expression of soluble phototransduction-associated proteins in

Expression of Soluble Phototransduction-AssociatedProteins in Ground Squirrel Retina

Malcolm von Schantz*% Agoston Szel,*~\ Theo van Veen* andDebora B. Farber\

Purpose. This study describes the expression and distribution of arrestin, phosducin, andrecoverin in the cone-dominant retina of the ground squirrel Spermophilus tridecemlineatus.

Methods. mRNA expression was studied by blot hybridization of ground squirrel retinal RNA,with human and murine RNA as controls. The distribution of the gene products in theground squirrel retina was investigated by immunocytochemistry using radial and consecutivetangential sections.

Results. Northern blot hybridization showed messages for arrestin (1.9 kb), phosducin (1.4kb), and recoverin (1.2 kb) in ground squirrel retinal RNA. Both controls showed transcriptsof the same or similar sizes. Rod-like cells and blue cones were stained by antibodies againstarrestin and phosducin. The arrestin antiserum stained the whole cell bodies, most intenselyin the myoid region, whereas phosducin immunoreactivity was confined to the outer andinner segments, which were stained with approximately equal intensity. The strongest immu-noreaction was found in the photoreceptor plasma membrane. Recoverin antibodies recog-nized the entire soma of all photoreceptor cells. The myoid region and the synaptic pedicleswere most heavily stained. No light-dependent migration was observed with either antiserumin any photoreceptor type.

Conclusion. The presence of arrestin immunoreactivity in rod-like cells and blue cones isconsistent with previous reports on other mammals. However, it has not been reported pre-viously that phosducin immunoreactivity is distributed in the same way. The colocalization ofarrestin and phosducin in rod-like cells and blue cones is yet another trait distinguishing bluecones from red and green cones. Invest Ophthalmol Vis Sci. 1994;35:3922-3930.

A he retina of the 13-lined ground squirrel Spermophi-lus tridecemlineatus contains approximately 90.6%green cones, 6.1% rod-like cells, and 3.1% bluecones.1 In contrast to the rod-like cells found in theEuropean ground squirrel Spermophilus citellus,2 thosein S. tridecemlineatus appear to be a homogeneousgroup with respect to structural traits and biochemical

From the *Department of Zoology, University of Gb'teborg, Sweden; %the Jules SteinEye Institute, University of California at Los Angeles School of Medicine, LosAngeles, California; and the '[Second Department of Anatomy, Histology, andEmbryology, Semmelweis University of Medicine, Budapest, Hungary.Supported by grants from the Paul and Marie Berghaus fund, the Count KnutPosse fund (MvS), Hungarian Ministry of Welfare grant T-243, HungarianScience Research Fund grant 1134 (AS), Swedish Natural Science Research councilgrant 4644-311 (TvV), the Retinitis Pigmentosa Foundation Fighting Blindness(TvV and DBF), and National Institutes of Health grant EY02651 (DBF). DBF isthe recipient of a Research to Prevent Blindness Senior Scientific Investigator Award.This article is dedicated to professor RolfElofsson on the occasion of his retirement.Submitted for publication March 7, 1994; revised May 19, 1994; accepted May 25,1994.Proprietary interest category: N.Reprint requests: Dr. Debora B. Farber, Jules Stein Eye Institute, UCLA School ofMedicine, 100 Stein Plaza, Los Angeles, CA 90024.

markers. The photopigment in the rod-like cells of S.tridecemlineatus consistently cross-reacts with an anti-body against blue cone pigment. We have previouslystudied the expression and distribution of the mem-brane-associated components of the phototransduc-tion cascade in the ground squirrel retina.1 Other in-vestigators have taken advantage of the dominanceof cones in the ground squirrel retina to prove thatinterphotoreceptor retinoid-binding protein (IRBP)3

and peripherin/rds4 are expressed in cones as well asin rods.

The traditional subdivision of retinal photorecep-tors into rods and cones was originally based on differ-ences in general structure and function. It was laterfound that rod outer segment disks are separate mem-brane enclosures within the cell membrane, whereascone outer segment disks are imaginations of the cellmembrane. Moreover, cones are equipped with theirown phototransduction cascade, which consists of pro-teins similar to but distinct from those found in rods.1

3922Investigative Ophthalmology & Visual Science, October 1994, Vol. 35, No. 11Copyright © Association for Research in Vision and Ophthalmology

Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933401/ on 02/11/2018

Phototransduction Proteins in the Ground Squirrel 3923

There are, however, several traits that distinguishcones sensitive to short wavelengths (i.e., blue-sensi-tive cones) from those sensitive to long or mediumwavelengths (i.e., red-sensitive or green-sensitivecones). The most prominent of these is the greaterdiameter of blue cones. Biochemical distinctions in-clude differences in fucose incorporation5 and instaining by procion yellow6 and by antibodies againstcarbonic anhydrase.7 In the ground squirrel retina, wehave also reported a difference in peanut agglutininbinding to the interphotoreceptor matrices of blueand green cones.8

Arrestin, phosducin, and recoverin are all solubleproteins that are expressed in relatively high amountsin the mammalian retina. Initially discovered and de-scribed in the retina, where they are thought to inter-act with the proteins of the phototransduction cas-cade, they were all subsequently found to be expressedin other tissues as well. The three proteins are alsoknown to be members of larger families.

Arrestin (S-antigen, 48-kd protein) is a stronglyimmunogenic, soluble retinal protein9 that takes partin the quenching of the light signal by binding tophosphorylated rhodopsin, thereby preventing its in-teraction with a-transducin.10 Arrestin immunoreactiv-ity is almost ubiquitous in photoreceptors throughoutthe animal kingdom.11 The mammalian pineal organcontains a form of arrestin12 that is encoded by anmRNA identical to that of retinal arrestin.13 The ar-restin promoter also directs expression into the lensand the brain.14 Arrestin is now known to be part ofa family of at least four homologous proteins.15"17

Phosducin (MEKA, 33-kd phosphoprotein) is an-other abundant soluble retinal protein that can forma heterotrimeric complex with the /3y-transducin het-erodimer.18 Similar to arrestin, it is present in thepineal organ.19 It appears to be widely expressedamong vertebrates, including teleosts.20 A recent studyhas described the inhibition of the interaction be-tween the /3-adrenergic receptor and its G-protein bya 33-kd protein.21 This protein was found to be similaror identical to the retinal phosducin. There are nodata available to indicate whether a similar mechanisminvolving phosducin is at work in the retina. However,phosducin seems to interact with the proteins of thephototransduction cascade competing with a-trans-ducin for binding to the /?y-transducin subunits.22 Inanother study, a molecule that has sequence similari-ties with phosducin has been demonstrated in a widevariety of tissues.

Recoverin is a 26-kd calcium-binding protein ini-tially observed in the bovine retina24 and presumedto promote a recovery of the photoreceptor cGMPconcentration after illumination by acting on guanyl-ate cyclase. Recent data, however, demonstrate thatrecoverin does not activate guanylate cyclase, as pre-

viously thought.25 Another recent report shows thatin the frog retina, the regulator protein S-modulinprolongs the lifetime of activated phosphodiesteraseat high calcium ion concentrations by inhibiting rho-dopsin phosphorylation.26 Recoverin is thought to bea mammalian homologue of S-modulin, and it maytherefore play the same role. Recoverin immunoreac-tivity has also been found in cone bipolar neurons27

and in the pineal organ.28 Again, several similar mole-cules have been reported to be present in the nervous

on

system/3

In this investigation, we have studied the expres-sion of arrestin, phosducin, and recoverin in theground squirrel retina using RNA blot hybridizationand immunocytochemistry. The immunocytochemicalexperiments were performed using radial sections orconsecutive tangential sections. In the latter, we useda battery of photopigment antibodies to discriminateamong the three types of photoreceptors, as describedpreviously.1 Briefly, green cones are identified by theantibody COS-1. Both blue cones and rod-like cellsare recognized by the blue-cone pigment antibody OS-2 and are distinguished by their shapes and by thefact that only rod-like cells are labeled by the opsinantiserum AO. Part of this investigation has previouslybeen published in abstract form.30

MATERIALS AND METHODS

Animals and Tissue Preparation

All animals were maintained and killed in accordancewith the ARVO Statement for the Use of Animals inOphthalmic and Vision Research.

Wild ground squirrels (S. tridecemlineatus) weretrapped by commercial vendors in Illinois and Wiscon-sin, shipped to our laboratory, and maintained on agrain-nut diet under a 12-hour light/12-hour darkcycle. Animals were killed and the retinas prepared forRNA analysis or immunocytochemistry as describedpreviously.1 Light-adapted animals were killed at 9AM; dark-adapted animals were killed and dissectedunder infrared light after 3 hours of darkness.

Human eyes were enucleated 2 to 12 hours afterdeath and kept at — 80°C until thawed and dissectedfor RNA extraction. The tenets of the Declaration ofHelsinki were followed, with informed consent andapproval by the institutional human experimentationcommittee obtained.

RNA Analysis

RNA was purified as described previously.31 Poly-ARNA was obtained using the Mini-Oligo(dT) CelluloseSpin Column kit (5 Prime —• 3 Prime, Boulder, CO).Total RNA (10 fig) or poly-A RNA (1 to 2 fig) wasseparated in a 1.2% agarose-formaldehyde gel. RNA


3924 Investigative Ophthalmology 8c Visual Science, October 1994, Vol. 35, No. 11

TABLE l. Antibodies Used for Immunocytochemistry

Designation

AOOS-2COS-14/17/85Anti-32S6K22

Antigen

OpsinBlue opsinGreen opsinArrestinPhosducinRecoverin

Origin

BovineChickenChickenBovineBovineBovine

Host

RatMouseMouseRabbitRabbitRabbit

Type

PMMPPP

Dilution

1:10,0001:5,0001:501:1,0001:1,0001:300

Source

Szel and Rohlich32

Szel and Rohlich32

Szel and Rohlich32

T. Shinohara33

T. Shinohara34

A. S. Polans35

P = Polyclonal; M = monoclonal.

was blotted overnight onto Hybond N+ membranes(Amersham, Little Chalfont, Bucks, UK) in 20 X SSC.Cross-linking of RNA to the membrane and strippingof the probe between hybridizations were performedaccording to the manufacturer's instructions. The fol-lowing cDNA clones were used for hybridization:pHL1301, a 1.2-kb Eco RI fragment encoding humana-tubulin36; BSC70, a 1.6-kb Eco RI fragment encod-ing bovine arrestin37; zr.414, a 1.1-kb Eco RI fragmentencoding murine phosducin (Bowes-Rickman C, Ra-poport A, Farber DB, unpublished data, 1994); andRecV, a 0.5-kb Pst I-Eco RI fragment encoding bovinerecoverin.24 DNA (50 ng) was labeled38 using theKlenow fragment of DNA polymerase (Promega, Mad-ison, WI) and a-32P-dCTP (NEN Research Products,Boston, MA). Hybridization was carried out overnightin a hybridization oven (Bellco Biotechnology, Vine-land, NJ) at 65°C in a phosphate-buffered 1% BSAsolution containing 7% SDS. Washing and visualiza-tion were performed as described previously.1 Eachblot was initially hybridized with the tubulin probe toensure that equal amounts of intact RNA were presentin all lanes.

Immunocytochemistry

Enucleated eyecups were immediately immersed in4% paraformaldehyde and 0.1% glutaraldehyde dis-solved in 0.1 M Sorensen phosphate buffer (pH 7.2)for 4 hours. The material was embedded either inaraldite (Durcupan ACM; Fluka, Buchs, Switzerland),or in diethylene glycol distearate wax (DGD; Polys-ciences, Warrington, PA) ,39 The araldite and the DGDblocks were sectioned on an ultramicrotome (1.5 fxm)with glass knives. Serial tangential and radial sectionswere obtained. Immunoreaction, visualization withthe avidin-biotin-peroxidase system (Vectastain;Vector Laboratories, Burlingame, CA), and mountingof slides were performed as described previously.1 Theantibodies used are summarized in Table 1. Themonospecificity of all antibodies had been checkedby Western blotting and by blocking with competingantigen (see references in Table 1). All antibodieshave been found to perform consistendy in a widevariety of mammalian species. The results of the im-

munocytochemistry were visualized with an Axiophotphotomicroscope (Zeiss, Oberkochen, Germany) us-ing Nomarski optics. For demonstration of colocaliza-tion of photoreceptor-specific proteins, correspond-ing areas of adjacent tangential sections were photo-graphed.

The embedding techniques were selected ac-cording to the specific properties of the primary anti-bodies. Although the antibodies exhibited the samestaining pattern regardless of the method chosen, theintensity of the immunoreaction was subject to consid-erable variation depending on the embedding me-dium used.

RESULTS

Single 1.9-kb arrestin transcripts were detected in allthree species (Fig. 1A). The mouse phosducin probealso hybridized to single transcripts, sized 1.2 kb inmurine, 1.4 kb in ground squirrel, and 1.1 kb in hu-man retina (Fig. IB). The signal was considerablystronger in the lane containing mouse retinal mRNA.The recoverin probe visualized one abundant tran-script in all three species (Fig. 1C). The transcript sizewas 1.0 kb in murine and ground squirrel RNA and1.2 kb in human RNA.

A strong labeling with the arrestin antibody wasobserved in the whole soma of a minor subpopulationof photoreceptors (Fig. 2A). The most intense label-ing was localized to the vitreal half of the inner seg-ment (myoid). The blue pigment antibody OS-2 rec-ognized blue cones, which were less abundant andhad a larger outer segment diameter than did greencones, and rod-like cells, identified by their smallerdiameter (Fig. 3A). Only the latter cell type was recog-nized by the rhodopsin antibody AO (Fig. 3B). Immu-nolabeling of consecutive sections showed that all blueopsin-positive elements were also positive for arrestin,whereas no green cones were recognized by the ar-restin antibody (Fig. 3E).

The phosducin antiserum labeled rod-like cells,as well as a subpopulation of cone cells (Fig. 2B). Aswith arrestin, the labeling was present in the entirecell soma; unlike arrestin, however, the labeling was



ical staining pattern within the photoreceptor mosaicwas the same with light-adapted and dark-adapted ani-mals.

DISCUSSION

The intensity of arrestin cDNA hybridization wasroughly equal in all three species studied—a surpris-ing result, because arrestin immunoreactivity is foundin fewer cells in the ground squirrel retina than inthe murine and human retinas. The phosducin cDNAhybridized more strongly with murine retinal mRNAthan with mRNA from other species. This is to beexpected, because the probe is of murine origin. Thesignal was, however, weaker in the ground squirrellane than in the lane containing human mRNA, con-sistent with the fact that ground squirrel rod-like cellsare much less abundant than human rods.

Antibodies against arrestin and phosducin labeled

B

FIGURE l. Northern blots hybridized with cDNA probes forarrestin (A), phosducin (B), and recoverin (C). On all blots,murine RNA is seen in the left lane, human RNA in themiddle lane, and ground squirrel RNA in the right lane.(A) and (C) contain 1.0 fig of poly-A RNA per lane, and(B) contains 2.0 ^g per lane. (A) was exposed for 4 days.(B) is a composite, showing a 4-day exposure of the groundsquirrel and human lanes, whereas the mouse lane has onlybeen exposed for 6 hours. (C) was exposed for 16 hours.

equally intense in outer (Fig. 3C) and inner segments(Fig. 3D) and was most intense in the photoreceptormembrane. The strength of the immunoreaction wasconsiderably lower than that found with the arrestinantibody. Immunolabeling of consecutive sectionsshowed all identifiable blue opsin-positive elements(Fig. 3A) to be positive for phosducin (Figs. 3C, 3D).No immunoreaction was detected in green cones.

The recoverin antiserum labeled all photorecep-tor cells (Fig. 4). The label intensity was found to bethe strongest in the myoids and in the pedicles ofthe photoreceptors. The outer segments were stainedweakly, and a diffuse label was observed in the entireouter nuclear layer. A few cells in the inner nuclearlayer were also recognized (Fig. 4A). Tangential sec-tions (Fig. 4B) were used to show that no photorecep-tor cell was left unstained. All elements encounteredat either the outer segment level or the myoid levelwere recognized by the recoverin antibody.

For all the antibodies used, the immunocytochem-

BFIGURE 2. Consecutive radial araldite sections of the groundsquirrel retina reacted with arrestin (A) and phosducin (B)antibodies. Both antibodies stain only a few photoreceptors.In these cells, anti-arrestin stained the complete cell body,showing the most intensive reaction in the myoid. The phos-ducin antibody recognized an equally small subpopulationof photoreceptors. Note that the reaction is most intensivein the cell membrane. Some of the stained cells are markedwith black arrows. The outer limiting membrane (OLM)and the retinal pigment epithelium (RPE) are marked withopen arrow and asterisk (*), respectively. Magnification,X825.


3926 Investigative Ophthalmology & Visual Science, October 1994, Vol. 35, No. 11

FIGURE 3. Parallell tangential araldite sections showing the same retinal area reacted withantibodies against blue opsin (A), rhodopsin (B), phosducin <C» D), and arrestin (E). (F)is a diagram made on figure (A) to show the individual photoreceptor types. Sections (A),(B), and (C) represent the outer segment level; at the lower edge of the sections, however,inner segments also appear. Sections (D) and (E) were taken from a more vitreal level;therefore, mainly inner segments are encountered. The honeycomb structure produced bythe apical processes of the RPE cells is present only at the outer segment level. The antibodyCOS-1 labeled green cones (not shown). The wider blue cone outer segments are stainedby OS-2 (A, thick arrow). The rod-like cells, which have a considerably smaller outer segmentdiameter, are also stained weakly by this antibody (A, thin arrow). Anti-rhodopsin exclusivelystains rod-like cells (B, thin arrow). The phosducin antibody stains both blue cones {thickarrows) and rod-like cells (thin arrows) at both the outer (C) and inner (D) segment levels.Similarly, anti-arrestin reaction is found in both photoreceptor types (E). Note that thelatter antibody gives the highest staining intensity in the vitreal level of the inner segments.Green cones are not stained by any of the antibodies used. The scheme in (F) shows thethree photoreceptor types that can be identified in (A) through (E). Blue cones are black,green cones are left unmarked, and rod-like cells are labeled by small black dots. Magnifica-tion, X840.

rod-like photoreceptors and a subpopulation of thecone photoreceptors that were identified as bluecones by immunocytochemistry. Previous reports arenot conclusive regarding arrestin localization. Earlierstudies demonstrated arrestin immunoreactivity inrods as well as in cones in several vertebrates,40'41 but

other investigators could not detect arrestin in rhesusmonkey or squirrel cones,42 human cone outer seg-ments,43 or porcine cones.44 Positive immunoreactionwas observed in the tree shrew in a subpopulation ofcones that was probably identical to blue cones,45 simi-lar to what has been found in the ground squirrel



• »••

FIGURE4. Sections of DGD-embedded ground squirrel retinareacted with recoverin antibody. On radial section (A) theRPE and the OLM are marked between the outer nuclear(ONL) and outer plexiform layers. The thick and thin linesmark the OS-IS junction and the myoids. The pedicles ofthe photoreceptors are recognized by anti-recoverin, as isthe ONL. Some cells in the inner nuclear layer are alsolabeled (arrowheads). The tangential section (B) was cut atthe photoreceptor level; the left and right sides of the sec-tion represent the outer and inner segment levels, respec-tively. The thick and thin lines mark the same levels as thosein (A). Note that all photoreceptors are stained at both theOS and IS levels. Magnification, X840.

Spermophitus beecheyi.4G A recent study, performed onthe baboon retina with monoclonal antibodies againstdifferent epitopes on the arrestin molecule, suggeststhat at least two immunologically different types ofretinal arrestin exist, because some epitopes were pres-ent in blue cones but not in red or green cones.47 Theexistence of several retinal arrestin subspecies withvarying isoelectric focusing properties has been re-ported in another investigation.48 It is now known thathomologous molecules such as arrestin3S, /?-ar-restin,15 X-arrestin,16 and C-arrestin17 are also ex-pressed in the retina. A variable degree of cross-reac-tivity between different homologues could explain thedifferences in the previously reported results.

To our knowledge, it has not hitherto been re-ported that phosducin immunoreactivity also displaysa differential distribution among different cone types.Although no reports are available that show immuno-reactivity with other phosducin antisera to red/greencones, it is possible that these photoreceptors areequipped with a homologue with a different primarystructure. At least two phosducin genes with some dif-ferences in the 5'-flanking region are present in themouse genome, and two are expressed in the retina.49

There is also evidence for expression of phosducin ora similar polypeptide in several different tissues of thebody.21

As mentioned, phosducin is associated with the/?/y-transducin complex.18 It is now known from stud-ies on cow and monkey retinas that rods and cones areequipped with different forms of fi- and y-transducin.Rods contain the subunits fil and yl, whereas conescontain the homologous /?3 and -y2.50'51 We have pre-viously found a different distribution of the /3-trans-ducin subunits in the ground squirrel retina.1 Wefound /?l-transducin immunoreactivity in rod-likecells and blue cones, whereas a /?3-transducin anti-body labeled all photoreceptor types (y-transducinwas not studied). The apparent colocalization of phos-ducin and /31-transducin in the same photoreceptortypes in the ground squirrel raises the question ofwhether this corresponds to a specific biochemicalassociation as well. This is not supported by publisheddata from the bovine retina, where the phosducin -transducing/y complex reacted with antibodiesagainst both 01- and /?3-transducin.50 However, pre-liminary evidence indicates that bovine cones mayhave a specific phosducin with a slightly higher molec-ular weight.52 It should also be noted that the phos-ducin localization reported here differs from thatfound in the mouse retina, where a phosducin anti-body labeled the entire cell but was more intense inthe inner segments.53

The localization of arrestin in blue cones and inrods reported here indicates that a mechanism of de-activation of the visual pigment by phosphorylationmay be present in both blue cones and rods. It remainsto be determined whether the interaction betweenarrestin- and phosducin-like proteins reported inbrain preparations is also present in the retina. It is,however, established that long-wavelength cones alsohave a photopigment phosphorylation mechanism, atleast in nonmammalian vertebrates. Thus, in the ret-ina of the lizard Anolis carolinensis, which contains ex-clusively red and green cones, the opsins undergo avery strong light-induced phosphorylation.54 Inchicken, the phosphorylation of the red opsin mole-cule has been specifically demonstrated.55 This phos-phorylation is presumably part of an alternative mech-anism for quenching of the visual pigment, possibly



involving a red and green cone-specific arrestin.47 Thegene for an arrestin that is a likely candidate for thisrole has recently been identified.17

The blue cone photopigment sequence56 is oneof several criteria that place blue cones as an interme-diate between rods and red/green cones, sometimeswith more similarities to the former than to the latter.The findings described here also indicate a closer evo-lutionary relationship between rods and blue cones.

We found an arrestin immunolabeling that wasdistributed throughout the entire cell soma but wasstrongest in the myoid region of the inner segmentin rod-like cells and blue cones. In the light-adaptedbaboon retina, arrestin immunolabeling has been re-ported to be localized to the cone outer segment,47

whereas in rat and mouse cones,57 arrestin immunore-activity remains in the inner segments for an hourafter illumination. These discrepancies may emanatefrom interspecific differences.

Previous reports have described a rapid light-de-pendent migration of phototransduction proteins be-tween inner and outer segments. Thus, arrestin immu-noreactivity is found in light-adapted outer segmentsof rat,58 toad,59 and mouse60 retina, and it migratesinto inner segments on dark adaptation. Moreover,the levels of arrestin mRNA are at their peak just be-fore the onset of the dark period, as opposed to themRNAs encoding the proteins of the phototransduc-tion cascade.61 In contrast, phosducin and transducinare found in light-adapted inner segments and mi-grate to the outer segments in the dark. However,none of these reports describes the situation in conephotoreceptors. The results presented here indicatethat migration of arrestin or phosducin does not takeplace in rod-like cells or in cones. We were also unableto detect any difference in a- or /3-transducin or iny-phosphodiesterase immunolabeling between light-and dark-adapted animals.1 These findings are consis-tent with those reported in other studies on rat andmouse retina.57 Apparently, cone photoreceptors donot have the same mechanism for rapid movement ofphototransduction proteins that rods do. Interest-ingly, no light-dependent migration in rod-like cellswas observed; the lack of such migration is yet anothertrait of these cells more reminiscent of cones than ofrods. These issues would merit a special study withmaterial collected at different time points after illumi-nation.

The recoverin message hybridized strongly withthe corresponding cDNA in all three species, and wefound recoverin immunoreactivity in all photorecep-tor types of the ground squirrel retina. These findingsare in accordance with those in the rod-dominant bo-vine retina.24 A similar immunocytochemical staininghas been found in human retina using the same anti-body as in the present investigation.35 This is in con-

trast to the related protein visinin, which has beenreported to be present only in cones.62 The immuno-reactive cells found in the inner nuclear layer are inaccordance with published reports27'63 and may ex-press either recoverin or a related molecule.

In a previous study, we focused on the expressionof the proteins of the phototransduction cascade inthe retina and found the rod-like cells of the groundsquirrel to be interesting from an evolutionary per-spective, because they appear to be not only structural,but also biochemical intermediates between rods andcones.1 Here, we report new differences between blueand longer-wavelength cones. Our results indicate thatthe classification of visual cells according to structureis not always in agreement with the distribution of theisoforms of their biochemical machinery.

Key Words

cone and rod-like photoreceptors, Spermophilus tridecemlinea-tus, arrestin, phosducin, recoverin

Acknowledgments

The authors thank Artur S. Polans, Winston A. Salser, andToshimishi Shinohara for the gift of DNA probes and anti-bodies. They also thank Clyde K. Yamashita for his invalu-able help with the ground squirrel colony and with the han-dling and dissection of retinas, Katharina Ryden for techni-cal assistance, and Andrea Viczian for sharing human retinalRNA samples.

References

1. von Schantz M, Szel A, van Veen T, Farber DB. Expres-sion of phototransduction genes in the ground squir-rel retina. Invest Ophthalmol Vis Sci. 1994; 34:2558-2566.

2. Szel A, Rohlich P. Four photoreceptor types in theground squirrel retina as evidenced by immunocyto-chemistry. Vision Res. 1988;28:1297-1302.

3. Porrello K, Bhat SP, Bok D. Detection of interphotore-ceptor retinoid binding protein (IRBP) mRNA in hu-man and cone-dominant squirrel retinas by in situhybridization. / Histochem Cytochem. 1991; 39:171 -176.

4. Arikawa K, Molday LL, Molday RS, Williams DS. Local-ization of peripherin/rds in the disk membranes ofcone and rod photoreceptors: Relationship to diskmembrane morphogenesis and retinal degeneration./ Cell Biol. 1992; 116:659-667.

5. Bunt AH, Klock IB. Fine structure and radioautogra-phy of retinal cone outer segments in goldfish andcarp. Invest Ophthalmol Vis Sci. 1980; 19:707-719.

6. de Monasterio FM, Schein SJ, McCrane EP. Stainingof blue-sensitive cones of die macaque retina by afluorescent dye. Science. 1981;213:1278-1281.

7. Nork TM, McCormick SA, Chao G-M, Odom JV. Dis-tribution of carbonic anhydrase among human photo-receptors. Invest Ophthalmol Vis Sci. 1990; 31:1451-1458.

8. Szel A, von Schantz M, Farber DB, van Veen T. Differ-



ence in PNA-label intensity between short- and mid-dle-wavelength sensitive cones in the ground squirrelretina. Invest Ophthalmol Vis Sd. 1993;34:3641-3645.

9. Kalsow CM, Wacker WB. Localization of a uveitogenicsoluble retinal antigen in the normal guinea pig eyeby an indirect fluorescent antibody technique. IntArchAllergy. 1973; 44:11-20.

10. Wilden U, Hall SW, Kilhn H. Phosphodiesterase acti-vation by photoexcited rhodopsin is quenched whenrhodopsin is phosphorylated and binds the intrinsic48-kDa protein of rod outer segments. Proc Natl AcadSd USA. 1986; 83:1174-1178.

11. van Veen T, Elofsson R, Hartwig H-G, et al. Retinal S-antigen: Immunocytochemical and immunochemicalstudies on distribution in animal photoreceptors andpineal organs. Exp Biol. 1986;45:15-25.

12. Kalsow CM, Wacker WB. Pineal reactivity of anti-retinasera. Invest Ophthalmol. 1977; 16:181-184.

13. Abe T, Shinohara T. S-antigen from the retina andpineal gland have identical sequences. Exp Eye Res.

14. Breitman ML, Tsuda M, Usukura J, et al. Expressionof S-antigen in retina, pineal gland, lens, and brain isdirected by 5'-flanking sequences. / Biol Chem.1991;266:15505-15510.

15. Sterne-Marr R, Gurevich V, Goldsmith P, et al. Poly-peptide variants of /?-arrestin and arrestin3. / BiolChem. 1993;21:15640-15648.

16. Murakami A, Yajima T, Sakuma H, McLaren MJ, InanaG. X-arrestin: A new retinal arrestin mapping to theX chromosome. FEBS Lett. 1993;334:203-209.

17. Craft CM, Whitmore DH, Wiechmann AF. Cone ar-restin identified by targeting expression of a func-tional family. J Biol Chem. 1994; 269:4613-4619.

18. Lee RH, Lieberman BS, Lolley RN. A novel complexfrom bovine visual cells of a 33 000-Dalton phospho-protein with /?- and y-transducin: Purification andsubunit structure. Biochemistry. 1987;26:3983-3990.

19. Reig JA, Klein DC. Pineal transduction: Adrenergic— Cyclic AMP-dependent phosphorylation of cyto-plasmic 33-kDa protein (MEKA) which binds beta,gamma-complex of transducin. / Biol Chem. 1990;265:5816-5824.

20. Pagh-Roehl K, Han E, Burnside B. Identification ofcyclic nucleotide-regulated phosphoproteins, includ-ing phosducin, in motile rod inner-outer segments ofteleosts. Exp Eye Res. 1993;57:679-691.

21. Bauer PH, Miiller S, Puzicha M, et al. Phosducin is aprotein kinase A-regulated G-protein regulator. Na-ture. 1992; 358:73-76.

22. Lee RH, Ting TD, Lieberman BS, et al. Regulationof retinal cGMP cascade by phosducin in bovine rodphotoreceptor cells. / Biol Chem. 1992; 267:25104-25112.

23. Miles MF, Barhite S, Sganga M, Elliott M. Phosducin-like protein: An ethanol-responsive potential modula-tor of guanine nucleotide-binding protein function.Proc Natl Acad Sd USA. 1993;90:10831-10835.

24. Dizhoor AM, Ray S, Kumar S, et al. Recoverin: A cal-cium sensitive activator of retinal rod guanylate cy-clase. Sdence. 1991;251:915-918.

25. Gray-Keller MP, Polans AS, Palczewski K, Detwiler PB.The effect of recoverin-like calcium-binding proteinson the photoresponse of retinal rods. Neuron.1993; 10:523-531.

26. Kawamura S. Rhodopsin phophorylation as a mecha-nism of cyclic GMP phosphodiesterase regulation byS-modulin. Nature. 1993;362:855-857.

27. Milam A, Dacey DM, Dizhoor AM. Recoverin immuno-reactivity in mammalian cone bipolar cells. Vis Neu-rosd. 1993; 10:1-22.

28. Korf H-W, White BH, Schaad NC, Klein DC. Recov-erin in pineal organs and retinae of various vertebratespecies including man. Brain Res. 1992;595:57-66.

29. Flaherty KM, Zozulya S, Stryer L, McKay DB. Three-dimensional structure of recoverin, a calcium sensorin vision. Cell. 1993;75:709-716.

30. von Schantz M, van Veen T, Farber DB. Expressionof photoreceptor mRNAs and proteins in a cone-dom-inant retina. ARVO Abstracts. Invest Ophthalmol Vis Sd.1992; 33:1006.

31. Farber DB, Seager DancigerJ, Organisciak DT. Levelsof mRNA encoding proteins of the cGMP cascade asa function of light environment. Exp Eye Res.1991;53:781-786.

32. Rohlich P, Szel A. Binding sites of photoreceptor-spe-cific antibodies COS-1, OS-2 and AO. Curr Eye Res.1993; 12:935-944.

33. Cantera R, von Schantz M, Chader GJ, et al. Postnataldevelopment of photoreceptor-specific proteins inmice with hereditary retinal degeneration: An immu-nocytochemical study. Exp Biol. 1990;48:305-312.

34. Abe T, Nakabayashi H, Tamada H, et al. Analysis ofthe human, bovine, and rat 33-kDa proteins and cDNAin retina and pineal gland. Gene. 1990;91:109-215.

35. Polans AS, Crabb J, Palczewski K. Calcium-bindingproteins in the retina. Methods Neurosd. 1993; 15:248-259.

36. Davis RC, Thompson AR, Fuller ML, et al. mRNA spe-cies regulated during the differentiation of HL-60cells to macrophages and neutrophils. Dev Biol.1987; 119:164-174.

37. Shinohara T, Dietzschold B, Craft CM, et al. Primaryand secondary structure of bovine retinal S-antigen(48-kDa protein). Proc Natl Acad Sd USA. 1987;84:6975-6979.

38. Feinberg AP, Vogelstein B. A technique for radiolabel-ing DNA restriction endonuclease fragments to highspecific activity. Anal Biochem. 1983; 132:6-13.

39. Huang JC, Mieziewska K, Philp N, van Veen T, AguirreGD. Diethylene glycol distearate (DGD): A versatileembedding medium for retinal cytochemistry. / Neu-rosd Methods. 1993;47:227-234.

40. Faure J-P, Mirshahi M, Dorey C, et al. Productionand specificity of monoclonal antibodies to retinal Santigen. Curr Eye Res. 1984; 3:867-872.

41. Mirshahi M, Boucheix C, Collenot G, Thillaye B,Faure J-P. Retinal S-antigen epitopes in vertebrateand invertebrate photoreceptors. Invest Ophthalmol VisSd. 1985;26:1016-1021.

42. Broekhuyse RM, Winkens HJ. Photoreceptor cell-spe-



cific localization of S-antigen in retina. Curr Eye Res.1985;4:703-706.

43. McKechnie NM, Al-Mahdawi S, Dutton G, ForresterJV. Ultrastructural localization of retinal S antigen inthe human retina. Exp Eye Res. 1986;42:479-487.

44. Reid DM, Loeffler KU, Campbell AM, Forrester JV.Electron immunocytochemical localization of retinalS-antigen with a rat monoclonal antibody. Exp Eye Res.1987;45:731-745.

45. Milller B, Peichl L, De Grip WJ, Gery I, Korf H-W.Opsin- and S-antigen-like immunoreactions in photo-receptors of the tree shrew retina. Invest OphthalmolVisSci. 1989; 30:530-535.

46. Long KO, Fisher SK. The distribution of photorecep-tors and ganglion cells in the California ground squir-rel, Spermophilus beecheyi. J Comp Neurol. 1983; 221:329-340.

47. Nir I, Ransom N. S-antigen in rods and cones of theprimate retina: Different labeling patterns are re-vealed with antibodies directed against specific do-mains in the molecule. / Histochem Cytochem. 1992;40:343-352.

48. Weyand I, Kuhn H. Subspecies of arrestin from bovineretina: Equal functional binding to photoexcited rho-dopsin but various isoelectric focusing phenotypes inindividuals. EurJ Biochem. 1990; 193:459-467.

49. Abe T, Kikuchi T, Chang T, Shinohara T. The se-quence of the mouse phosducin-encoding gene andits 5'-flanking region. Gene. 1993; 133:179-186.

50. Lee RH, Lieberman BS, Yamane HK, Bok D, FungBK-K. A third form of the G protein ft subunit: 1:Immunochemical identification and localization tocone photoreceptors. / Biol Chem. 1992; 34:24776-24781.

51. Peng Y-W, Robishaw JD, Levine MA, Yau K-W. Reti-nal rods and cones have distinct G protein (5 and7 subunits. Proc Natl Acad Sci USA. 1992;89:10882-10886.

52. Lee R, Fung B, Yamane H. Cone photoreceptors con-tain a unique cone F-transducin and a putative conephosducin. ARVO Abstracts. Invest Ophthalmol Vis Sci.1993; 34:1360.

53. Lee RH, Whelan JP, Lolley RN, McGinnis JF. Thephotoreceptor-specific 33 kDa phosphoprotein ofmammalian retina: Generation of monospecific anti-bodies and localization by immunocytochemistry. ExpEye Res. 1988;46:829-840.

54. Walter AE, Shuster TA, Farber DB. Light-inducedphosphorylation of proteins from the all-cone retinaof the lizard, Anolis carolinensis. Invest Ophthalmol VisSci. 1986;27:1609-1614.

55. Fukada Y, Kokame K, Okano T, et al. Phosphorylationof iodopsin, chicken red-sensitive cone visual pigment.Biochemistry. 1990; 29:10102-10106.

56. NathansJ, Thomas D, Hogness DS. Molecular geneticsof human color vision: The genes encoding blue,green, and red pigments. Science. 1986;232:193-202.

57. Mirshahi M, Thillaye B, Tarraf M, de Kozak Y, FaureJ-P. Light-induced changes in S-antigen (arrestin) lo-calization in retinal photoreceptors: Differences be-tween rods and cones and defective process in RCSrat retinal dystrophy. EurJ Cell Biol. 1994;63:61-67.

58. Broekhuyse RM, Tolhuizen EFJ, Janssen APM, Win-kens HJ. Light induced shift and binding of S-antigenin retinal rods. Curr Eye Res. 1985;4:613-618.

59. Mangini N, Pepperberg DR. Localization of retinal"48K" (S-antigen) by electron microscopy. JpnJ Oph-thalmol. 1987; 31:207-217.

60. Whelan JP, McGinnis JF. Light-dependent subcellularmovement of photoreceptor proteins. / Neurosci Res.1988; 20:263-270.

61. Bowes C, van Veen T, Farber DB. Opsin, G-proteinand 48-kDa protein in normal and rd mouse retinas:Developmental expression of mRNAS and proteinsand light/dark cycling of mRNAs. Exp Eye Res.1988; 47:369-390.

62. Hatakenata S, Kiyama H, Tohyama M, Miki N. Immu-nohistochemical localization of chick retinal 24 kdal-ton protein (visinin) in various vertebrate retinae.Brain Res. 1985;331:209-215.

63. Wiechmann AF, Hammarback JA. Molecular cloningand nucleotide sequence of a cDNA encoding recov-erin from human retina. Exp Eye Res. 1993; 56:463-470.


expression of soluble phototransduction-associated proteins in

Documents