|
Towards a Behavioral Genetics of Zebrafish The zebrafish is a well-characterized genetic organism that is currently being exploited primarily to analyze vertebrate development. The utility of this organism for understanding the genetics of behavior and disorders with behavioral components, such as addiction, sensory deficits, or neurological and psychiatric disorders, is not known because of the lack of behavioral screens. The purpose of this meeting was to bring together investigators who are zebrafish geneticists with investigators who have developed methodologies for studying behavior in a wide variety of fish species. It is anticipated that the outcome will be new genetic screens for mutations in zebrafish that affect behavior. This two-day meeting was held in conjunction with the MBL course "Neural Development and Genetics of Zebrafish" and was attended by about 90 people. The meeting was organized by Susan Volman and Jonathan Pollock, NIH/National Institute on Drug Abuse and sponsored by many of the NIH Institutes involved in the Trans-NIH Zebrafish Initiative with additional assistance from Aquatic Habitats for the poster session. List of Speakers and Poster Presentations LIST OF SPEAKERS AND POSTER PRESENTATIONS
SPEAKER ABSTRACTS
Establishing the Genetic Basis of Vertebrate Oculomotor Behaviors in Zebrafish
The central nervous system of all vertebrate embryos is derived from a series of conspicuous embryonic segments, called neuromeres, that are particularly visible in the mid- and hindbrain areas that give rise to the brainstem sensory and motor nuclei. This presentation focuses on a series of eight hindbrain rhombomeric segments that represent domains of unique gene expression and lineage restriction responsible for the development of equally unique neuronal subgroups producing eye motion. In all vertebrates, columns of vestibular and reticular subnuclei delineated by Hox gene clusters extend throughout this segmented hindbrain. The function of each oculomotor related subgroup is hypothesized to be determined by specific genetic and molecular signaling mechanisms. Establishing the role of single gene loci in the development of physiologically-characterized oculomotor related phenotypes will be addressed based on a genetic analysis of the anatomical framework underlying three dimensional oculomotor behavior in zebrafish. For example, the detailed neuronal map of the circuitry underlying both vestibuloocular and optokinetic reflexes including eye fixation will be shown to consist of subnuclei so precisely segregated within distinct rhombomeric segments that one must conclude that, each "functional" neuronal group is likely produced by spatially restricted expression of developmental regulatory genes. The zebrafish brainstem is, therefore, well suited for a genetic analysis of neuronal circuits responsible for oculomotor behavior. Individual physiological phenotypes identified within unique rhombomeric morphogenetic units will be used to show how a structural and behavioral analysis of single gene mutations in the zebrafish might distinguish single genes correlated to single behaviors. A prospective strategy for identifying single genes that act to determine unique neuronal signal processing elements contributing to a particular behavior will be illustrated by the oculomotor neural integrator which is a model system to investigate the genetic basis of persistent neural activity in vertebrates. In conclusion, the goal is to demonstrate that genes identified by disruption of oculomotor behaviors in zebrafish can provide an efficient way of assigning roles to orthologous human genes as they become recognized by sequence from the human genome project. Neuroethology of social behavior and reproductive plasticity among teleost fish.
Vertebrate social behavior may be characterized by several types of sexual plasticity. Perhaps the most frequently described among these behaviors are alternative mating tactics where one sex, usually males, displays two or more classes of social reproductive behaviors. Among vertebrates, teleost fishes are perhaps the "champions" of alternative tactics, because of the extreme range in reproductive plasticity they exhibit . In some cases, alternative male phenotypes or morphs originate from two distinct life history trajectories. For other species that exhibit adult sex reversal, alternative male tactics represent sequential life history stages for an individual that is initially either a reproductively active male (role change) or female (sex change); yet other sex changing species exhibit a single male reproductive tactic. Lastly, some teleosts may exhibit some form of reproductive suppression where a non-reproductive individual can transform into a reproductive one given the appropriate social conditions. All of these examples represent points on a continuum of sexual plasticity with yet other species exhibiting intermediate conditions. A Behavioral Screen for Zebrafish Circadian Clock Mutants.
A wide variety of biological processes are regulated by circadian clocks. These clocks are systems of cellular oscillators that can self-generate circa-24 hour rhythmicity in the absence of external timing cues. Under natural conditions, the timing of a circadian clock is set by the environmental light:dark cycle. This results in appropriately-timed, daily rhythms in gene expression, metabolism, neural and hormonal activity, and many aspects of behavior. Recently, there has been rapid progress in identification of the molecules that make up circadian clocks, and in our understanding of how these molecules interact to produce circadian rhythmicity. Most of this progress has come, directly or indirectly, from studies of genes that were originally identified by mutations. Clock mutant screens approaching saturation have been performed in cyanobacteria, Arabidopsis, Neurospora, and Drosophila. Screens for clock mutants that affect behavior in vertebrates have been more limited, but have already contributed much to our understanding of clock mechanisms. In each of these systems, the core of the clock mechanism is made up of clock-specific proteins that interact in negative feedback loops to produce rhythms in gene expression. The specific molecules that make up these loops differ across kingdoms. Vertebrate circadian oscillator mechanisms clearly include molecular elements that are evolutionarily conserved with Drosophila. However, the vertebrate system is more complex and differs in several ways from the Drosophila system. Clock mutant screens in zebrafish are likely to identify new vertebrate circadian clock genes, as well as provide insights into the functions of known clock genes. Before zebrafish genetics could be exploited in studies of circadian clock mechanisms, it was necessary to develop efficient measures of circadian rhythmicity that can be used to screen for mutants, as well as information on the organization of the circadian system that can be used to interpret mutant phenotypes. Studies from this laboratory and others on molecular, cellular, and behavioral circadian rhythmicity in zebrafish have begun to fulfill these requirements. The most efficient measure of circadian clock function that we have found in zebrafish is the circadian rhythm in spontaneous swimming activity of 9-18 day old larvae. We measure these rhythms with an automated infrared video image analysis system that tracks the movements of fish housed in individual 0.7 ml wells. With this system, a single video camera can monitor the activity of up to 150 individuals simultaneously for up to a week in constant conditions. Over 95% of larval zebrafish express robust circadian behavioral rhythms under these conditions, with highest activity during the subjective day. For our purposes, the time of the activity peak at the end of a week in constant conditions is the most precise measure of the circadian timing of individual fish. We have also examined locomotor activity rhythms of adult zebrafish, using a recording system based on infrared beam detectors. We observe more variability in the activity patterns of adults than those of larval zebrafish. Patterns range from a single, robust circadian rhythm, to splitting of the activity rhythm into two rhythms with different periods, to complete arrhythmicity. Under the best conditions that we have devised, this system can detect significant circadian rhythmicity in ~70% of adult zebrafish. With this kind of variability, the adult behavioral rhythm would not be useful for screening purposes, but we have found it to be useful for some physiological experiments. Zebrafish pineals, cultured in a flow-through superfusion system in constant darkness, produce exceptionally robust circadian rhythms of melatonin release for at least seven days. This organ, like the pineals of many other non-mammalian vertebrates, is directly photosensitive; exposure to light in vitro suppresses pineal melatonin production and resets the phase of the circadian oscillator. Therefore, the cultured pineal performs all of the basic functions of a circadian clock, and it provides an easy and reproducible physiological assay for cell- and tissue-level rhythmicity. The zebrafish retina also contains a circadian oscillator that regulates melatonin synthesis. Although pineal and retinal circadian clocks contribute to the regulation of behavioral rhythms in other species, we have been unable to detect any effect of pineal or ocular ablation on the swimming rhythms of zebrafish. Therefore, the behavioral rhythm and the melatonin rhythms of cultured pineal and retina are independent measures of circadian clock function in this species. We have begun a screen for ENU-induced mutations that alter the timing of the larval swimming rhythm. Because recording of these activity rhythms is largely automated, it requires relatively little human effort to test a few hundred animals every week. Measurement of circadian rhythmicity is time-consuming, however, and the limiting factor in the screen is the rate at which animals can be tested. Therefore, we chose to screen for dominant mutations, which can be done with a relatively small breeding facility and limited screening capacity. We test the progeny of ENU-treated males, each of which is heterozygous for a unique set of mutations. Individuals with out-of-phase activity peaks after a week in constant conditions are selected and raised, and their progeny are tested to determine whether they are authentic mutants. In a pilot screen of 1275 mutagenized genomes, we identified two semi-dominant mutations that shorten the free running period of the larval behavioral rhythm by 0.5-0.8 h in heterozygotes and 1.0-1.5 h in homozygotes. We have so far confirmed that one of these mutations also shortens the period of the melatonin release rhythms measured from cultured pineal glands, indicating that the mutant gene product affects tissue-level rhythmicity as well as behavior. We have used microsatellite markers to map these mutations, and found that they identify two genes that are located on different linkage groups.
We are now working toward high-resolution mapping and phenotypic characterization of these mutations, and we are mapping candidate cloned genes to determine whether any are linked to our mutations. We are also continuing to screen for new mutants, with the goal of identifying additional alleles of these genes, as well as mutations in other vertebrate clock genes. Our experience so far indicates that this will be a productive approach to identification of vertebrate circadian clock genes. Searching for Visual System Mutations in Zebrafish
Zebrafish are highly visual animals, exhibiting light responses after just 3 days of development, making them ideal for the genetic analysis of visual behavior. They have large eyes and are tetrachromatic, possessing ultraviolet- sensitive cones as well as red-, green -, and blue-sensitive cones. They also have abundant rods, and like other fish, their retinas continue to grow for the life of the animal. Our group has recently developed two behavioral tests that can be used to uncover visual system specific mutations in zebrafish. One test, based on the optokinetic reflex, enables us to isolate recessive visual system mutations in larval fish. The optokinetic response is first evident at 3-4 days; by 5 days of age 98% of wild-type fish will respond optokinetically to a moving stripe pattern with a smooth pursuit eye movement followed by a rapid saccade in the opposite direction. Analysis of the optokinetic responses of larval fish takes on average no more than one minute which includes time spent placing the fish in the test apparatus, aligning and observing them, and returning them to their tank. Thus, one investigator can examine up to 500 larvae per day. The second behavioral test is based on the escape response exhibited by zebrafish when they encounter a threatening object. This test is used to isolate dominant visual system mutations in adult fish. Individual fish are placed in a round container with clear sides that has a post in the middle. Surrounding the container is a rotating drum on which is marked a black segment that serves as the threatening object. The fish will respond positively to the threatening object about 85% of the time when the drum is rotating at a rate such that the fish encounter the threatening object 20-25 times a minute. In other words, with this test we can evaluate whether a fish is seeing the black segment in 5-10 seconds, which enables us not only to measure absolute thresholds for the rod and cone systems, but also to measure the course of dark adaptation of zebrafish. Once an animal with defective vision is identified behaviorally, the next task is to localize the mutation within the visual system. Since we are primarily interested in ocular abnormalities, we use electrophysiological recordings from the eye as secondary screens to localize mutations within the retina. The electroretinogram, a field potential recorded from the surface of the eye, provides information with regard to outer retinal mutations, whereas ganglion cell recordings made from the optic nerve can indicate both outer and inner retinal abnormalities. Once a mutation is localized to a particular retinal locus, singe cell recordings using patch electrodes and retinal slices, can be made. In this talk, I shall describe these techniques and several of the mutants we have isolated. References Brockerhoff, S. E., Hurley, J. B., Janssen-Bienhold, U., Neuhauss, S. C., Driever, W. and Dowling, J. E. A behavioral screen for isolating zebrafish mutants with visual system defects. Proc. Natl. Acad. Sci., 92, 10545-10549, 1995. Brockerhoff, S. E., Hurley, J. B., Niemi, G. A. and Dowling, J. E. A new form of inherited red-blindness in zebrafish. J. Neurosci., 17, 4236-4242, 1997. Li, L. and Dowling, J. E. A dominant form of inherited retinal degeneration caused by a non-photoreceptor cell-specific mutation, Proc. Natl. Acad. Sci., 94, 11645-11650, 1997. Li, L. and Dowling, J. E. Zebrafish visual sensitivity is regulated by a circadian clock, Visual Neurosci., 15, 851-857, 1998. The Sense of Hearing in Fishes: Methodologies and Results
The methods of psychophysics have been formally applied to nonhuman animals for over 75 years. Psychophysics measures the performance of an organism in detecting, discriminating, and otherwise obtaining information from sensory stimuli, and can be used to determine the stimulus values required to produce a given level of behavioral performance. In animal psychophysics, performance is usually measured as the latency, magnitude, or probability of a conditioned response that is signaled or controlled by a sensory stimulus. The focus of psychophysical studies is to determine what the animal can detect and discriminate rather than how the animal normally behaves with respect to the stimuli. Thus, psychophysics can reveal what values and dimensions of a stimulus could be used as information, but it does not predict behavior in the animal’s usual world. This presentation describes some of the conditioning and psychophysical methods that have been used in studies of the sense of hearing of goldfish, that like zebrafish, are otophysan "hearing specialists." Fish detect sound using one or more of the otolith organs (saccule, lagena, and utricle) found in all fish species. These organs contain a patch of hair cell receptors overlain by a solid otolith having a density of about 3. As sound passes through a fish and brings its tissues into motion, the otoliths move at a different amplitude due to their greater inertia. In this way, a relative displacement of the otolith occurs that is in proportion to acoustic particle motion. All otolith organs in all species will tend to respond to sound-induced motions of the fish's body. Otolith organs can respond with reasonable sensitivity to accelerations due to gravity (zero Hz), and to translatory motions up to several hundred Hz. Behavioral and neurophysiological thresholds for whole-body, sinusoidal motions range between 0.1 and 1 nanometer at 100 Hz. In many fish species, including goldfish and zebrafish, the otoliths may also receive a displacement input from the swimbladder, or other gas-filled chamber near the ears. Since motions of the swimbladder wall are created by changes in the bladder's volume as sound pressure fluctuates, this input to the ears is proportional to sound pressure. Thus, many fishes may respond to both acoustic pressure and particle motion. Species having a particularly efficient mechanical coupling between the gasbladder and the otolith organs (e.g., the Weberian ossicles of goldfish and zebrafish) may have very high sensitivity to sound pressure and may hear in a frequency range up to 3-5 kHz. We study the sense of hearing in the goldfish using two methodologies: Classical, or Pavlovian, conditioning and various psychophysical methods. In these experiments, goldfish are gently restrained in a cloth bag in the center of a small, cylindrical water tank with an underwater loudspeaker at the bottom. A mild electric shock through a restrained fish's body causes a brief, unconditioned suppression of respiration and bradycardia. An auditory signal that terminates with the shock becomes a conditioned stimulus after 10 to 20 delay conditioning trials, evoking a conditioned respiratory suppression. Respiration is measured with a thermistor near the fish's mouth that responds to the cooling effects of respiratory water flow. Respiratory suppression is quantified as a ratio of the respiration during the stimulus divided by the activity preceding and during the stimulus. The experiments are entirely computer-automated. Psychophysical methods are selected depending on the goals of the experiment. These include a method of limits, automated staircase tracking, a method of constant stimuli giving psychometric functions, and novel methods such as stimulus generalization. Stimulus protocols are selected according to the goals of the experiment. In detection studies, a pure tone of 6 sec duration (continuous or in bursts) is presented against a background of silence, or of some controlled masking noise. In discrimination studies, a background sound is presented as brief, repeating bursts and the conditioning stimulus is a change in some acoustic feature (e.g., frequency, level, temporal pulse pattern). In stimulus generalization studies, animals are given conditioning trials to a single stimulus, and then tested for response to novel stimuli. The usefulness of the generalization method is that goldfish apparently learn the characteristics of conditioned stimuli very specifically, and tend not to generalize (respond) to novel stimuli differing in frequency and other stimulus dimensions that lead to qualitatively different perceptions in human listeners. This pervasive failure to generalize permits the identification and analysis of stimulus dimensions that are normally salient or "information-bearing" for goldfish. These, in turn, permit the analysis of corresponding perceptual dimensions. Psychophysical studies have been carried out on sensitivity and the frequency range of hearing, the effects of signal duration and noise masking on signal detection, frequency, intensity, and temporal pattern discrimination, amplitude modulation detection and discrimination, and other discriminations among spectrally and temporally complex sounds. Using stimulus generalization paradigms, goldfish have been shown to behave as if they had internal perceptual dimensions corresponding to pure tone pitch, complex pitch, timbre, and roughness. In addition, goldfish have been shown capable of analytic and synthetic listening, and of auditory scene analysis for simultaneous complex sound sources. Thus, these methods have shown the sense of hearing in goldfish to be qualitatively indistinguishable from that revealed in psychophysical studies of avian and mammalian listeners. We have tentatively concluded that in spite of the fact that the goldfish’s primary auditory receptor organs are the saccular otolith organs, its central auditory system and sense of hearing appear to be primitive vertebrate characters in the sense that they are generally shared with all other vertebrates investigated, including humans. Essentially nothing is known about the sense of hearing of zebrafish. However, since their peripheral auditory system is essentially similar to that of goldfish, it would be expected that their sense of hearing is similar to that of goldfish. In this sense, goldfish would probably be a good model for zebrafish hearing. At the same time, the behavioral methods outlined here would seem to be potentially useful for studies on the sense of hearing (and other sensory modalities) of zebrafish. One potential problem in applying these particular methods to zebrafish, however, is animal size. The goldfish used in these experiments are three inches or more in standard length, and zebrafish, particularly juveniles, tend to be much smaller than this. Animal restraint and measurement of respiration could be a new problem for the considerably smaller animals. I would suggest attempting to measure respiration or the EKG by monitoring the electric field surrounding the animal. Imaging, ablations and behavior: Optical studies of neuronal circuits in zebrafish
One of the key problems in neurobiology is to monitor activity in single neurons non-invasively during behavior, so that the pattern of active cells can be correlated with the behavior. We have taken advantage of the transparency of larval zebrafish and used calcium imaging and confocal microscopy to study which neurons are active during escape behaviors. We have also developed approaches for using lasers to kill individual neurons in intact fish so that we could study the behavioral consequences of these ablations. Our work has focused on descending reticulospinal neurons that interact with spinal circuits to produce the rapid escape movements fish use to avoid predators. The reticulospinal neurons we studied include the Mauthner cell, MiD2cm and MiD3cm, which form a serially repeated set of neurons in hindbrain segments 4, 5 and 6. Our functional imaging and ablation experiments support the hypothesis that high performance escape movements are produced by this segmentally repeated set of hindbrain neurons. Many hindbrain neurons are segmentally arranged, so it is likely that there are other segmentally repeated functional groups. The approaches we have used to study hindbrain cells can be applied to studies of the behavioral roles of neurons throughout the brain and spinal cord of both normal and mutant lines of zebrafish. Genetic analysis of neural circuit formation in the zebrafish embryo
Behavior, including complex behaviors such as curiosity, passion and aggression, rely on the functionality of the nervous system. In vertebrates, this functionality is provided by a large number of neuronal connections which are organized in intricate neural circuits. Formation of neural circuits occurs extensively during embryogenesis and includes a variety of specialized processes, such as neural specification, axonal guidance and synapse formation. Although molecular insights into each of these processes are available, it remains a major challenge to understand the molecular mechanisms by which neural circuits underlying a particular behavior, be it even very simple, are established during embryogenesis.
Our lab is interested to understand the molecular and cellular mechanisms by which simple neural circuits develop during embryogenesis. We are specifically interested in circuits generating locomotion, in particular alternating, rhythmic movements. Rhythmic movements are widespread among vertebrates, and provide the basis for more complex movements, including swimming, crawling and walking. We focus on mutations in 10 zebrafish genes, which we identified in a large scale genetic screen. The initial phenotypic analysis of the mutant phenotypes suggests that these 10 genes as crucial components for the proper development of the circuits producing undulating, rhythmic tail movements of the zebrafish embryo and larvae. The ten genes can be subdivided in two group, according to their mutant phenotypes. Mutant embryos for any of the 7 ‘accordion’ group genes do not display the normal, alternating pattern of left/right tail movements, but instead the embryos contract and expand along their anterior- posterior body axis (like an ‘accordion’). This abnormal phenotype can be mimicked in wild type embryos by drugs known to antagonize the reciprocal inhibition circuit, indicating that the ‘accordion’ group genes are essential components of the neural circuit underlying reciprocal inhibition. Mutant larvae for any of the 3 ‘twitch twice’ group genes do not display the normal, alternating pattern of left/right tail movements, but instead the mutant larvae flip their tail repetitively to the same side, causing the larvae to rotate around it’s own axis. The mutant phenotype indicates that the ‘twitch twice’ genes are required to generate alternating, rhythmic tail movements during swimming. Moreover, separating the left half from the right half of the hindbrain causes an identical ‘twitch twice’ phenotype in wild type larvae. This suggests that the neural circuits underlying coordinated tail movements are at least in part located in the hindbrain, and that the three ‘twitch twice’ group genes are essential components of this neural circuit. We will present progress on how some of the genes contributes to the formation of the neural circuit, and on the molecular cloning of the affected genes. Zebrafish Embryos: Conservation and Research Resource Applications
Our long range goal is to cryopreserve teleost embryos successfully, specifically zebrafish embryos. The availability of viable embryos after cryopreservation could have a profound influence on the conservation of rare or threatened species and on the extension of medical research resources. In medicine, the zebrafish has become one of the more important vertebrate models for the study of development and genetics. The preservation of important genetic lines is essential, because these lines will play an important role in future studies on human health and disease. Systematic germ plasm cryopreservation can have a major impact on the management of NIH resources, such as the Zebrafish Stock Center by: i) reducing the size and production costs of facilities; ii) allowing the maintenance of large gene pools and reducing inbreeding, while minimizing the amount of space required to hold living animals; iii) reducing pressure on wild populations from collection activities; and iv) facilitating global, regional, and institution- to-institution transport of genetic material. For conservation, the development of frozen or 'insurance' populations would preserve genetic diversity and assist efforts to prevent extinction of wild fish species in natural aquatic ecosystems. Although freezing teleost spermatozoa is commonly practiced, to date, fish embryos have never been cryopreserved successfully. The zebrafish (Brachydanio rerio) is an excellent model for basic studies of cryobiology because they breed regularly providing embryos daily, and a great deal is known about their normal development and physiology. For successful cryopreservation, cellular permeability to water and cryoprotectants must be understood. Ideally, water must exit, and an appropriate cryoprotectant enter all the cells. Using a wide variety of techniques, such as electron-spin resonance, biophysical modeling, and magnetic resonance imaging and spectroscopy, we have found a major permeability barrier in the zebrafish embryo: the yolk syncytial layer (YSL), which develops between the yolk and blastoderm. Due to its low permeability, the YSL blocks water exit from, and entry of some cryoprotectants into, the yolk. Specifically, electron microscopy has shown that cryopreservation destroys the YSL, presumably because ice-crystals form in the yolk and destroy the YSL. Thus, standard techniques of immersing the embryos in various cryoprotectants allowing passive permeation over time may not be practical in this system. Therefore, we are examining ways to modify membrane permeability to water and cryoprotectants using molecular and biochemical techniques. This work was supported by grants from the National Institutes of Health (R01 RR08769), the Smithsonian Institution, and Maryland Sea Grant College. Asymmetric Gene Expression in the Developing Zebrafish Brain
Analyses of zebrafish mutations may provide insight into the molecular mechanisms that underlie left-right differences in the vertebrate brain. Zebrafish cyclops (cyc) mutants lack the ventral brain and floor plate. The cyc gene was recently found to encode a nodal-related TGF-b family member that is expressed in the midline mesendoderm and prechordal plate during gastrulation but is down regulated by early somite stages. However, a new wave of asymmetric expression appears in the left lateral plate mesoderm and in the dorsal diencephalon during mid-somitogenesis. Expression in the dorsal brain may correlate with the left habenular nucleus and the epiphysis (pineal organ) also derives from this approximate region. To better characterize the patterning of the diencephalon, we are examining the relationship of cyc expression with other developmental (i.e. floating head, one-eyed pinhead) and pineal-specific (i.e. visual pigments) genes that are expressed in the dorsal forebrain. Zebrafish antivin (similar to mouse lefty) is transcribed in the left side of the dorsal diencephalon, as is the bicoid homeobox gene Pitx2. Together with cyc, these genes are bilaterally expressed in the brains of zebrafish mutants that exhibit midline or other gastrulation defects, consistent with an early specification of asymmetry that is later manifested in the dorsal brain. We are exploring how asymmetric gene expression is regulated in the diencephalon through RNA misexpression studies. By transient rescue of mutant embryos through gastrulation, we aim to uncover the function of left-sided Nodal signalling in the developing zebrafish forebrain and the anatomical and/or behavioral consequences on the adult. Supported by the NSF (MEH), NIH (J-C I-P and JOL) and the Pew Scholars Program (MEH and J-C I-P). Odorant receptors and olfactory system function in goldfish and zebrafish
The detection and discrimination of the multitude of environmental stimuli by the vertebrate olfactory system results from the activation of olfactory neurons in the nose. The first step in olfactory processing resides at the level of the interaction of odorous ligands with odorant receptors. A large multigene family thought to encode odorant receptors was initially identified in the rat by Buck and Axel in 1991. These receptors are predicted to exhibit a seven transmembrane domain topology characteristic of the superfamily of G protein-coupled receptors. The sizes of the receptor repertoires of different vertebrate species are extremely large and are estimated to contain between 100 and 1000 individual genes. These observations suggest that olfactory discrimination is accomplished by the integration of signals from a large number of specific receptors, each capable of binding only a small number of structurally-related odorants. Other olfactory G protein-coupled receptors unrelated to the first family of odorant receptors described have been identified in the vomeronasal organ (VNO) of mammals as well as in the olfactory epithelium of fish. These receptors are encoded by two unrelated gene families: the VNR family and the V2R family; the V2R receptors are structurally related to the CaSR and mGluR families. While it has been proposed that the VNR and V2R receptors are pheromone receptors (based on their expression in the mammalian VNO), the actual function of these orphan receptors awaits a direct demonstration of their molecular specificities. As an approach toward identifying ligands for olfactory receptors, we have pursued an expression cloning strategy using the goldfish as a model system. The odorants that fish detect are water soluble, and include amino acids (feeding cues), bile acids (nonreproductive social cues with possible roles in migration), and sex steroids and prostaglandins (pheromonal cues). Electrophysiological studies have characterized the sensitivities of fish olfactory systems to specific ligands, demonstrating, for example, thresholds for detection in the picomolar (for sex steroids) to nanomolar (for amino acids) range. Thus, the defined properties of odorant-evoked pathways in vivo provide an excellent starting point for the molecular and biochemical characterization of fish odorant receptors. We have recently succeeded in expression cloning a cDNA encoding a goldfish odorant receptor preferentially tuned to recognize basic amino acids. This receptor, called receptor 5.24, that shares significant similarity to receptor families that include the CaSR, mGluR, and V2R class of VNO receptors. The affinity and specificity of the cloned goldfish odorant receptor for basic amino acids are remarkably similar to basic amino acid sensing pathways characterized in vivo. Degenerate polymerase chain reaction (PCR) reveals other related sequences that are expressed in the goldfish olfactory epithelium. Our results provide the first direct evidence that these receptors in fact comprise a family of odorant receptors. A number of questions can now be pursued based on our demonstration of receptor 5.24’s functional characteristics. For example, can receptor expression or the behavior of cells expressing this receptor be altered by odorant-induced activity or experience? What are the molecular specificities of other olfactory CaSR-like receptors related to receptor 5.24? Do closely related receptors recognize other amino acid odorants? Do more distantly related receptors recognize other classes of odorants, such as pheromones or migratory cues? Future studies will address these questions using both the goldfish as well as the closely related zebrafish as model systems. Development and Early Behavior in Fishes: a comparative approach
Fishes have a greater variety of species, and hence early development and behavior, than other vertebrates. Reproduction and early development are the key features to understand this variation. I will propose a unifying scheme to incorporate this variation, with examples including zebrafish. The scheme integrates early development (ontogeny), evolution (function, phylogeny), and causation (physiological mechanisms). The life history of any species consists of a maximum of four developmental intervals. The number of these developmental intervals for any species, and the features of each interval are the keys to understanding the patterns of early development in fishes. The timing of key events during development characterizes this pattern of ontogeny. I will explain this scheme with examples of avoidance and preference tests, social and feeding behavior, and sexual development. Key words: toxicology, avoidance responses, locomotor behavior, choices, receptor development, social behavior, genetics, feeding, sex Zebrafish Touch-Insensitive Mutants
Developmental changes in neuronal connectivity, synaptic function and excitable membrane properties underlie stage-specific appearance of embryonic behaviors. Several zebrafish embryonic mutants display abnormal motility (*Granato et al., 1996) and provide opportunities for identification of genetic, molecular and cellular mechanisms generating specific behaviors. One group of mutants swims spontaneously but not in response to touch (touch-insensitive mutants). The specificity of the behavioral phenotype suggests that the defect arises at the level of the relevant sensory neurons, mechanosensory Rohon-Beard cells. Whole cell recording using patch clamp techniques was performed to examine the excitable membrane properties of Rohon-Beard cells. Wildtype zebrafish embryos respond to touch at 27 hours post fertilization. Electrophysiological analysis of Rohon-Beard cells of wild type embryos indicates that, during the transition from a touch-insensitive to a touch-sensitive embryo, action potentials acquire larger overshoots and briefer durations. During the same period, amplitudes of sodium and potassium currents both increase. In contrast, Rohon-Beard cells of several homozygous touch insensitive mutant embryos (macho, mao; alligator, ali; steifftier, ste) fail to fire action potentials. The strongest defects are found in homozygous mao mutants. Similar to wildtype embryos, mechanosensory neurons of unaffected sibling embryos of all mutant lines fire action potentials with overshoots of normal amplitude. General measures of membrane properties, such as resting membrane potential and input resistance, are not affected in homozygous mutants excluding the possibility of nonspecific effects. Examination of voltage-dependent currents that underlie action potentials demonstrates that impulses are not generated because of a specific reduction in sodium current. The amplitude of voltage-dependent potassium current increases normally, consistent with the fact that developmental regulation of the duration of the action potential is unaffected in mechanosensory neurons of mutants. However, sodium current remains small, thereby preventing the normal increase in overshoot of the action potential. These results indicate that developmental regulation of sensory neuron sodium current plays an essential role in acquisition of embryonic touch sensitivity. *Granato et al., 1996, Development 123: 399-413. Chemically-mediated predator-prey interactions in Ostariophysan fishes
Animal behavior evolves, like any other trait, by natural selection rewarding individuals with appropriate behavior with greater rates of reproductive success than individuals with inappropriate or ill-timed behavior. Genes that contribute to the physiology (sensory receptors) and inclination (neural processing) to perform effective behavior are thusly promoted at the expense of genes for less effective behavior. Antipredator behavior serves as a good model for the effect of environmental selection gradients on the evolution of behavior. Predation exerts a steep selection gradient on shaping behavior. The expression of behavior is an interaction between the genes that code for the behavior and environmental cues that signal the appropriate timing and degree of response. In aquatic environments, visual information is not available under conditions of high turbidity, areas of structural complexity, or at night. Chemical cues are not affected by these limitations. In addition, water is the universal solvent and ideal for the solution and dispersal of chemical cues. Chemical cues are released during predation events - at the point of initial detection, during attack and capture and even post-ingestion of prey. Thus, publicly available chemical information is a reliable indicator of predation risk and should be used widely by aquatic animals for risk assessment. For example, larval damselflies (Enallagma sp.) are aquatic insects. They respond with antipredator behavior to injury-released chemical cues from conspecifics. These cues are released when a predator grasps its prey. Fishes too, including zebra danios (Brachydanio rerio), exhibit clear antipredator responses to injury-released chemical cues from their own species. Typically, these cues elicit a reduction in overall activity, movement to the substratum, increased shoal cohesion and area avoidance (when possible). Injury-released chemical cues occur in relatively late stages of a predation event. These cues indicate that the predator is foraging actively. However, prey can detect the predator's presence at a much earlier stage by detecting the odor of the predator itself. This affords prey more time to initiate crypsis or escape behavior. How do prey come to recognize the odor of their predators? This is not a trivial question because the suite of predators to which a prey population is subjected can vary widely over space and time. A growing body of evidence indicates that genetically-based recognition templates for every potential predator was not the parsimonious solution favored by natural selection. Instead, a robust learning paradigm has evolved whereby novel stimuli, such as an odor or image, come t o be associated with predation risk when presented simultaneously with injury-released chemical cues. Thereafter, the novel stimulus on its own elicits a full antipredator response. This phenomenon has been demonstrated in a number of aquatic taxa, including, damselfly larvae, fathead minnows (Pimephales promelas) and zebra fish. Recognition learning of indicators of predation risk is remarkable in that only a single trial is required. Memory of the novel cue's riskiness is retained for at least a year by minnows. Minnows and zebra fish are so adept at recognition learning that they can be easily tricked to fear non-biological stimuli. Recent data indicate that recognition learning occurs even if there is a slight delay between introduction of injury-released cues and the novel cue. Zebra fish are in the superorder Ostariophysi. This group includes the cyprinidae (minnows, of which the zebra fish is one), the characidae ("tetras"), the siluridae (catfish), the catastomidae (suckers) and sundry other minor families. Altogether, they represent 64% of all freshwater fish species. One defining character of this group is the presence of specialized club cells in their epidermis, known as alarm substance cells (ASC). They are fragile, thin-walled cells on the surface of the epithelium and are easily ruptured when the prey is grasped by a predator. Alarm substance, thought to be hypoxanthine-3(N)-oxide or some close derivative, greatly enhances the general phenomena associated with injury-released chemical cues, including learning. Production and maintenance of ASCs have a high metabolic cost. Minnows invest facultatively into ASCs in response to perceived levels of predation risk. For example, minnows fed high rations develop a thicker epidermis with disproportionately more ASCs than minnows fed a low ration. This cost is offset by the fitness advantage of attracting additional predators to a predation event in progress. This helps the prey because predators threaten and bully each other providing the prey with an opportunity for escape. Shoalmates from a minnow's own shoal ("familiar shoalmates") execute more effective group-level antipredator responses than those executed by non-familiar shoalmates. High-ration minnows make more ASCs when held with non-familiar shoalmates than when held with familiar shoalmates. This latter finding suggests a trade-off between the cost of making ASCs and predation risk. When in the company of familiar shoalmates minnows rely more on effective group antipredator behavior and spend less energy making ASCs. When in the company of non-familiar shoalmates, minnows invest more in ASCs to better attract secondary predators. This system is ripe for developmental biologists using zebra fish as a model. Virtually nothing is known about the olfactory receptors involved in this response, the neural processing for learned recognition of novel stimuli, the production of ASCs nor the nature of the cue itself. POSTER ABSTRACTS
Computer-Assisted Visualizations of Neural Networks: Expanding the Field of View using Seamless Confocal Montaging
Microscopic analysis of structure/function relationships within the neural networks of adult and developing tissues often requires visualization of large regions of neuronal architecture. To accomplish this, there are two visualization approaches: (1) image the entire area at once with low spatial resolution; or (2) image small areas at higher magnification/resolution, and then piece the regions back together using a mosaic reconstruction (i.e. photomontaging). Low magnification imaging is relatively rapid to perform, resulting in a visualization that encompasses a large field of view with an extended depth of field. However, for fluorescence microscopy, low magnification visualization is often plagued by poor spatial resolution. Although high magnification imaging has superior spatial resolution, it produces a visualization with limited depth of field. Moreover, when creating a larger field of view, the visualization is fragmented at the seams where multiple images must be stitched together. Using confocal microscopy as well as common image-processing functionalities, we outline a new visualization approach that transforms a montage of spatially-contiguous z-series (i.e. vertical optical sections) into a large visualization with a seamless field of view and an extended depth of field. We illustrate our method for visualizing neural networks using tissues from the adult gastropod mollusc, Tritonia diomedea, and the developing zebrafish, Danio rerio. Crossed Modulation of Inhibitory Synaptic Inputs in Left-Right Decision Neurons
Continuous fluctuations of membrane potential or synaptic noise in central neurons are caused by spontaneous impulses in presynaptic afferents. The paired Mauthner (M-) cells of teleosts are a part of the brain stem escape network and the direction of escape of fish is determined by which one reaches threshold first. In these cells, synaptic noise is mostly inhibitory (ISN) and it influences the neuron's input-output relation. We found that in the adult zebrafish M-cells ISN exhibits two distinct patterns, i.e. a "noisy" state, made of bursts of fast strychnine-sensitive inhibitory postsynaptic potentials, and a "quiet" state, with less inhibition, both of which can last from 14 msec to several minutes, and then spontaneously flip-flop. Simultaneous recordings have revealed that these patterns are complementary in the left and right M-cells. An action potential induced by a brief current injection transiently turns on the noisy state in the activated neuron, while it suppresses the ISN in the other side. Finally the noisy state contains periodic components the frequency of which is also up-regulated after an action potential. Our data suggest that this asymmetric modulation of ISN contributes to the "choice" of the direction of the escape most appropriate for animal. The Anatomical Organization of the Locus Coeruleus in the Zebrafish
The locus coeruleus (LC) is a brainstem noradrenergic nucleus present in all vertebrates. In the zebrafish, this nucleus contains an average of only 6.8 ± 1.5 (range 3 - 10) neurons. Both genetic and epigenetic factors are implicated in the regulation of LC cell number. The low neuron number offers an opportunity for elucidating the organization, development and function of the entire nucleus. Preliminary results suggest that subsets of distinct LC neurons exist. (i) On the basis of dendritic morphology, three types of neuron can be distinguished. The relative proportion of the three types of cells appears to depend on the total number of neurons in the LC. (ii) Retrograde tracing experiments show that LC neurons project to many targets in the forebrain, midbrain and cerebellum. Double-labelling studies reveal the existence of neurons with distinct projection patterns. No target receives input from all neurons, and each neuron projects to a combination of targets. The combination of targets each neuron projects to also appears to depend on the total number of cell in the LC. These observations suggest that in the LC, anatomical organization is subject to adjustment based on cell number. Behavioral screening techniques for larval and adult zebrafish with special reference to behavioral lateralization
We have developed behavioral tests, which allow screening of large numbers of zebrafish larvae and systematic testing of adults. These are suitable to identify mutations affecting visual mechanisms as well as behavioral lateralization. In adult tests, fish live singly in a small tank, with a test compartment, where they are fed, and which they visit frequently and spontaneously. They respond to colored beads initially with bites and after habituation across trials with viewing alone. The right eye is used when a bite will be performed, and there is a shift to left eye use in viewing after habituation. Dishabituation is clear as a result of color change between trials. This would allow measurement of the ability to discriminate hue and of perceptual memory, as well for screening fish with reversed bias. The screening tests for fry uses a small swimway, divided into compartments, each with an exit to the next trough a vertical slit. Positive phototaxis is used to cause the larva to move to the next compartment, by darkening the one the larva is in, and illuminating the next. Larvae aged 4, 6, and 8 days show a turning bias which in wild type is leftward on day 4 but shifts toward rightward by day 8. These age-dependent changes are very similar to shift described in chicks; the underlying mechanism is unknown but their discovery in a teleost confirms evidence that they may be widespread in vertebrate development. In a further application of this procedure a black stripe (1.2 cm) was placed on the side walls of the compartments. We found that naive, 6 day old larvae turned away from the stripes (avoidance), however larvae that were exposed to the stripes on day 5 in their home environment showed either no avoidance or even some preference (approach) to the stripes. This test presents an good opportunity to screen for mutations affecting early avoidance and approach responses in larvae, as well as long term memory and lateralization. Locomotive Repertoire of the Larval Zebrafish: Swimming, Turning and Prey Capture
Larval zebrafish are a popular model system because of their transparency and relative simplicity. They have a total of about 160 neurons in the brainstem that project to the spinal cord, many of which can be individually identified and laser-ablated in intact larvae. This should facilitate cellular-level characterization of the descending control of larval behaviors. As a first step, we attempted to delimit the range of locomotive behaviors exhibited by zebrafish larvae. Using high-speed digital imaging, a variety of swimming and turning behaviors were analyzed in 6-to-9 day old larval fish. Swimming episodes appeared to fall into two categories, with the center of bend of the larva’s body occurring either near the mid-body (burst swims) or closer to the caudal fin (slow swims). Burst swims also involved larger amplitude bending, faster speeds and greater yaw then slow swims. Turning behaviors clearly fell into two distinct categories: fast, large-angle escape turns characteristic of escape responses, and much slower routine turns lacking the characteristic "return tail flip" that accompanied the escape turns. Prey-capture behaviors were also recorded and appeared to be comprised of two more elemental locomotive behaviors: routine turns and slow swims. The different behaviors observed were analyzed with regard to possible underlying neural control systems. Our analysis suggests the existence of discrete sets of controlling neurons and helps to explain the need for the 160 or so spinal-projecting nerve cells in the larval zebrafish brainstem. Cloning and Characterization of a Gene Homologous to the Delta Opioid Receptor in Zebrafish
A full-length cDNA, ZFOR1, has been isolated from the teleost zebrafish (Danio rerio) using a probe from rat m opioid receptor. ZFOR1 encodes a 373 amino acid protein with seven potential transmembrane domains that shows a high degree of homology to mammalian d opioid receptor. We have also isolated a genomic clone which contains two exons of ZFOR1, homologous to exons 2 and 3 in mouse and human d opioid receptor. Expression of ZFOR1 appears to be restricted to nervous tissue as assessed by Northern blot. In situ hybridization in zebrafish brain with specific probes revealed several discrete areas of ZFOR1 expression; higher levels are detected in dorsal telencephalic areas, periventricular layer of the optic tectum, and granular layer of the cerebellum. To characterize the receptor and to compare its pharmacological profile with the delta opioid receptor we stably expressed ZFOR1 in HEK 293 cells. Competitive binding studies with several opioid ligands presented low affinity for the classical opioid selective ligands , but showed high affinity for the d -agonist BW373U86 (Ki 89nM). In functional GTPgS studies, BW373U86was found to be a good agonist. These are the first molecular evidences on the presence of a functional d opioid receptor-like in the zebrafish. Involvement of Adhesion Molecules in Plasticity of Zebrafish Brain After Avoidance Conditioning
We have developed an active shock avoidance paradigm for goldfish and zebrafish that favorably lends itself towards behavioral analyses in theses cyprinids. Application of [ 14C] -deoxyglucose revealed a significantly increased energy demand in the zebrafish optic tectum after learning of the avoidance response. In situ hybridizations with antisense probes against the cell adhesion molecules L1.1 and NCAM located both messages to specific neuronal populations in several distinct layers of the optic tectum. Expression of L1.1 and NCAM mRNA was significantly enhanced 3 hours after learning. In order to analyze, whether induction of these cell adhesion molecules is a prerequisite for long-term memory formation, HNK-1 antibodies, directed against a specific carbohydrate epitope on L1.1, NCAM and ependymins, were injected between training and test. Various other antibodies served as controls. In the test, fish injected with HNK-1 antibody exhibited significantly reduced retention scores (RS = 0.30) as compared with controls (RS = 0.77). The HNK-1 antibody had no influence on the performance of the avoidance response, when injected into overtrained animals (RS = 0.80). Similar results where obtained with specific antibodies directed against the deglycosylated (i.e., HNK-1-free) form of ependymins (RS = 0.24). Our results point towards a pivotal role of cell adhesion molecules in long-term memory formation in the cyprinid brain. - Financial Support by the Deutsche Forschungsgemeinschaft is gratefully acknowledged (Schm 478/10-1). Fighting Behavior in Zebrafish, Danio rerio
Zebrafish establish dominant hierarchies that are maintained for long periods of time. To study the behavioral elements that lead to these social hierarchies, we analyzed the interactions of groups of two fish of similar body mass and size. Behavioral patterns such as lateral displays, frontal displays, nipping, and chasing were observed and recorded over a 30 min period. Hierarchies appeared to be established within a fighting period of up to 5 min. This fighting period is characterized by frequent lateral and frontal displays that are accompanied by nipping. Following this fighting period, dominant fish chased subordinate fish for the rest of the observational period. A second encounter of the same pair of fish on the following day (day 2) showed similar characteristics. The winner of the fight on day 1 also was the winner on day 2. Our studies indicate that zebrafish can establish dominant hierarchies during a short period of intense fighting. Whether the social status of zebrafish leads to molecular or physiological changes in the nervous system will be addressed in future studies. Supported by Center for Research, WPUNJ. Innervation of Fin Muscles in Zebrafish, Danio rerio
Sets of fin muscles control movements of dorsal, caudal, and pectoral fins in adult zebrafish. In order to gain insight into the control of these fin muscles, we characterized the organization of fin motorneurons (MNs) anatomically. Application of neurobiotin to dorsal fin depressor and erector muscles labeled 10 to 12 secondary MNs per hemisegment. These MNs (10 - 15 µm in diameter) are located in a ventro-lateral motor column and tend to appear in groups of up to 4 MNs. Double labeling with fluorescent tracers showed that dorsal fin MNs can be distinguished from secondary axial MNs by position of cell bodies. Dye injections into the protractor muscle revealed a new type of zebrafish neuron with contralateral projections. Caudal fin muscles are innervated by nerves that originate in the last 5 segments of the spinal cord. Six of eight caudal fin muscles are located in a dorsal, median, and ventral region of the tail. Dye injections into these three regions yielded staining of up to 66 MNs for the dorsal group, up to 71 MNs for the medial group, and up to 33 MNs for the ventral group. The MNs innervating pectoral fin muscles are located in the first segments of the spinal cord. Application of tracers to adductor and abductor muscles yielded up to 40 neurons for each muscle. MNs of each pectoral fin muscle are located in separate clusters. Future investigations will be concerned with physiological activities of mapped fin MNs. Supported by Center for Research, WPUNJ. Learned Recognition by Zebra Fish (Brachydanio Rerio) of Novel Predator Odor Following Non-Simultaneous Presentation of Alarm Pheromone in Skin Extract and Predator Odor
Fishes in the superorder Ostariophysi (minnows, characins, catfish, etc) possess specialized epidermal cells that contain an alarm pheromone. These cells lack a duct to the skin surface. The alarm pheromone is released when the skin is damaged, such as occurs when the fish is grasped by a predator. Thus, the pheromone serves as a reliable indicator of predation risk. Fish can learn to recognize and associate novel odors, such as the odor of a predator, when a novel odor is encountered simultaneously with alarm pheromone. Thereafter, the novel cue is recognized as an indicator of risk and induces antipredator behavior. In nature, the odor of a novel predator may not be detected simultaneously with alarm pheromone. Can a delay between the presentation of alarm pheromone and novel predator odor still result in associative learning? We presented zebra danios with the odor of northern pike (Esox lucius) 5 min after presenting them with either alarm substance or water (control). During a predation event, 5 min is a long time. When later retested with pike odor alone, zebra fish conditioned with alarm substance significantly increased antipredator behavior (z = 1.91, P = 0.024) but control fish did not (z = 0.10, P = 0.464). These data show 1) that learned recognition of predation risk is robust enough to accommodate ecologically realistic temporal shifts in stimulus presentation, and 2) confirm that zebra danios are good test organisms for learned risk aversion and potentially the developmental and genetic mechanisms that support these behaviors. |
