87:608-614, 2002. J Neurophysiol

Ethan Gahtan, Nagarajan Sankrithi, Jeanette B. Campos and Donald M. O'Malley

You might find this additional information useful...

35 articles, 10 of which you can access free at: This article cites http://jn.physiology.org/cgi/content/full/87/1/608#BIBL

17 other HighWire hosted articles, the first 5 are: This article has been cited by

  [PDF]  [Full Text]  [Abstract]

, May 2, 2007; 27 (18): 4984-4994. J. Neurosci.H. A. Burgess and M. Granato

Sensorimotor Gating in Larval Zebrafish 

[PDF]  [Full Text]  [Abstract], May 16, 2007; 27 (20): 5271-5279. J. Neurosci.

T. Sato, T. Hamaoka, H. Aizawa, T. Hosoya and H. Okamoto from the Posterior Tectum to the Hindbrain in Zebrafish

Genetic Single-Cell Mosaic Analysis Implicates ephrinB2 Reverse Signaling in Projections 

  [Full Text]  [Abstract], July 1, 2007; 2007 (16): pdb.prot4781-pdb.prot4781. CSH Protocols

J. R. Fetcho Imaging Neuronal Activity with Calcium Indicators in Larval Zebrafish

  [PDF]  [Full Text]  [Abstract]

, August 1, 2007; 134 (15): 2771-2781. DevelopmentCui, W. Zhou and J. Y. Kuwada H. Hirata, T. Watanabe, J. Hatakeyama, S. M. Sprague, L. Saint-Amant, A. Nagashima, W. W.

swimming and provide a model of multi-minicore diseaseZebrafish relatively relaxed mutants have a ryanodine receptor defect, show slow 

[PDF]  [Full Text]  [Abstract], October 15, 2008; 28 (42): 10641-10653. J. Neurosci.

T. Kohashi and Y. Oda Modes of Sensory Input

Initiation of Mauthner- or Non-Mauthner-Mediated Fast Escape Evoked by Different

on the following topics: http://highwire.stanford.edu/lists/artbytopic.dtlcan be found at Medline items on this article's topics

Physiology .. Action Potential Veterinary Science .. Spinal Cord Veterinary Science .. Brain Stem Physiology .. Axons

including high-resolution figures, can be found at: Updated information and services http://jn.physiology.org/cgi/content/full/87/1/608

can be found at: Journal of Neurophysiologyabout Additional material and information http://www.the-aps.org/publications/jn

This information is current as of November 25, 2008 .  

http://www.the-aps.org/.American Physiological Society. ISSN: 0022-3077, ESSN: 1522-1598. Visit our website at (monthly) by the American Physiological Society, 9650 Rockville Pike, Bethesda MD 20814-3991. Copyright © 2005 by the

publishes original articles on the function of the nervous system. It is published 12 times a yearJournal of Neurophysiology

on Novem

ber 25, 2008 jn.physiology.org

Dow

nloaded from

Evidence for a Widespread Brain Stem Escape Networkin Larval Zebrafish

ETHAN GAHTAN, NAGARAJAN SANKRITHI, JEANETTE B. CAMPOS, AND DONALD M. O’MALLEYDepartment of Biology, Northeastern University, Boston, Massachusetts 02115

Received 19 July 2001; accepted in final form 2 October 2001

Gahtan, Ethan, Nagarajan Sankrithi, Jeanette B. Campos, andDonald M. O’Malley. Evidence for a widespread brain stem escapenetwork in larval zebrafish. J Neurophysiol 87: 608–614, 2002;10.1152/jn.00596.2001. Zebrafish escape behaviors, which typicallyconsist of a C bend, a counter-turn, and a bout of rapid swimming, areinitiated by firing of the Mauthner cell and two segmental homologs.However, after laser-ablation of the Mauthner cell and its homologs,escape-like behaviors still occur, albeit at a much longer latency. Thismight suggest that additional neurons contribute to this behavior. Wetherefore recorded the activity of other descending neurons in thebrain stem using confocal imaging of cells retrogradely labeled withfluorescent calcium indicators. A large majority of identified descend-ing neurons present in the larval zebrafish, including both ipsilaterallyand contralaterally projecting reticulospinal neurons, as well as neu-rons from the nucleus of the medial longitudinal fasciculus, showedshort-latency calcium responses after gentle taps to the head of thelarva—a stimulus that reliably evokes an escape behavior. Previousstudies had associated such in vivo calcium responses with the firingof action potentials, and because all responding cells have axonsprojecting into to spinal cord, this suggests that these cells are relayingescape-related information to spinal cord. Other identified neuronsfailed to show consistent calcium responses to escape-eliciting stim-uli. In conjunction with previous lesion studies, these results indicatethat the neural control systems for turning and swimming behaviorsare widely distributed in the larval zebrafish brain stem. The degree ofrobustness or redundancy of this system has implications for thedescending control of vertebrate locomotion.

I N T R O D U C T I O N

Involvement of the Mauthner cell (M cell) in vertebrateescape behaviors was hypothesized many years ago (reviewedby Eaton et al. 2001; Zottoli and Faber 2000). Electricalrecording of M cells in free-swimming goldfish (Zottoli 1977)showed that their firing was correlated with a C-start behaviorthat seems to be an important maneuver by which cyprinidfishes escape from predators. It was subsequently shown thatfiring of the M cell is sufficient to initiate a C bend andcounter-turn (Nissanov et al. 1990). Other brain stem neuronsalso appear to be involved in the initiation and control ofC-start type escapes as indicated by lesion studies of the Mcell, which failed to eliminate the behavior (Eaton et al. 1982,1984; Kimmel et al. 1980; Zottoli et al. 1999). Anatomicalstudies of neurons projecting into the spinal cord of larvalzebrafish revealed a large array of neurons of which two cells

(MiD2cm and MiD3cm) were suggested to be segmental ho-mologs of the M cell (Metcalfe et al. 1986) and were proposedto provide directional control over the escape behavior (Fore-man and Eaton 1993). Subsequent in vivo calcium imagingstudies supported this hypothesis: caudal stimuli activated justthe M cell, while rostral stimuli activated the “Mauthner ar-ray”, i.e., the M cell plus the two segmental homologs(O’Malley et al. 1996). Dramatic confirmation of this hypoth-esis came with laser ablation of the M cell and its homologs:lesioning just the M cell increased response latencies to caudalstimuli, while lesioning the entire Mauthner array sharplyincreased response latencies to rostral stimuli (Liu and Fetcho1999). An unexpected result, however, was that after an ab-normally long latency, fairly normal looking “escape” behav-iors occurred that approached wild-type C starts in terms ofbend angle and angular velocity (Budick and O’Malley 2000a;Liu and Fetcho 1999). Because the long latencies of theseresponses would bode poorly in terms of escape from preda-tors, these abnormal behaviors might be referred to as “large,delayed” turns.

From the preceding data, we might infer that at least twoadditional classes of neurons contribute to teleost C-start es-cape behaviors. First, some neurons in the brain stem mustmediate the large, delayed turns observed after ablation of theMauthner array. Second, additional neurons are predicted tomediate the counter-turn that follows the initial c-bend (termedA2 neurons by Foreman and Eaton 1993). The identities ofthese postulated neurons are unknown, but this group is com-posed, at least in part, of neurons descending from brain intospinal cord. The descending neurons in the larval zebrafish areanatomically well defined, numbering about 220 in total andincluding reticulospinal, vestibulospinal, T-reticular, and ICneurons as well as cells of the nucleus of the medial longitu-dinal fasciculus (nMLF) (Gahtan and O’Malley 2001; Kimmelet al. 1985; Metcalfe et al. 1986). Using confocal calciumimaging, we set out to determine which neurons would respondto escape-eliciting stimuli. Unexpectedly, we found that asubstantial majority of descending neurons, spanning the ros-tral-caudal extent of the brain stem, exhibit large calciumresponses indicative of the firing of action potentials and thetransmission of escape-related information to spinal cord.

Address for reprint requests: D. M. O’Malley, Dept. of Biology, 414 MugarHall, Northeastern University, Boston, MA 02115 (E-mail: d.omalley@neu.edu).

The costs of publication of this article were defrayed in part by the paymentof page charges. The article must therefore be hereby marked ‘‘advertisement’’in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

RAPID COMMUNICATIONJ Neurophysiol87: 608–614, 2002; 10.1152/jn.00596.2001.

608 0022-3077/02 $5.00 Copyright © 2002 The American Physiological Society www.jn.org

on Novem

ber 25, 2008 jn.physiology.org

Dow

nloaded from

M E T H O D S

A 75% solution of the fluorescent calcium indicator Oregon-greendextran bis-(o-aminophenoxy)-N,N,N�,N�-tetraacetic acid (BAPTA,10,000 molecular weight) was injected into the caudal spinal cord of2 to 4 days posthatching larval zebrafish (Danio rerio) that had beenanesthetized with 0.02% 3-aminobenzoic acid ethyl ester (MS222,Sigma). A glass microelectrode, broken back to approximately halfthe diameter of the spinal cord, was used for injections. Labeling ofhindbrain neurons is believed to occur by the severing of their spinalaxons (Gahtan and O’Malley 2001). A substantial but variable numberof descending neurons are typically labeled after the injections, whichare aimed at the ventral portion of the cord. After injection, larvaewere placed individually in circular wells (1.7-cm diam) containing10% Hanks solution (Westerfield 1995). They were housed in anincubator (approximately 27°C, 14 h/10 h light/dark cycle) overnightto allow for recovery and transport of the indicator from spinal axonsto the soma of reticulospinal neurons. For confocal imaging, larvaewere again anesthetized, placed on their backs on a glass coverslip,and restrained in agar, after which the anesthetic was replaced with10% Hanks solution (see O’Malley and Fetcho 2000 for furtherdetails). A Zeiss Axiovert microscope with a �40, 0.75 NA objectiveand BioRad MRC600 laser scanning confocal microscope were usedfor imaging reticulospinal neurons in intact larvae. Reticulospinal andother descending neurons were identified morphologically based onsoma location and other anatomical features (Kimmel et al. 1985;Metcalfe et al. 1986).

In each larva, a reticulospinal neuron was selected for recording anda glass “tapper” was positioned near the ear on the side opposite to thespinal projection of that neuron’s axon. An electrical pulse to apiezoelectric crystal attached to the tapper caused the glass tip togently contact the head of the larvae; this stimulus reliably induces anescape turn contralateral to the tap (Fetcho and O’Malley 1995;O’Malley and Fetcho 2000; O’Malley et al. 1996). On each trial, a tapwas delivered during collection of a sequence of images of the cell,and analysis of fluorescence values before (baseline) and after the tapwas made off-line (as described in Fetcho and O’Malley 1995). Atleast 2 min were allowed to elapse between trials as this frequency ofstimulation tends to minimize any habituation of the escape response(Fetcho and O’Malley 1995; Fetcho et al. 1998). Cells were imagedusing either frame scans (24 scans at �500 ms/scan), slow line scans(512 scans at 6 ms/scan), or fast line scans (512 scans at 2 ms/scan).Tap-evoked behaviors caused cells to move briefly out of the plane offocus. This “behavioral” movement artifact was clearly distinguish-able from the much smaller movement artifact produced by the tapperalone and served as a rough marker of the presence and duration ofbehaviors. However, these movement artifacts precluded imaging ofcellular calcium levels during the behavior. In some fish, after col-lecting a series of trials, the fish was paralyzed by bath application ofcurare (3 mg/ml), and further trials were collected. Such trials allowedbetter estimation of the latency of tap-evoked fluorescence changesbecause of a reduction in the duration of the movement artifact. Areticulospinal cell type was classified “responsive” during escapes if afluorescence increase (posttap peak/pretap baseline) of �10% oc-curred on at least three of five trials (in noncurarized fish) for fourdifferent cells in at least three different larvae. Cell types wereclassified “nonresponsive” only if they failed to meet the responsecriteria and other cell types in the same larvae showed responses tohead taps. Cells that were nonresponsive to head taps contralateral totheir spinal axon projection were also tested for their responsivenessto ipsilateral taps.

R E S U L T S

Escape-eliciting stimuli generated calcium responses in neu-rons in both the medulla and midbrain. Figure 1 shows exam-ples of responses in three specific cell types (right) and their

location in the brain stem (left). Cells examined in the nMLF,the most rostral group of neurons that projects from brain tospinal cord in larval zebrafish, responded to taps to the head. Amedial-lateral (cMeL) cell of the nMLF (top) shows a largecalcium response just after the tap (*), which decays nearly tobaseline by the end of the imaging sequence. Examples ofcalcium responses in a rostral and a caudal medullary neuronare also shown (middle and bottom). In each case, the calciumdynamics are typical of somatic calcium signals (insets: noterapid rise, gradual recovery) observed in vertebrate neuronsboth in culture and in vivo (Fetcho and O’Malley 1995, 1997;O’Donovan et al. 1993; O’Malley 1994; O’Malley et al. 1996,1999). These prior studies demonstrated that somatic calciumsignals of this size occur only after the firing of one or moreaction potentials: in no case have we observed somatic calciumsignals in response to subthreshold stimuli.

To examine calcium responses with higher temporal resolu-tion, line-scans (i.e., rapid scanning of a single line across acell) were used to measure calcium dynamics with 2-ms reso-lution. A line-scan of a medullary cell type called RoM3L isshown in Fig. 2. A and B show the anatomical location of thecell and the position of a line that was repetitively scannedacross the soma (rotated 90° in 2B). In the line-scan image inFig. 2C, the green band is the cell body, and it shows a rapidfluorescence increase just after the end of the movement elic-ited by the head-tap (tap indicated by2). The underlying plotshows that fluorescence was near maximal by the time the fishstopped moving, indicating that calcium had begun enteringthe cell before the end of the movement. Because fluorescencecannot be measured when the larva is moving, we also exam-ined tap-elicited responses in curare-paralyzed fish (Fig. 2D).Here a similar calcium response is seen, but at a shorterlatency, in this instance occurring 12 ms after the onset of thestimulus artifact produced by movement of the tapper itself.

Calcium responses were examined in 15 cell types in 22different paralyzed fish. In the great majority of cases, fluores-cence increases were detected at latencies ranging between 5and 40 ms after the onset of the tapper movement (Fig. 2E).The variability in latency determination was due primarily todifferences in the duration of the tapper-movement artifact,which varied from preparation to preparation. Thus we cannotstate the actual latency of the onset of calcium increase but canonly provide an upper limit before which the response mayhave begun. The histogram is plotted in conjunction with anexample escape behavior, and it can be seen that most calciumresponses occurred by either the end of the C bend or duringthe counter-turn. The specific cells in which paralyzed-laten-cies were measured are noted in Fig. 3.

To assess the extent of activation of descending neurons, weimaged calcium responses in most of the distinct cell typesprojecting from brain into spinal cord. In attempting this sur-vey, our goal was simply to determine whether or not thesecells were responding to escape-eliciting stimuli rather than toextensively evaluate the response properties of one or just afew cell types. For this purpose, the criterion chosen was toexamine at least four individual cells of each type, distributedamong at least three different fish. Cell types that consistentlyexhibited calcium responses at the end of the movement arti-fact are color-coded green, while cells that usually showed nodetectable calcium response are coded red (Fig. 3A). All ofthese cell types are present in bilateral pairs but are only shown

609BRAIN STEM ESCAPE NETWORK IN ZEBRAFISH

J Neurophysiol • VOL 87 • JANUARY 2002 • www.jn.org

on Novem

ber 25, 2008 jn.physiology.org

Dow

nloaded from

unilaterally in this illustration. A small fraction of cell typeswith inconsistent responses or incomplete results are codedyellow; cells in the image that were not examined were leftwhite. Only a few cell types failed to respond to contralateralhead-taps, so ipsilateral stimuli were also tested in these cellsto confirm that they would not respond to head taps. The moststriking observation is that a majority of cells, and cell types,consistently responded to head taps.

Based on earlier categorizations, we distinguish 27 types ofreticulospinal cell (Metcalfe et al. 1986), plus 4 large nMLFcells, IC cells, T-reticular neurons, and vestibulospinal cells(Kimmel et al. 1985). Of these cell types, 17 or 55% consis-tently responded to head taps (Fig. 3B, green rows). An addi-tional seven cell types showed calcium responses but did notreach criterion (yellow rows): in some cases cells were encoun-

tered only infrequently (e.g., RoL2) and in other cases, theyexhibited calcium responses but not often enough to reach thecriterion established for inclusion in the responding category(e.g., MiV1). Nonetheless, all six of these equivocal cell typesshowed at least some clear calcium responses, and if we addthem to the responding classes, it would amount to a total of 24responding cell types. This would constitute 77% of all distin-guishable cell types that project from brain into spinal cord, butmore significantly, it amounts to 92% of the cell types exam-ined in this study. Only two cell types (6%) failed to respondconsistently, and five reticulospinal cell types were not exam-ined. Cell types that responded included: cells in all sevenhindbrain segments (ranging from RoL1 to CaD), all four ofthe larger cell types in nMLF (smaller nMLF cells were notexamined) and T-reticular cells. The IC cells were the largest

FIG. 1. Examples of calcium responsive neurons. Left: a composite of 2 confocal image stacks, showing the rostral-caudal extentof neurons that project from the brain into the spinal cord. The segmental arrangement of the neurons is indicated alongside theimage. Right: examples of calcium responses from 3 neurons located in midbrain or hindbrain, as indicated. Each neuron showsa robust calcium response (plotted in the inset) in response to a tap to the head (during the frame marked by an *). Frames wereacquired at 450-ms intervals and are plotted from left to right in consecutive rows. Fluorescence plots are background subtractedand normalized to the preceding baseline. Relative fluorescence intensity is indicated by the color scale at the bottom, which alsoapplies to Fig. 2.

610 GAHTAN, SANKRITHI, CAMPOS, AND O’MALLEY

J Neurophysiol • VOL 87 • JANUARY 2002 • www.jn.org

on Novem

ber 25, 2008 jn.physiology.org

Dow

nloaded from

group of cells that did not show tap-evoked calcium responses(vestibulospinal cells were not examined in this study). Thelarge majority of cells studied exhibited no spontaneous cal-cium activity when the larva was restrained on the stage of theconfocal microscope, so the observed calcium responses wereabsolutely correlated with the escape-eliciting stimulus.

D I S C U S S I O N

Previously, there was no physiological data on any of the 21cell types in the brain stem of larval zebrafish that we observedhere to respond to escape-eliciting stimuli. This number in-cludes the seven “equivocal” cell types but does not include thethree Mauthner-array cell types previously linked to C-startescape behaviors. In addition to constituting most of the ana-tomically distinct cell types that project from the brain intospinal cord, this group also includes a large majority of the

total number of descending neurons. All cell types are presentin bilateral pairs and there are multiple “copies” of many ofthem, as noted in Fig. 3B. We estimate that in larval zebrafishthere are, in total, �220 neurons that project from brain intospinal cord. If we assume that our recordings of an individualcell type are representative of all cells of that type, then we canestimate, for purposes of discussion, that we have sampledresponses of 140 neurons of the total population. Of these 140neurons, 94 (or 63%) consistently respond to escape elicitingstimuli and an additional 28 neurons (19%) exhibit calciumresponses at least some of the time. To summarize, our bestestimate is that, in larval zebrafish, escape-related activity canbe observed in 82% of the neurons projecting from brain intospinal cord. This result has several ramifications.

The brain stem escape network (BEN), as originally envi-sioned by Eaton and colleagues (Eaton et al. 2001; Foreman

FIG. 2. Example of line scans before and after curare. A: cell RoM3L (circled) is located rostral to and slightly medial to theMauthner cell (M). B: RoM3L is rotated 90° to show the orientation used in the line scans in C and D. The red scan line was scannedrepeatedly at 2-ms intervals. C and D: plotting of the line scans, using a 10-ms running average, shows the basal fluorescence inthat region of the cell, which is followed by an abrupt fluorescence increase immediately after the end of the movement artifact.Each line scan is 1-s long, running from left to right. The time of the head-tap is indicated (2). The calcium response is comparablebefore (C) and after (D) treatment with 3 mg/ml curare, but paralysis eliminates active movements of the fish, leaving only a briefermovement artifact caused by the tapper. In this example, fluorescence was elevated as soon as the cell had re-stabilized, reaching�10% above pretap baseline at 12 ms post-tap. E: summary of latencies to calcium responses. The upper plot shows a sampleC-start response to a puff of fluid aimed at the head of a free-swimming larva. The orientation of the rostral midline of the larvais plotted at 1-ms intervals and shows the C bend, the counter-turn, and the beginning of a burst swim bout. Shown below are thepooled latencies to calcium responses, grouped in 5-ms bins from all cells for which response latencies were determined inparalyzed larvae. These latencies are determined primarily by the time for the tapper movement artifact to cease, and so these valuesshould be taken only as an upper limit of the times at which the calcium responses began.

611BRAIN STEM ESCAPE NETWORK IN ZEBRAFISH

J Neurophysiol • VOL 87 • JANUARY 2002 • www.jn.org

on Novem

ber 25, 2008 jn.physiology.org

Dow

nloaded from

and Eaton 1993), included a small number of both ipsilaterallyand contralaterally projecting neurons. Liu and Fetcho (1999)’sresults indicated that additional neurons in brain stem arecapable of generating escape-like turns, i.e., turns that aremuch faster and larger than the other main type of larvalturning behavior, referred to as “routine” turns (Budick andO’Malley 2000b). From the Liu and Fetcho study, it seemedcertain that brain stem neurons in array-lesioned fish werefiring after an abnormally long delay (i.e., after the C bend andcounter-turn would normally be over), but it was not known ifthese putative escape-controlling neurons would normally fireconcurrently with the Mauthner array in nonlesioned fish. Ourresults show that many brain stem neurons exhibit escape-related calcium responses during either the C bend, or at latest,during the counter-turn. The large numbers of neurons acti-vated adds considerably to the possible anatomical foundationof the BEN. While it is not necessarily the case that all of theseneurons are directly involved with the generation and controlof escape locomotion, they all appear to be sending informa-

tion to spinal cord that is time-locked to the escape behavior.Because the neuroanatomical substrate underlying the escapebehavior appears to be quite similar for both larval and adultzebrafish and goldfish as well (Eaton et al. 2001; Lee and Eaton1991; Lee et al. 1993; Metcalfe et al. 1986), the calcium-imaging data may be generally applicable to teleost locomo-tion.

While we do not know the specific functional roles of theescape-responsive neurons, lesion studies suggest that they areindeed important for escapes. We have made repeated effortsto eliminate swimming and turning behaviors by laser-ablatingeither the Mauthner array � nMLF (Budick and O’Malley2000a) or larger groupings of neurons (Gahtan and O’Malley2000) and have not been able to eliminate any behavior,although we have observed Mauthner cell ablations to delaythe escape behavior (unpublished data in agreement with Eatonet al. 1982; Liu and Fetcho 1999). In addition, severing theaxons of yet larger numbers of reticulospinal neurons disruptsthe normal kinematics of the C start, producing an initially

FIG. 3. Summary of calcium responses in 26 distinct types of descending neurons. A: a majority of reticulospinal neurons andother cell types are activated by escape-eliciting stimuli. In this illustration, cells are color coded as responding (green), equivocal(yellow), or nonresponding (red). Both ipsilaterally projecting neurons (left side of image) and contralaterally projecting neurons(right side) respond to head taps. B: responses are summarized by cell type, with the total (unilateral) number of neurons of eachcell type present in the larva indicated in parenthesis. We distinguish here 30 distinct cell types, although nucleus of the mediallongitudinal fasciculus (nMLF) has additional cell types that were not examined. Vestibulospinal cells were also not included. Inthe third column, the number of responding cells of that cell type is shown followed by the total number of cells of that typeexamined. The fourth column shows the number of trials in which calcium responses occurred followed by the total number of trialsrecorded for all cells of that type. In next column, the average peak fluorescence response for all trials of that cell type is indicated.Average response latency (time after onset of movement to reach 10% above baseline) is shown under the last column for all celltypes where the effects of curare on latency were recorded. The cell types in B are color-coded according to the scheme used in A.

612 GAHTAN, SANKRITHI, CAMPOS, AND O’MALLEY

J Neurophysiol • VOL 87 • JANUARY 2002 • www.jn.org

on Novem

ber 25, 2008 jn.physiology.org

Dow

nloaded from

S-shaped bend (Gahtan and O’Malley 2001), which is consis-tent with the present imaging data indicating the involvementof a large fraction of descending neurons. One possibility isthat controls for different components of the escape behaviorare distributed throughout the brain stem. Some fraction ofthese neurons may contribute to directional control of theescape trajectory, which is quite variable (Budick andO’Malley 2000b; Foreman and Eaton 1993). Other neuronsmay help coordinate the sequencing of different components ofthe behavior, i.e., C bend, counter-turn, and burst swim. Si-multaneous high-speed behavioral imaging and confocal cal-cium imaging of neurons, in partially restrained larval ze-brafish, may prove useful for relating neuronal activity to thesespecific locomotive components (Bhatt et al. 2000). The factthat these neurons are distributed throughout all segments ofthe hindbrain has additional significance.

Vertebrate brains appear to develop by a common geneticplan that involves segmentation of the hindbrain (Keynes andLumsden 1990; Prince et al. 1998). This segmental organiza-tion appears important for a variety of neural operations thatunderlie rhythmic behaviors (Bass and Baker 1997). Perhapsthe clearest example of a segmentally organized system con-cerns the escape behavior, where the M cell and its twosegmental homologs appear to provide directional control overthe behavior (Foreman and Eaton 1993; Liu and Fetcho 1999;O’Malley et al. 1996). Because the much larger population ofcalcium-responsive neurons identified in the current study aredistributed across all hindbrain segments, these neurons mightbe viewed as part of a multisegmental network that collectivelycoordinates the larva’s escape behavior. This idea is consistentwith the proposal that there are multiple sets of segmentalhomologs (Metcalfe et al. 1986), which might generate seg-mental codes to shape different aspects of the behavior. Theresults of the laser-ablation and axon-severing experimentsalso support the idea that multisegmental networks of neuronscoordinate motor control. The extent to which this occurs inhigher vertebrate behaviors is currently not well understood inlarge part because of the difficulties in establishing the cellular-level organization of motor control in such animals (Kiehn etal. 1998).

A detailed cellular-level description is important not only foraddressing issues such as segmental coding but perhaps moreimportantly for defining the type or class of neural architecturebeing implemented. Aside from the Mauthner array’s controlof escape latency, the descending motor control system doesnot appear to be of the “dedicated” type, where specific neu-rons are exclusively involved in the control of specific behav-iors. Other types of neural architectures, e.g., “distributed” and“reorganizing” (reviewed in Morton and Chiel 1994; see alsoEaton and DiDomenico 1985; Leonard 2000), seem morelikely to fit with our behavioral and confocal observations.Indeed, our results are reminiscent of the widespread activationof hindbrain neurons by vestibular stimuli in lamprey (Delia-gina et al. 1992; Orlovsky et al. 1992) and the widespreadactivation of abdominal ganglion neurons in Aplysia (Wu et al.1994; Zochowski et al. 2000). The resiliency of the larvalbehaviors to extensive laser-ablations might also suggest aredundant type of neural architecture. The BEN is not a simplesystem. It is capable not only of coordinating the differentcomponents of the escape behavior but also of executing asensorimotor transformation to produce correctly sized and

timed C bends and counter-turns so as to generate an appro-priate escape trajectory (Eaton et al. 1991, 2001). More re-cently evolved vertebrates have far greater numbers and typesof descending neurons which are often intermingled (Brodal1981; Siegel and Tomaszewski 1983) and so, experimentally,pose a far greater challenge (see e.g., Kiehn et al. 1998). Webelieve that to accurately determine the type of neural archi-tecture mediating a behavior, it is essential to identify most (ifnot all) of the cell types involved in that behavior and to learnwhat each cell type contributes.

We appreciate the helpful comments of two anonymous reviewers.This work was supported by National Institute of Neurological Disorders

and Stroke Grants NS-37789 (D. M. O’Malley) and NS-11127 (E. Gahtan) anda fellowship from the US Department of Education/Advanced Scientific Com-putation Center at Northeastern University (N. Sankrithi).NOTE ADDED IN PROOF

A recent study on adult rainbow trout, using a complementarymethodology, also reported widespread activation of neurons in thebrain stem by startle responses (Bosch et al. 2001).

REFERENCES

BASS AH AND BAKER R. Phenotypic specification of hindbrain rhombomeresand the origins of rhythmic circuits in vertebrates. Brain Behav Evol 50,Suppl 1: 3–16, 1997.

BHATT DH, LIU KS, AND FETCHO JR. Identifying spinal interneurons involvedin motor behaviors of zebrafish by simultaneous calcium imaging and highspeed motion analysis. Soc Neurosci Abstr 26: 1995, 2000.

BOSCH TJ, MASLAM S, AND ROBERTS BL. Fos-like immunohistochemicalidentification of neurons active during the startle response of the rainbowtrout. J Comp Neurol 439: 306–314, 2001.

BRODAL A. Neurological Anatomy (3rd ed.). New York: Oxford Univ. Press,1981.

BUDICK SA AND O’MALLEY DM. Minimal behavioral deficits are observedafter laser-ablation of the nMLF in larval zebrafish. Soc Neurosci Abstr 26:158, 2000a.

BUDICK SA AND O’MALLEY DM. The behavioral repertoire of larval zebrafish:swimming, escaping and prey capture. J Exp Biol 203: 2565–2579, 2000b.

DELIAGINA TG, ORLOVSKY GN, GRILLNER S, AND WALLEN P. Vestibularcontrol of swimming in lamprey. II. Characteristics of spatial sensitivity ofreticulospinal neurons. Exp Brain Res 90: 489–498, 1992.

EATON RC AND DIDOMENICO R. Command and the neural causation of behav-ior: a theoretical analysis of the necessity and sufficiency paradigm. BrainBehav Evol 27: 132–164, 1985.

EATON RC, DIDOMENICO R, AND NISSANOV J. Role of the Mauthner cell insensorimotor integration by the brainstem escape network. Brain Behav Evol37: 272–285, 1991.

EATON RC, LAVENDER WA, AND WIELAND CM. Alternative neural pathwaysinitiate fast-start responses following lesions of the Mauthner neuron ingoldfish. J Comp Physiol 145: 485–496, 1982.

EATON RC, LEE RKK, AND FOREMAN MB. The Mauthner cell and otheridentified neurons of the brainstem escape network of fish. Prog Neurobiol63: 467–485, 2001.

EATON RC, NISSANOV J, AND WIELAND CM. Differential activation of Mau-thner and non-Mauthner startle circuits in the zebrafish: implications forfunctional substitution. J Comp Physiol [A] 155: 813–820, 1984.

FETCHO JR, COX K, AND O’MALLEY DM. Monitoring activity in neuronalpopulations with single-cell resolution in a behaving vertebrate. HistochemJ 30: 153–167, 1998.

FETCHO JR AND O’MALLEY DM. Visualization of active neural circuitry in thespinal cord of intact zebrafish. J Neurophys 73: 399–406, 1995.

FETCHO JR AND O’MALLEY DM. Imaging neuronal networks in behavinganimals. Curr Opin Neurobiol 7: 832–838, 1997.

FOREMAN MB AND EATON RC. The direction change concept for reticulospinalcontrol of goldfish escape. J Neurosci 13: 4101–4113, 1993.

GAHTAN E AND O’MALLEY DM. Analysis of spontaneous activity in controland reticulospinal (RS) ablated zebrafish larvae. Soc Neurosci Abstr 26: 158,2000.

GAHTAN E AND O’MALLEY DM. Rapid lesioning of large numbers of identifiedvertebrate neurons: applications in zebrafish. J Neurosci Methods 108:97–110, 2001.

613BRAIN STEM ESCAPE NETWORK IN ZEBRAFISH

J Neurophysiol • VOL 87 • JANUARY 2002 • www.jn.org

on Novem

ber 25, 2008 jn.physiology.org

Dow

nloaded from

KEYNES R AND LUMSDEN A. Segmentation and the origin of regional diversityin the vertebrate central nervous system. Neuron 2: 1–9, 1990.

KIEHN O, HARRIS-WARRICK RM, JORDAN LM, HULTBORN H, AND KUDO N.Neuronal mechanisms for generating locomotor activity. Ann NY Acad Sci860, 1998.

KIMMEL CB, EATON RC, AND POWELL SL. Decreased fast-start performance ofzebrafish larvae lacking Mauthner neurons. J Comp Physiol [A] 140: 343–350, 1980.

KIMMEL CB, METCALFE WK, AND SCHABTACH E. T-reticular interneurons: aclass of serially repeating cells in the zebrafish hindbrain. J Comp Neurol233: 365–376, 1985.

LEE RKK AND EATON RC. Identifiable reticulospinal neurons of the adultzebrafish Brachydanio rerio. J Comp Neurol 304: 34–52, 1991.

LEE RKK, EATON RC, AND ZOTTOLI SJ. Segmental arrangement of reticulospi-nal neurons in the goldfish hindbrain. J Comp Neurol 329: 539–556, 1993.

LEONARD JL. Network architectures and circuit function: testing alternativehypotheses in multifunctional networks. Brain Behav Evol 55: 248–255,2000.

LIU K AND FETCHO JR. Laser ablation reveals functional relationships ofsegmental hindbrain neurons in zebrafish. Neuron 23: 325–335, 1999.

METCALFE WK, MENDELSON B, AND KIMMEL CB. Segmental homologiesamong reticulospinal neurons in the hindbrain of the zebrafish larva. J CompNeurol 251: 147–159, 1986.

MORTON DW AND CHIEL HJ. Neural architectures for adaptive behavior.Trends Neurosci 17: 413–420, 1994.

NISSANOV J, EATON RC, AND DIDOMENICO R. The motor output of the Mauth-ner cell, a reticulospinal command neuron. Brain Res 517: 88–98, 1990.

O’DONOVAN MJ, HO S, SHOLOMENKO G, AND YEE W. Real-time imaging ofneurons retrogradely and anterogradely labeled with calcium sensitive dyes.J Neurosci Methods 46: 91–106, 1993.

O’MALLEY DM. Calcium permeability of the neuronal nuclear envelope:evaluation using confocal volumes and intracellular perfusion. J Neurosci14: 5741–5758, 1994.

O’MALLEY DM AND FETCHO JR. The zebrafish hindbrain: a transparent systemfor imaging motor circuitry. In: Imaging Living Cells, edited by Yuste R,

Lanni F, and Konnerth A. Cold Spring Harbor, NY: Cold Spring HarborLaboratory Press, 2000, p. 14.1–14.12.

O’MALLEY DM, BURBACH BJ, AND ADAMS PR. Fluorescent calcium indicators:subcellular behavior and use in confocal imaging. In: Confocal Microscopy:Methods and Protocols, edited by Paddock S and Totowa T. Clifton, NJ:Humana, 1999, p. 261–303.

O’MALLEY DM, KAO Y-H, AND FETCHO JR. Imaging the functional organiza-tion of zebrafish hindbrain segments. Neuron 17: 1145–1155, 1996.

ORLOVSKY GN, DELIAGINA TG, AND WALLEN P. Vestibular control of swim-ming in lamprey. I. Responses of reticulospinal neurons to roll and pitch.Exp Eye Res 90: 479–488, 1992.

PRINCE VE, MOENS CB, KIMMEL CB, AND HO RK. Zebrafish hox genes:expression in the hindbrain region of wild-type and mutants of the segmen-tation gene valentino. Development 125: 393–406, 1998.

SIEGEL JM AND TOMASZEWSKI KS. Behavioral organization of reticular forma-tion: studies in the unrestrained cat. I. Cells related to axial, limb, eye andother movements. J Neurophysiol 50: 696–716, 1983.

WESTERFIELD M. The Zebrafish Book: A Guide for the Laboratory Use ofZebrafish (Brachydanio rerio) (3rd ed.). Eugene, OR: University of OregonPress, 1995.

WU J-Y, COHEN LB, AND FALK CX. Neuronal activity during differentbehaviors in Aplysia: a distributed organization? Science 263: 820–823,1994.

ZOCHOWSKI M, COHEN LB, FUHRMANN G, AND KLEINFELD D. Distributed andpartially separate pools of neurons are correlated with two different com-ponents of the gill-withdrawal reflex in Aplysia. J Neurosci 20: 8485–8492,2000.

ZOTTOLI SJ. Correlation of the startle reflex and Mauthner cell auditoryresponses in unrestrained goldfish. J Exp Biol 66: 243–254, 1977.

ZOTTOLI SJ AND FABER DS. The Mauthner cell: what has it taught us?Neuroscientist 6: 26–38, 2000.

ZOTTOLI SJ, NEWMAN BC, RIEFF HI, AND WINTERS DC. Decrease in occurrenceof fast startle responses after selective Mauthner cell ablation in goldfish(Carassius auratus). J Comp Physiol [A] 184: 207–218, 1999.

614 GAHTAN, SANKRITHI, CAMPOS, AND O’MALLEY

J Neurophysiol • VOL 87 • JANUARY 2002 • www.jn.org

on Novem

ber 25, 2008 jn.physiology.org

Dow

nloaded from