Spectral dimorphism of crayfish visual pigment in solution

Vision Res. Vol. 22. pp. 727 Printed in Great Britain

to 737.

1$03.00/0 Pergamon Press Ltd

0C142-6989/82/070727-I

1982

SPECTRAL

DIMORPHISM OF CRAYFISH PIGMENT IN SOLUTION

VISUAL

DENIS LARRIVEE* and TIMOTHYH. GOLDSMITH Department of Biology, Yale University, New Haven, CT 06511, U.S.A. (Received 19 Augusr 1981; in reuisedform 24 Nocemher 1981) Abstractxrayfish (Procambarus) photoreceptor membranes were prepared by a new technique. Digitonin extracts made at 0” contain two rhodopsin-like pigments, PSI2 and Ps,2, in a ratio of about 5:4. On exposure to light at 10°C. P,,, is converted to a metarhodopsin-like photoproduct with I,,, at 515 nm (M,,,), and PSI2 is converted to a metarhodopsin-like photoproduct with i,,, at 475 nm (M475). When warmed to 22°C. both photoproducts bleach in minutes to retinal and opsin. Neither Psdl nor Ps,2 is attacked by hydroxylamine, but both are destroyed by the detergent Ammonyx LO and by reduction of disulfide bridges with 2-mercaptoethanol. Neither is altered by changing the concentration of Cl-. Ps6* is more susceptible than P,,, to thermal denaturation and attack by OH-, and its spectrum is distinctly narrower than predicted by an Ebrey-Honig nomogram. Ps62 and PsL2 do not interconvert, either in the dark or in the presence of light. PsL2 is distinct from M sls; it is substantially more stable at 22°C and has a 23% smaller molar extinction coefficient. On the other hand, M,,, is spectrally very similar to the single metarhodopsin that is formed by irradiation of rhabdoms from dark-adapted crayfish. As neither Pse2 nor PslZ is found in crayfish rhabdoms, we conclude that both pigments are created in digitonin micelles from a spectrally and kinetically homogeneous population of rhodopsin molecules with I,,, near 530 nm in the rhabdom.

INTRODUCTION With a few exceptions (Ostroy, 1978; earlier studies reviewed in Goldsmith, 1972), there is relatively little information on arthropod visual pigments in detergent solution. Crayfish visual pigments are of particular interest because there have been some unusual problems in relating spectral sensitivity data from the eye to the pigments in digitonin extracts and to microspectrophotometric (MSP) data from isolated rhabdoms. In brief, the spectral sensitivity of the ERG (Kennedy and Bruno. 1962; Goldsmith and Fernandez, 1968; Wald, 1968) and of single retinular cells of the main rhabdom (Kong and Goldsmith, 1977) is maximal at 560 nm; while digitonin extracts contain two pigments with 1,,, at about 560 and 510 nm (Fernandez, 1965; Fujimoto er al., 1966; Wald, 1967): and MSP of isolated rhabdoms indicates a rhodopsin with imax at 530nm (Goldsmith, 1978a). The recent discovery that crayfish metarhodopsin fluoresces (Cronin and Goldsmith, 1981) has made it possible to show that the 530 absorption band in the rhabdom is due to a single rhodopsin and not a pigment mixture (Cronin and Goldsmith, 1982). In principle, the 30 nm shift in spectral sensitivity to 560 nm can be accounted for quantitatively by accessory pigment screens (Goldsmith, 1978b), and even larger shifts are known (Nosaki, Waterman and Fernandez, 1970). Because of the remaining discrepancy involving the pigment extracts, and because of the small amount of

data on them that has been published, we have reinvestigated the properties of crayfish visual pigment in digitonin solution. In this study we have used a relatively simple procedure to obtain a substantially enriched preparation of crayfish rhabdoms, and we confirm and enlarge upon Wald’s (1967) observation of two rhodopsin-like pigments in extracts of crayfish rhabdoms. We explore several explanations for the origin of both pigments and show that: (1) neither pigment can be interconverted with its companion pigment either by selective chromatic irradiation or by thermal means, (2) neither pigment originates in measurable quantities from sources other than the rhabdoms, (3) neither pigment is affected by addition of Cl- ions, and (4) the molar proportion of each pigment does not change when the animals are darkadapted for periods up to 5-r days. Our data, together with that presented elsewhere (Cronin and Goldsmith, 1982), indicate the formation of two spectrally distinguishable pigments in solution from a spectrally homogeneous population of pigment molecules in the native membrane. Isolation of the pigments in solution has also permitted us to explore the nature and degree of protein-chromophore interaction by exposing the pigments to various reagents, and thus to propose tentative explanations for the derivation of the two pigments in solution. METHODS Experimental

animals and extraction

of tisuul pigment

Procumbarus clurkii (Carolina Biological Supply) were used in all experiments. The crayfish were kept

*Present address: Department of Biology, Benedictine College, Atchison, KA 66001, U.S.A. 727

DENIS LARRIVEEand TIMOTHYH. GOLDSMITH

728

at room temperature on a 12 hour L:D cycle before subsequent experimental manipulation. Two crayfish were dark-adapted at room temperature and then chilled on ice for 10 min before removing the eye stalks. Subsequent steps were carried out under dim red light as described in Larrivee and Goldsmith (1982). The severed eyestalks were placed in a drop of phosphate buffer (0.01 M. pH 7.0) and the corneas cut along the entire circumference at the junction with the hard chitin of the eye stalk. The corneas were gently detached, together with numerous crystalline cones, without disrupting the pigmented photoreceptor layer. The retinas were dissected free of underlying neuropile and frozen in a drop of 0.01 M phosphate buffer, pH 7 in a plastic capsule of the sort used for embedding specimens for electron microscopy. The frozen cubes were thawed and the tissue gently teased over a 210 pm mesh metal screen while adding about 1 ml phosphate buffer. This released individual rhabdoms and fine pigmented debris into suspension. and the filtrate was layered over a column of 139, sucrose 2 cm long by 1 cm wide. The rhabdoms were allowed to settle through the sucrose cushion for 30min at 0 After removing the upper pigmented fraction, the lower, rhabdomrich fraction was centrifuged for 10min at 8ooO x g, forming a lightly colored pellet. The pellet was washed with buffer, recentrifuged, and extracted with 200 ~1 of 2% digitonin in 0.01 M phosphate buffer for 1.5 hr and at 0°C before a final centrifugation for 15 min at 8000 x g. The supernatant was retained for spectrophotometric analysis. In most cases the supernatant solution appeared clear. Spectra were recorded in 0.3 ml quartz cuvettes in a Cary 14 spectrophotometer, with a sensitivity of 0.1 optical density units full scale. In order to regulate temperature, the cuvettes were encased in a hollowed aluminum block (with a suitable aperture for the analytical light beam), through which a temperature-controlled solution of methanol-water (1: 1, v: v) was circulated. The temperature in the block was calibrated by means of a chromel-alumel thermocouple and a Keithley 616 digital voltmeter. Temperature was controlled to + 1 C. Spectral data were obtained in the form of ditlerence spectra. An initial absorbance spectrum was obtained prior to irradiation or addition of reagent. and a second absorbance spectrum was recorded following treatment of the sample and subtracted from the initial absorbance spectrum. High pressure

liquid chromotograph)

The procedure of Groenendijk et (II. (1980) was used for the extraction of vitamin A aldehyde from the retina. A suspension of retinas from six animals in 150~1 of buffer was combined with 5 ~1 buffered (pH 6.8) 0.5 M hydroxylamine, 300 ~1 methanol, 300 ~1 dichloromethane, and 200~1 water, and rapidly vortexed for 1 min. The aqueous-organic suspension was centrifuged. and the lower, organic layer collected and

evaporated under reduced pressure. The residue was dissolved in hexane-dioxane (100:6, v:v) to a known volume, typically 40~1, and sealed under nitrogen in glass ampules in preparation for chromatographic analysis. A liquid chromatograph unit equipped with an Altex 100 dual head reciprocating pump was used in conjunction with a 25 x 0.46cm silica column packed with 10pm diameter Spherisorb. Effluent from the column was monitored at 365 nm by means of an LDC Spectromonitor III detector. Flow rate through the column was 2.0 ml min- ‘. RESULTS

Digitonin

extracts

of

crayfish

rhahdoms

contain

two

pigments Partial bleaching of extracts. Two rhodopsin-like pigments are found in digitonin extracts of rhabdom membranes made at O’C. consistent with previous observations by Wald (1967). These pigments and their photoproducts are revealed by partial bleaching, as exemplified by the difference spectra illustrated in Figs 1 and 2. The absorption spectrum of an extract was recorded at 10°C (1st spectrum). The extract was then irradiated with red light (2s > 630 nm), and the spectrum rerecorded (2nd spectrum). The sample was then warmed to 25’C, and when the absorbance was again stable, a third spectrum was recorded. The 1st spectrum minus the 3rd is the difference spectrum for‘the

-0.05 300

’ j 400

’

’ ’ 500

WAVELENGTH

’

’

’ 600

’

’ 700

(nm)

Fig. 1. Difference spectra for the partial bleaching of the long wavelength-absorbing pigment (Psc2) and its metarhodopsin (M,,,). A digitonin extract of rhabdoms was irradiated at 10°C for 15 min with red light (I’s > 630 nm) to form Msls; the solution was then warmed to 25” for 1 hr to allow decay of the metarhodopsin. The difference spectrum for the combined effects of red light and warming (initial spectrum minus final spectrum) is shown by the filled circles and solid curve. A pigment with E.,,, at about 560 has bleached to a final photoproduct with A,,,,, at about 375nm. The difference spectrum for the metarhodopsin was obtained by subtracting the final spectrum from one recorded after the exposure to red light but before the solution was warmed to 25°C. and is shown with the unfilled circles and broken curve. The metarhodopsin has i,,, at 515 nm.

729

Crayfish visual pigment

o”‘3 -0.05300

”

600

”

1

WAVELENGTH (nm) 400

500

WAVELENGTH

600

700

(nm)

Fig. 2. Difference spectra for the bleaching of the pigment (Psi2) that remains after exposure of the solution to red light (filled circles, solid curve) and its metarhodopsin (M4,& These spectra were obtained after those of Fig. 1. The sample was retooled to to” and the experimental procedure repeated, but with an orange bleaching light (2s r 550 nm).

effects of red light and warming (Fig. 1, filled circles): it indicates loss of absorption maximal at about XiOnm, and a photoproduct with i,,,,, at about 375nm. The 2nd spectrum minus the 3rd (Fig. 1, unfilled circles) is the difference spectrum for the thermal bleaching of the metarhodopsin* that was formed by red light; it has ii,,, at 515 nm and indicates that the metarhodopsin has a somewhat greater molar extinction coefficient than the original pigment. The procedure was repeated with a bleaching fight containing shorter wavelengths. The same extract was cooled again to lO”C, and the 4th spectrum recorded. Following irradiation with orange light (2s > SSOnm), the 5th spectrum was measured, and after rewarming to 25’C and waiting for all spectral changes to occur, the 6th spectrum was finally recorded. The 4th spectrum minus the 6th (Fig. 2, filled circles) shows the combined effects of the second (orange) actinic exposure, followed by warming; it describes the bleaching of a pigment with i,,, at about 512 nm to a photoproduct with &,._ at about 37.5nm. The 5th spectrum minus the 6th (Fig. 2, unfilled circles) is the difference spectrum for the thermal decay of the corresponding metarhodopsin. with i,,, at 475 nm. Despite their spectral similarity. P,,, is different from Ms 15 (evidence summarized below). Acerage spectra of the pigments. Normalized averaged spectra of the two rhodopsin-hike pigments are combined

*A word about nomenclature. We use the word rhodopsin to indicate the visual pigment of crayfish and the term metarhodopsin to refer to the thermally stable photoproduct with which rhodopsin is photointerconverted. Although for some workers this very general usage may represent an etymological abuse, it serves a useful purpose. generally introduces no ambiguity. and is firmly embedded in the literature of invertebrate visual pigments. The properties of Pse2 and Psll are rhodopsinlike, and it is convenient to refer to their photoproducts M s 15 and M4, s as metarhodopsins. V.R.22 7-B

““a

500

Fig. 3. Average, normalized difference spectra for P,,, (large filled circles) and Psi, (large unfilled circles) from 6 experiments similar to that of Figs 1 and 2. Error bars represent f 1 SEM. The small filled and unfilled circles show the spectra of the corresponding pigments of the crayfish Orconectes as reported by Wald (1967). Broken curves were drawn through the present data points by eye. The solid curves are drawn from the Ebrey and Honig (1977) nomogram for pigments with i.,,, at 562 and 512nm. The fit to Psi2 is good, but P562 has a narrower absorption spectrum than the vertebrate visual pigments on which the nomogram is based.

shown in Fig. 3 by the Iarge circles. There is a close similarity to the pigment spectra reported by Wald (1967) for Orconecres (small circles). The dashed curves are drawn through our data points by eye. The unbroken curves are Ebrey and Honig (1977) nomogram pigments with A,,,, at 562 nm and 512 nm. Crayfish P5,2 fits the Ebrey-Honig nomogram we11 at wavelengths greater than 480nm: below 480 it may be narrowed somewhat by the distorting effect of the 375 nm-absorbing photoproduct. but the general fit with Wald’s data, which were obtained with hydroxylamine, is nevertheless good. PSh2, on the other hand. is distinctly narrower than predicted by the nomogram. Average, normalized spectra of the corresponding metarhodopsins are shown in Fig. 4. Ma,5 decays

Fig. 4. Averaged. normalized spectra for the metarhodopsins M5,5 (filled circles, solid curve) and M.+,s (unfilled circles. broken curve), based on 6 experiments similar to Figs l-2. Error bars indicate + 1SEM.

DENS LARRIVEE and TIMOTHY H. GOLDSMITH

730 A

0.00

A

,

400

AA-A

.

,

fA*AA

600

500

700

WAVELENGTH (nm) Fig. 6. Absorbance changes following the same protocol described in Figs 1 and 2 for samples of rhabdomal and extra-rhabdomal material in the extract. In contrast with rhabdomal samples, which yielded P,,, and P,,, (circles), the extra-rhabdomal samples contained many fewer microscopically recognizable photoreceptors and were much more deeply pigmented with contaminating ommochromes. The smaller spectral changes that were observed in extracts of extra-rhabdomal samples (filIed triangfesf are consistent with photooxidation of xanthommatin.

a

I Icir

I

MINUTES Fig. 5. High pressure oxime from the retinas daylight and prepared pared with samples of

liquid chromatography of retina1 of 6 crayfish exposed for 1 hr to as described in the text (A), comknown isomeric composition (B).

The abscissa is a time axis; the ordinate is absorption at 365 nm. more rapidly Wald (1967).

than

M5 ,s and was not observed

by

all-trans retinal. Anifor 1 hr in daylight and the eyes excised. The retinas were treated with methanol and extracted with dichloromethane in the dark. HPLC separation of the extract (Fig. 5) indicated the presence of retinal, principally as the trans-isomer. No dehydroretinal was observed. This finding supports the interpretation that the final photoproduct in Figs 1 and 2 is retinal and is consistent with the best available evidence on the identity of the rhodopsin chromophore in arthropods (Paulsen and Schwemer, ~igh~-ada~fed

mals

were

light

retinas

Molar extinction coeficients of the pigments. have the of and George adding to sample and the change to forof oxime bleaching yield measure the coefficient each and its photoproduct 1). extinction of and 2 values squid Hubbard St 1958; Hagins, and other (Gold1972). coefficients the dopsin also similar those squid of Drosophila (Ostroy, 1978), and are higher than those of the photopigments. consistent with all other observations made on invertebrate

contain

adapted

In order

or from

we have

of the

by following

It 6) that in extracts of rhabdom-

1972).

1. Molar (1 mol- ’ cm-‘, + 1 SD) of the two pigments and their photoproducts that are present in digitonin

extracts of crayfish photoreceptor P 562

MS15

37,800 +_4030 (n = 13)

54,450 + 6810 (n = 4)

to

determine

P 512

membranes M 475

42,000 rt 2110 ‘55.600 + 4970 fn = 9) (n = 71

Crayfish visual pigment 1.0

t

l

0.7 E i

I

I

Thermal Bleach ot 32-Z f

731

show an appropriate semi log-plot of the decay of absorbance at 560 nm vs time. Points after 15 min are well fit by g linear regression with a rate constant of 2.3 x lo-” min-’ (time constant of decay = 42 min), and if this curve is subtracted from the data, the earlier portion of the decay (unfilled circles) is described by a single rate constant of 0.138 min- 1 (time constant of decay = 7.25 min). P 512 is diRerent from M,,,. As described above, Ps, 2 can be obtained in the absence of P562, either by extracting at 22°C or by photoconverting Pse2 to M 515 with red light at 1o’C and then warming the sample to 22’C until MS 15 dqays. M,,, decays in minutes at 22°C. while P, i 2 is stable for hours. These observations on the thermal lability of M5,5 demonthe same as M,,,. strate that P, , 2 is 512 is

Pm

0.01

0

40

20

600

60

MINUTES

Fig. 7. Kinetics of thermal decay of a mixture of P56L and P si2 at 32°C (solid circles) can be resolved into two firstorder processes occurring in parallel. The last four data points were fit to a curve by linear regression, and the resulting straight line subtracted from the earlier part of the data to yield the open circles. On this semi-log plot, the open circles are fit by another straight line of steeper slope. Error bars represent f 1 SEM; n = 4. Inset: Difference spectra for the decay of absorption at 560nm that had occurred at 10, 15 and 240 min in a representative experiment. The shift in I,,, to shorter wavelengths with time

indicates that P,,z is more susceptible to thermal denaturation than Ps12. Do, initial absorbance; D,,final absorbance; D,. absorbance at time t. served following irradiation of rhabdom-poor extracts are small, relatively ill defined, and probably reflect photo-oxidation of xanthommatin (Marak et ai., 1970; Denys 1982). Both P562 and PSI2 therefore appear to be derived from rhabdomal membranes. Thermal lability of P5,,. If digitonin extracts of rhabdom preparations are made at 22°C (rather than O’C), only P, I 2 is observed, and in normal amounts. This observation suggests that Pse2 is more sensitive to thermal denaturation than Ps12, an inference that was confirmed by the following additional observations. Extracts obtained at 0°C bleach when warmed to 32°C in the dark (Fig. 7). A significant absorbance change that is maximal at 560 nm can be seen within 5 min. Longer warming results in a cumulative absorbance change that is centered near 560 nm at 15 min but which shifts gradually to shorter wavelengths with progressive pigment decay (Fig. 7, inset). No increase in absorbance is seen anywhere in the visible region of the spectrum, as would be expected if P 562 were thermally converted to PJlz. Thus in solutions above room temperature, there is a rapid decay of P562, together with a slower decay of Ps12. Kinetic analysis of the decay can be resolved into two first-order processes and is therefore consistent with this interpretation. The filled circles in Fig. 7

solution, P5,2 does not appear as an intermediate radiation of P5 Lz with blue light does not P 562 or M5,5. A sample of PSlr was by partial bleaching followed by warming to destroy M,,,. Subsequent irradiation produced a photoprodu~t spectrum was by warming the sample to permit decay (Fig. 8, filled circles). ’

470nm

;

irradiation

’

’

of

I-O-

s ,o

a 0.6’

0.6-

? F 4

0.4-

400

600

500 WAVELENGTH

(nm)

Fig. 8. When irradiated with blue light, Ps,* forms M4,$, with no trace of P se2 or Ms,,. Rhabdoms were extracted with 2 per cent digitonin at room temperature for one hour and the extract was then irradiated with red light for 5 min to destroy any remaining P,,,. The sample was cooled to 5°C and then irradiated with blue light (470nm narrow band interference filter) for 10 min. After recording the absorbance spectrum, the sample was warmed to 25°C for 0.5 hr to allow the early photoproduct to decay fully. A final absorbance spectrum was recorded, which was subtracted from the first to yield a difference spectrum. Because the absorbance changes were small, the sample was retooled to 5°C and the procedure repeated. The difference spectra were summed, and three experiments were averaged to produce the normalized spectrum shown by the unfilled circles kl SEM. The spectrum of Mb,* from Fig. 4 is.replotted (solid curve) for comparison. The fit is good. Had any Ps6,, and M,,, been formed from Psll by blue light, the difference spectrum would have had &,, at longer wavelengths than 475 nm.

732

DENIS LARRIVEE and TIMOTHY H.GOLDSMITH 1.8 -

1

I

I

1

I

'CD-' Irradiation

l '362

0~:

1.6-

/

0 PSI2

I

w

1

of

s ’

0

I

’

’

:

rhobdoms

.>630nm 0>550nm

I'0 &'

-0-o” 0.8

"-;-

,

I

2

I 3

DAYS

IN THE

I 4

I 5

6

0.2

DARK

t

Fig. 9. Gradual increase in pigment level in the rhabdoms as a function of days of dark adaptation of the animal. The concentration of P 562 relative to the value at 12 hr is shown by the large filled circles. There is less Pslz than Ps6s (lower unfilled circles and dashed curve). If the data for PSI2 are expressed relative to the value for PSI2 at 12 hr (upper unfilled circles), it becomes apparent that the 5:4 proportions of Pse2 and P,,, do not change with time. The total pigment increases, most likely due to enlargement of the rhabdom. The pigments from two animals were extracted to obtain the data at each time. is spectrally indistinguishable from M,,, (Fig. 8, solid curve) obtained by irradiating P,,, with orange light. In digitonin solution, therefore, P5 f 2 and P,,, and their metarhodopsins appear to be distinct molecular entities. photoproduct

&j&s

qf previous

light

ml

durk

adaptation

qf the

rhabdom Change in pigment

keel

with prior

dark udaptation.

We have attempted to learn whether either pigment represents a transitional component of the rhabdom. Crayfish were dark-adapted for varying lengths of time up to 5) days, rhabdoms prepared, and the pigment composition assayed as in Figs 1 and 2. The results are shown in Fig. 9. The molar proportions of P562 and Ps,, in the extracts did not change with time in the dark, but there was a steady increase in the total amount of pigment present per eye, probably reflecting an increase in the volume of the rhabdom (Kong and Goldsmith. 1977; Eguchi and Waterman, 1979). Extraction

of ~etarhodopsin

following

irradiation

of

membranes. Rhabdomat membranes were irradiated to convert much of the rhodopsin to metarhodopsin and then extracted with digitonin solution at O’C. The spectrum was measured, the sample warmed to allow decay of metarhodopsin, and a final spectrum recorded. The difference spectra for the metarhodopsin produced in situ by red (tilled circles) and orange (unfilled circles) isomerizing lights are plotted in Fig. 10. along with M,, s (solid curve) taken from Fig. 4. Both red and orange light produce a metarhodopsin with i.,,, at about 515 nm. This is the same result as obtained by microspectrophotometry (Goldsmith, 1978a), where the metarhodopsin is measured in single isolated rhabdoms. The tluorrhabdo~

0 400

I

.hri’cs**‘* 500 WAVELENGTH

600 tnm)

Fig. 10. When the visual pigment is irradiated in the membranes, Msrs is subsequently extracted. Rhabdoms were irradiated at 0°C and then extracted for 1.5 hrs with 2% digitonin. The spectrum of the cold extract was measured, the solution warmed to 25°C until the metarhodopsin had decayed, and a final spectrum recorded. Difference spectra were formed by subtraction. and the results for red (i’s > 630 nm. n = 2) and orange (2.‘~> 550nm, n = 3) actinic exposures, shown by the filled and open circles respectively, have the same i,,, at 515 nm. Error bars on the open circles indicate k 1 SE. The spectrum of MSls (from Fig. 4) is shown by the solid curve for comparison.

escence excitation spectrum of metarhodopsin in the rhabdom is also similar (Cronin and Goldsmith. 1981). This result differs, however, from what might have been expected from irradiation of the pigments in vitro, where the same orange light should produce an approximately equal mixture of M,,, and M,,5 with an intermediate ;i,., at 490-500 nm. (There is an apparent divergence of the titled and unfilled circles at short wavelengths. On the basis of a two-way analysis of variance in the spectral region 450-5OOnm. the measurements represented by the filled circles are not significantly different from the data on which the solid

W

a

I.0

x 2

0.8

5 vt 2

0.6

:

WAVELENGTH

(nm 1

Fig. 11. Difference spectrum for the spontaneous decay of pigment at pH 9 and 10°C (filled circles, normalized average of 3 experiments + 1 SE). The spectrum of P,,, (from Fig. 3) is shown for comparison (solid curve). Alkahne pH leads to the differentiat destruction of PshZ.

Crayfish visual pigment I

-

’

’

*

I

*

2-mtrcoptocihanol

o.dWAVELENGTH

(nmf

Fig. 12. Difference spectra for the decay of absorption in pigment extracts containing 2.5% 2-mercaptoethanol. Successive spectra were measured at 15, 30 and 45 min at 10°C. Both PSe2 and P,,, are attacked. curve is based. Although the unfilled circles fall significantly above the solid curve, the difference is small, and there is no displacement of the I,,,. Moreover, difference spectra are most subject to systematic error in this spectral region. Compare also with Fig. 15.) The interpretation of this experiment will be expanded in the Discussion. Treafment

of thr pigments

with reagenfs

Alkaline pN destroys Ps,z. Raising the pH to 9 with Tris (final concentration of 150 mM) causes a slow loss of absorbance at 10 C. The difference spectrum for this bleaching is maximal near 560 nm (Fig. 11). Comparison with the difference spectrum for PsBz (Fig. 1I, curve taken from Fig. 3) shows that the bleaching is due primarily to a loss of Psb2. P,,, was also reported by Wald (1967) to be destroyed by an alkaline environment. Re~~~~ri~n of ~l~~~~~le bridges

hieuches

both

pig-

menrs. Addition

of 2-mercapthoethanol to a final concentration of 2.5”,, causes a slow bleaching of digitonin extracts held at 10 C (Fig. 12). Comparison of difference spectra recorded at 15. 30 and 45 min indicates that both pigments are attacked at about the same rate. After 45 min no further absorbance changes were produced when the sample was irradiated with either red or orange light. A concencaused slower tration of 0.5”,, 2mercaptoethanol bleaching, and after 1 hr. 45”,, of the pigment could still be bleached by orange light. Srcrhilir~ IO hydroxylamine. Both pigments are relatively insensitive to the addition of hydroxyiamine, even in moderately high concentrations. The filled triangles in Fig. 13 show a difference spectrum recorded following 1 hr of incubation with 0.1 M NHzOH at 5 C and pH 6.5. Only a slight change in absorbance is evident during this period. When the solution is irradiated with red light and warmed to

room temperature, however. a very large absorbance change is observed. maximal at 560 nm (filled circles). Additional irradiation with orange light at room temperatures induces a further change in absorbance that

733

is maximal at 512 nm (unfilled circles). Both pigments therefore undergo differential bleaching as though hydroxylamine were not present in solution. Our finding that Psta is stable to 0.1 M hydroxylamine differs from the report of Wald (1967). When the temperature of the solution is kept low and the cuvette is irradiated with red light however, the absorbance of the photoproduct M,, 5 is variably but substantially reduced in magnitude from that observed in the absence of hydroxylamine. The absorbance of MAT5is even more severely decreased. Absence of eficts of Cl-. Because chicken iodopsin (Knowles, 1976; Fager and Fager, 1979) and gecko rhodopsin (Crescitelli. 1980) undergo bathochromic shifts when chloride ions are added to chloride-free extracts, we looked for similar effects on the crayfish pigments. Addition of NaCl (to a final concentration of 100 mM) to chloride-free extracts caused insignificant changes in absorbance. Moreover, the same absorbance changes were seen following exposure to light in the presence and absence of Cl-. We therefore conclude that Cl- has little effect on the absorption spectra of Ps62 and P5t2. Instability in Ammonyx LO. When extracts of rhabdoms made with Ammonyx LO are irradiated, there are no significant changes in absorbance. Both Pse2 and P,,z therefore appear to be unstable in the presence of this detergent (cf. bovine rhodopsin: Ebrey, 1971: Applebury et a/., 19%; Hargrave. 1977). DISCUStON

Compurison

with edier

ohserrutions

Our findings confirm earlier observations of two rhodopsin-like pigments in digitonin extracts of crayfish retinas but differ from previous work in some details (Table 2). Our results are in closest agreement with Wald’s (1967) report for Orconectes. As extracts made at room temperature contain only the short

400

500 WAVELENGTH

600 (nm)

Fig. 13. Difference spectra for the addition of hydroxylamine to a final concentration of 0.1 M. pH 7. 10°C (filled triangles). compared with difference spectra for the subsequent analysis of PShZ and P,,, (performed as in Figs 1 and 2). Both pigments are stable to NH,OH (although there may be some small bleaching of Pso2 in 1hr), but MJls and M,,, are both attacked, even at 10°C (not shown).

734

DENISLARRIVEE and

TIMOTHY H. GOLDSMITH

Table 2. Reported wavelengths of maximal absorption (nm) and approximate relative amounts (PI/P*) of crayfish photoreceptor pigments in digitonin solution PI

M,

P,

M,

PI/P2

556

515-520

556 562 562

515 515 515

512 505 525 510 512

478 475

0.43/l ?/l 311 1.1/l 1.2/l

Genus

Authors

Procambarus Procambarus Procambarus

Fernandez, 1965* Fujimoto er al.. 1966 Wald, 1967t Wald, 1967 Present study

Orconectes Procambarus

*Fernandez states a A,,, of 50+506 nm for the short wavelength pigment. but from his Fig. 25 the I,,, appears to be very close to 512 nm. Furthermore, his difference spectrum for P562 was pulled to shorter wavelengths by a small amount of bleaching of P512 (see his Fig. 23). Although he observed a photoproduct with i.,., at 478 nm. due to his bleaching protocol he did not recognize it as the metarhodopsin of P512 and attributed it to a secondary photoproduct of M,,,. With the benefit of hindsight, his spectral data are therefore compatible with our findings. tWald’s (1967) extract of Procambarus was more heavily contaminated with ommochromes than his extract of Orconectes, and the absorbance changes at long wavelengths indicate that there was likely some photooxidation of xanthommatin. This may account for some of the difference he reported between Procambarus and Orconectes.

wavelength pigment, some of the differences found in the proportions of the two pigments probably reflect the greater thermal lability of the long wavelength member, the results of Fujimoto et al. (1966) representing the extreme example. With the exception of Fujimoto et al. (1966) there thus appears to be a consensus that digitonin extracts of crayfish retinas contain two rhodopsins. With the possible exception of Palaenwnetes (Femandez, 1965), however, crayfish is unique among crustacea in yielding dual pigments from the main rhabdom. The pigments are like invertebrate rhodopsins in that retinal rather than dehydroretinal is the final product of bleaching, in having molar extinction coefficients around 40,0001 mol-’ cm-‘, and in giving rise to metarhodopsins with somewhat larger molar extinction coefficients. We believe, however, that neither of these pigments is native crayfish rhodopsin. Origin of the two pigments in extracts

Kinetic analysis of absorption and fluorescence changes in isolated rhabdoms shows that (i) there is a single rhodopsin that is reversibly photoconverted to its thermally stable metarhodopsin, and (ii) there is no metarhodopsin in the dark-adapted eye (Cronin and Goldsmith, 1982). Morebver, (iii) the rhodopsin has I,,, at 525-530nm (Goldsmith, 1978a; Cronin and Goldsmith, 1982), and (iv) the metarhodopsin has A,,, at about 515nm (Goldsmith, 1978a; Cronin and Goldsmith, 1981). The pigment composition of the rhabdom and the effects of light can therefore be summarized as R530 +i MS1s. Our principal task is to relate the presence of Ps,, and P5,2 in digitonin extracts to the finding of a single rhodopsin in situ. In other words, an interpretation of the present experiments must be consistent with the results of the MSP and microfluorimetric studies enumerated i-iv above. Let us consider one obvious example that violates this stricture. The similarity between Ps,, and the spectral sensitivity function for either the ERG (Ken-

nedy and Bruno, 1961: Goldsmith and Fernandez, 1968; Wald, 1968) or single retinular cells (Kong and Goldsmith, 1977) might suggest that P562 is the visual pigment of the rhabdom. This interpretation is not compatible with the rhodopsin photosensitivity maximum at 530nm (point iii), however. In addition, one must hypothesize that the 530nm absorption maximum seen in the rhabdom represents a mixture of Psh2 and another molecular species absorbing at shorter wavelengths. The present results show that P, 12 is different from M, 15 in having a significantly smaller molar extinction coefficient (Table 1) and being considerably less sensitive to thermal denaturation (Results: P5 12 is different from M, I 5). Moreover, there is no MS15 in the dark-adapted rhabdom (point ii). The second species in the pigment mixture would therefore have to be Ps 12. In order for a mixture of PseZ and PSI2 to account for the 530 nm photosensitivity spectrum of isolated rhabdoms (point iii), PsbZ and P, 12 would have to be interconvertible through a common metarhodopsin or give rise to spectrally identical metarhodopsins (Cronin and Goldsmith, 1982). Neither condition is observed (Figs 7, 8 and 2). Moreover, participation of two rhodopsins in the photosensitivity spectrum is at variance with the kinetic analysis (point i). And finally, P,,, makes up a larger fraction of the total pigment in the extract (56%) than is needed in mixtures with PSI2 to approximate the 525-530 nm absorption maximum in the rhabdom (35%) (Cronin and Goldsmith, 1982). As Psb2 is the more labile member of the pair and thus more easily lost, the discrepant proportions are not likely due to differential denaturation during extraction. The hypothesis that a mixture of Ps6* and PSI2 is responsible for the absorption spectrum of the dark-adapted rhabdom is therefore not consistent with the relative amounts of the pigments present in extracts. These considerations suggest that Pse2 and P5,* are the result of digitonin-mediated changes in the

Crayfish

visual

~ p i

.! ‘..digitonin ..

idigitonin

“4

512

‘562

‘C

hv M475 T

retint( + opsin

Fig. 14. Postulated scheme of interrelationships to account for Pse2 and Psz2 in extracts, and a single rhodopsin, RszO. in rhabdoms. The laminated structure at the top represents a single rhabdom. Wavy arrows indicate light reactions; broken arrows show the action of digitonin. The straight arrows represent the thermal decay of the metarhodopsins; the thermal denaturations of PJe2 and P,,, (Fig. 7) are not shown. See the text for further discussion.

pigment

735

two different conformations on extraction. Once made in the membrane, M,,, behaves more conventionally than its parent rhodopsin, however, for it does not undergo any spectral shifts when it is extracted. Moreover, as the metarhodopsin made from Pse2 is spectrally very similar to the metarhodopsin formed and measured microspectrophotometrically in the rhabdom (Fig. IS), Fig. 14 depicts them as the same pigment. By contrast, the only way to produce M,,=, is from PJLZ. which in turn requires prior treatment of the membranes with digitonin. The manner in which the detergent forms two spectrally distinct pigments from a pigment of apparent spectral homogeneity is obscure. Whether the rhabdom contains two sub~pulations of rhodopsin differentiated on some basis other than obvious spectral and kinetic properties, or whether Pse2 and Ps,2 represent different conformations or modifications of a protein with the same chemical structure in the photoreceptor membranes, it is clear that some readjustment of the protein has occurred in order to produce two pigments spectrally distinct from one another and from their precursor(s) within the rhabdom. The readjustment apparently does not entail breakage of disulfide bridges, however, since the addition of 2-mercaptoethanol leads to bleaching of both pigments (Fig. 12). Nor does there appear to be substantial unfoIding of the protein in the vicinity of the chromophore-opsin link, since penetration to, and attack of. the SchitT base linkage by the relatively small hydroxylamine molecule does not occur in either pigment. Presumably PSBL and P,,, differ at least in the configuration of the protein in the region of the aliphatic chain of the chromophore (Kropf and Hubbard, 1958; Ebrey and Honig, 1975; Matsumoto and Yoshizawa, 1978: Warshel. 1978; Honig er al., 1979). On the other hand, a digitonin-mediated change in

properties of a single rhodopsin that has 1,,, at about 530 nm in the membranes of the rhabdom. These relationships are summarized in Fig. 14. On extraction with digitonin (broken arrows), a spectrally homogeneous population of rhodopsin molecules (R& gives rise to two pigments (P562 and P512), each of which is converted by light to a spectraily distinct metarhodopsin (Ms15 and Md,Sf. Ps62 and P, I z both appear to be of rhabdomal origin (Fig. 6), and once they exist in solution, neither is converted into the other (Figs 7, 8 and related text). They might arise from two populations of RSJo. but Rs3* is seemingly homogeneous with respect to absorption spectrum, photochemical kinetics (Cronin and Goldsmith, 19821, and molecular weight (Larrivee and Goldsmith, unpublished observations). RsaO might nevertheless exist in two forms, differing in primary structure, or, perhaps more likely, in some secondary modification such as phosphorylation or nature of carbohydrate side chain. Alternatively, either some fraction of Rs30 might become chemically modified on extraction, or R,,, might slip into two alternative conformations when its native association with the membrane is changed by digitonin. Although crab rhodopsin seems to undergo a hy~ochromic shift on extraction from the rhabdom (Bruno and Goldsmith, 1972), the simultaneous appearance of two conformers, one batho chromicaliyand the other hypsochromically shifted, has not been reported previously for any visual pigment. 0[.."1".""""' When metarhodopsin is made by irradiating RSfO 400 450 500 550 600 WAVELENGTH tnm) in the rhabdom, the I,,, lies at about 515 nm, with no evidence for Ma,5 (Goldsmith. 1978a: Cronin and Fig. 15. Comparison of Mst5 (from Fig. 4) as measured in solution (solid curve) with microspectroohotometric Goldsmith, 1981). Furthermore, when this metarhodopsin is extracted it is recovered in digitonin sol- measurements of metarhodopsin in the-rhabdom (dashed ution as Msts (Fig. 10). This experiment provides curve, from Goldsmith 1978al. The MSP measurement is the average difference spectrum for the photobleaching of further evidence against the presence of both PshZ M,,, at pH9 in 1OY.bformaldehyde-fixed rhabdoms. The and Psl Z in the rhabdom. and is most readily comcurve for Msls in sift appears to be slightly broader than absorption

patible with a single spectral

form (RssO) that adopts

in solution.

736

DENISLARRIVEE and TIMOTHYH.

conformation does not occur in the vicinity of the metarhodopsin chromophore once metarhodopsin is formed in the rhabdom prior to extraction. Furthermore, the metarhodopsin formed in situ from RsJ, and subsequently extracted appears to be the same as that formed in vitro from PsbZ (Fig. 10). Whatever changes digitonin imposes on the mole&le in the R s3@--+ Ps6* transition are therefore relieved when the chromophore isomerizes to the all-rruns configuration. The same cannot be said for the digitoninimposed R530-+ P,,, transition, because following isomerization of the chromophore. Ms15 does not appear. Instead the photoproduct is displaced still further hypsochromically, with i,,, at 475 nm. Some intimacy of opsin~hromophore interaction is present in the PsG2-M sls system that is not present in the P,,2-M47s pair. At the same time, the relationship is more fragile in the first instance, because Ps,, is more sensitive to denaturation by heat (Fig. 7) or an alkaline environment (Fig. 11) than is P,,,. In this interpretation, PslZ appears to have assumed a tertiary structure from which the conversion to M515 is improbable. The reasons for this are not clear. M, 15 has a more open structure than P,,, in the sense that it is more susceptible to attack at the chromophore binding site by NH,OH. But the protein in P,,Z appears to have lost some capacity for interacting with the all-trrms chromophore, because folIowing isomerization, the spectrum of the metarhodopsin (MkT5) is shifted towards that of an unenhanced Schiff base. We have suggested that the digitonin micelle provides a supportive matrix in which crayfish opsin adopts two distinct conformations that are different from the conformation that exists in the rhabdom. In this respect digitonin differs from the detergent Ammonyx LO. which destabilizes the pigment and leads to a loss of absorption. Paramagnetic resonance spectra of spin labeled phospholipids solubilized in digitonin display much less molecular motion than in other detergents, indicating a rigid structure for the digitonin micelle (Hong and Hubbell. 1972). This rigidity appears to be a prerequisite for the regeneration of vertebrate rhodopsin in detergent solution (Hong and Hubbell. 1973). Moreover, the relative immunity of the absorption spectrum of rhodopsin to elevated pressure indicates a rigid internal environment in the chromophore region of the protein (Lamola at u!., 1974). That a rigid environment may be required for the structural integrity of crayfish rhodopsin is further suggested by the absence of translational motion and relatively restricted rotational movement in the membranes of the photoreceptor (Goldsmith and Wehner, 1977).

Acknowledyemenfs-This work was supported by a grant from the National Institutes of Health (EY-00222). D.L. was a Postdoctoral Trainee on National Institutes of

Health Training Grant T32-EY07000. We are grateful to Professor Allen Kropf, in whose laboratory the high pressure

liquid

chromatography

of

retinaldehyde

was

per-

GOLDSMITH

formed. and to Thomas Cronin. for numerous discussions of the data and a critical reading of the manuscript.

REFERENCES Applebury M. L.. Tuckerman D. M.. Lamola A. A. and Jovin T. M. t 1974) Rhodoosin ourification and recombi-

nation with phospholipids

assayed by the Metarhodop-

sin I - Metarhodopsin II transition. Biochem~.~rr~ 13. 3448-3458. Bruno M. and Goldsmith T. H. (1972) Rhodopsin of the blue crab Caflinrcres: Evidence for absorption differences in vitro and in viva. Vision Rex 14. 653-658. Crescitelli F. (1980) The gecko visual pigments. The nitrate effect. Vision Res. 20, 937- 945. Cronin T. and Goldsmith T. H. (1981) Chromophore fluorescence of crayfish metarhodopsin studied in single rhabdoms. Biop~~~. J. 35.6.53-664. Cronin T. and Goldsmith T. H. (1982) The photosensitivity spectrum of crayfish rhodopsin measured using fluorescence of metarhodopsin. 1. yen. Physio/. 79, 313-332. Denys C. J. (1982) Ommochrome pigments in the eyes of Euphuusiu superha (Crustacea. Euphausiacea). Polar Biol. In press. Ebrey T. (I971 1The use of Ammonyx LO in the purification of rhodopsin and rod outer segments. Vision Res. II, 1007-1009. Ebrey T. and Honig B. (1975) Molecular aspects of photoredeptor function. Q. Rev. Biophys. 8, 129-184. Ebrev T. and Honie. B. (19771 New wavelength denendent visual pigment &mo&ams: Vision Rex 1% 147-i 51. Eguchi E. and Waterman T. H. (1979) Longterm dark induced fine structural changes in crayfish uhotoreceotor membrane. J. camp. Physiof: 131, l9i-203.’ ’ Faeer L. Y. and Faeer R. S. (19791 Halide control of color of the chicken coie pigment iodopsin. Expl Eye Res. 29, 401-408. Fernandez H. (1965) A survey of the visual pigments of decapod crustacea of south Florida. Ph.D. Thesis. University of Miami, Coral Gables, Florida. Fujimoto K., Yanase T. and Ishizuka I. (1966) The visual substance of the crayfish, Procambarus &&ii. Mem. Osaka Gakugei Uniu. B. 15, 109-114. Goldsmith T. H. (1972) The natural history of invertebrate visual pigments. In Handbook of Sensory Physiol, Vol. VII/l Photochemistry of Vision (Edited by Dartnall H. J. A.). pp. 685-719. Springer-Verlag. Berlin. Goldsmith T. H. (1978a) The spectral absorption of crayfish rhabdoms: pigment, photoproduct and pH sensitivity. Vision Rex 18, 463-473. Goldsmith T. H. (1978b) The effects of screening pigments on the spectral sensitivity of some crustacea with scotopit (superposition) eyes. Vision Rex 18. 475-482. Goldsmith T. H. and Fernandez H. R. (1968) Comparative studies of crustacean spectral sensitivity. 2. rrrgl. Physiol. 60, 156175. Goldsmith T. H. and Wehner R. (1977) Restrictions on rotational and translational diffusion of pigment in the membranes of a rhabdomeric photoreceptor. J. yen. Phy siol. 70, 453490. Groenendijk G. W. T., Jacobs C. W. M.. Bonting S. L. and Daeman F. J. M. (1980) Dark isomerization of retinals in the presence of phosphat~dyl ethanolamine. Eur. f. Bio&em. 106, 119-12.8. Hagins F. (1973) Purification and partial characterization of the motein comnonent of sauid rhodo&n. J. hiol. Chem. i48, 3298-3304. Hararave P. (1976) The lame scale oreoaration of rhodopsin. Vision Rex. 16, lOl3-ylOl4. ’ Hijglund G., Hamdorf K., Langer H.. Paulsen R. and Schwemer J. (1973) The photopigments in an insect retina. In Bi~Jc~~e~nj.sfr~ and P~z~s~~J/~)~~ 0/ Vistriil Pi~rn~~f.~

Crayfish visual pigment te$;;d

by Langer H.), pp. 167-175. Springer-Verlag.

Hong K. and Hubbell W. L. (1972) Preparation and properties of lipid bilayers containing rhodopsin. Proc. natn. Acad. Sci. U.S.A. 69, 2617-2621. Hong K. and Hubbell W. L. (1973) Lipid requirements for rhodonsin reaenerabilitv. Biochemistrr 12, 45174523. Honig B:, Din”; U.. Nakanishi K.. Balogh-Nair V.. Gawinowicz M. A., Arnaboldi M. and Motto M. G. (1979) An external point-charge model for wavelength regulation in visual pigments. J. Am. them. Sot. 101, 7084-7086. Hubbard R. and St George R. C. C. (1958) The rhodopsin system of the squid. J. gen. Physiol. 41, 501-528. Kennedy D. and Bruno M. S. (1961) The spectral sensitivity of crayfish and lobster vision. .I. gen. Physiol. 44, 1089-I 102. Knowles A. (1976) The effects of chloride ion upon chicken visual pigments. Biochem. hiophys. Res. Commun. 73, 5662. Kong K., Goldsmith T. H. (1977) Photosensitivity of retinular cells in white-eyed crayfish (Procomharus clarkii). d. camp. Phrsiol. 122, 273-288. Kropf A. and Hubbard R. (1958) The mechanism of bleaching rhodopsin. Arm. N.Y. Acad Sci. 74, 266-280. Lamola A. A., Yamane T. and Zipp A. (1974) Effects of detergents and high pressures upon the metarhodopsin I* metarhodopsin II equilibrium. Biochemistry 13, 738-745. Larrivee D. C. and Goldsmith T. H. (1982) Isolation of

737

crayfish rhabdoms. In Methods in Enzymology, Biomemhrunes (Edited by Packer L.), Academic Press, New York. In press. Marak G. E., Gallik M. S. and Cornesky R. A. (1970) Light sensitive pigments in insect heads. Ophthal. Rex 1, 65-71. Matsumoto H. and Yoshizawa T. (1978) Recognition of opsin to the longitudinal length of retinal isomers in the formation of rhodopsin. Vision Res. 18, 607-609. Nosaki H. (1969) Electrophysiological study of color encoding in the compound eye of crayfish, Procamharus c/a&ii. 2. oergl. Physiol. 64, 318-323. Ostroy S. E. (1978) Characteristics of Drosophila rhodopsin in wild type and Norp A vision transduction mutants. .I. gen. Physiol. 72, 717-723. Paulsen R. and Schwemer J. (1972) Studies on the insect visual pigment sensitive to ultraviolet light. Retinal as the chromophoric group. Biochim. biophys. Acta 283, 520-529. Wald G. (1967) Visual pigments of crayfish. Nature 215, 1131-I 133. Wald G. (1968) Single and multiple visual systems in arth-: ropods. .I. gen. Physiol. 51, 125156. Warshel A. (1978) Charge stabilization mechanism in the visual and purple membrane pigments. Proc. natn Acad. Sci. U.S.A. 75. 2558-2562.

Waterman T. H. and Fernandez H. R. (1970) E-vector and wavelength discrimination by retinular cells of the crayfish Procambarus. Z. rergl. Physiol. 68, 154174.