welte, W.;
Hofmann, K. P. FEBS Lett. 1989, 257, 163-166. (h) Franke, R. R.; Sakmar,
T. P.; Oprian, D. D.; Khorana, H. G. J. Biol. Chem. 1988, 263, 2119-
2123.
(3) (a) Birge, R. R. Ann. ReV. Biophys. Bioeng. 1981, 10, 315-354. (b)
Mathies, R. Methods Enzymol. 1982, 88, 633-643. (c) Cooper, A. Nature
1979, 282, 531-533. Yoshizawa, T.; Wald, G. Nature 1963, 197, 1279-
1286. (d) Emeis, D.; Kuhn, H.; Reichert, J.; Hofmann, K. P. FEBS Lett.
1982, 143, 29-34. (e) Bennett, N.; Michel-Villaz, M.; Kuhn, H. Eur. J.
Biochem. 1982, 127, 97-103.
The visual cells of cephalopods contain in addition to
rhodopsin, a photosensitive isomerase, retinochrome, which
performs the all-trans-retinal f 11-cis-retinal regeneration.10
In view of the efficiency of retinochrome in isomerization of
(7) (a) Han, M.; Lin, S. W.; Minkova, M.; Smith, S. O.; Sakmar, T. P.
J. Biol. Chem. 1996, 271, 32337-32342. (b) Sheikh, S. P.; Zvyaga, T. A.;
Lichtarge, O.; Sakmar, T. P.; Bourne, H. R. Nature 1996, 383, 347-350.
(8) Preliminary experiments have shown that photoactivation of the
diazoketo group in 1 does take place at the low temperature of -140 °C.
(9) (a) Negishi, E. I.; Owczarczyk, Z. Tetrahedron Lett. 1991, 32, 6683-
6686. (b) Wada, A.; Tanaka, Y.; Fujioka, N.; Ito, M. Bioorg. Med. Chem.
Lett. 1996, 6, 2049-2052. (c) Sporn, M. B.; Roberts, A. B.; Goodman, D.
S. The Retinoids: biology, chemistry, and medicine, 2nd ed.; Raven Press,
Ltd.: New York, 1994. (d) Hosoda, A.; Taguchi, T.; Kobayashi, Y.
Tetrahedron Lett. 1987, 28, 65-68. (e) Mead, D.; Asato, A. E.; Denny,
M.; Liu, R. S. H.; Hanzawa, Y.; Taguchi, T.; Yamada, A.; Kobayashi, N.;
Hosoda, A.; Kobayashi, Y. Tetrahedron Lett. 1987, 28, 259-262. (f) Trehan,
A.; Liu, R. S. H. Tetrahedron Lett. 1988, 29, 419-422. (g) Knudsen, C.
G.; Chandraratna, R. A. S.; Walkeapaa, L. P.; Chauhan, Y. S.; Carey, S.
C.; Cooper, T. M.; Birge, R. R.; Okamura, W. H. J. Am. Chem. Soc. 1983,
105, 1626-1631.
(4) (a) Zhang, H. Z.; Lerro, K. A.; Yamamoto, T.; Lien, T. H.; Sastry,
L.; Gawinowicz, M. A.; Nakanishi, K. J. Am. Chem. Soc. 1994, 116, 10165-
10173. (b) Nakanishi, K.; Zhang, H.; Lerro, K. A.; Takekuma, S.;
Yamamoto, T.; Lien, T. H.; Sastry, L.; Baek, D. J.; Moquin-Pattey, C.;
Boehm, M. F. Biophys. Chem. 1995, 56, 13-22.
(5) Nakayama, T. A.; Khorana, H. G. J. Biol. Chem. 1990, 265, 15762-
15769.
(10) (a) Tokunaga, F.; Yoshihara, K. Isr. J. Chem. 1995, 35, 357-364.
(b) Hara, T.; Hara, R. In Handbook of Sensory Physiology; Dartnall, H. J.
A., Ed.; Springer: Berlin, Heidelberg, New York, 1972; Vol. VII, pp 720-
746. (b) Hara, T.; Hara, R.; Nat. New Biol. 1973, 242.
(6) Farrens, D. L.; Altenbach, C.; Yang, K.; Hubbell, W. L.; Khorona,
G. Science 1996, 274, 768-770.
S0002-7863(97)00956-6 CCC: $14.00 © 1997 American Chemical Society

Communications to the Editor
J. Am. Chem. Soc., Vol. 119, No. 24, 1997 5759
Scheme 2a
Figure 1. UV spectrum of retinochrome pigment of 2 formed when
apo-retinochrome was incubated with substiochiometric amount of
retinal analog 2 (1:0.8) in phosphate buffer prior to irradiation in the
dark. The pigment absorbs at 474 nm. Photoisomerization of 2 to the
11-cis isomer was effected by irradiation with a Xenon lamp behind a
540 nm “cutoff” filter.
a Conditions: (a) TESCl, Im, DMF; (b) cyanomethyldiphenylphos-
phate oxide, NaH, THF; (c) Dibal-H, ether; (d) triethyl 3-methyl-4-
phosphonocrotonate, NaH, THF; (e) nBu4NF, THF; (f) TBSCl, Im,
DMF; (g) MnO2, CH2Cl2; (h) ethyl formate, NaH, THF; (i) 4-carboxy-
benzenesulfonazide, NaH, THF.
Table 1. Percentage of Isomers Isolated from Photoisomerization
of 2 and 9 with apo-Retinochrome
all-trans-retinal, the possibility of using it for introducing the
11-cis geometry at the last synthetic step was investigated. All-
trans-3-diazo-4-oxoretinal (2), made in a straightforward man-
ner, was thus subjected to isomerization by retinochrome isolated
from the retina of the squid Todarodes pacificus.11
all-trans retinal
11-cis
13-cis
all-trans
Synthesis of 2 (Scheme 2) was achieved by slight modifica-
tion of a previous route.12 The hydroxyl group of 4-hydroxy-
â-ionone (4)13 was protected with triethylsilyl chloride (TESCl).
HWE olefination with the sterically hindered cyanomethyl-
diphenylphosphate oxide14 led to the isolation of 5 in 72% yield
(91% E-isomer). Subsequent Dibal-H (diisobutylaluminum
hydride) reduction of the cyano and further HWE olefination
with triethyl 3-methyl-4-phosphonocrotonate yielded 6 with all
carbons in place in 50% yield for both steps. The functional
group interconversions were easily performed by Dibal-H reduc-
tion of 6 followed by removal of the TES protecting group and
selective protection of the primary alcohol. The 4-hydroxyl was
oxidized by MnO2, and the resultant 4-oxo species was formyl-
ated at C-3. Next, the silyl protecting group was removed and
diazotization of 8 was accomplished by 4-carboxybenzene-
sulfonazide15 and 2 was obtained by MnO2 oxidation of the
primary alcohol.
2
9
75
83
5
6
20
11
easily identified by its expected H-12 cis J coupling (11.2 Hz),
as compared to the H-12 trans J coupling of 2 (15.4 Hz).
To demonstrate the utility of this protocol, all-trans-4-oxo-
retinal (9), prepared by MnO2 oxidation of 7, was isomerized
under similar conditions. Retinochrome photoisomerization of
9 led to a 83:6:11 ratio of 11-cis/13-cis/all-trans isomers (Table
1). The all-trans isomer can be readily recycled (especially
since it is the last step of synthesis) in order to increase the
efficiency of this scheme.
All-trans-3-diazo-4-oxo-retinal (2) was added to a suspension
of apo-retinochrome at pH 6.5 to yield a stable pigment (Figure
1) with kinetics similar to all-trans retinal.16 Formation of the
protonated Schiff base (PSB) chromophore within the protein
red-shifted the absorption to 474 nm. The photoisomerization
of 2 was accomplished by irradiation of a suspension of 2 in
phosphate buffer pH 6.5 at 4 °C in the presence of catalytic
amounts of apo-retinochrome (2 mol%) with a xenon lamp
behind a 540 nm wavelength “cutoff” filter for 40 min.16
Successive hexane extractions yielded a mixture containing 75%
11-cis-3-diazo-4-oxo-retinal (1) (Table 1). HPLC separation
of this mixture afforded pure 11-cis product in 45% yield.17 As
seen from Table 1, the retinochrome photoisomerization of 2
led to a 75:5:20 ratio of 11-cis/13-cis/all-trans isomers (ratio
determined by HPLC analysis). The 11-cis isomer (1) was
It is also well-known that apo-retinochrome can isomerize
various geometrical isomers of retinal to 11-cis-retinal. Al-
though not pertinent to our synthesis, possibly other retinal
analogs could be isomerized to 11-cis. It is also noteworthy
that the photoisomerization with retinochrome does not effect
the photocleavable diazoketone functional group in 2, as
evidenced by the strong 2072 cm-1 IR band of the 11-cis product
1 obtained following the isomerization.
In conclusion, photoisomerization of all-trans-retinal analogs
to the 11-cis isomer with retinochrome can be utilized as an
efficient synthetic method. This is because the 11-cis geometry
is established at the last stage of synthesis with high stereose-
lective output; moreover, it is applicable to radioactive analogs
containing, e.g., tritium, which would be essential for sequencing
the cross-linked sites. The low-temperature photolytic cross-
linking of 1 to rhodopsin and sequencing of the cross-linked
amino acid(s) will be reported in due course.
(11) Hara, T.; Hara, R. Methods Enzymol. 1982, 81, 827-833.
(12) Boehm, M. F.; Gawinowicz, M. A.; Foucault, A.; Derguini, F.;
Nakanishi, K. J. Am. Chem. Soc. 1990, 112, 7779-7782.
(13) Henbest, H. B. J. Chem. Soc. 1951, 1074-1078.
Acknowledgment. The studies were supported in part by NIH grant
GM34509.
JA970956I
(14) Loupy, A.; Sogadji, K.; Seyden-Penne, J. Synthesis 1977, 126.
(15) Hendrickson, J. B.; Wolf, W. A. J. Org. Chem. 1968, 33, 3610-
3618.
(16) All-trans-4-oxo-3-diazo-retinal (0.107 mg, 0.33 µmol) as a solution
in ethanol (0.2 mL) was added to the suspension of apo-retinochrome (2.2
mol %, 0.0071 µmol) in phosphate buffer (24 mL, pH 6.5) containing 0.1%
BSA at 4 °C. The suspension was stirred at this temperature and irradiated
with a xenon lamp (Ushio, 300 W) equipped with a 540 nm wavelength
cutoff filter for 40 min. The mixture was then extracted with cold hexane
(5 × 10 mL). The organic layer was separated by centrifugation at 10 000
rpm for 10 min. The pellets of apo-retinochrome separated under these
conditions and could be reused, and the aqueous phase was discarded.
Combined organics were dried with anhydrous sodium sulfate, and the
solvent was evaporated to yield the crude mixture of isomers.
(17) Normal phase HPLC (Cosmosil Si-60, 4.6 × 250 mm) purification
of the crude isomeric mixture (80% hexane, 20% ethyl acetate, 3 mL/min,
detection at 370 nm) led to the isolation of 1: 1H NMR (CDCl3, 400 MHz)
1.22 (6H, s), 1.93 (3H, s), 2.01 (3H, s), 2.35 (3H, d, J ) 1.2 Hz), 2.67 (2H,
s), 6.03 (1H, d, J ) 11.2 Hz), 6.07 (1H, d, J ) 8.3 Hz), 6.25 (1H, d, J )
16.1 Hz), 6.30 (1H, d, J ) 16.1 Hz), 6.60 (1H, d, J ) 12.7 Hz), 6.67 (1H,
dd, J1 ) 11.2 Hz, J2 ) 12.7 Hz), 10.09 (1H, d, J ) 8.3 Hz); IR (cm-1
)
1354.3, 1614.7, 1656.2, 2072.3, 2855.9, 2926.2; UV (hexane) 361 nm,
(methanol) 377 nm; HRMS calcd for C20H25O2N2 [M + 1]+ 325.1916, obsd
325.1918.