Cellular and Subcellular Structureof Anterior Sensory Pathways inPhestilla sibogae (Gastropoda,

Nudibranchia)

DMITRI Y. BOUDKO,1 MARILYN SWITZER-DUNLAP,2 AND MICHAEL G. HADFIELD1*1Kewalo Marine Laboratory, University of Hawaii, Honolulu, Hawaii 96813

2Biological EM Facility, Pacific Biomedical Research Center, University of Hawaii,Honolulu, Hawaii 96822

ABSTRACTTwo sensory-cell types, subepithelial sensory cells (SSCs) and intraepithelial sensory

cells (ISCs), were identified in the anterior sensory organs (ASO: pairs of rhinophores and oraltentacles, and the anterior field formed by the oral plate and cephalic shield) of thenudibranch Phestilla sibogae after filling through anterior nerves with the neuronal tracersbiocytin and Lucifer Yellow. A third type of sensory cells, with subepithelial somata and tuftsof stiff-cilia (TSCs, presumably rheoreceptors), was identified after uptake of the mitochon-drial dye DASPEI. Each sensory-cell type has a specific spatial distribution in the ASO. Thehighest density of ISCs is in the oral tentacles (<1,200/mm2), SSCs in the middle parts of therhinophores (.4,000/mm2), and TSCs in the tips of cephalic tentacles (100/mm2). Thesemorphologic data, together with electrophysiologic evidence for greater chemical sensitivity ofthe rhinophores than the oral tentacles (Murphy and Hadfield [1997] Comp. Biochem. Physiol.118A:727–735; Boudko et al. [1997] Soc. Neurosci. Abstr. 23:1787), led us to conclude that thetwo pairs of chemosensory tentacles serve different chemosensory functions in P. sibogae; i.e.,ISCs and the oral tentacles serve contact- or short-distance chemoreception, and SSCs and therhinophores function for long-distance chemoreception or olfaction. If this is true, then theISC subsystem probably represents an earlier stage in the evolution and adaptations ofgastropod chemosensory biology, whereas among the opisthobranchs, the SSC subsystemevolved with the rhinophores from ancestral cephalaspidean opisthobranchs. J. Comp.Neurol. 403:39–52, 1999. r 1999 Wiley-Liss, Inc.

Indexing terms: mollusca; chemoreception; sensory-cell ultrastructure; fluorescent tracers;

neuronal integration

Gastropod molluscs have well developed chemosensorycapacities, including both contact and distance chemorecep-tion (olfaction), which are critically important for theirbehavioral coordination and orientation to prey, conspe-cific individuals, and potential predators in the environ-ment (Gelperin, 1974; Audesirk, 1975; Emery and Aude-sirk, 1978; Chase, 1982). Behavioral and electrophysiologicalstudies reveal that the epithelia of the anterior sensoryorgans (ASO: cephalic tentacles, oral tentacles, and partsof the cephalic shield) in different gastropod species aresensitive to chemical stimuli (Arey, 1918; Jahan-Parwar,1972; Bicker et al., 1982a; Boudko et al., 1997). The ASO ofsome molluscs rival the olfactory systems of vertebratesand arthropods in their olfactory sensitivity (Gurin andCarr, 1971), odor spectra recognized, and density of recep-tor cells (Chase and Tolloczko, 1993).

The ultrastructure and innervation of the ASO in gastro-pod molluscs have been extensively studied and reviewed(Jahan-Parwar, 1975; Dorsett, 1986; Emery, 1992). Subepi-thelial and intraepithelial sensory cells have been specifi-cally identified in different chemosensory areas of gastro-pod chemoreceptive organs. Subepithelial sensory cells,those whose cell bodies lie below the basal lamina of theepidermis and extend sensory dendrites to the surface,

Grant sponsor: Office of Naval Research; Grant number: N00014–94–1-0524; Grant sponsor: National Institutes of Health RCMI; Grant number:RR03061.

*Correspondence to: Michael G. Hadfield, Kewalo Marine Laboratory,University of Hawaii, 41 Ahui Street, Honolulu, HI 96813.E-mail: hadfield@hawaii.edu

Received 21 July 1997; Revised 21 July 1998; Accepted 14 August 1998

THE JOURNAL OF COMPARATIVE NEUROLOGY 403:39–52 (1999)

r 1999 WILEY-LISS, INC.

have been reported in the olfactory tentacles of terrestrialsnails (Chase and Tolloczko, 1993) and slugs (Wright,1974; Kataoka, 1976), in tentacles of the sea hares Aplysiacalifornica and A. brasiliana (Emery and Audesirk, 1978),and in rhinophores of larvae of the nudibranch Rostangapulchra (Chia and Koss, 1982). Chemosensory cells clus-tered in small subepithelial ganglia in the rhinophoreswere described also in the nudibranch Tethys leporina(Merton, 1920). Morphological diversity and specificity indistribution of different types of subepithelial receptorcells in the tentacles and areas of the oral plate offreshwater pulmonate molluscs suggest that this cellcategory includes receptors of several sensory modalities(Zylstra, 1972; Zaitseva and Bocharova, 1981; Yi andEmery, 1991).

Only intraepithelial sensory cells were described fromultrastructural examination of the rhinophores of severalother nudibranch species (Storch and Welsch, 1969). De-spite the absence of identified afferent pathways, thesecells were argued to be chemoreceptors and rheoreceptorsbased on their morphological features. No afferent path-ways were found to arise from morphologically similarintraepithelial cells in the tentacles of other opistho-branchs (Emery and Audesirk, 1978) or prosobranchs(Storch and Welsch, 1969), nor in the lip of pulmonates(Zylstra, 1972). It was suggested that these peripheralcells may not be sensory (see Discussion section of Zylstra,1972, and review by Emery, 1992). However, similarprimary intraepithelial sensory cells have been clearlyidentified in chemosensory epithelia of pulmonate andprosobranch osphradia (Crisp, 1973; Haszprunar, 1985,1987; Nezlin et al., 1994).

To clarify the morphologic properties, distribution, andafferent organization of the anterior sensory structures ofgastropod molluscs, we used the tropical sea slug Phestillasibogae (Bergh, 1905) as an abundantly available modelthat can be propagated easily in the laboratory. The diet ofP. sibogae is limited to species of the hermatypic coralgenus Porites, mostly P. compressa in Hawaii (Harris,1975). Soluble factors secreted by Porites species stronglyinfluence the behavior and life cycle of P. sibogae. Thesemolluscs find their prey at a distance by means of chemore-ception (Harris, 1971). Their veliger larvae are unequivo-cally stimulated to undergo settlement and metamorpho-sis in response to living P. compressa and aqueous extractsof the coral (Hadfield and Scheuer, 1985; Hadfield andPennington, 1990). Although P. sibogae has not been asextensively used in neurobiological investigations as someother gastropod species, its brain has equivalent proper-ties for intracellular studies of neuronal networks, includ-ing large, rapidly identifiable neurons and neuronal clus-ters easily accessed in semi-intact preparations (Willows,

1985). At least some neurons in the brain of P. sibogae arehomologous with cells in other nudibranch molluscs andmay be used as electrophysiological monitors for specificbehavioral processes (Dickinson, 1979). Recent investiga-tions of chemosensory pathways in P. sibogae demon-strated the potential to differentiate responses in chemo-sensory cells and study their pharmacological propertiesin situ and in vitro (Murphy and Hadfield, 1997; Boudko etal., 1997).

As a preamble to intensive investigations of the physi-ological and molecular bases of chemoreception in P.sibogae, we conducted a morphological analysis of thespecies’ ASO with the following goals: (1) to characterizethe ultrastructure of cells composing the chemosensoryepithelia, (2) to determine the spatial distribution ofsensory cells in the ASO, and (3) to analyze the afferentstructure of different types of sensory cells in ASO.

MATERIALS AND METHODS

Animals

Laboratory-grown individuals of P. sibogae were takenfrom stocks maintained at the Kewalo Marine Laboratory(University of Hawaii; methods for propagation weredescribed by Miller and Hadfield, 1986). Adult animals,each 15–20 mm long and weighing 0.6 6 0.2 g, wereselected for most examinations (over 100). Only a fewcomparative observations (n 5 10) were performed onsmaller, juvenile specimens.

Preparation of ASO

An individual P. sibogae was bathed in low-calciumartificial sea water (442 mM NaCl, 9.6 mM KCl, 25.5 mMMgSO4, 32.5 mM MgCl2, 0.1 mM CaCl2, 2 mM N-2-hydroxyethylpiperazine-N’-2-ethanesulfonic acid, 0.1 mMethyleneglycoltetraacetic acid, pH 8.3) for about 10 min-utes for immobilization. The ASO of the slug were carefullyremoved, singly or in groups, from the animal and isolatedwith proximal portions of their peripheral nerves (Fig. 1).For different purposes, these preparations included asingle rhinophore or oral tentacle or the entire ASO. Thesepreparations were then rinsed several times and placed innatural, 0.22-µm filtered seawater (FSW). For backfillingexperiments, rhinophoral nerves (RhN) were dissectedeither with or without the rhinophoral ganglia (RhG, Fig.1B). Preparations of the oral tentacles included half of theoral plate and the oral tentacle nerve (OtN, Fig. 1).

Dye tracing (backfilling) of nerves

Nerves of ASO were backfilled through glass pipettesaccording to the method described by Chase and Tolloczko(1993). Pipettes for backfilling were prepared from borosili-cate glass capillaries (IB 100–4, WPI) with a P-87 micropi-pette puller (Sutter Instruments Co., Novato, CA). Glasspipettes were pulled from capillaries with 150-µm tips.Then, each pipette was constricted approximately 150–300µm back from its tip with a loop of current-heated plati-num wire, so that it tightly fit the nerve under study.

Electrophysiological recording was used to select nervepreparations for backfilling that were appropriately com-pressed into the pipette. If the signal-to-noise ratio ofrecorded spontaneous discharges was greater than 1,000,then backfilling was good predictably and labeling was

Abbreviations

ASO anterior sensory organsDAB diaminobenzidineFSW 0.22 µm filtered seawaterISC intraepithelial sensory cellOtN oral tentacle nerveRhG rhinophoral ganglionRhN rhinophoral nerveSEM scanning electron micrographySSC subepithelial sensory cellSTR streptavidin Texas red conjugateTSC subepithelial sensory cell with tuft of stiff ciliaTEM transmission electron micrography-

40 D.Y. BOUDKO ET AL.

repeatable among preparations (1700 differential AC am-plifier AM Systems, Inc., Everett, WA; band filter-set,0.1–1,000 Hz; and WPI acquisition system MPW100,acquisition rate was 4 kHz).

Neuronal tracers

Lucifer Yellow (LY) and biocytin (Molecular Probes, Inc.,Eugene, OR) were used as 3–6% solutions. Horseradishperoxidase (HRP; Sigma, St. Louis, MO) was used as a 4%solution in 0.5 M Tris-HCl buffer. ASO, backfilled for 2, 4,6, 8, 12, and 24 hours, were fixed in 4% paraformaldehydein 0.5 M sodium phosphate buffer solution (PBS; adjustedosmolarity 1.0 M/kg, pH 7.4) for 4 hours at 4°C, and rinsedin several changes of PBS. Some preparations were whole-mounted (n 5 102), and others were cryoprotected (n 5 35)in 20% sucrose in 0.4 M PBS for about 4 hours, embeddedin Tissue-Tek (Miles Laboratories, Elkhart, IN) on dry ice,sectioned at about 30 µm with a cryostat microtome, andmounted on polylysine-coated glass slides. Biocytin-backfilled preparations were incubated with a 1:100 solu-tion of streptavidin–Texas red conjugate (STR, MolecularProbes, Inc.) in 1% Triton X-100/0.4 M PBS. Incubationtimes were 40 hours for whole-mounts and 6 hours forcryostat sections. The preparations were rinsed, dehy-drated in a graded ethanol series, cleared in methylsalicylate, and mounted with Pro-Tex (American HospitalSupply Co., McGav Park, IL). Backfilled HRP was detectedhistochemically by the light- and electron-dense product

formed in its reaction with diaminobenzidine (DAB, afterthe method of Kunkel et al., 1993). Control preparations ofASO, bathed in the backfilling chamber and subsequentlytreated as the backfilled preparations, were used to detecta nonspecific staining. Selected frames from three-dimensional stacks were selected in TIFF bitmap format toprepare final plates and line drawings using Corel Drawsoftware (Corel Co., Ontario, Canada). Brightness andcontrast intensities of computer-generated figures wereadjusted for maximum clarity. Additionally, a digital noise-reduction algorithm was used in generating Figures 3Cand 4C.

DASPEI labeling

The mitochondrial styryl dye DASPEI (2-(4-dimethyl-aminostyryl)-N-ethylpyridinium iodide; Molecular Probes,Inc.) has been used as a selective vital marker of sensoryterminals and cells in several species of vertebrates andinvertebrates (Nurse and Farraway, 1989; Balak et al.,1990; Leise, 1996). We used DASPEI to determine thedistribution of sensory cells that did not label with ortho-grade tracers. DASPEI was applied to ASO in threedifferent ways to achieve different goals. (1) To obtain alow-resolution whole picture of anterior sensory pathways,an isolated ASO preparation was immersed in 0.1% DAS-PEI in FSW for 1 minute, rinsed, and replaced in FSW for2 hours to allow proximal diffusion of the dye into sensoryaxons. (2) To obtain high-resolution confocal images of

Fig. 1. Anterior sensory organs (ASO) of Phestilla sibogae.A: Scanning electron photomicrograph of a 6-mm juvenile specimen,anterior view; dotted lines demarcate ASO as they were isolated forbackfilling experiments. B: Isolated central nervous system andperipheral nerves of P. sibogae; rhinophoral and oral tentacle nerveswere cut at the positions indicated by the dotted lines. Cc, cerebral

commissure; Ce, cerata; Cs, cephalic shield; l/r CPlG, left and rightcerebropleural ganglia; E, eye; Ol, oral-labial region; Ot, oral tentacle;l/s-OtN, large and small oral-tentacle nerves; l/r PeG, left and rightpedal ganglia; Pel, pedal loop; l/r RhG, left and right rhinophoralganglia; Rh, rhinophore; RhN, rhinophoral nerve. Scale bar 5 500 µmin A,B.

SENSORY PATHWAYS IN A NUDIBRANCH 41

sensory cells, an ASO was compressed beneath a coverslipand perfused with a drop of 0.01–0.5% DASPEI applied atone side of the coverslip; after 20–120 seconds, the DAS-PEI was mostly removed by applying FSW to one side ofthe coverslip and drawing the FSW under the cover slip byabsorbing water with filter paper at the opposite side. (3)To visualize dye diffusion into afferent nerves or intoadjacent cells by means of intercellular coupling, 0.5%DASPEI was applied for a few seconds to a small patch ofthe epithelial surface of a compressed ASO through a10-µm plastic capillary.

Fluorescent and confocal microscopy andimage analysis

A Zeiss Axophot fluorescence microscope was used toexamine labeling patterns. An LY cube (excitation 430,barrier 540) was used to examine DASPEI-stained prepa-rations. Images (TIFF, bitmap) were captured with a CCDcamera (EDC-1000U, Electrim Co., Princeton, NJ). Mochaimage analysis software (Jandel Scientific, San Rafael,CA) was used to determine morphometric properties,numbers, and densities of stained cells. Confocal imageswere obtained on a BioRad 1024 confocal microscope byusing Confocal Assistant 4.02 software for three-dimen-sional reconstruction of (1) 16- to 64-frame stacks acquiredin 1- to 1.5-µm focal steps for biocytin-STR labeled prepara-tions, and (2) two- to three-frame stacks acquired in 2-µmfocal steps for DASPEI-stained preparations.

Electron microscopy

For transmission electron microscopy (TEM), rhino-phores were placed in a primary fixative of 2.5% glutaral-dehyde in 0.2 M sodium cacodylate buffer (pH 7.2–7.4)containing 0.1 M NaCl and 0.35 M sucrose for 1 hour(modified from Eisenman and Alfert, 1982). Tissues werethen rinsed in buffer (0.2 M sodium cacodylate, 0.3 MNaCl) three times for 15 minutes each, and immersed for 1hour in 1.0% OsO4 in 0.2 M sodium cacodylate, 0.3 M NaCl.Tissues were next rinsed in distilled water and dehydratedthrough an ethanol series (15 minutes each step), passedthrough propylene oxide and embedded in LX-112 resin.Resin blocks were sectioned with a diamond knife on aReichert Ultra-Cut E ultramicrotome. Sections were col-lected on Formvar-coated slot grids, contrasted with ura-nyl acetate and lead citrate, examined, and photographedin a Zeiss 10/A-transmission electron microscope.

For scanning electron microscopy (SEM), tissues werefixed and dehydrated as described for standard TEM. Topermit examination of internal structure, some of therhinophores were sliced open with a razor blade in the 70%ethanol dehydration step. After dehydration in ethanol,tissues were critical-point dried in a Tousimis Autosamdri810 Critical Point Dryer, mounted on aluminum stubs,coated with gold-palladium in a Hummer II Sputter Coater,and examined with a Hitachi S-800 field emission scan-ning electron microscope.

RESULTS

Anatomy and innervation of the ASO

The head of P. sibogae bears two pairs of well-developedsensory tentacles, distinguished as the rhinophores (dor-sal) and oral tentacles (Fig. 1A). The oral tentacles arebilateral extensions from the oral field that reach 5–7 mm

in length in adult specimens. The oral tentacles areslightly flattened in the dorsal-ventral plane, and each hasa medial groove not seen in young individuals. The rhino-phores project above the head and grow to approximately5-mm long in adults. Transverse and longitudinal muscles,fibrous connective tissue, and nerves fill much of thehemocoelic spaces of the rhinophores and oral tentacles.

Bilateral rhinophoral nerves arise from the anterodorsalsurface of cerebropleural ganglia and run a short distanceto a rhinophoral ganglion in the base of each rhinophore(RhG, Figs. 1B, 2A). From each RhG, four smaller nerves,each about 20 µm in diameter in adult animals, plus a fewfiner neuronal fibers, emerge and run into the respectiverhinophore. Two nerves, one about twice the diameter ofthe other, arise from the ventrolateral surface of eachcerebral ganglion and run into the oral tentacle on eachside (Figs. 1B, 2A). In large specimens, the larger oraltentacle nerves are about 60 µm in diameter and thesmaller ones about 25 µm. Additional fine nerves arisefrom the anterior surfaces of both cerebropleural andpedal ganglia and pass to different parts of the cephalicshield.

Labeling of ASO by centrifugal backfilling ofnerves

Dense populations of sensory cells in the epithelia of therhinophores, oral tentacles, and oral plate were backfilledthrough the rhinophoral and large oral tentacle nerves.Only a few epithelial sensory cells on the dorsolateralsurface of the oral plate and in the ventrobasal region ofthe oral tentacles were backfilled through the smaller oraltentacle nerves. Backfilling through anteriorly directednerves rising from the pedal ganglia and pleural portionsof the cerebropleural ganglia did not fill cells in theepithelia of the ASO.

Results of backfilling ASO with biocytin and LY differedtemporally and in the number of sensory cells filled (Fig.2B). When biocytin was applied for less than 8 hours, theresults were similar to those obtained with LY applied foreven much longer intervals (primary labeling, Fig. 2B),and the location of biocytin- and LY-backfilled cells wassimilar. However, in contrast to biocytin, the labelingpattern with LY did not change in 8-, 12-, and 24-hourpreparations, with densities of labeled cells remaining low(Fig. 2B). After 24 hours, LY produced nonspecific stainingof ASO tissues.

The first stained sensory cells were identified in differ-ent portions of ASO after 2–4 hours of backfilling withbiocytin. The densities of labeled cells increased dramati-cally between 8 and 12 hours of backfilling (secondarylabeling, Fig. 2B). Preparations backfilled 12 and 24 hoursdisplayed similar patterns and local densities of labeledcells, indicating that all cells connected to the nerve werefilled within 12 hours.

Backfilling with HRP was slow compared with LY,typically requiring more than 12 hours, and resulted insome nonspecific staining associated with endogenousperoxidase activity and transmembrane leakage. Neverthe-less, HRP coupled with TEM was a useful tracer foridentification and localization of sensory plexi exterior tothe basement membrane.

Two morphological sensory-cell types were discerned bybackfilling: (1) intraepithelial sensory cells (ISCs, Fig.3A–D,F) that lie entirely within the epithelium of rhino-phores or oral tentacles and have sensory dendrites 15- to

42 D.Y. BOUDKO ET AL.

25-µm long; and (2) subepithelial sensory cells whosesomata often were concentrated beneath the basal lamina,in the hemocoel of rhinophores, and have 40- to 100-µm-long fine dendrites extending to the surface of the epithe-lium (subepithelial sensory cells, SSCs; Fig. 3G, H).

Labeling sensory cells with DASPEI(centripetal tracing)

Different populations of cells in theASO epithelia staineddifferentially with DASPEI according to the duration ofapplication. Ampullar sensory cells were the first struc-tures labeled when a DASPEI solution contacted theepithelia (Fig. 4). Sensory dendrites of these cells werecolored within 20–60 seconds and sensory cell somatawithin 60–120 seconds. Epithelial cells bearing motile-cilia were colored within approximately 120 seconds, but

less intensely than the sensory cells. After 5 minutes ofapplication, supporting and motile-cilia cells were in-tensely labeled and background staining increased dramati-cally, making it difficult to identify sensory terminals.Different concentrations of the dye had different labelingkinetics. The concentration of DASPEI providing the bestcontrast in labeling of the sensory cells was about 0.1% inFSW.

The uptake of DASPEI by the different types of sensorycells was similar. After 60 seconds of exposure to DASPEI,ISCs, and sensory plexi formed by their basal extensionswere clearly visible (n 5 25, Fig. 4B, left and middle). Inthe experiments with local DASPEI application (n 5 7),the dye diffused rapidly into basal processes of ISCs andstained ISCs over 200–500 µm distant from the applica-tion point. At these distances, however, the labeling se-quence was inverted; the somata labeled first, followed bythe dendrites. Retrograde diffusion of DASPEI into proxi-mal parts of sensory pathways required 20–30 minutes(Fig. 4 B). Confocal microscopy of DASPEI-stained prepa-rations produced high-resolution (603) images of den-drites, somata, and axons of SSCs in the medial third partof rhinophores (n 5 8; Fig. 4C).

The number and distribution of ISCs and SSCs observedwith bath-applied DASPEI were comparable to thoseobtained after long-term backfilling with biocytin (.12hours). In addition, DASPEI stained sensory cells at thetips of the rhinophores and oral tentacles that did notbackfill with LY or biocytin (Fig. 4A). Tufts of inflexiblecilia, abundant at the tips of the tentacles and rhino-phores, were associated with the DASPEI-stained den-drites of these cells (n 5 23). The distribution and morphol-ogy of these sensory cells bearing tufts of stiff cilia (TSCs)distinguished them from both SSCs and ISCs (Fig. 5).

Distribution and afferent organization ofidentified sensory cells

Each type of identified sensory cells is characterized by aspecific organization of afferent pathways and distributionin the ASO (summarized at Fig. 5A,B). The greatestdensities of biocytin-backfilled ISCs were found in thedorsal sensory fields of the oral tentacles (approximately1,200/mm2), the anterior sensory fields of the medial andbasal portions of the rhinophores (about 700/mm2), and thedorsal margin of the cephalic shield (about 800/mm2). Lowdensities of ISCs, uniformly distributed on the ventralfield of the oral disk, were backfilled through the OtN(,20/mm2). Backfilling into the rhinophoral sensory cellswas slower through the RhG (i.e., proximal to the RhG)than directly into the RhN (i.e., distal to the RhG);however, after backfilling for 24 hours, equivalent patternsof cell distribution and density were observed.

An extensive basal plexus (located near the basal lamina)formed by basal extensions of ISCs was observed in dorsalfields of the ASO (Fig. 3B,C). Additionally, branches ofsome ISCs penetrate the basal lamina and join to formafferent pathways; the proportion of ISCs with centripetalprocesses is higher in the oral tentacles (1:10) than in therhinophores (approximately, 1:30). In contrast, every ISCin the ventral fields of the ASO bears processes thatpenetrate the basal membrane and form fine plexi insidethe hemocoel (Fig. 3F).

The highest density of SSCs was found in the medialthird of each rhinophore. In contrast to the relativelyuniform distribution of ISCs, the SSCs are clustered in

Fig. 2. A: Diagrammatic reconstruction of sensory pathways in theanterior sensory organs (ASO) of P. sibogae. CG, cerebral portions ofcerebropleural ganglion; Cs, cephalic shield; E, eye; Es, esophagus; Ot,oral tentacle; OtN, large oral tentacle nerve; Rh, rhinophore; RhG,rhinophoral ganglion; RhN, rhinophoral nerve. B: Rate of biocytin andLucifer Yellow (LY) backfilling into anterior sensory pathways. PL,primary labeling through afferent axons with both biocytin and LY;SL, secondary labeling with biocytin, but not LY. Error bars indicatestandard error.

SENSORY PATHWAYS IN A NUDIBRANCH 43

Fig. 3. Biocytin backfilling of anterior sensory pathways of P. sibogaerecorded by confocal microscopy (refer to the inset in G for sites inA–H). A:Cluster of intraepithelial sensory cells (ISCs) in the basal-lateral edge of anoral tentacle (whole-mount); arrowheads indicate axonal extensions ofISCs; bidirectional arrow outlines the boundaries of the epithelium; dottedline denotes outer surface. B: Cluster of ISCs on the basal-medial edge ofan oral tentacle showing the plexus formed by basal extensions of ISCs(whole-mount), the faint images in the upper part of the photo are the edgeof cephalic shield. C: ISCs in the rhinophore (whole-mount); asterisksindicate dendrites; arrowheads point to basal extensions of ISCs; bidirec-tional arrow outlines the boundaries of the epithelium; dotted line denotesouter surface. D: The pattern of distribution of ISCs in basal areas of therhinophores (whole-mount). E:Axons (arrowheads) in the proximal portion

of the sensory pathway of a rhinophore (frozen section). F: Sensory cell(large arrow) and sensory plexus (arrowheads) in the ventral surface of anoral tentacle (frozen section); asterisk indicates outer surface of the sensoryepithelium. G: Group of subepithelial sensory cells (SSCs, arrows) in themedial portion of a rhinophore (frozen section); asterisks indicate den-drites; arrowheads point to axons; bidirectional arrow outlines the bound-aries of the epithelial layer. Inset: diagram showing approximate locationsof patches of sensory epithelia displayed in A–H. H: SSCs (arrows) withaxonal branches (small arrowheads), which pass into the dense entangle-ment of nerve fibers (large arrowhead); asterisks indicate dendrites;bidirectional arrow outlines the boundaries of the epithelial layer (frozensection). Scale bars 5 25 µm inA–H.

44 D.Y. BOUDKO ET AL.

patches containing 50–100 cells. The somata of SSCs insome areas appear to be layered beneath the basal lamina.The estimated mean density of SSCs in the middle third ofthe rhinophores was about 950/mm2. However, local cellu-

lar densities are higher in some clusters (e.g., the esti-mated density of SSCs in the cluster shown in Figure 3Hwas .4,000/mm2). Clusters of SSCs were not revealed bybackfilling in other areas of the ASO, although individual

Fig. 4. Sensory cells in the anterior sensory organ epithelia of P.sibogae stained in vitro with DASPEI. A: Three fluorescence photomi-crographs displaying progressive uptake (left to right) of DASPEI bysensory terminals after 20, 40, and 60 seconds of exposure to a 0.1%DASPEI solution applied at the tip of the rhinophore. Inset: Greatermagnification of a subepithelial sensory cell with tuft of stiff cilia(TSC) indicated by arrow in the right-hand photograph. B: Threefluorescence photomicrographs showing sensory pathways in a rhino-phore bathed in 0.1% DASPEI for 1 minute, washed, and thenmaintained in the dark for 2 hours before examination. Left, cluster of

intraepithelial sensory cells (ISCs); middle, sensory cells in a basalportion of a rhinophore; right, DASPEI fluorescence in proximalregions of a rhinophoral nerve. C: Confocal photomicrographs ofDASPEI-stained subepithelial sensory cells (SSCs) (distal third of arhinophore, slightly compressed beneath a cover slip in vitro prepara-tion). Left, montage of three frames about 25 µm below the epithelialsurface; right, montage of three frames about 35 µm beneath theepithelial surface. Arrows point to cell bodies of SSCs, small arrow-heads point to the sensory axons of these cells, and asterisks indicatesensory dendrites. Scale bars 5 25 µm.

SENSORY PATHWAYS IN A NUDIBRANCH 45

cells similar to the SSCs in the rhinophores were detectedon the edge of the cephalic shield and the middle dorsalparts of oral tentacles. All identified SSCs were clearlybipolar with a single sensory dendrite and a centripetalaxon. Afferent pathways from several SSC groups appearto converge into glomerulus-like dense entanglements ofnerve fibers (Fig. 3H) as revealed in the rhinophores byboth LY and biocytin backfilling. However, in contrast toLY, biocytin backfilled large numbers of sensory cells distalto the glomerulus-like structures.

The highest densities of TSCs were found in the tips ofrhinophores and oral tentacles (<100 cells/mm2 in adults,

based on counting tufts of cilia and DASPEI-stained TSCs,n 5 12; Fig. 5). Tufts of stiff cilia were also observed alongthe anterior edge (<20 cells/mm2) of the oral plate andinfrequently in other areas of the ASO (,10 cells/mm2).TSCs never backfilled through a RhN or OtN. However,short, fine axons of TSCs were visualized with DASPEI ina few preparations of the ASO (n 5 3).

Ultrastructure of cells in the ASO epithelium

Many cell types were discerned in the rhinophoralepithelium at the ultrastructural level. The most abun-

Fig. 5. A: The relative distribution of intraepithelial sensory cells (ISCs) and subepithelial sensorycells (SSCs) revealed by backfilling with biocytin. B: The spatial distribution of populations of ISCs, SSCs,and subepithelial sensory cell with tuft of stiff cilia (TSC) in the anterior sensory organs of P. sibogae afterlabeling with different neuronal tracers. Error bars indicate standard error.

46 D.Y. BOUDKO ET AL.

dant epithelial cells are characterized by dense brushborders of microvilli on their apical surfaces, large vacu-oles with amphidiskoidal interior structures, and heteroge-neously granulated nuclei lying in their basal region.Abundant rough endoplasmic reticulum, Golgi bodies, andlipid vesicles in these cells are indicative of intense meta-bolic and secretory activity. Typically, just a few, small(,0.5 µm in the long axis) oval mitochondria are present.Cellular interactions between supporting cells and otherepithelial cells of different types apparently are limited byseptate desmosomes.

Cells bearing active cilia were also abundant in the ASOepithelia. These cells occur at a very low density in siteswhere TSCs and SSCs are found. Epithelial cells withactive cilia are characterized by a long (<20 µm) tuft of30–100 cilia, each with a typical 9-plus-2 microtubularcore. The nuclei of these cells are centrally located andsurrounded by rough endoplasmic reticulum. Large (,2.5µm) spherical mitochondria and numerous lipid vesiclestypically occur in their basal portions. Basal bodies of thecilia and a dense layer of microfilaments are seen in theapical regions. In contrast to the supporting cells, theactive-cilia cells lack large vacuoles. All of the sensory-celltypes recognized in the epithelia of the ASO are readilydistinguished at the ultrastructural level from the twoclasses of epithelial cells described above.

Intraepithelial sensory cells

The ISCs are distinguished by having their surfacescovered by microvilli fused to form swollen, vesicle-likestructures with randomly arranged microtubules (Fig. 6).The few cilia (0–6) that extend from the ISC dendrites varyfrom 2 to10 µm in length (Fig. 6C–F). The longest cilia are,in some instances, drawn into thin extensions, similar indiameter to microvilli. The dendrites of ISCs vary inthickness and contain dense arrays of longitudinal micro-filaments that pass into the basal regions of the cells (Fig.6A). Elongate mitochondria lie between the microfila-ments, and spherical mitochondria are found in basalregions of the cells. Backfilling with biocytin revealedbasal plexi between the proximal regions of these cells(Fig. 3C), and ultrastructural examination of HRP-backfilled ASO (n 5 8) showed that the basal plexi of theISCs are located outside the basal lamina (Fig. 6B).

Subepithelial sensory cells

It was difficult to obtain a complete image of a SSC in asingle, thin section for TEM. Fine dendritic endings andsubepithelial somata of SSCs were identified in specificareas after backfilling the RhN with HRP. The surface andinterior ultrastructure of such endings are very similar tothe ultrastructure of thin ISC dendrites. However, unlikethe dendrites of ISCs, the dendrites of the SSCs are ofuniform thickness.

Sensory cells with tufts of stiff cilia

Ultrastructural examination of rhinophores revealedthat the somata of most TSCs are so tightly packed intosmall subepithelial clusters that it is difficult to identifyseparate cells except for their distinctive nuclei (Fig.7A,D). However, in TEM serial sections, observed at highmagnification, it could be seen that the TSC clustersusually include three to five cell bodies encapsulated by aconnective tissue membrane. The few cilia (three to six)that originate from the surfaces of TSC dendrites are

slightly larger in diameter than the motile-cilia on thesensory tentacles and are characterized by a rough sur-face. TSC dendrites contain elongate striated ciliary root-lets that are much longer than the ciliary rootlets in otherciliated cells of the ASO epithelia (Fig. 7B). Dendrites ofthe TSCs are also filled with microtubules and microfila-ments, and typically include large elongate mitochondria(Fig. 7B). The cytoplasm of the somata of the TSCscontains numerous small spherical mitochondria. At leasttwo types of synaptic vesicles were identified in theneuropile of TSC clusters: (1) spherical, electron-transpar-ent synaptic vesicles 15–20 nm in diameter, and (2) large,electron-dense microvesicles 50–70 nm in diameter (Fig.7C). The sensory cilia of each TSC form inflexible tufts(Fig. 7E).

DISCUSSION

Anterior sensory organs

The morphology of the ASO varies among the nudi-branchs, with most of the variation lying in the degree ofdevelopment and the specific morphology of the oraltentacles and rhinophores. Enlargement of these tentaclepairs produces an increase in the sizes of sensory fields andan increase in the number of sensory-receptor cells, allow-ing better resolution of chemical gradients. Thus, it may bethat the relative size of each pair of tentacles reflects theselective pressure on the evolution of each of these chemo-sensory structures and indicates their relative sensorysignificance.

In most aeolid nudibranchs, including P. sibogae, bothoral tentacles and rhinophores are prominent (Fig. 1A),whereas in degenerate forms like Pseudovermis both oraltentacles and rhinophores are absent. In dorid nudi-branchs, oral tentacles are small or absent, but rhino-phores are well developed. The oral veil of dendronotidslike Tritonia diomedea is highly sensory (Willows, 1978)and appears to be homologous to the combined areas wehave termed ‘‘oral-labial disk’’ and ‘‘cephalic shield’’ in P.sibogae, which are also rich in sensory cells (Fig. 5B). Thissame area is spread into the large, flattened oral tentaclesof Aplysia spp., which play major chemosensory roles inthese animals (Jahan-Parwar, 1972).

Anterior sensory cells

Sensory cells resembling the ISCs, SSCs, and TSCs of P.sibogae have been widely reported in gastropods (seereview by Emery, 1992). In nudibranchs, subepithelialsensory cells are found in the rhinophores of larval Ros-tanga pulchra (Chia and Koss, 1982) and associated withrepugnatorial glands in the cerata of Melibe leonina (Bick-ell-Page, 1991). However, SSCs apparently do not occur inthe ASO of some adult nudibranchs (Storch and Welsch,1969) or the notaspidean opisthobranch Pleurobranchaeacalifornica (Bicker et al., 1982b). The present studiesrevealed two different types of peripheral sensory cellswith subepithelial somata in the ASO of P. sibogae, re-ferred to here as SSCs and TSCs.

The SSCs are abundant in the middle of the rhinophores(local cellular densities . 4,000/mm2) and rare in otherparts of the ASO. The morphological organization of theSSCs appears to be similar to that of receptors in theolfactory tentacles of the terrestrial snail Achatina fulica,where some sensory axons converge into glomerular struc-tures beneath the sensory epithelium (Chase and Tolloc-

SENSORY PATHWAYS IN A NUDIBRANCH 47

Fig. 6. Electron photomicrographs of intraepithelial sensory cells(ISCs) of the rhinophores. A: Transmission electron photomicrograph(TEM) of two ISCs. B: TEM of a rhinophore backfilled with horserad-ish peroxidase (HRP) and treated with diaminobenzidine (DAB). Thedark, HRP-DRB reaction product is localized in basal extensions ofISCs (black arrows), exterior to the basal lamina (arrowheads). Thelarge open arrow indicates direction toward the external surface of the

epithelium. C–F: scanning electron photomicrographs of the surfacesof rhinophores showing clumps of cilia on motile-cilia cells andvariations of the external dendritic surfaces of ISCs (arrows). Bl, basallamina; Ci, cilia; D, dendrite; Ex, exterior of cell; In, interior side ofcell; Mi, mitochondria; Mt, microtubules; Mv, microvilli; N, nucleus; V,vacuoles. Scale bars 5 5 µm in A,D-F; 2.5 µm in B; 10 µm in C.

Fig. 7. Electron photomicrographs of subepithelial sensory cells withtufts of stiff cilia (TSCs) in the tips of the rhinophores. A: Transmissionelectron photomicrograph (TEM) showing TSC somata below the basallamina and a dendrite extending between the epithelial support cells to thesurface. B: TEM of the distal end of a TSC dendrite with ciliary basalbodies, long ciliary rootlets, and elongate mitochondria. C: Magnified TEMof a TSC cluster showing different synaptic vesicles.Arrows 1 and 3, groupsof small electron transparent vesicles; arrow 2, large granular vesicles.D: Scanning electron photomicrograph (SEM) of fractured rhinophore

showing a TSC cluster; surface of the rhinophore is to the right. E: SEM oftuft of stiff cilia. Bb, basal body of cilium; Bl, basal lamina; Ci, motile cilia;Cr, ciliary rootlets; Ct, tuft of stiff cilia; D, dendrite; M, mitochondria; Mi,long mitochondrion in sensory dendrites; Mv, microvilli; Mt, microtubules(neurotubules) in dendrite; Nf, nerve fibers; S, cluster of TSC somata (see asection through a cluster of TSC somata in 7A); N, nuclei; Nf, nerve fibers;Os, osmiophilic inclusions; V, vacuole of support cell. Scale bars 5 10 µm inA,D; 2.5 µm in B; 1.25 µm in C; 5.0 µm in E.

zko, 1986). In P. sibogae, the sensitivity of the rhinophoresto natural scents is approximately 1,000-fold higher com-pared with the oral tentacles and the oral-disk and cephalic-shield areas (Budko and Hadfield, 1995; Boudko et al.,1997; Murphy and Hadfield, 1997). We conclude that SSCsare olfactory receptors, because they are the most abun-dant sensory-cell type in the rhinophores of P. sibogae, theorgans considered the major olfactory structures of nudi-branchs based on behavioral and electrophysiological data(Willows, 1978; Murphy and Hadfield, 1997, and personalobservations). However, the importance of the differentgastropodASO structures in olfaction varies (Emery, 1992),and our findings on the spatial distribution of olfactoryreceptor cells in P. sibogae are not universal. For example,the anterior (oral) tentacles of A. californica are mostsensitive to food odors and perhaps serve in olfaction inparallel with the rhinophores (Frings and Frings, 1965;Jahan-Parwar, 1972; Audesirk, 1975). Subepithelial recep-tor cells similar to the SSCs of P. sibogae are abundant inboth pairs of tentacles in A. californica (Emery andAudesirk, 1978).

The second type of receptors with subepithelial cellbodies, TSCs, those bearing tufts of stiff-cilia, are abun-dant in the tips of the tentacles (<100/mm2) in P. sibogae.The structure and distributions of TSCs make them primecandidates to be rheoreceptors; no alternative structureslikely to serve this role were found. The probability of arheotactic role for the TSCs is supported by behavioral andelectrophysiological data that indicate most sea slugs areable to determine flow directions and lose that ability afterremoval of rhinophores or severing of the anterior nerves(Field and Macmillan, 1973; Willows, 1978; Murray andWillows, 1996). Individuals of P. sibogae display upstreamorientation, a behavior that disappears if the tentacles’tips are removed (personal observations). Afferent activityin both rhinophoral and oral tentacle nerves of P. sibogae isaffected by currents around the ASO (Boudko et al., 1997).Axons of the TSCs stained with DASPEI but were neverbackfilled through anterior nerves, suggesting that thesecells lack direct projections into the brain. Perhaps, theysynapse with interneurons that extend to the brain. Thepresence of synaptic vesicles in the interiors of the neuro-piles of TSC clusters (Fig. 7C) suggests efferent or recipro-cal synaptic regulation of these cells.

The structure of intraepithelial receptors varies amonggastropods (reviewed by Dorsett, 1986; Emery, 1992). TheISCs found in adult P. sibogae differ from the disco-ciliacells in the ASO of P. californica (Matera and Davis, 1982;Davis and Matera, 1982) and the ampullar cells in theapical sensory organ of larvae of P. sibogae (Bonar, 1978).They are similar to ISCs found in the skin of pulmonates(Zylstra, 1972), in prosobranch osphradia (Haszprunar,1985; Nezlin et al., 1994), in tentacles of two sea harespecies (Emery and Audesirk, 1978), and in rhinophores ofsome other nudibranch species (Storch and Welsch, 1969).For most of these animals, the centripetal connections ofISCs remain unclear. However, this is not the case for P.sibogae, in which the ISCs were backfilled with neuronaltracers, suggesting direct projection of their axons into theCNS. ISCs were most abundant in the dorsal surface of theoral tentacles (1,200/mm2). The oral tentacles in P. sibogaeare less chemosensitive than the rhinophores and areimportant for trail following, food finding, and courtshipbehaviors, all of which involve detection of signal mol-ecules near their sources (Murphy and Hadfield, 1997;

Boudko et al., 1997; personal observations). Because ISCsare the major sensory cells of the oral tentacles of P.sibogae, we conclude that the ISCs function in contact orshort-distance chemoreception.

Anterior afferent pathways

The data provided here indicate that some but not allperipheral sensory cells project to the brain of P. sibogae.This finding is consistent with observations on othergastropod molluscs (Bicker et al., 1982b; Chase and Kamil,1983; Yi and Emery, 1991). However, differences do occur.Centrifugal backfilling of the RhN proximal and distal tothe RhG displayed similar patterns and densities of sen-sory cells, suggesting that afferent pathways generallycross the RhG. In contrast, anterior sensory pathways in P.californica mostly terminate in rhinophoral and tentacu-lar ganglia (Bicker et al., 1982b). Thus, the role of RhG inP. sibogae may be efferent modulation of olfactory input,rather than processing of olfactory information, as in P.californica (Bicker et al., 1982b).

Larger numbers of sensory cells were backfilled withbiocytin than with LY, suggesting that some parts of theperipheral sensory network are selectively permeable tobiocytin and not to LY. Selective cell-cell permeability tobiocytin, and not LY, has been reported to occur betweenstrongly electrically coupled neurons in the CNS of somegastropod molluscs (Audesirk et al, 1982; Ewadinger et al.,1994). The data for P. sibogae presented in Figure 2Bstrongly suggest that backfilling with biocytin combinestwo processes, primary and secondary labeling, whichdiffer in their rates and points of origin; with LY, onlyprimary labeling occurred. In primary labeling, only nervecells whose axons are drawn into the backfilling pipetteare filled, whereas in secondary labeling the tracer crossescell-cell junctions (gap junctions or electrical synapse) andlabels other cells. If this distinction is valid, then all of thesensory cells labeled by both of these tracers have primaryafferents in the brains, whereas the cells that backfilledonly by biocytin are electrically coupled to the primarysensory cells. This hypothesis is supported by electrophysi-ologic evidence, not presented here, for type-specific andstrong (20–40%) electrical coupling of sensory cells in vitro(personal observations).

Two chemosensory systems of gastropodmolluscs

The function of different receptors in gastropod ASOshas been the subject of much controversy, but someconclusions can be drawn. Intraepithelial receptors arewidespread in the chemoreceptive structures of primitivegastropods, whereas subepithelial receptors prevail in thechemosensory structures of more derived gastropods (re-viewed by Emery, 1992). ISCs seem to serve primitivechemosensory systems, whereas advanced olfactory sys-tems are based on SSCs. The presence of SSCs in modernsea slugs indicates that SSCs did not arise as an adapta-tion to terrestrial or freshwater habitats. Close packingand vertical ‘‘stacking’’ of sensory-cell bodies allows a largeincrease in sensory-cell density and produces a pseu-dostratified appearance in the olfactory epithelia of verte-brates (Kessel and Kardon, 1979). Such a pattern, how-ever, is impossible in the single-layered epithelium ofgastropod molluscs, in which densities of ISCs are re-stricted by their size and the pressure of spatial competi-tion from other epithelial cells. Subepithelial displacement

50 D.Y. BOUDKO ET AL.

of sensory somata may be an adaptation, which hasallowed an overall increase the number of sensory cells.

The present investigations revealed that populations ofboth ISCs and SSCs occur in ASO epithelia of P. sibogae,but with significant variation in their spatial densities.Although both sensory cell types occur in all of the anteriorsensory organs, ISCs are predominantly associated withthe putative contact/short-distance chemosensory struc-tures (oral tentacles), and SSCs are most abundantlyfound in the olfactory structures (rhinophores). Thesefindings, plus observations in the literature lead us toconclude that two chemosensory subsystems, based onISCs and SSCs, serve different chemosensory modalities(contact/short distance chemoreception and olfaction) ingastropods. These subsystems may be present in specificproportions and spatial patterns in the different chemosen-sory structures of different species. Populations of ISCsand SSCs in the sensory tentacles of P. sibogae aredistinguished by their spatial distribution, axonal organi-zation, and sensitivity to natural scents; however, theyshare remarkable electrophysiological and pharmacologi-cal similarities (Boudko et al., 1997). Emery (1992) consid-ered ISCs ‘‘the simplest and probably most primitivesensory neuron...’’ of gastropods (p. 308), a conjecture, that,if true, suggests that the oral tentacles and contact orshort-distance chemoreception are the more primitiveorgans and functions in sea slugs like P. sibogae. Additionof the SSC-laden rhinophores would then represent a laterevolutionary step, as has been suggested by others (Rud-man and Willan, 1998). However, the near ubiquity ofcephalic rhinophores, as well as oral tentacles, across mostopisthobranch orders (only some members of the orderCephalaspidea lack them [Thompson, 1976; Emery, 1992;Rudman and Willan, 1998]) indicates that this evolution-ary step must have been taken in ancestors that gave riseto most of the existing orders. Most authors accept thatthese ancestors were cephalaspideans (e.g., Schmekel,1985; Rudman and Willan, 1998). It thus remains to bediscovered whether ISCs are the major sensory cells foundin the Hancock’s organs, which lie on the lateral areas ofthe cephalic shield of many cephalaspideans and thus maybe the evolutionary sources of oral tentacles, and whetherSSCs occur on the heads of tentacle-less cephalaspideansand on the rhinophores of the cephalaspideans that havethem.

ACKNOWLEDGMENTS

We thank Brendan Yee and John R. Jordan of BioRad,Inc., for their assistance in obtaining confocal imagesduring a demonstration of the BioRad 1024 confocalsystem at the Kewalo Marine Laboratory. Drs. Eric Holmand Roger Croll provided valuable comments on variousdrafts of this manuscript, and colleagues at the KewaloMarine Laboratory rendered invaluable assistance in main-taining the stock cultures of P. sibogae. We thank EllaMeleshkevitch for technical assistance in the backfillingand tissue culture experiments and Sofia Boudko forassistance in preparing Figures 2A and 5B. This researchwas supported by Office of Naval Research Grant N00014-1–0524 to M.G.H. and by National Institutes of HealthGrant RR03061 to the Biological EM Facility.

LITERATURE CITED

Arey LB. 1918. The multiple sensory activities of the so-called rhinophoresof nudibranchs. Am J Physiol 46:526–532.

Audesirk TE. 1975. Chemoreception in Aplysia californica. I. Behaviorallocalization of distance chemoreceptors used in food finding. Behav Biol15:45–55.

Audesirk G, Audesirk TE, Bowsher P. 1982. Variability and frequent failureof Lucifer yellow to pass between two electrically coupled neurons inLymnaea stagnalis. J Neurobiol 13:369–375.

Balak KJ, Corwin JT, Jones JE. 1990. Regenerated hair cells can originatefrom supporting cell progeny: Evidence from phototoxicity and laserablation experiments in the lateral line system. J Neurosci 10:2502–2512.

Bickell-Page LR. 1991. Repugnatorial glands with associated striatedmuscle and sensory cells in Melibe leonina (Mollusca, Nudibranchia).Zoomorphology 110:281–291.

Bicker G, Davis WJ, Matera EM, Kovac MP, Stormo-Gipson DJ. 1982a.Chemoreception and mechanoreception in the gastropod mollusc Pleu-robranchaea californica. I. Extracellular analysis of afferent pathways.J Comp Physiol 149:221–234.

Bicker G, Davis WJ, Matera EM, Kovac MP, Stormo-Gipson DJ. 1982b.Chemoreception and mechanoreception in the gastropod mollusc Pleu-robranchaea californica. II. Neuroanatomical and intracellular analy-sis of afferent pathways. J Comp Physiol 149:235–250.

Bonar DB. 1978. Ultrastructure of a cephalic sensory organ in the larvae ofthe gastropod Phestilla sibogae (Aeolidacea, Nudibranchia). Tissue Cell10:153–165.

Boudko DY, Switzer-Dunlap M, Hadfield MG. 1997. Morphology, electro-physiology and pharmacology of anterior sensory pathways in thenudibranch mollusc, Phestilla sibogae. Soc Neurosci Abstr 23:1787.

Budko DY, Hadfield MG. 1995. Pharmacological analysis of chemoreceptionin the nudibranch Phestilla sibogae. Amer Zool 35:27A.

Chase R. 1982. The olfactory sensitivity of the snail, Achatina fulica. JComp Physiol 148A:225–235.

Chase R, Kamil R. 1983. Neuronal elements in snail tentacles as revealedby horseradish peroxidase backfilling. J Neurobiol 14:29–42.

Chase R, Tolloczko B. 1986. Synaptic glomeruli in the olfactory system of asnail, Achatina fulica. Cell Tissue Res 246:567–571.

Chase R, Tolloczko B. 1993. Tracing neural pathways in snail olfaction:From the tip of the tentacles to the brain and beyond. Microsc Res Tech24:214–230.

Chia F-S, Koss R. 1982. Fine structure of the larval rhinophores of thenudibranch Rostanga pulchra, with emphasis on the sensory receptorcells. Cell Tissue Res 225:235–248.

Crisp M. 1973. Fine structure of some prosobranch osphradia. Mar Biol22:231–240.

Davis WJ, Matera EM. 1982. Chemoreception in gastropod molluscs:Electron microscopy of putative receptor cells. J Neurobiol 13:79–84.

Dickinson P. 1979. Homologous neurons control movements of diverse gilltypes in nudibranch molluscs. J Comp Physiol 131:277–283.

Dorsett DA. 1986. Brains to cells: The neuroanatomy of selected gastropodspecies. VI. Epithelial receptor. In KM Wilbur and AOD Willows (eds):The Mollusca. Neurobiology and Behavior. Orlando: Academic Press,pp.150–187.

Eisenman E, Alfert M. 1982. A new fixation procedure for preserving theultrastructure of marine invertebrate tissues. J Microsc 125:117–120.

Emery DJ. 1992. Fine structure of olfactory epithelia of gastropod molluscs.Microsc Res Tech 22:307–324.

Emery DG, Audesirk TE. 1978. Sensory cells in Aplysia. J Neurobiol9:173–179.

Ewadinger N, Syed N, Lukowiak K, Bulloch A. 1994. Differential tracercoupling between pairs of identified neurones of the mollusc Lymnaeastagnalis. J Exp Biol 192:291–297.

Field LH, Macmillan DL. 1973. An electrophysiological and behavioralstudy of sensory responses in Tritonia (Gastropoda, Nudibranchia).Mar Behav Physiol 2:171–185.

Frings H, Frings C. 1965. Chemosensory bases of food-finding and feedingin Aplysia juliana (Mollusca, Opisthobranchia). Biol Bull 128:211–217.

Gelperin A. 1974. Olfactory basis of homing behavior in the giant gardenslug, Limax maximus. Proc Natl Acad Sci USA 71:966–970.

SENSORY PATHWAYS IN A NUDIBRANCH 51

Gurin S, Carr WE. 1971. Chemoreception in Nassarius obsoletus: The roleof specific stimulatory proteins. Science 174:293–295.

Hadfield MG, Pennington JT. 1990. Nature of the metamorphic signal andits internal transduction in larvae of the nudibranch, Phestilla sibogae.Bull Mar Sci 46:455–464.

Hadfield MG, Scheuer D. 1985. Evidence for a soluble metamorphic inducerin Phestilla sibogae: Ecological, chemical, and biological data. Bull MarSci 37:556–566.

Harris LG. 1971. Nudibranch associations as symbioses. In TC Cheng (ed.):Aspects of the Biology of Symbiosis. London: University Park PressEdition, pp. 77–90.

Harris LG. 1975. Studies on the life history of two coral-eating nudibranchsof the genus Phestilla. Biol Bull 149:539–550.

Haszprunar G. 1985. The fine morphology of the osphradial sense organs ofthe Mollusca: I. Gastropoda, Prosobranchia. Philos Trans R Soc Lond BBiol Sci 307:457–496.

Haszprunar G. 1987. The fine morphology of the osphradial sense organs ofthe Mollusca: III. Placophora and Bivalvia. Philos Trans R Soc Lond BBiol Sci 307:37–61.

Jahan-Parwar B. 1972. Behavioral and electrophysiological studies onchemoreception in Aplysia. Am Zool 12:27–537.

Jahan-Parwar B. 1975. Chemoreception in gastropods, In VDA Denton andJP Coghlan (eds): Olfaction and Taste, Vol. V. New York: AcademicPress, pp. 133–140.

Kataoka S. 1976. Fine structure of the epidermis of the optic tentacle in aslug, Limax flavus L Tissue Cell 8:47–60.

Kessel R, Kardon RH. 1979. Tissues and Organs: A Text-Atlas of ScanningElectron Microscopy. San Francisco: W.H. Freeman and Co, pp. 129–134.

Kunkel DD, Scharfman HE, Schmiege DL, Schwartzkroin PA. 1993.Electron microscopy of intracellularly labeled neurons in the hippocam-pal slice preparation. Microsc Res Tech 24:67–84.

Leise EL. 1996. Selective retention of the fluorescent dye DASPEI in alarval gastropod mollusc after paraformaldehyde fixation. Microsc ResTech 33:496–500.

Matera EM, Davis WJ. 1982. Paddle cilia (discocilia) in chemosensitivestructures of the gastropod mollusk Pleurobranchaea californica. CellTissue Res 222:25–40.

Merton H. 1920. Untersuchungen uber die Hautssinnesograine der Mol-lusken. I. Opisthobranchia. Abh Senckenb Naturforsch Ges 36:447–473.

Miller SE, Hadfield MG. 1986. Ontogeny of phototaxis and metamorphiccompetence in larvae of the nudibranch Phestilla sibogae Bergh (Gas-tropoda: Opisthobranchia). J Exp Mar Biol Ecol 97:95–112.

Murphy BF, Hadfield MG. 1997. Chemoreception in the nudibranchgastropod Phestilla sibogae. Comp Biochem Physiol 118A:727–735.

Murray JA, Willows AOD. 1996. Function of identified nerves in orientationto water flow in Tritonia diomedea. J Comp Physiol 176A:201–209.

Nezlin LP, Elofsson R, Sakharov DA. 1994. Transmitter-specific subsets ofsensory elements in the prosobranch osphradium. Biol Bull 187:174–184.

Nurse CA, Farraway L. 1989. Characterization of Merkel cells and mecha-nosensory axons of the rat by styryl Pyridium dyes. Cell Tissue Res255:125–128.

Rudman WB, Willan RC. 1998. Introduction to Opisthobranchia. In:Mollusca, The Southern Synthesis: Part B. Fauna of Australia, Vol. 5.Canberra: Australian Biological Resources Study, Australia. CSIRO.pp. 915–942.

Schmekel L. 1985. Aspects of evolution within the opisthobranchs. In ERTrueman and MR Clarke (eds): The Mollusca, Vol. 10, Evolution. NewYork: Academic Press, pp. 221–267.

Storch V, Welsch U. 1969. Uber bau und funktion der nudibranchier-rhinophoren. Z Zellforsch 97:528–536.

Thompson TE. 1976. Biology of Opisthobranch Molluscs, Vol. 1. London:Ray Society, pp. 1–206.

Willows AOD. 1978. Physiology and feeding in Tritonia: I. Behaviour andmechanics. Mar Behav Physiol 5:115–135.

Willows AOD. 1985. Neural control of behavioral responses in the nudi-branch mollusc Phestilla sibogae. J Neurobiol 16:157–170.

Wright BR. 1974. Sensory structure of the tentacles of the slug, Arion ater(Pulmonata, Mollusca): 1. Ultrastructure of the distal epithelium,receptor cells and tentacular ganglion. Cell Tissue Res 151:229–244.

Yi H, Emery DG. 1991. Histology and ultrastructure of olfactory organ ofthe freshwater pulmonate, Helisoma trivolvis. Cell Tissue Res 265:335–344.

Zaitseva OV, Bocharova LS. 1981. Sensory cells in the head skin of pondsnails. Cell Tissue Res 220:797–807.

Zylstra U. 1972. Distribution and ultrastructure of epidermal sensory cellsin the freshwater snails Lymnaea stagnalis and Biomphalaria pleilleri.Neth J Zool 22:283–298.

52 D.Y. BOUDKO ET AL.