Vision Research 50 (2010) 2055–2061

Contents lists available at ScienceDirect

Vision Research

journal homepage: www.elsevier .com/locate /v isres

Spectral sensitivity of single cones in rainbow trout (Oncorhynchus mykiss):A whole-cell voltage clamp study

Leslie G. Anderson a, Shai Sabbah b, Craig W. Hawryshyn b,*

a Department of Biology, University of Victoria, P.O. Box 3020 Station CSC, Victoria, British Columbia, Canada V8W 3N5b Department of Biology and Center for Neuroscience Studies, Queen’s University, Kingston, Ontario, Canada K7L 3N6

a r t i c l e i n f o

Article history:Received 30 May 2010Received in revised form 12 July 2010

Keywords:UltravioletOpsin gene expressionFishVisual pigmentsPhotocurrent responses

0042-6989/$ - see front matter � 2010 Elsevier Ltd. Adoi:10.1016/j.visres.2010.07.013

* Corresponding author.E-mail address: craig.hawryshyn@queensu.ca (C.W

a b s t r a c t

The UVS cone mechanism is known to light adapt at low intensities in comparison to other cones. Wewere interested in whether this property was related to higher sensitivity in UVS cones or to networkadjustments in sensitivity. We investigated spectral sensitivity of 107 individual cone photoreceptorsin rainbow trout (Oncorhynchus mykiss) using a whole-cell voltage clamp technique. Mean time-to-peakresponse was 339 ± 90 ms and flash sensitivity for a 100 ms flash was 4.37 � 10�3 ± 2.50 � 10�3 pA pho-tons�1 lm2, with no significant differences between the UVS, SWS, MWS and LWS cone classes. The spec-tral sensitivity of the UVS, SWS and LWS cones conformed to the expression of SWS1, SWS2 and LWSopsin genes. The spectral sensitivity of MWS cones, however, showed clear evidence of co-expressionof RH2a and RH2b opsin pigments. The fish used in this study bridged the ontogenetic stage where theMWS cones shift their expression from RH2b to RH2a.

� 2010 Elsevier Ltd. All rights reserved.

1. Introduction

The visual system of rainbow trout as well as of other Pacific sal-monids is currently of interest to those studying ultraviolet (UV)photoreception and its functional implications (Hawryshyn, 1992,2000). Examination of sensitivity to UV light in salmonids hasincorporated psychophysical techniques (Hawryshyn, Arnold, Bow-ering, & Cole, 1990), microspectrophotometry ( Hawryshyn, Haim-berger, & Deutschlander, 2001; Hawryshyn & Hárosi, 1994; Templeet al., 2006), compound action potential recordings (Beaudet, Brow-man, & Hawryshyn, 1993; Deutschlander, Greaves, Haimberger, &Hawryshyn, 2001; Parkyn & Hawryshyn, 2000), single unit record-ing (Coughlin & Hawryshyn, 1994) electroretinograms (Allison,Dann, Veldhoen, & Hawryshyn, 2006; Allison et al., 2003) and morerecently, molecular work (Dann, Allison, Levin, Taylor, & Hawry-shyn, 2004; Veldhoen et al., 2006).

UVS cones contain a visual pigment produced by the SWS1 op-sin gene and are normally found in small single cone photorecep-tors. The SWS cones contain a visual pigment produced by theSWS2 opsin gene and are found in slightly larger single cone pho-toreceptors. Double cones are made up of two photoreceptor mem-bers that are fused along the inner segment region forming apartitioning membrane. Double cones possess two outer segments,the MWS cone outer segment containing a visual pigment pro-duced by RH2a/RH2b opsin genes and the LWS cone outer segment

ll rights reserved.

. Hawryshyn).

containing a visual pigment produced by the LWS opsin gene. Rodphotoreceptors, not examined in this study, contain a visual pig-ment produced by the RH1 opsin gene (Allison et al., 2003, 2006;Dann et al., 2004; Hawryshyn et al., 2001; Temple, Veldhoen, Phe-lan, Veldhoen, & Hawryshyn, 2008b).

Whole-cell voltage clamp electrophysiology has emerged as apowerful tool in the characterization of cellular behaviour and,when applied to cells in a neural network that are alive and intact,it can be a useful tool in the elucidation of intercellular signaling/processing. However, few studies have used the whole-cell volt-age clamp recording to assess spectral sensitivity of individualphotoreceptors. Whole-cell voltage clamp recording has been re-cently used to assess the photoreceptor responses in an inverte-brate (cuttlefish, Sepia officinalis, Chrachi, Nelson, & Willisamson,2005) and the action spectra of negative feedback of horizontalcells on three cone classes in the goldfish: short wavelength-sen-sitive (SWS), mid wavelength-sensitive (MWS) and long wave-length-sensitive (LWS) cones (Kamermans, Kraaij, & Spekreijse,2001; Kraaij & Kamermans, 2000; Kraaij, Kamermans, & Spe-kreijse, 1998; Kraaij, Spekreijse, & Kamermans, 2000). Suction pip-ette electrodes have been used to measure membranephotocurrents of single isolated cone photoreceptors in two cypri-nids (goldfish, Carassius auratus, Palacios, Varela, Srivastava, &Goldsmith, 1998; and the giant danio, Danio aequipinnatus, Pala-cios, Goldsmith, & Bernard, 1996) as well as in the monkey Maca-ca fascicularis (Baylor, Nunn, & Schnapf, 1987) but suction pipetterecording, unlike whole-cell voltage clamp recording, is not con-ducive to recording from intact retinas.

2056 L.G. Anderson et al. / Vision Research 50 (2010) 2055–2061

Investigations into the transduction and neural processing ofindividual UVS cones are remarkably few (Palacios et al., 1996).Measurements of goldfish visual sensitivity show that the UVScone mechanism functions at a range of background intensity,which is a considerably dimmer (2–3 log units) than for other conemechanisms (SWS, MWS, LWS) (Hawryshyn, 1991). These proper-ties could be related to either individual UVS cone photoreceptorshaving intrinsically higher sensitivity characteristics or UVS conesignals being processed by different retinal network mechanisms.Here we examine the photocurrent responses of individual UVScones and other cone photoreceptors in an isolated retinal prepara-tion and describe their spectral sensitivity characteristics by fittingvisual pigment templates for the different opsin genes expressed inrainbow trout. This is the first study to successfully use whole-cellvoltage clamp recording.

2. Materials and methods

2.1. Animals

Juvenile rainbow trout (Oncorhynchus mykiss) between 3 and10 g body mass were obtained from the Fraser Valley Trout Hatch-ery, Abbottsford, BC, Canada. Fish were held on a 12L:12D photope-riod and maintained at 15 ± 1 �C, and fed trout pellets once daily.Lighting in the holding facility was provided by full spectrum fluo-rescent bulbs (colour temperature 6500 K). All experimental proce-dures and care were approved by the University of Victoria AnimalCare Committee under the auspices of the Canadian Council forAnimal Care.

2.2. Preparation of the tissue

Rainbow trout were dark adapted for 8 ± 1 min to facilitate theisolation of the retina from the pigment epithelium and retainingmost of the cone population. Fish were anesthetized in a solutionof 250 mg l�1 tricaine methanesulphonate (MS-222) during thelast 2 min of this dark phase and then euthanized by cervical tran-section. Under dim deep red illumination (680 nm interference fil-ter), an eye was enucleated, hemisected and most of the vitreouswas removed with filter paper. The retina was isolated, placed pho-toreceptor side up in a superfusion chamber and superfused con-tinuously (1.5 ml min�1) with oxygenated Ringer’s solution (pH7.7, 17 �C).

The superfusion chamber was mounted on a Nikon Eclipse600FN fixed stage microscope (Nikon, Toronto, Ontario). The prep-aration was illuminated with infrared light (k > 850 nm) andviewed with a Nikon 60� water immersion objective, and a videoCCD camera (Canadian Photonic Laboratory, Minnedosa,Manitoba).

2.3. Optical system

A 150 W xenon lamp and a monochromator (Photon Technol-ogy International, London, Ontario) were used to supply spectrallynarrow band stimuli (10 nm FWHM; full width half maximum).Flash duration was controlled by an electronic shutter (Uniblitz,Vincent Associates, Rochester, NY), and intensity was controlledby a neutral density filter wedge (4.0 neutral density, Inconel onfused silica, Melles Griot, Rochester, NY). Sixty micrometer diame-ter spots were projected onto the isolated retina.

2.4. Electrodes and solutions

Electrodes were pulled from glass capillaries (Schott #8250, 1.5/1.0 (OD/ID mm)) (WPI, Sarasota, Florida) with a Sutter P-97 micro-

pipette puller (Sutter Instruments Company, Novato, CA) and hadimpedances between 6 and 9 MX when filled with pipette mediumand measured in Ringer’s solution. Electrodes were mounted on athree-axis piezio-electric EXFO micromanipulator (Burleigh, Con-cord, Ontario) and connected to an Axon Multiclamp 700B ampli-fier (Molecular Devices, Sunnyvale, CA).

The Ringer’s solution contained 102.0 NaCl, 2.6 KCl, 1.0 CaCl2,1.0 MgCl2, 28.0 NaHCO3, and 5.0 glucose (concentrations expressedin mM) and was continuously bubbled with approximately 2.5%CO2 and 97.5% O2 yielding a pH of 7.8. The pipette medium con-tained 11.7 KCl, 60.0 KGlu, 1.0 MgCl2, 0.1 CaCl2, 5.0 EGTA, 5.0HEPES, 3.9 ATPNa, 0.1 GTPNa, 0.2 cGMPNA, 20.0 Phosphocreatine,and 0.6 Creatine phosphokinase (concentrations in mM). The pHof the pipette medium was adjusted to 7.25 with KOH. All chemi-cals were obtained from Sigma–Aldrich (Oakville, Ontario).

The liquid junction potential was measured with a patch pipettefilled with pipette medium and positioned in a bath containingpipette medium. The potential was adjusted to zero and the bathsolution was replaced with Ringer’s solution. The resulting changein voltage potential was 9 mV and considered to be the junctionpotential. All photocurrent data were corrected accordingly.

2.5. Recording procedure

Individual photoreceptors were visually identified in the iso-lated retina preparation and recordings were obtained by loweringa patch electrode directly onto the inner segment of the cell whilemonitoring current response to 5 mV voltage pulses and thenapplying suction to the cell to produce a greater than 1 GX seal.To measure photocurrent responses, whole-cell recordings weremade under voltage clamped (V = �60 mV) conditions, while theretina was stimulated with 100 ms flashes of various wavelengthsand intensities. The inter-stimulus interval was always longer than5000 ms. The stimulus and recording protocols were derived fromprevious whole-cell electrophysiology studies of spectral sensitiv-ity in other species of fishes (Kraaij et al., 1998; Palacios et al.,1998). Series resistance compensation was employed during themeasurements to reduce voltage error in membrane voltage andto increase temporal resolution of the voltage clamp.

The test wavelengths used to obtain spectral sensitivity mea-surements for a cell were based on the visual identification of celltype as being either a single or double cone member. For each cell,several wavelengths (spaced 10 nm apart) were used that corre-sponded to the expected kmax values for that cell type and then ran-domly between 300 and 680 nm for double cones, and between300 and 520 nm for single cones (spaced 20 nm apart).

2.6. Analysis of kinetics and flash sensitivity

For the analysis of cone response kinetics and flash sensitivity,four to seven representative cones for each cone type were chosen.Individual photocurrent responses to flash stimuli were smoothedusing a 79-point boxcar smoothing function (time window of0.53 ms for data acquisition at 150 kHz). Then, for several wave-lengths near the peak spectral sensitivity for each cone, two to fourindividual photocurrent responses to flash stimuli (stimulus inten-sity of 103–104 photons lm�2 flash�1) were averaged to calculatethe peak amplitude and time-to-peak of photocurrent responsesfor each cone. The peak photocurrent response (p) was determinedby taking the difference between the peak photocurrent (pA –picoamperes) and the baseline photocurrent (pA) between 100and 400 ms after light onset. These results were combined toestablish averages (±1 standard deviation) for each cone type. Flashsensitivity to 100 ms flash stimuli at selected wavelengths was de-fined as the peak photocurrent response (pA)/intensity (pho-tons lm�2 flash�1) or pA photon�1 lm2 (Palacios & Goldsmith,

L.G. Anderson et al. / Vision Research 50 (2010) 2055–2061 2057

1993). Comparison between the mean time-to-peak and flash sen-sitivity of the various cone classes was made using a one-way anal-ysis of variance (ANOVA) with a = 0.05.

2.7. Analysis of spectral sensitivity data

For each cell recorded, we measured the photocurrent responseto several stimulus intensities at several test wavelengths. For eachtest wavelength, the photocurrent responses were plotted againststimulus intensity. It is important to note that the stimulus inten-sity was not sufficient to obtain photocurrent responses at satura-tion, for all wavelengths. Therefore we could not analyze theresponse versus intensity curves using the Naka Ruston functionrather the RI curves were fitted with a third order polynomial usingthe least squares method. A criterion value of 15 pA was used tointerpolate threshold intensity. Sensitivity (spectral) was definedas the reciprocal of threshold intensity for a given test wavelength.The spectral sensitivity data of each cell were normalized to unityto minimize sensitivity variation between cells. The quality of theseal between the cell membrane and the electrode was a signifi-cant factor in producing such a variation. Mean normalized sensi-tivity was plotted against test wavelength to give spectralsensitivity.

2.8. Visual pigment templates

Salmonid fishes have a paired pigment retina with mixtures ofA1 and A2 chromophores in both rod and cone outer segments(Temple et al., 2008a, 2008b), with an A1:A2 ratio varying season-

Fig. 1. Photocurrent response properties of a LWS cone to five wavelengths, each atseven stimulus intensities (equal quanta stimuli). For this figure, photon irradianceof 107 photons lm�2 s�1 corresponds to an intensity of 0 log. The stimuli were100 ms in duration and the vertical lines indicate the stimulus onset point. Scaling,timing, and stimulus wavelength are indicated on the figure.

Table 1Time-to-peak and flash sensitivity of cone classes.

Cone class

LWS (n = 4) MWS (n = 7

Time-to-peak (ms) 326.98 ± 112.2 322.43 ± 74Flash sensitivity* 0.00461 ± 0.00253 0.00473 ± 0

Stimulus wavelengths (nm) – UVS k = 340, 360, 380; SWS k = 400, 440, 46* Flash sensitivity for a 100 ms flash = pA photons-1 lm2; range represent

ally (Temple et al., 2006). The visual pigment templates that werefitted to spectral sensitivity curves were generated by combiningabsorption spectra for the A1 and A2 chromophores derived fromGovardovskii, Fyhrquist, Reuter, Kuzmin, and Donner (2000). TheA1 and A2 visual pigment equations (A1 – Eqs. (1) and (2) in Gov-ardovskii et al., 2000 and A2 – Eqs. (1) and (6) in Govardovskii etal., 2000) were added together with a variable fraction parameteri.e. (1 � a) � A1 + a � A2. Visual pigment templates were fitted tomean normalized sensitivity curves using a least squares fit. Theleast squares fit was performed while leaving the kmax and the afraction parameter free, allowing the software to find a visual pig-ment template that best describes the mean normalized sensitivityfor each cone type.

3. Results

3.1. Cone photocurrent responses

Individual photoreceptors, when held at a membrane potentialof �60 mV under voltage clamp conditions, responded to a seriesof light flashes of increasing intensity with photocurrents ofincreasing amplitude and duration. Representative responses of aLWS cone to increasing equal quanta intensities of light stimuliat five wavelengths are shown in Fig. 1. The average time-to-peakwas 339 ± 90 ms and flash sensitivity for a 100 ms flash was4.37 � 10�3 ± 2.50 � 10�3 pA photons�1 lm2, with no significantdifferences between UVS, SWS, MWS and LWS cone classes (kinet-ics: ANOVA, F(4, 19) = 0.590, P = 0.674; flash sensitivity: ANOVA,F(4, 19) = 0.156, P = 0.958, Table 1). Fig. 2 illustrates the responsedynamics of an LWS cone for five test wavelengths. The photocur-rent response versus intensity points for each of the five test wave-lengths were fitted with third order polynomial functions (blacklines) (400 nm (filled squares; R2 = 0.94), 460 nm (open circles;R2 = 0.99), 500 nm (filled diamonds; R2 = 0.99), 560 nm (filled tri-angles; R2 = 0.99), 620 nm (open squares; R2 = 0.97).

3.2. Cone spectral sensitivity

Fig. 3 shows that the peak sensitivity (kmax) values of individualcones recorded in this study fell into one of four clusters. A total of107 cones were examined, 20 in the LWS cone cluster, 61 in theMWS cone cluster, 13 in the SWS cone cluster, and 13 in the UVScone cluster (see Fig. 3 legend for the spectral bounds defining acluster).

Fig. 4 shows the mean normalized sensitivity and visual tem-plate curve for four-cone classes in rainbow trout. Visual pigmenttemplates were fitted to the mean normalized sensitivity curvesusing a least squares fit with the kmax and proportion of A2 as freeparameters. Table 2 lists the results of this analysis. The visual pig-ment templates fitted to the cone classes resulted in the followingfits: (i) UVS cones showed a kmax of 373 nm, an A2 proportion of 1.0,and a R2 value of 0.894; (ii) SWS cones showed a kmax of 423 nm, anA2 proportion of 1.0, and a R2 value of 0.756; (iii) MWS conesshowed two sensitivity peaks corresponding to the RH2a andRH2b opsin pigments. The trout used in this study showed differen-

) SWS (n = 4) UVS (n = 5)

.12 374.63 ± 114.79 390.52 ± 91.93

.00289 0.00460 ± 0.00199 0.00377 ± 0.00357

0; MWS k = 500, 520, 540; LWS k = 560, 570, 580.s average ± 1 SD.

Fig. 2. Photocurrent response versus intensity curves of a LWS cone for fivewavelengths. Photocurrent response versus intensity curves were fitted with thirdorder polynomials (solid lines). Stimulus wavelengths with R2 values for polynomialfit: 400 nm (filled squares; R2 = 0.94), 460 nm (filled circles; R2 = 0.99), 500 nm(filled diamonds; R2 = 0.99), 560 nm (filled triangles; R2 = 0.99), 620 nm (filledinverted triangle; R2 = 0.97).

Fig. 3. Spectral distribution of cone peak sensitivity values for the 107 conesrecorded in this study. Spectral bounds for the cone classes: UVS cones 360–385 nm, SWS cones 410–440 nm, MWS cones 500–550 nm, and LWS cones 560–585 nm.

2058 L.G. Anderson et al. / Vision Research 50 (2010) 2055–2061

tial expression of the RH2a and RH2b opsin genes (Temple et al.,2008a, 2008b), which resulted in shifts in spectral sensitivity ofthe MWS cones. The mean normalized spectral sensitivity curvefor the MWS cones represents a composite of both the RH2a andRH2b visual pigments (kmax RH2b < kmax RH2a) and it was assumedthat the concentration of the two pigments in cones with co-expression was approximately the same. Thus, the breadth of thespectral sensitivity curve was larger than could be explained byone visual pigment template, rather both visual pigment templatespectra were required to provide an adequate fit. The overall spec-tral sensitivity of the MWS cones represents an envelope of theRH2a and RH2b based visual pigments. For the short wavelengthside of the spectral sensitivity curve, which was mediated by theRH2b based visual pigment, MWS cones showed a kmax of500 nm, an A2 proportion of 1.0, and a R2 value of 0.835. For thelong wavelength side of the spectral sensitivity curve, which wasmediated by the RH2a visual pigment, MWS cones showed a kmax

of 542 nm, an A2 proportion of 0.98, and a R2 value of 0.97; (iv)LWS cones showed a kmax of 585 nm, an A2 proportion of 0.33,and a R2 value of 0.857. It is likely that variability in the LWS conesensitivity data resulted in an A2 proportion of 0.33, which differsfrom the A2 proportion of 1.0 for the UVS, SWS and MWS cones.Table S1 provides the mean normalized data, standard error andsample size for all cone classes.

Table 3 shows that there were small differences in FWHM band-width for the UVS, SWS and LWS cones. The FWHM values in com-parison to those obtained from other studies using salmonids(Table 3) revealed the presence of two opsins in the MWS conessampled. The FWHM value for the MWS cone spectral sensitivitywas 7149 cm�1 in comparison to microspectrophotometry studieswhere the absorbance spectra showed a FWHM of 3600–4440 cm�1 and it was argued that one opsin was present in MWScones (Hawryshyn & Hárosi, 1994; Hawryshyn et al., 2001). Recentevidence has shown that salmonid MWS cones can coexpress RH2aand RH2b and express RH2a or RH2b alone (Temple et al., 2008a,2008b). Fig. 3 shows that there were MWS cones that had peaksensitivity kmax values indicating that MWS cones were coexpress-ing RH2A and RH2b, and others that were expressing either RH2a orRH2b alone.

The kmax values for the four-cone classes appear to be in goodagreement with those reported in previous studies of this species(Table 4). Two of the previous studies (Hawryshyn, Arnold, Chais-son, & Martin, 1989; Parkyn & Hawryshyn, 2000) showed kmax val-ues consistent with a RH2a opsin pigment and one study (Beaudetet al., 1993) showed a kmax value consistent with a RH2b opsin pig-ment, while the majority of studies had kmax values reflecting a co-expression of RH2a and RH2b (Douglas, 1983; Hawryshyn & Hárosi,1994; Hawryshyn et al., 2001; Kusmic, Barsanti, Passarelli, & Gual-tieri, 1993).

4. Discussion

We examined the properties of 107 cones that fell into four spec-tral clusters in terms of peak sensitivity. Analysis of the mean nor-malized spectral sensitivity through visual pigment template fittingrevealed four-cone classes: UVS cones (kmax = 373 nm), SWS cones(kmax = 423 nm), MWS cones (kmax = 500/540 nm), LWS cones(kmax = 585 nm). The average time-to-peak was 339 ± 90 ms andflash sensitivity for a 100 ms flash was 4.37 � 10�3 ± 2.50 �10�3 pA photons�1 lm2, with no significant differences betweenUVS, SWS, MWS and LWS cone classes. These cone flash sensitivityvalues are comparable to those seen in Palacios et al. (1996) andKraft (1988). The spectral sensitivity of the MWS cones showed aspectral breadth larger than would be expected for a single visualpigment indicating the expression of both RH2a and RH2b opsinpigments.

4.1. Comparison with earlier work on different spectral cone classes

Whole-cell voltage clamp recordings from intact cone photore-ceptors in an isolated retina preparation have permitted measure-ment of cone spectral sensitivity in rainbow trout. The flashsensitivity of the UVS cones was comparable to that of the SWS,MWS and LWS cones. Thus the large difference in the backgroundintensity range over which UVS cones function in comparison tothe other cones seen in goldfish (Hawryshyn, 1991) is very likelynot due to differences in UVS cone flash sensitivity but rather mustbe related to differences in network processing of the UVS cone sig-nal. This could be related to differences in the post-receptor adap-tation mechanism for UVS cones that occurs as signals are relayedfrom cone bipolar cells to ganglion cells (Dunn, Lankheet, & Reike,2007).

The cone visual pigments of O. mykiss have been previouslymeasured by MSP and behavioural methods (Tables 3 and 4). Theresults of our study are generally in good agreement with previouswork, for all four-cone classes. The variance in kmax values amongstthese studies reflect differences in measurement techniques, modeof analysis and variability attributable to visual pigment composi-tion through A1/A2 chromophore shifting (Temple et al., 2006) and

Fig. 4. The mean normalized spectral sensitivity of rainbow trout cones: (A) UVS, (B) SWS, (C) MWS, and (D) LWS cones. Filled circles represent the mean normalizedsensitivity ± one standard error. Solid lines represent the least squares fit of visual pigment templates to the mean normalized sensitivity points. kmax values for visualpigment template fits and R2 values: UVS cones = 373 nm (R2 = 0.894), SWS cones = 423 nm (R2 = 0.756), MWS cones = 500 nm (RH2b)(R2 = 0.835) and = 542 nm (RH2a)(R2 = 0.97), LWS cones = 585 nm (R2 = 0.857). See Table 2 for additional information on visual pigment template fitting.

Table 2Visual pigment template fits for rainbow trout cone spectral sensitivity.

Cone type

UVS SWS RH2a RH2b LWS

kmax (nm) 373 423 500 542 585A2 proportion 1.00 1.00 1.00 0.98 0.33Fit range (nm) 340–460 340–520 320–540 560–660 380–660R2 0.894 0.756 0.835 0.970 0.857

L.G. Anderson et al. / Vision Research 50 (2010) 2055–2061 2059

the presence of two RH2 opsin pigments in the MWS cones (Tem-ple et al., 2008a, 2008b).

4.2. Multiple cone opsin subtypes

Until recently, classes of cone photoreceptors within a specieswere commonly assumed to contain a single opsin type. Recent re-search has demonstrated that more than one opsin subtype may beexpressed not only within a class of photoreceptors but withinindividual photoreceptors (Cheng, Novales Flamarique, Hárosi,Rickers-Haunerland, & Haunerland, 2006; Temple et al., 2008a,2008b). Previous research has proposed the presence of multipleopsins within a cone class on the basis of sensitivity spectra that

are broader than a single visual pigment template (Govardovskiiet al., 2000; Makino & Dodd, 1996; Palacios et al., 1998). InOncorhynchus spp. it has been shown that the single UVS andSWS cones express the SWS1 and SWS2 opsin classes while thetwo outer segment members of double cones express RH2a orRH2b or both and LWS opsin classes (Allison et al., 2003; Templeet al., 2008a, 2008b). A recent MSP investigation in coho salmon(Oncorhynchus kisutch) identified a range of kmax values in MWScones that was greater than could be explained by a single opsin(RH2a), taking into account chromophore shifting. This led to thediscovery and identification of a second RH2 opsin gene referredto as RH2b (Temple et al., 2008a). The variation in the peak sensi-tivity of MWS cones evident in Fig. 3 argues that some of thesecones were expressing either RH2a or RH2b pigments, and otherswere coexpressing both pigments. The expression pattern in sal-monids is related to the ontogenetic stage, where RH2b is ex-pressed in parr and the RH2a is expressed in the ocean phasesmolts (Temple et al., 2008b). The fish used in our study bridgedthe ontogenetic stage where the MWS cones shift their expressionfrom RH2b to RH2a. Thus the MWS cone mean spectral sensitivityshown in Fig. 4C appears much broader (see FWHM values in Table3) than the UVS, SWS and LWS spectral sensitivity (Fig. 4a, b, and drespectively) indicating the presence of both RH2b (kmax range

Table 3Comparison of full width half maximum (FWHM; average ± 1 SD cm�1) values of rainbow trout spectral sensitivity curves for all cone classes.

Cone class

UVS SWS MWS LWS Species Reference

1 5303 6121 7149 4054 O. mykiss Present work2 5051 ± 130 4329 ± 112 3604 ± 62 3256 ± 79 (3709 ± 89)* O. mykiss Hawryshyn et al. (2001)3 4700 ± 1200 6600 ± 1500 4400 ± 500 3700 ± 200 O. mykiss Hawryshyn and Hárosi (1994)4 5702 ± 715 5092 ± 538 4165 ± 382 3645 ± 326 O. kisutch Cheng et al. (2006)5 5845 ± 783 5119 ± 248 4101 ± 254 4036 ± 203 O. keta Cheng et al. (2006)6 5915 ± 684 5187 ± 304 4188 ± 265 3983 ± 252 O. tschawytscha Cheng et al. (2006)7 6105 ± 831 4982 ± 444 4201 ± 305 3660 ± 298 O. gorbuscha Cheng et al. (2006)8 5902 ± 850 5406 ± 681 4351 ± 306 4149 ± 355 Salmo salar Cheng et al. (2006)

* Values for A1 and A2 (in parentheses) based pigments.

Table 4Comparison of kmax (nm) values obtained from electrophysiological (present study),microspectrophotometric and behavioural observations of rainbow trout for all coneclasses.

Cone class

UVS SWS MWS LWS Method Reference

1 373 423 500(RH2b)/542(RH2a)

585 WC Present work

2 370–380

440 540 580 GCR Parkyn andHawryshyn(2000)

3 390 420 510 n/a GCR Beaudet et al.(1993)

4 371 432 519 574(623)* MSP Hawryshyn et al.(2001)

5 400 453 530 598 MSP Kusmic et al.(1993)

6 365 434 531 576 MSP Hawryshyn andHárosi (1994)

7 n/a 440 535 630 Behaviour Douglas (1983)8 360 430 540 620 Behaviour Hawryshyn et al.

(1989)

WC, whole-cell voltage clamp; GCR, ganglion cell responses; MSPmicrospectrophotometry.* Values for A1 and A2 (in parentheses) based pigments.

2060 L.G. Anderson et al. / Vision Research 50 (2010) 2055–2061

495–523 nm) and RH2a (kmax range 512–548 nm) opsin pigments.The spectral sensitivity can only be described using two visual pig-ment templates representing both RH2b (kmax = 500 nm) and RH2a(kmax = 542 nm). Zebrafish also show the expression of multipleRH2 opsin genes except they show up to four RH2 subtype opsingenes (Takechi & Kawamura, 2005).

The capacity to shift A1/A2 ratios and change opsin gene expres-sion provides salmonids with highly dynamic spectral tuningmechanisms. The two mechanisms together may allow for a shiftof approximately 60 nm in the MWS cones and nearly 90 nm inthe LWS cones (Temple et al., 2008b). Precise spectral tuning tovariable spectral environments of salmonids (Flamarique & Hawry-shyn, 1993; Novales Flamarique, Hendry, & Hawryshyn, 1992) al-lows for maintenance of an optimum signal-to-noise ratio indynamic target detection scenarios.

Acknowledgments

We would like to thank the Fraser Valley Trout Hatchery,Abbottsford, British Columbia for the donation of the fish used inthis study. This work was supported by US Air Force Office of Sci-entific Research operating Grant #F9550-05-1-0070 and the Can-ada Research Chairs program (CWH). Shai Sabbah was supportedby an NSERC Vanier Canada Graduate Scholarship.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.visres.2010.07.013.

References

Allison, W. T., Dann, S. G., Helvik, J. V., Bradley, C., Moyer, H. D., & Hawryshyn, C. W.(2003). Ontogeny of ultraviolet-sensitive cones in the retina of rainbow trout.Journal of Comparative Neurology, 461, 294–306.

Allison, W. T., Dann, S. G., Veldhoen, K. M., & Hawryshyn, C. W. (2006). Degenerationand regeneration of ultraviolet cone photoreceptors during development inrainbow trout. Journal of Comparative Neurology, 499, 702–715.

Baylor, D. A., Nunn, B. J., & Schnapf, J. L. (1987). Spectral sensitivity of cones of themonkey Macaca fascicularis. Journal of Physiology, 390, 145–160.

Beaudet, L., Browman, H. I., & Hawryshyn, C. W. (1993). Optic nerve response andretinal structure in rainbow trout of different sizes. Vision Research, 33,1739–1746.

Cheng, A., Novales Flamarique, I., Hárosi, F. I., Rickers-Haunerland, J., & Haunerland,N. H. (2006). Photoreceptor layer loss of salmonid fishes: Transformation andloss of single cones in juvenile fish. Journal of Comparative Neurology, 495,213–235.

Chrachi, A., Nelson, L., & Willisamson, R. (2005). Whole-cell recording of light-evoked photoreceptor responses in a slice preparation of the cuttlefish retina.Visual Neuroscience, 22, 359–370.

Coughlin, D. J., & Hawryshyn, C. W. (1994). Ultraviolet sensitivity in the torussemicircularis of juvenile rainbow trout (Oncorhynchus mykiss). Vision Research,34, 1407–1413.

Dann, S. G., Allison, W. T., Levin, D. B., Taylor, J. S., & Hawryshyn, C. W. (2004).Salmonid opsin gene sequences undergo positive selection and indicate analternate evolutionary relationship in Onchorhynchus. Journal of MolecularEvolution, 58, 401–412.

Deutschlander, M. E., Greaves, D. K., Haimberger, T. J., & Hawryshyn, C. W. (2001).Functional mapping of ultraviolet photosensitivity during metamorphictransitions in a salmonid fish, Oncorhynchus mykiss. Journal of ExperimentalBiology, 204, 2401–2413.

Douglas, R. H. (1983). Spectral sensitivity in rainbow trout (Salmo gairdneri). RevueCanadienne de Biologie Experimentale, 42, 117–122.

Dunn, F., Lankheet, M. J., & Reike, F. (2007). Light adaptation in cone vision involvesswitching between receptor and post-receptor sites. Nature, 449, 603–607.

Flamarique, I., & Hawryshyn, C. W. (1993). Spectral characteristics of salmonidmigratory routes from Southern Vancouver-Island (British-Columbia). CanadianJournal of Fisheries and Aquatic Sciences, 50, 1706–1716.

Govardovskii, V. I., Fyhrquist, N., Reuter, T., Kuzmin, D. G., & Donner, K. (2000). Insearch of the visual pigment template. Visual Neuroscience, 17, 509–528.

Hawryshyn, C. W. (1991). Light-adaptation properties of the ultraviolet-sensitivecone mechanism in comparison to the other receptor mechanisms in goldfish.Visual Neuroscience, 6, 293–301.

Hawryshyn, C. W. (1992). Polarization vision of fishes. American Scientist, 80,164–175.

Hawryshyn, C. W. (2000). Ultraviolet polarization vision in fishes: Possiblemechanisms for coding e-vector. Philosophical Transactions of the Royal Societyof London Series B-Biological Sciences, 335, 1187–1190.

Hawryshyn, C. W., Arnold, M. G., Bowering, E., & Cole, R. L. (1990). Spatialorientation of rainbow-trout to plane-polarized light – The ontogeny of e-vectordiscrimination and spectral sensitivity characteristics. Journal of ComparativePhysiology A – Sensory Neural and Behavioral Physiology, 166, 565–574.

Hawryshyn, C. W., Arnold, M. G., Chaisson, D. J., & Martin, P. C. (1989). The ontogenyof ultraviolet photosensitivity in rainbow trout (Salmo gairdneri). VisualNeuroscience, 2, 247–254.

Hawryshyn, C. W., Haimberger, T. J., & Deutschlander, M. E. (2001).Microspectrophotometric measurements of vertebrate photoreceptors usingCCD-based detection technology. Journal of Experimental Biology, 204,2431–2438.

Hawryshyn, C. W., & Hárosi, F. I. (1994). Spectral characteristics of visualpigments in rainbow trout (Oncorhynchus mykiss). Vision Research, 34,1385–1392.

Kamermans, M., Kraaij, D., & Spekreijse, H. (2001). The dynamic characteristics ofthe feedback signal from horizontal cells to cones in the goldfish retina. Journalof Physiology, 534, 489–500.

Kraaij, D. A., & Kamermans, M. (2000). The nature of surround-induceddepolarizing responses in goldfish cones. Journal of General Physiology, 115,3–15.

L.G. Anderson et al. / Vision Research 50 (2010) 2055–2061 2061

Kraaij, D. A., Kamermans, M., & Spekreijse, H. (1998). Spectral sensitivity of thefeedback signal from horizontal cells to cones in goldfish retina. VisualNeuroscience, 15, 799–808.

Kraaij, D. A., Spekreijse, H., & Kamermans, M. (2000). The open- and closed-loopgain-characteristics of the cone/horizontal cell synapse in goldfish retina.Journal of Neurophysiology, 84, 1256–1265.

Kraft, T. W. (1988). Photocurrents of cone photoreceptors of the golden-mantledground squirrel. Journal of Physiology, 404, 199–213.

Kusmic, C. L., Barsanti, L., Passarelli, V., & Gualtieri, P. (1993). Photoreceptormorphology and visual pigment content in the pineal organ and in the retina ofjuvenile and adult trout, Salmo irideus. Micron, 24, 279–286.

Makino, C. L., & Dodd, R. L. (1996). Multiple visual pigments in a photoreceptor ofthe salamander retina. Journal of General Physiology, 108, 27–34.

Novales Flamarique, I., Hendry, A., & Hawryshyn, C. W. (1992). The photicenvironment of a salmonid nursery lake. Journal of Experimental Biology, 169,121–141.

Palacios, A. G., & Goldsmith, T. H. (1993). Photocurrents in retinal rods of pigeons(Columba-livia) – Kinetics and spectral sensitivity. Journal of Physiology, 471,817–829.

Palacios, A. G., Goldsmith, T. H., & Bernard, G. D. (1996). Sensitivity of cones from acyprinid fish (Danio aequipinnatus) to ultraviolet and visible light. VisualNeuroscience, 13, 411–421.

Palacios, A. G., Varela, F. J., Srivastava, R., & Goldsmith, T. H. (1998). Spectralsensitivity of cones in the goldfish, Carassius auratus. Vision Research, 38,2135–2146.

Parkyn, D. C., & Hawryshyn, C. W. (2000). Spectral and polarization sensitivity ofsalmonids: A comparative analysis using electrophysiology. Journal ofExperimental Biology, 203, 1173–1191.

Takechi, M., & Kawamura, S. (2005). Temporal and spatial changes in the expressionpattern of multiple red and green subtype opsin genes during zebrafishdevelopment. Journal of Experimental Biology, 208, 1337–1345.

Temple, S. E., Plate, E. M., Ramsden, S., Haimberger, T. J., Roth, W.-M., & Hawryshyn,C. W. (2006). Seasonal cycle in vitamin A1/A2-based visual pigmentcomposition during the life history of coho salmon (Oncorhynchus kisutch).Journal of Comparative Physiology A, 192, 301–313.

Temple, S. E., Ramsden, S. D., Haimberger, T. J., Veldhoen, K. M., Veldhoen, N. J.,Carter, N. L., et al. (2008a). Effects of exogenous thyroid hormones on visualpigment composition in coho salmon (Oncorhynchus kisutch). Journal ofExperimental Biology, 211, 2134–2143.

Temple, S. E., Veldhoen, K. M., Phelan, J. T., Veldhoen, N. J., & Hawryshyn, C. W.(2008b). Ontogenetic changes in photoreceptor opsin gene expression in cohosalmon (Oncorhynchus kisutch, Walbaum). Journal of Experimental Biology, 211,3879–3888.

Veldhoen, K., Allison, W. T., Veldhoen, N., Anholt, B. R., Helbing, C. C., & Hawryshyn,C. W. (2006). Spatio-temporal characterization of retinal opsin gene expressionduring thyroid hormone-induced and natural development of rainbow trout.Visual Neuroscience, 23, 169–179.