restoration of visual function following optic nerve regeneration …amensing/publications/callahan...

Restoration of visual function following optic nerveregeneration in bluegill ~Lepomis macrochirus!� pumpkinseed ~Lepomis gibbosus! hybrid sunfish

MICHAEL P. CALLAHAN1,2 and ALLEN F. MENSINGER1

1Department of Biology, University of Minnesota, Duluth, Duluth, Minnesota2Biology Department, University of Southern Maine, Portland, Maine

(Received October 3, 2006; Accepted March 16, 2007!

Abstract

Simple ~dorsal light reflex! and complex ~predator-prey interactions! visually mediated behaviors were usedconcurrently with morphological examination to assess restoration of visual function following optic nerve crush inbluegill ~Lepomis macrochirus! � pumpkinseed ~Lepomis gibbosus! hybrid sunfish. Regenerating optic nerve axonsprojected into the stratum opticum-stratum fibrosum et griseum superficiale by week 2, the stratum griseum centraleby week 4, and stratum album centrale by week 6. Initial projections into the laminae were diffuse and lessstratified compared to controls. By week 12, the projection pattern of regenerating nerve fibers closely resembledthe innervation of normal tecta. Visual improvements were correlated with increasing projections into the tectum.The dorsal light reflex improved from a 458 vertical deviation following nerve crush to 4.58 by week 16. Initialpredator-prey interactions were exclusively mediated by the control eye. As regeneration progressed, there wasa gradual expansion of the visual field. The reaction distance and attack angles within the visual field ofthe experimental eye were initially less than controls, however, these differences disappeared by week 10.Improvements in visual function were closely correlated with an increase of regenerating ganglion cell axonsinto the optic tectum indicating sufficient synaptogenesis to mediate both simple and complex visual behavior.

Keywords: Optic nerve regeneration, Dorsal light reflex, Predator-prey interactions, Fish vision, Visually mediatedbehavior

Introduction

Teleost fish historically have been used as models for nerveregeneration because they possess the ability to regenerate nervesthroughout life ~Sperry, 1948!. A number of studies have addressedthe ability of the optic nerve to regenerate to anatomical andfunctional completion ~Schwassmann & Kruger, 1965; Meyer,1980; Northmore, 1989!. Within two weeks of optic nerve tran-section, retinal axons re-invade and branch across the optic tectum~Stuermer & Easter, 1984; Rankin & Cook, 1986; Hayes & Meyer,1988; Schmidt et al., 1988!. Synaptogenesis has been documentedwithin three weeks after optic nerve crush in the goldfish, withsubsequent regeneration proceeding across the tectum in an ante-rior to posterior gradient ~Schmidt & Edwards, 1983; Hayes &Meyer, 1989; Meyer & Kageyama, 1999!. There is often anoverproduction of optic nerve axons early in regeneration ~Schmidtet al., 1988!, and although the retinotopic map is roughly restoredby week 5, refinement of the axons and synapses continues for

several more weeks ~Meyer, 1980; Meyer & Kageyama, 1999;Meyer et al., 1985!.

Sunfish are primarily visual predators and feed throughout theday on a wide variety of prey ~Mittelbach, 1984; Janssen &Corcoran, 1993!. Sunfish visual projections have been mapped~Striedter & Northcutt, 1989; Northcutt & Butler, 1991!, and thisfamily of fishes has served as a model for physiological andbehavioral assessment of optic nerve regeneration ~Schwassmann& Kruger, 1965; Northmore, 1981, 1991!. However, previousstudies used simple visual tests such as light detection ~Northmore,1989; Northmore & Masino, 1984! or contrast sensitivity ~North-more & Celenza, 1992! to document visual recovery. Therefore,the extent that regenerated fibers can mediate complex visualbehaviors, such as prey tracking, remains largely unknown.

The ability of fish to detect, track, and attack prey provides amore complicated challenge than contrast sensitivity and a morenaturalistic assessment of visual function. Predator-prey inter-actions have been used to test a number of sensory modalities~Howick & O’Brien, 1983; McMahon & Holanov, 1995; Rich-mond et al., 2004!. Zooplankton have been used predominantly asprey ~Werner & Hall, 1974; Batty et al., 1990; Walton et al., 1992!in fish trials because these small targets allow assessment of visual

Address correspondence and reprint requests to: Michael P. Callahan,Biology Department, University of Southern Maine, 96 Falmouth Street,Portland, ME. E-mail: [email protected]

Visual Neuroscience ~2007!, 24, 309–317. Printed in the USA.Copyright © 2007 Cambridge University Press 0952-5238007 $25.00DOI: 10.10170S0952523807070289

309

acuity. However, plankton possess relatively modest escape abili-ties, whereas active prey such as minnows provide a more chal-lenging target that allows examination of visual integration withmotor behavior.

The purpose of this study was to measure the rate and accuracyof optic nerve regeneration in hybrid sunfish using anatomicaland behavioral methods. Histological examination of optic nerveprojections was used concurrently with the dorsal light reflex~DLR! and predator-prey interactions to determine the time courseand extent of regeneration and the restoration of normal visualfunction.

Materials and methods

Animal collection and care

Bluegill ~Lepomis macrochirus!� pumpkinseed ~Lepomis gibbo-sus! hybrid sunfish were collected by seine ~0.5-cm mesh! fromJune through September 2002 at the Rice Lake Dam, Rice Lake,Minnesota. The hybrid sunfish were chosen because they were themost abundant sunfish ~�70%! in the collection area. Fish weretransported to the University of Minnesota, Duluth ~UMD! in acooler with lake water and a portable aerator. At UMD, all sunfishwere maintained in a 760 L recirculating aquarium system. Thissystem consisted of multiple tanks that were equipped with me-chanical, ultraviolet, and biological filtration. The water was sup-plied with supplemental aeration and maintained at 228C. Sunfishwere maintained on a 12 h light0dark cycle and fed frozen brineshrimp daily except for two days prior to predator-prey experi-ments. All fish were cared for in accordance with University ofMinnesota and the American Physiological Society animal careguidelines.

Optic nerve crush

Sunfish ~n � 24; 7 to 9 cm standard length! were anesthetized byimmersion in a buffered solution of 0.02% MS-222 ~pH 7.3! untila stage III level of surgical anesthesia was attained ~Jolly et al.,1972!. The fish were then placed between moist sponges in a40 cm �11 cm aquarium. A small tube was inserted into the buccalcavity to provide continual circulation of the anesthesia over thegills and maintain a deep level of anesthesia.

The connective tissue around the dorsum of the right eye wascut with microscissors, with care taken to preserve the integrity ofthe ocular muscles. The dorsal part of the eye was depressed toexpose the optic nerve and the nerve was crushed between a pairof fine forceps for 5 to 10 seconds within 2 mm of its projectionfrom the eyecup. The procedure left the optic nerve sheath intact.The efficacy of the crush was determined by visual inspection ofthe sheath to insure there was complete disruption of the fibers. Ifany nerve fibers were observed transversing the crush site, theprocedure was repeated. The eye and connective tissue werereturned to their normal positions.

The sunfish were revived by irrigating the gills with freshaquarium water. After recovery from the anesthetic, the fish wasmoved into a 38 L recovery tank and treated with antibiotics~Furan 2: 60 mg Nitrofurazone, 25 mg Furazolidone, and 2 mgMethylene Blue Trihydrate per capsule; one capsule per 38 L for48 h with an 80% water change and reapplication of the antibioticafter 24 h!. After two days, the fish was returned to its individualcompartment within the recirculating aquarium system.

Dorsal light reflex

Dorsal light reflex experiments were performed weekly on exper-imental fish and unoperated control fish ~N � 4!. Fish were placedin a light tight container and transferred to a dark room for aminimum of 45 min to allow dark adaptation. They then weretransferred to a 7 L test aquarium ~26 cm � 17 cm � 21 cm! in thedark and allowed to acclimate for 15 min. Following acclimation,the aquarium was illuminated with a 100 W, 12 V incandescentbulb that provided broadband light at an intensity of approximately50 lux at the water surface. The fish were videotaped using a Sonydigital video camera for up to two minutes and then returned totheir home aquaria.

The videotape was analyzed using DVShelf frame grabbing andScion Image ~Scion Corporation! software. Single frames werecaptured and analyzed at 15-s intervals following light onset. Allexperimental fish tilted their control eye away from the downwell-ing light. The angle that the fish tilted its ventral to dorsal bodyaxis from vertical was calculated in each frame. The angle for thefirst four time-points was averaged to determine the DLR.

Predator-prey interaction experiments

A 50-cm diameter aquarium was used for predator-prey inter-actions. The arena was filled with 30 l of water to a depth of 10 cm.The arena was surrounded by black cloth to shield the fish fromobservers and was illuminated to 8 lux at the water surface via a100 W incandescent light bulb. The light intensity is within therange that crepuscular predators such as the sunfish would encoun-ter during sunrise and sunset when foraging reaches its maximum~Richmond et al., 2004!. The interactions were filmed from abovewith the video camera. Either normal ~unoperated control! orexperimental ~optic nerve crush! fish were transferred to the arenaand allowed to acclimate for 15 min. A single fathead minnow~Pimephales promelas; 2.4 to 2.7 cm standard length! was intro-duced into the tank below the water surface via a 2.5 cm diametertube. If the minnow was consumed, another one was introduceduntil the fish either refused to approach the prey for 10 min or the30-min trial was completed. Fish also were filmed in the dark~0 lux! using infrared light.

Reaction distance and strike angle were measured from themidpoint between the eyes of the sunfish to the midpoint of thebody of the prey at the moment that the sunfish detected the prey.Prey detection was defined as the point the sunfish initiatedorientation towards the prey prior to an attack at the prey. Strikesuccess was any strike that resulted in the sunfish contacting theprey with its mouth, however not all successful strikes resulted incapture of the minnow.

Nerve labeling

Optic nerve regeneration was analyzed by bulk labeling the nervewith biocytin ~e-biotinoyl-L-lysine!. The intact optic nerves offour sunfish were labeled and served as controls. Three additionalsunfish were used to determine that the efficacy of the nerve crushand the optic nerve was labeled 48 h after surgery. Four experi-mental fish ~total n � 24! were sampled every two weeks forhistological examination.

For bulk nerve labeling, fish were anesthetized, a moistenedabsorbent cloth placed over the body and the gills continuallyperfused with anesthetic solution. The cornea and lens were re-moved and the vitreous humor was wicked away to expose the

310 M.P. Callahan and A.P. Mensinger

optic nerve head. The nerve fibers surrounding the optic nervewere transected with glass microneedles and crystals of biocytinwere applied to the transected fibers for 15 min. A thin sheet ofcyanoacrylate gel ~Krazy Glue! was used to seal the eye. The sheetwas immediately solidified by the addition of a catalyst ~ZipKicker!. The fish were revived and placed in a recovery tank for48 h. The fish were then deeply anesthetized ~0.050% MS-222! tosurgical level III ~Jolly et al., 1972!, and both eyes and the dorsalcranium were removed. The tissue was preserved by immersion in3% paraformaldehyde-1% glutaraldehyde in phosphate bufferedsaline ~PBS!. The brains were removed after a minimum of 72 h inthe fixative, then cryoprotected in 30% sucrose in 0.1 M PBSovernight. The brain was then transferred to 0.1 M PBS.

The optic tectum was cut into 50-µm coronal sections with acryostat ~Microm 505HM!. The sections were rinsed with 0.1 M

PO4 buffer and then washed with 0.4% triton-X 100 in 0.1 M PBSfor 30 min. The sections were then incubated in ABC solution~Vectastain! for three hours followed by rinses in 0.1 M PO4

buffer. The sections were reacted in a 50% solution of diamino-benzidine ~DAB! in deionized water and 0.1 M PO4 buffer with30% H2O2 for 2 to 10 min. The reaction was halted by removal ofthe DAB-H2O2 and addition of 0.1 M PO4 buffer to the wells.Tissue was rinsed a final time with 0.1 M PO4. The optic tectumslices were mounted on chrome-alum coated slides and stained in0.3% cresyl violet, dehydrated, cleared, mounted with Permount,and cover-slipped.

Identical sites were sampled consistently between animals byreferring to cytoarchitectural landmarks such as nuclear bordersand cell morphology. Regeneration was qualitatively analyzed bycomparing the extent of labeled optic nerve axons in the stratum

Fig. 1. Light photomicrographs of coronal sections through the optic tectum showing the projection of bulk-labeled optic nerve fibersat various time points following regeneration. All sections are approximately 900 mm from the anterior margin of the optic tectum. Thetectal lamina are labeled in the control panel ~A!; SO-SFGS � stratum opticum-stratum fibrosum et griseum superficiale ~white arrow!;SGC � stratum griseum centrale ~brackets!; SAC � stratum album centrale ~black arrow!. Subsequent panels represent time followingoptic nerve crush: ~B! two weeks ~arrow indicates projections!, ~C! six weeks, ~D! eight weeks, ~E! 10 weeks, ~F! 12 weeks. All scalebars � 100 mm. V � ventral; M � medial. Section thickness � 50 mm.

Optic nerve regeneration in hybrid sunfish 311

opticum-stratum fibrosum et griseum superficiale ~SO-SFGS!, stra-tum griseum centrale ~SGC! and stratum album centrale ~SAC!laminae ~Fig. 1A! of regenerating fish with controls.

Statistical analysis

Data analysis was performed using GraphPad Software ~San Di-ego, CA!. Samples were tested for normality using the method ofKolmogorov and Smirnov. Bartlett test was used to assure that thestandard deviations ~SD! of the groups were equal in order to useparametric statistical tests. If the SDs were unequal, non-parametrictests were employed. Oriana software ~version 2.0! was used toperform Watson Williams’s tests for circular statistics.

Results

Histology

The efficacy of the crush was examined in sunfish ~N � 3! whoseoptic nerve was bulk labeled within 48 hours of crush. Labeledaxons were not detected in the optic nerve ~or the tectum! medialto the crush site, indicating the technique completely crushed alloptic nerve axons. Fig. 1 shows light photomicrographs of coronalcross sections through the left optic tectum of the sunfish atvariable times following nerve crush. All sections were takenapproximately 900 mm from the anterior margin of the optictectum. At two weeks post-crush ~Fig. 1B!, many axons hadregenerated past the crush site in the optic nerve, however; only afew labeled axons were visible in the extreme ventro-lateral por-tion of the SO-SFGS. By week 4, there was an increase in thelabeled axons in the SO-SFGS, a few projections into the SGC andnone into the SAC. By week 6 ~Fig. 1C!, all tectal laminaecontained regenerating axons. All laminae showed increasing nervelabel through week 12, however organization of the regeneratedfibers in each lamina remained less distinct than controls.

Dorsal light reflex

Control sunfish maintained an upright swimming position whenthe illumination source was directly above the tank. Followingoptic nerve crush, experimental fish exhibited a pronounced DLRthroughout the day and swam with their control eye tilted awayfrom downwelling light. Although the tilt was seen to dissipateduring the day by week 6, it persisted when dark-adapted fish werefirst exposed to light. Dark-adapted fish displayed an average tiltof 45.58 6 2.28 SE from vertical during the first two weekspost-crush and little improvement was observed through week 6.However, the DLR decreased significantly ~ANOVA; P � 0.001!over the next six weeks ~Fig. 2! and was reduced to 4.58 61.38 byweek 16.

Predator-prey interactions

Fig. 3 plots the distribution of predator-prey interactions through-out the study. All experimental fish resumed feeding within oneweek following surgery. Most of the attacks by the control sunfishwere at prey located in the visual field of the left eye ~; 65%! witha strike success rate of approximately 28%. Throughout the firstfour weeks in the experimental fish, greater than 90% of the strikeswere at prey within the view of the left ~control! eye. As regen-eration progressed, the range of the strikes expanded into the visual

field of the experimental eye until the distribution of strikes byweek 10 closely resembled controls.

Individual strikes are plotted in Fig. 4. There was no significantdifference in distance or strike angle for control fish in the lightbetween left and right visual fields ~Tables 1 and 2!. During thefirst two weeks post-surgery ~Fig. 4B!, the majority of strikes wereinitiated by prey in the visual field of the control eye. Strikedistance continued to be greater on the control side throughweek 8, however by weeks 9 and 10 there was no significantdifference in distance between the two eyes ~Table 1!. Strike angleaveraged approximately 458 degrees in control eyes. The averagestrike angle in the right visual field increased from 6.38 in weeksone and two to 32.88 by week 10 ~Table 2!. By Week 9–10, themean strike angle was no longer significantly different than controls.

To account for visual field overlap @108 binocular overlap foreach eye in fish ~Walls, 1942!# , the strike angles were replotted

Fig. 2. ~A! The dorsal light reflex angle of rotation ~degrees! was plottedversus time ~weeks! following nerve crush in two experimental and acontrol fish. ~B! The average angle of rotation ~degrees! is plotted versustime ~weeks! following optic nerve crush. Data are binned in two-weekintervals to reduce variation in regeneration. Sample size ~trials!: Weeks1–2 � 44; Weeks 3–4 � 40; Weeks 5–6 � 32; Weeks 7–8 � 24; Weeks9–10 � 14; Weeks 11–12 � 6; Weeks 15–16 � 6. Error bars � 1 standarderror. Bars with different letters indicate significantly different means ~P �

0.05; ANOVA!.


~Fig. 5! into three sections: 108 to 180 to the left ~control eye!,6108 ~binocular overlap!, and 108 to 1808 to the right ~experimen-tal!. Approximately 50% of the strikes of control fish occurredin the left visual field and 20% fell within the binocular field.During the first two weeks following nerve crush, over 80% ofthe strikes were to prey on the control side and 18% fell within thebinocular cone. By weeks 7 and 8, there was a large increase in thepercentage of strikes towards the right visual field and by weeks 9and 10, the attack distribution approximated the presurgical strikedistribution.

Discussion

The utility of behavioral techniques is that they provide a directmeasure of synaptic regeneration, whereas electrophysiological

studies typically measure only presynaptic impulses of retinalganglion cells ~Springer & Agranoff, 1977!. Behavioral assays arealso useful because they are non-invasive measures that allow formultiple experiments at different stages of regeneration ~Springer& Agranoff, 1977!.

Histology

The optic nerve regenerated quickly with axons projecting into theoptic tectum within two weeks of nerve crush. Central projectionswere first seen in the SO-SFGS at two weeks and were detected inthe SGC and SAC at weeks 4 and 6, respectively. The number ofprojections and refinement of tectal laminae continued to increasethroughout the experiment. The optic tectum of the experimentalfish at 12 weeks strongly resembled controls in the number of

Fig. 3. The sunfish reaction angle towards the prey was binned in 308 intervals to the left and right, centered on the front of the fish.The optic nerve was always crushed on the right eye of the fish. Missed strikes are indicated by the solid portion of each bar. Samplesize ~trials!: A � 5; B � 5; C � 20; D � 19; E � 11; F � 7


projections, however, the central projection pattern appeared lessdefined than in controls. Previous studies found optic nerve regen-eration and refinement remained below control levels up to 5months or longer post-transection ~Schmidt et al., 1988; Meyeret al., 1985!.

Behavior (simple)

The DLR is a natural teleost behavior during which fish attempt tokeep their lighter ventral surface parallel with downwelling lightby tilting about their longitudinal axis to make it more difficult forpredators to discern the silhouette of the fish. The behavior ismediated by fish rotating their bodies to equalize downwellinglight in both eyes. Although vestibular input will eventually coun-teract the reflex, post-operative tilts in excess of 60 degrees werecommon.

Springer et al. ~1977! demonstrated by tectal ablation that theoptic tectum was unnecessary for the DLR and other non-tectalnuclei may be the center for this behavior ~Springer & Agranoff,1977!, and Yanagihara et al. ~1993! abolished the DLR by lesion-ing the lateral valvula cerebelli. Gibbs and Northmore ~1996!demonstrated that the torus longitudinalis was responsible for theDLR. Regardless of its epicenter, the DLR is dependent on equal-izing light to both eyes and if vision is compromised in one eye,fish will swim with their control eye downward to equalize thelight to both eyes ~Mensinger & Powers, 1999!. The improvementin the DLR provides strong evidence for the synaptogenesis of theregenerating axons to their correct central targets.

Fish may eventually habituate to the unequal light distributionand resume normal swimming during the day as observed with the

sunfish following week 6. However, the reflex remains an accuratetest during the first few minutes of light exposure for dark-adaptedfish and allows differentiation between habituation and regenera-tion. Maximum tilts were observed soon after nerve crush, how-ever, there was a significant improvement after week 6, concurrentwith the projections and increased lamination in all tectal layers. Asmall 4.58 tilt that continued to persist at 15–16 weeks post-crushindicates that refinement may still be ongoing. Hayes and Meyer~1989! indicated that synaptic refinement in the goldfish couldpersist through 34 weeks post-transection.

Behavior (complex)

Historically, most optic nerve regeneration studies have pursuedrelatively simple behaviors such as LED flash response ~North-more, 1989; Northmore & Masino, 1984! and contrast sensitivity~Northmore & Celenza, 1992!. The DLR provided an estimate ofthe reestablishment of functional connections, however it did notrelay information on the spatial accuracy of these connections.Vision plays a critical role in sunfish and mediates predator-preyinteractions ~Howick & O’Brien, 1983; Nyberg, 1971!. The use ofsmaller prey, such as Daphnia spp., might have given informationon the return of visual acuity or synaptic refinement. However,minnows were chosen as they proved a more challenging target~only 33% of strikes in the control sunfish were successful! thatallowed assessment of the integration of visual function and motorbehavior. In addition, because the sunfish approached and con-sumed the minnows, prey were not beyond the gape limitation of

Table 1. Sunfish reaction distance

Time post-crush~weeks!

Left side~control!

Right side~experimental!

control 16.16 2.6 17.4 6 2.5~19! ~10!

1 to 2 19.3 6 1.7 8.3 6 5.9#

~50! ~3!3 to 4 20.7 6 0.7 14.6 6 1.6*

~225! ~26!5 to 6 20.16 0.5 17.0 6 1.0*

~410! ~74!7 to 8 23.0 6 0.9 18.3 6 1.3*

~111! ~36!9 to 10 20.7 6 1.0 19.8 6 1.2

~101! ~50!

Values represent mean reaction distance ~cm! 6 1 SE.The number of strikes analyzed is in parentheses. Thecontrol and experimental distances were compared usingMann Whitney unpaired t-test and * indicate signifi-cantly different means for each time interval. # indicatesinsufficient sample size for statistical analysis.

Fig. 4. The reaction distance ~cm! of sunfish was plotted as polar coordinates at various times ~weeks! following optic nerve crush. Thehatched oval represents the body of the sunfish and all optic nerve crushes were performed on the right eye. Symbols representindividual strikes with solid symbols representing missed strikes ~MS! and open symbols indicating successful strikes ~SS!. The thickcircle around the center denotes the maximum range of the lateral line ~5.7 cm!. L � light, D � dark. Sample size is the same as Fig. 3.

Table 2. Sunfish strike angles

Time post-crush~weeks!

Left side~control!

Right side~experimental! Combined

control 37.7 6 8.2 43.4 6 10.1 9.5 6 9.4a

~19! ~10! ~29!Week 1–2 51.8 6 6.4 6.3 6 2.7# 48.5 6 6.1b

~50! ~3! ~53!Week 3–4 49.0 6 2.7 12.7 6 4.0* 36.0 6 2.9b

~225! ~26! ~251!Week 5–6 49.0 6 2.0 19.2 6 2.9* 35.9 6 2.0b

~410! ~74! ~484!Week 7–8 55.0 6 3.8 34.4 6 5.5* 33.4 6 4.8b

~111! ~36! ~147!Week 9–10 48.16 3.9 32.8 6 7.6 18.9 6 4.2a

~101! ~50! ~151!

Values represent the absolute mean strike angle ~8! 6 1 SE. The number ofstrikes analyzed is in parentheses. The control and experimental angleswere compared using Mann Whitney unpaired t-test and * indicate signif-icantly different means for each time interval between these two columns.# indicates insufficient sample size for statistical analysis. The strike angleswere combined ~left �0–1808; right �0–1808! and analyzed using aWatson William F test. Significantly different means are indicated bydifferent letters.


the sunfish. Therefore, small variation in either minnows or sun-fish size were insignificant and played no role in the analysis of thepredator-prey interactions.

Most control sunfish attacks on minnows were forward ~6308!and biased to prey on the left. It may have been some anomalyin the aquarium ~differential light intensity, water currents, preyinsertion point!, or in the external lab environment ~noise, me-chanical vibration! that resulted in this bias. However, an alter-nate explanation for this bias could be a lateralization, or“handedness,” in feeding behavior or use of the left eye insunfish. Lateralization has been found in the stridulation soundproduction of channel catfish ~Fine et al., 1996! with 90% ofthe catfish exhibiting this behavior showing a preference to theright pectoral fin. Blue gourami demonstrated preferential use ofthe left ventral fin when obtaining tactile information aboutnovel inanimate objects introduced into their test aquaria ~Bisazzaet al., 2001!. Laterilization in feeding has been shown in ascale-eating cichlid ~Hori, 1993! and Macrorhamphosodes ura-doi ~Nakae & Sasaki, 2002!. Many studies have shown prefer-ential eye use in mirror image inspection, response to familiarand unfamiliar stimuli, and in social behavior in several speciesof fish ~Sovrano et al., 1999; Sovrano et al., 2001; Sovrano,2004; De Santi et al., 2002; Dadda & Bisazza, 2006!. Whereaslateralization has been established in fish, this would be the firstevidence of lateralization in feeding behavior or preferential eyeuse in sunfish. However, this potentially novel discovery re-quires further study for confirmation.

In sunfish predator prey interactions, lateral line input must beconsidered. Control sunfish successfully detected and attackedprey in total darkness, presumably using lateral line information.However, the range in the dark was limited, with an averagereaction distance of 3.2 cm and a maximum range of 5.7 cm.Therefore, any strikes at targets greater than 6 cm from thepredator were considered visually mediated. As regeneration pro-gressed, the average reaction distances exceeded the maximum

range of 5.7 cm of the lateral line providing evidence that themajority of the attacks were visually mediated.

There is a 108 binocular overlap for each eye of the sunfish, andtherefore prey-initiated strikes within the 6108 cone ~Walls, 1942!could possibly be viewed by either eye or both eyes. By factoringthis overlap, the percentage of strikes initiated by prey was heavilyweighted towards the control eye early in regeneration. Beginningin weeks 5 and 6 and continuing through weeks 9 and 10, there wasa gradual return of the strike distribution to control patterns. By theend of the experiment, the average attack angles were not statis-tically different to controls, indicating that sunfish were using theexperimental eye in prey capture similar to unoperated sunfish.This progression corresponds with the improved stratification inthe tectal laminae over the same time period. Although the refine-ment of optic projections might still be ongoing, restoration ofvisual function was sufficient to mediate predator prey interactionsthroughout the normal visual field.

Retinotectal topography indicated that points in the anterior andposterior visual fields correspond to points in the anterior andposterior tectum respectively ~Schwassmann & Kruger, 1965;Schwassmann & Krag, 1970; Cook, 1979!. Previous regenerationexperiments indicated that optic nerve regeneration proceededsequentially from anterior to posterior portions of the tectum~Stuermer & Easter, 1984; Hayes & Meyer, 1988!. The strikepattern indicated that the visual field of the experimental eyeexpanded in an anterior to posterior gradient. Gradual expansion ofthe visual field continued as the sunfish struck to the right atincreasing angles, from a range of 438 early in regeneration to over1208 late in regeneration.

Although previous studies have demonstrated that the teleostoptic nerve can regenerate and re-establish functional connections,the ability of the new connections to mediate complex visualbehavior remained unknown. These experiments demonstrate forthe first time that the regenerated optic nerve can mediate complexvisual behavior such as target detection and allow the animal tosuccessfully track and strike at the moving target.

Acknowledgments

The authors thank Dr. Thomas Hrabik and Dr. Carl Richards for construc-tive comments on the manuscript; Hazel Richmond, Mark Pranckus, RandyHedin, Lucy Palmer, and Beth Holbrook for help with collecting sunfish;Doug Schaff for his help in videotaping the behavior experiments andspecial thanks to Dr. Janet Fitzakerley for use of her cryostat and micro-scope. Funding was provided by a University of Minnesota Faculty Grantin Aid and the University of Minnesota, Duluth Biology Department, anda grant from the Visualization and Digital Imaging Laboratory.

References

Batty, R.S., Blaxter, J.H.S. & Richard, J.M. ~1990!. Light intensity andfeeding behaviour of herring, Clupea harengus. Marine Biology 107,383–388.

Bisazza, A., Lippolis, G. & Vallortigara, G. ~2001!. Lateralization ofventral fins use during object exploration in the blue gourami ~Tricho-gaster trichopterus!. Physiology & Behavior 72, 575–578.

Cook, J.E. ~1979!. Interactions between optic fibers controlling the loca-tions of their terminals in the goldfish optic tectum. Journal of Embry-ology and Experimental Morphology 52, 89–103.

Dadda, M. & Bisazza, A. ~2006!. Lateralized female topminnows canforage and attend to a harassing male simultaneously. BehavioralEcology 17, 358–363.

De Santi, A., Bisazza, A. & Vallortigara, G. ~2002!. Complementaryleft and right eye use during predator inspection and shoal-matescrutiny in minnows. Journal of fish Biology 60, 116–1125.

Fig. 5. The percentage of overall strikes towards the control eye ~left side,108 to 1808; closed circle!, in the binocular field ~6108, open circle! and tothe experimental eye ~right side, 108 to 1808; triangle! is plotted versus time~weeks! following optic nerve crush in the sunfish. The horizontal linesindicate control fish strike distributions ~left-solid; binocular-dotted; right-dashed!. Sample size is the same as Fig. 3.


Fine, M.L., McElroy, D., Rafi, J., King, C.B., Loesser, K.E. & New-ton, S. ~1996!. Lateralization of pectoral stridulation sound productionin the channel catfish. Physiology & Behavior 60, 753–757.

Gibbs, M.A. & Northmore, D.P.M. ~1996!. The role of torus longitudi-nalis in equilibrium orientation measure with the dorsal light reflex.Brain, Behavior and Evolution 48, 115–120.

Hayes, W.P. & Meyer, R.L. ~1988!. Normal and regenerating optic fibersin goldfish tectum: HRP-EM evidence for rapid synaptogenesis andoptic fiber-fiber affinity. Journal of Comparative Neurology 274,516–538.

Hayes, W.P. & Meyer, R.L. ~1989!. Normal numbers of retinotectalsynapses during the activity-sensitive period of optic regeneration ingoldfish: HRP-EM evidence implicating synapse rearrangement andcollateral elimination during map refinement. Journal of Neuroscience9, 1400–1413.

Hori, M. ~1993!. Frequency-dependent natural selection in the handednessof scale-eating cichlid fish. Science 260, 216–219.

Howick, G.L. & O’Brien, W.J. ~1983!. Piscivorous feeding behavior oflargemouth bass: An experimental analysis. Transactions of the Amer-ican Fisheries Society 112, 508–516.

Janssen, J. & Corcoran, J. ~1993!. Lateral line stimuli can override visionto determine sunfish strike trajectory. Journal of Experimental Biology176, 299–305.

Jolly, D.W., Mawdesley-Thomas, L.E. & Bucke, D. ~1972!. Anesthsiaof Fish. Veterinary Record 91, 424–426.

McMahon, T.E. & Holanov, S.H. ~1995!. Foraging success of large-mouth bass at different light intensities: Implications for time and depthof feeding. Journal of Fish Biology 46, 759–767.

Mensinger, A.F. & Powers, M.K. ~1999!. Visual function in regeneratingteleost retina following cytotoxic lesioning. Visual Neuroscience 16,241–251.

Meyer, R.L. ~1980!. Mapping the normal and regenerating retinotectalprojection of goldfish with autoradiographic methods. Journal of Com-parative Neurology 189, 175–190.

Meyer, R.L., Sakurai, K. & Schauwecker, E. ~1985!. Topographyof regenerating optic fibers in goldfish traced with local wheat germinjections into retina: Evidence for discontinuous microtopographyin retinotectal projection. Journal of Comparative Neurology 239,27–43.

Meyer, R.L. & Kageyama, G.H. ~1999!. Large-scale synaptic errorsduring map formation by regenerating optic axons in the goldfish.Journal of Comparative Neurology 409, 299–312.

Mittelbach, G.G. ~1984!. Predation and resource partitioning in twosunfishes ~Centrarchidae!. Ecology 65, 499–513.

Nakae, M. & Sasaki, K. ~2002!. A scale-eating triacanthodid, Macrorham-phosodes uradoi: Prey fishes and mouth “handedness” ~Tetraodonti-formes, Triacanthoidei!. Ichthyological Research 49, 7–14.

Northcutt, R.G. & Butler, A.B. ~1991!. Retinofugal and retinopetalprojections in the green sunfish, Lepomis cyanellus. Brain, Behaviorand Evolution 37, 333–354.

Northmore, D.P.M. ~1981!. Visual localization after rearrangement ofretinotectal maps in fish. Nature 293, 142–144.

Northmore, D.P.M. & Masino, T. ~1984!. Recovery of vision in fish afteroptic nerve crush: A behavioral and electrophysiological study. Exper-imental Neurology 84, 109–125.

Northmore, D.P.M. ~1989!. Quantitative electrophysiological studies ofregenerating visuotopic maps in goldfish I. Early recovery of dimmingsensitivity in tectum and torus longitudinalis. Neuroscience 32, 739–747.

Northmore, D.P.M. ~1991!. Visual responses of nucleus isthmi in a teleostfish ~Lepomis macrochirus!. Vision Research 31, 525–535.

Northmore, D.P.M. & Celenza, M.A. ~1992!. Recovery of contrastsensitivity during optic nerve regeneration in fish. Experimental Neu-rology 115, 69–72.

Nyberg, D.W. ~1971!. Prey capture in the largemouth bass. AmericanMidland Naturalist 86, 128–144.

Rankin, E.C.C. & Cook, J.E. ~1986!. Topographic refinement of regen-erating retinotectal projection of the goldfish in standard laboratoryconditions: A quantitative WGA-HRP study. Experimental Brain Re-search 63, 409–420.

Richmond, H.E., Hrabik, T.R. & Mensinger, A.F. ~2004!. Light intensity,prey detection and foraging mechanisms of age-0 year yellow perch~Perca flavescens!. Journal of Fish Biology 65, 195–205.

Schmidt, J.T. & Edwards, D.L. ~1983!. Activity sharpens the map duringthe regeneration of the retinotectal projection in goldfish. Brain Re-search 269, 29–39.

Schmidt, J.T., Turcotte, J.C., Buzzard, M. & Tieman, D.G. ~1988!.Staining of regenerated optic arbors in goldfish tectum: Progressivechanges in immature arbors and a comparison of mature regeneratedarbors with normal arbors. Journal of Comparative Neurology 269,565–591.

Schwassmann, H.O. & Kruger, L. ~1965!. Organization of the visualprojection upon the optic tectum of some freshwater fish. Journal ofComparative Neurology 124, 113–126.

Schwassmann, H.O. & Krag, M.H. ~1970!. The relation of visual fielddefects to retinotectal topography in teleost fish. Vision Research 10,29–42.

Sovrano, V.A., Rainoldi, C., Bisazza, A. & Vallortigara, G. ~1999!.Roots of brain specializations: Preferential left-eye use during mirror-image inspection in six species of teleost fish. Behavioural BrainResearch 106, 175–180.

Sovrano, V.A., Bisazza, A. & Vallortigara, G. ~2001!. Lateralizationof response to social stimuli in fishes: A comparison between differentmethods and species. Physiology & Behavior 74, 237–244.

Sovrano, V.A. ~2004!. Visual lateralization in response to familiar andunfamiliar stimuli in fish. Behavioural Brain Research 152, 385–391.

Sperry, R.W. ~1948!. Patterning of central synapses in regeneration of theoptic nerve in teleosts. Physiological Zoology 21, 351–361.

Springer, A.D., Easter, S.E. & Agranoff, B.W. ~1977!. The role of theoptic tectum in various visually mediated behaviors of goldfish. BrainResearch 128, 393–404.

Springer, A.D. & Agranoff, B.W. ~1977!. Effect of temperature on rateof goldfish optic nerve regeneration: A radioautographic and behavioralstudy. Brain Research 128, 405–415.

Striedter, G.F. & Northcutt, R.G. ~1989!. Two distinct visual pathwaysthrough the superficial pretectum in a percomorph teleost. Journal ofComparative Neurology 283, 342–354.

Stuermer, C.A.O. & Easter, S.S., Jr. ~1984!. A comparison of the normaland regenerated retinotectal pathways of goldfish. Journal of Compar-ative Neurology 223, 57–76.

Walls, G.L. ~1942!. The Vertebrate Eye and its Adaptive Radiation.Michigan: Cranbrook Institute of Science.

Walton, W.E., Hairston, N.G., Jr. & Wetterer, J.K. ~1992!. Growth-related constraints on diet selection by sunfish. Ecology 73, 429–437.

Werner, E.E. & Hall, D.J. ~1974!. Optimal foraging and the size selectionof prey by the bluegill sunfish ~Lepomis macrochirus!. Ecology 55,1042–1052.

Yanagihara, D., Watanabe, S. & Mitarai, G. ~1993!. Neuroanatomicalsubstrate for the dorsal light response I. Differential afferent connec-tions of the lateral lobe of the valvula cerebelli in goldfish ~Carassiusauratus!. Neuroscience Research 16, 25–32.


restoration of visual function following optic nerve regeneration …amensing/publications/callahan...

Documents