NMR studies of fluorinated visual pigment analogs

Vol.

179,

September

No. 30,

NMR

3, 1991

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

Pages

1991

Studies

of Fluorinated

Leticia U. Colmenares, Alfred Dennis Mead, J. P. Zingoni

Visual

Pigment

1337-1343

Analogs

E. Asato, Marlene Denny, and Robert S. H. Liu*

Department of Chemistry, University of Hawaii, 2545 The Mall, Honolulu, HI 96822

Received

July

9,

1991

Abstract. The 19F-nmr chemical shift data of isomeric pigments (1 Icis and 9-cis) of four vinyl fluororhodopsins and two trifluororhodopsins have been recorded. When compared with model protonated Schiff bases, a set of F-nmr opsin shift parameter (FOS) was obtained. The data revealed regiospecific protein perturbations on the F-resonances. They can be interpreted in terms of specific protein interactions such as the postulated second point charge and other polar interactions as well as the common hydrophobic protein 0 1991Academic Press, Inc. perturbation.

The possible use of fluorine label for studies of visual pigments was demonstrated in a preliminary study with 12-fluororhodopsin (1). Sensitivity of the fluorine nucleus was, however, shown to be partially negated by the excessive linewidth (approx. 1 ppm) associated with such a membrane protein, its solubilization in water necessarily being aided by detergents. Subsequently, it was shown in photobleaching studies that a Fsubstituent could lead to altered regiospecific photochemical properties (2), as a specific reporting group for suggesting possible use of fluorine information on protein-substrate interactions. Therefore, recently, we have renewed our effort in nmr studies of the F-labelled visual pigment analogs. In this paper, we wish to report results on a complete series of fluororhodopsins with the labels located on the polyene chain. We believe that the information provided by the fluorine reporting group, when compared with those of the corresponding model compounds, not only complements with data from the well established C-13 nmr studies (3) but possibly also provides a handle to information not readily provided by other nuclei or other methods.

1337

0006-291X/91 $1.50 Copyright 0 1991 by Academic Press. Inc. All righrs of reproducriort in arzy jtmn reserved.

Vol.

179, No. 3, 1991

BIOCHEMICAL

Materials

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

and

Methods

Procedures for preparation and characterization data of the fluorinated retinals are in the literature (4). All isomers were purified by preparative high pressure liquid chromatography immediately before use. Procedures for preparation of the butyl Schiff bases (SB) and the corresponding protonated Schiff bases (PSB) were essentially those in the literature (5). All uv/vis spectra were recorded on a PE-15 spectrometer, F-nmr spectra on a GE NT-300 spectrometer and H-nmr on a GE QE-300 spectrometer. Preparation of fluorine labelled rhodopsin analogs for nmr studies. Some minor modifications from published procedures (6) were introduced. Briefly, the crude Rod Outer Segment (ROS) extracted from 200 bovine retinae, isolated in the form of a pellet was homogenized in 90 ml HEPES buffer, and the ROS purified by flotation centrifugation. The ROS suspension was then photobleached (>420 nm, Corning 3-73 cut-off filter) in the presence of 1 ml of buffered (pH 7) 1M NH:!OH. The purified ROS pellet was repeatedly washed with HEPES buffer and hexane, and then solubilized in 10 ml of 2% CHAPS in water (50% D20). The clear solution (stored at -36eC) typically contained l-2 x 10-4 M opsin, as assayed by rhodopsin formation upon addition of 11 -cis-retinal. The rhodopsin analog was formed by mixing an ethanolic solution of a fluorinated retinal analog in O-2 fold excess with 6-8 ml of the opsin solution. A 0.1 ml aliquot was diluted with 2% CHAPS in a 0.5 ml cuvette Upon completion, the pigment solution for monitoring the binding process. was concentrated by ultrafiltration, under 40-45 psi of nitrogen gas, on an Omegacell assembly (MW cutoff, 10,000) to a final volume of approx. 2.5 ml. The concentrated solution, typically of the range l-5 x 10-h M, was transferred to a 10 mm NMR sample tube. F-nmr spectra of fluorinated rhodopsins. For recording the F-nmr spectra of the labelled rhodopsins, the probe temperature was kept at 15-2OOC. Typically, the spectrum of pigment was acquired with a 40 usec pulse and The a 50 msec delay between pulses for a total of 40,000-200,000 pulses. raw data were block-averaged, zero-filled and Fourier transformed using line broadening up to 80 Hz. The pigment sample was then photobleached using an orange light (Corning 3-71 or 3-68 cutoff filter) with the progress The F-nmr spectrum of of the reaction followed by uv/vis spectroscopy. the bleached sample was again recorded using the same parameters. F-nmr spectra of protonated Schiff bases (PSB). Previous reports have indicated that the retinylidene SB’s and PSB’s (7) are susceptible to isomerization either thermally or by nucleophilic catalysis. Therefore, wherever possible we have as solvent CD2C12, as recommended by Childs and Shaw (8), for preparation of the F-analogs of these compounds. For the protonating reagent, the choice was trichloroacetic acid. The probe was kept at OOC, if necessary, to minimize thermal temperature isomerization; however, certain amount of hydrolysis was found to be Parallel H-nmr spectra of the PSB’s were recorded for unavoidable. unambiguous assignment of the F-signals. 1338

Vol.

179, No. 3, 1991

BIOCHEMICAL

AND BIOPHYSICAL

RESEARCH COMMUNICATIONS

Extraction of chromophore from pigment analogs. The procedure for pigment denaturation and chromophore extraction was essentially that reported (9). For ease of comparison of hplc retention times, the chromophore was isolated in the form of the aldehyde. In addition to analysis by hplc, the extracted mixture was analyzed by their H-nmr and uvlvis spectra. Results

and Discussion

Fluorine labelled retinal analogs were known to exhibit a similar stereoselectivity as the parent retinal in their interaction with opsin (4). Hence, none of the all-trans and 13-cis isomers gave pigment analogs while four isomeric 8F-, six lOF-, three 12F- and two 14F-retinals, and three 19,19,19-F3(9-CF3) and two 2O,2O,2O-F3- (13-CF3) retinals gave pigment analogs in high yields. However, similar to the parent system, only the llcis and the 9-cis pigments are formed most readily (both in terms of ease of formation and pigment yields) (lo), suggesting only these centrally bent isomers are most compatible with the shape of the binding site of the native opsin. Hence, the F-nmr spectra of these isomers of five of these compounds (13-CF3 being the least stable) were recorded. Chromophore configuration. Retention of configuration of the pigment chromophore has now been demonstrated in selected cases by chromophore extraction/hplc analyses, a standard method for assaying chromophore purity (9). The case of 9-cis-14F-rhodopsin is shown in Figure 1 as a representative example. The retinal extract from the dena-

Y,l).dicis

Figure (upper)

1. Hplc chromatograms of retinal extracts and the irradiated pigment (lower).

1339

from

denatured

9-cis-14F-rhodopsin

Vol.

-50

179,

No.

3, 1991

-55

BIOCHEMICAL

-60

-65

AND

BIOPHYSICAL

-110

PPbA

RESEARCH

-120

-130

COMMUNICATIONS

-140

PPM

Figure 2. F-nmr spectra of 19,19,19-F3-rhodopsin (left, identical chemical shift all-trans and 11-cis of bleached sample) and 14-F-rhodopsin (right) solubilized CHAPS before and after irradiation with orange light.

-150

for in

tured pigment (upper) was shown to contain only the 9-cis and the 9,13dicis isomers, the latter likely to derive from catalyzed dark isomerization of the freed retinal, a complicating situation recently demonstrated in formation of 9,13-dicis rhodopsin (11). In contrast, the retinal extract from the irradiated sample was shown All these isomeric to contain primarily the all-trans isomer (lower). pigments exhibit distinct F-nmr signals (see figures below), and the spectrum of the 9-cis pigment is uncomplicated by signals from, e.g., the 9,13-dicis isomer (which should appear above -130 ppm). Hence, the nmr method is a useful, and direct method for determining isomeric purity of the pigment chromophore. Chromophore extraction was also conducted on SF-rhodopsins and 9-CFg-rhodopsins.

19F-nmr

spectra

of

PSB and

pigment

analogs.

The *9F-nmr

spectra of the 1 1-cis and 9-cis isomers of these four vinyl fluoro-pigments solubilized in D20 using CHAPS as the detergent and two CF3 -pigments, (12) before and after irradiation with orange light (> 470 nm), were recorded. Those of 14F-rhodopsin and 9-CFg-rhodopsin are shown in For comparison, the F-nmr spectra of Figure 2 as representative examples. the corresponding PSB’s of the fluorinated retinals are listed together with the pigment data in Table 1. In the Table are also listed the uv/vis absorption maxima of these Additionally, we have pigments, and uv/vis opsin shifts, OS, (13). calculated the fluorine opsin shift (FOS) values, defined as the chemical shift difference between of the fluorinated pigment signals and that of the corresponding PSB. (Notice that FOS is defined as pigment shift minus PSB shift instead of the reverse as in uv OS (13). This allows us to emphasize 1340

Vol.

BIOCHEMICAL

179, No. 3, 1991

Table

1.

W-

and F-opsin

11-cis 94x, 14-F 12-F 10-F 8-F

PsBa (ml

Rhodopsin (ml

391 459 446 a 42Sa 413

9-cis 9-CT, 14-F 12-F 10-F 8-F

386 451 436 420a 410

AND BIOPHYSICAL

shifts

of

RESEARCH COMMUNICATIONS

fluorinated

rhodopsins

W OSb (cm-‘)

PSB (-pm)

Pigmente (-mm)

FOSf (ppm)

456 527 507 499 463

3,600 2,800 2,700 3,300 2,600

58.2’ 125.5d 107.8’ 112.2c 116.8d

53.8 117.1 94.2 107.7 115.0

4.4 8.4 13.6 4.5 1.8

454 510 493 486 459

3,900 2,600 2,650 3,200 2,600

64.6d 131.4d 120.9d 119.7c 105.9d

60.2 123.5 114.3 114.9 99.3

4.4 7.9 6.6 4.8 6.6

a. In ethanol. b. Column c. In CEl,. d. In CD&l,. minus column 5.

2 minus column e. In CHAPS.

3 in cm-‘. f. Column 6

deviation of pigment shifts from those of the model PSB’s at the same time avoiding negative numbers.) A cursory examination of the FOS data of these pigments immediately reveals that most (eight out of ten) of the pigment peaks are shifted downfield by 4-8 ppm. Such a trend for a protein bound substrate is consistent with what is commonly known as the hydrophobic shift caused by the proximity of the protein side chains within the binding site (14), associated with the hydrophobic nature of the protein binding site. The two cases deviating from these values, 11 -cis-12F-rhodopsin (FOS = 13.6 ppm) and 11-cis-8F-rhodopsin (1.7 ppm), merit special attention. For the 12F-rhodopsins, it should be noted that the 9-cis pigment exhibit the normal 6.6 ppm downfield shift for the fluorine label. Hence, the unusual effect for the 1 1-cis isomer is not due to the unique location of the fluorine label on the polyene chromophore, but rather due to its orientation within the binding pocket. In fact, the trend is reminiscent of the binding result of 12-methylretinal where the low yield of the 11-cis isomer in contrast to the high yield of the unnatural 9-cis pigment was attributed to interference by a carboxylate group which served the dual function of the counter-ion to the imino nitrogen and the postulated second point charge (15). Apparently the relatively small size of the F-substituent (its van der Waals radius being about 20% larger than that of a hydrogen atom) (16) does not affect the relative yields of the two isomeric pigments. But the close proximity of a negative charge exerted a large deshielding effect (down field shift) of the 12F-signal in the 11-cis isomer (FOS = 13.6 On the other hand, the same 12F-substituent in the 9-cis isomer is mm). 1341

Vol.

179,

No.

3.

1991

BIOCHEMICAL

AND

BIOPHYSICAL

134s

Time

3. Hplc

COMMUNICATIONS

1 13-cis1 and 7,11-dicis 1 1‘

9-cis-8F-rhodopsin

Figure

RESEARCH

chromatograms

i 11 -cis-8F-rhodopsin

-

L

of retinal

extracts

after

irradiation.

known to be more distant from the postulated second point charge (17), thus insensitive to its perturbation (FOS = 6.6 ppm). The unusually small down field shift for 1 I-cis-8-F-rhodopsin (FOS = 1.7 ppm) was unexpected, especially in view of the normal FOS value for Parallel to the unusual F-shift is the abnormal its 9-cis isomer (6.6 ppm). Figure 3 shows the photochemical result of 1 I-cis-8-F-rhodopsin. chromatograms of the extracted retinal from irradiated solutions of 9-cisIt is clear that the 1 I-cis 8F-rhodopsin (a) and1 I-cis-8F-rhodopsin (b). isomer resulted isomerization at the 7,8 or 13,14 positions while the 9-cis We suspect that the isomer gave primarily the normal all-trans isomer. unusual chemical shift for the 11-cis isomer is related to its unique photochemistry. One possible explanation is a proximal polar amino acid Examination of a three dimensional molecular model of rhodopsin residue. constructed after a modified structure originally postulated by Cl@, Hargrave (19), indeed revealed the close proximity of cysteine-167 to C-8, However, at as pointed out to us by Mirzadegan (private communication). the exact nature of interaction between the highly this time, electronegative F-substituent and the polar residue is unclear, especially in view of the fact that in the lo-F-series where hydrogen bonding interaction was invoked for the unusual photochemistry (2) showed only a Experiments small downfield F-chemical shift (FOS = 4.4 and 4.7 ppm). with model compounds will have to be designed in order to clarify the exact nature of the polar interaction. ,4cknowledements. This work was supported by a grant from the U. S. Public Health Services (DK-17806) and partially by the UH Biomedical 1342

Vol.

179, No. 3, 1991

BIOCHEMICAL

AND BIOPHYSICAL

RESEARCH COMMUNICATIONS

program. We thank Drs. W. Niemczura and T. T. Bopp and Mr. W. Yoshida for their advice and technical assistance on F-nmr. LUC wishes to thank the East-west Center for a graduate fellowship (1987-1991). References 1. 2.

3.

4.

5.

6. 7.

8. 9.

10. 11. 12. 13.

14. 15. 16. 17.

18. 19.

Liu, R. S. H., Matsumoto, H., Asato, A. E., Denny,. M., Shichida, Y., Yoshizawa, T., Dahlquist, F. (1981) J. Am. Chem. Sot. 103, 7195-7201. (a) Liu, R. S. H., Crescitelli, F., Denny, M., Matsumoto, H., Asato, A. E. (1986) Biochemistry 25, 7026-7030; (b) Shichida, Y., Ono, T., Yoshizawa, T., Matsumoto, H., Asato, A. E., Zingoni, J., Liu, R. S. H. (1987) Biochemistry 26, 4422-4428. (a) Mollevanger, L. C. P. J., Kentgens, A. P. M., Pardoen, J. A., Veeman, W. S., Lugtenburg, J., deGrip, W. J. (1987) FEBS 163, 9-14; (b) Smith, S., Palings, I., Miley, M., Courtin, J., deGroot, H., Lugtenburg, J., Mathies, R., Griffin, R. (1990) Biochemistry 29, 8158-8164. (a) Asato, A. E., Matsumoto, H., Denny, M., Liu, R. S. H. (1978) J. Am. Chem. Sot. 100, 5957-5960; (b) Liu, R. S. H., Asato, A. E. (1990) in “Chemistry and Biology of Synthetic Retinoids” ed. Dawson, M., Okamura, W., CRC Press, p. 51-75. (a) Blatz, P. E., Mohler, J. H., Navangul, H. V. (1972) Biochemistry 11, 848-855; (b) Fukada, Y., Okano, T., Shichida, Y., Yoshizawa, T., Trehan, A., Mead, D., Denny, M., Asato, A. E., Liu, R. S. H. (1990) Biochemistry 29, 3131-3140. (a) Papermaster, D., Dryer, W. (1974) Biochemistry 13, 2438-2444; (b) Liu, R. S. H., Matsumoto, H. (1982) Methods Enzymol. 81, 694-698. Shriver, J., Mateescu, G., Abrahamson, E. (1979) J. Am. Chem. Sot. 18, 4785-4792; (b) Lukton, D., Rando, R. R. (1984) J. Am. Chem. Sot., 106, 4525-4531. Childs, R., Shaw, G. (1988) J. Am. Chem. Sot. 110, 3013-3018. (a) Pilkiewicz, F. G., Pettei, M. J., Yudd, A. P., Nakanishi, K., (1977) Expt. Eye Res. 24, 421-423. (b) Groenendek, C. W. T., deGrip, W. J., Daeman, F. J. M. (1979) Anal. Biochem. 99, 304-310. Asato, A. E., Liu, R. S. H. (1986) Tetrahedron Lett. 29, 3337-3340. Shichida, Y., Nakayama, K., Yoshizawa, T., Trehan, A., Denny, M., Liu, R. S. H. (1988) Biochemistry 27, 6495-6499. Kropf, A. (1982) Vision Res. 22, 495-497. (a) Derguini, F., Dunn, D., Eisenstein, L., Nakanishi, K., Odashima, K., Rao, V. J., Sastry, L., Termini, J. (1986) Pure Appl. Chem. 58, 719-724; (b) Lugtenburg, J., Muradin-Szweykowska, M., Heermans, C., Pardoen, J. A., Harbison, G.S., Herzfeld, J., Griffin, R. G., Smith, S. O., Mathies, R. A. (1986) 108, 3104-3105. Gerig, J. T. (1989) Methods Enzymol. 177, 3-22. Motto, M. G., Sheves, M., Tsujimoto, K., Balogh-Nair, V., Nakanishi, K. (1980) J. Am. Chem. Sot., 102, 7947-7949. Pauling, L., (1960) “Nature of Chemical Bonds”, Cornell University Press, New York. (a) Liu, R. S. H., Asato, A. E., Denny, M., Mead, D. (1984) J. Am. Chem. Sot. 106, 8298-8300; (b) Liu, R. S. H., Mirzadegan, T. (1988) J. Am. Chem. Sot. 110, 8617-8623. Mirzadegan, T., Liu, R. S. H. (1991) Prog. Retionoid Res. 11. Hargrave, P. A., McDowell, J. H., Feldmann, R. J., Atkinson, P. H., Rao, J. K. M., Argos, P. (1984) Vision Res. 24, 1487-1499. 1343