Eye position and eye velocity integrators reside in separate brainstem nuclei.

P oc. Na l. Acad. Sci. USA Vol. 91, pp. 807-811, Janua y 1994 Neu obiology Eye posi ion and eye eloci y in eg a o s eside in sepa a e b ains em nuclei (gaze s abiliza ion/mo o pe o mance/in eg a o hy hmici y/p eposi us nudeus/lidocaine inac i a ion) ANGEL M. PASTOR*, ROSA R. DE LA CRUZ*, AND ROBERT BAKER *Labo a o y o Neu oscience, Depa men o Physiology and Biology, Uni e si y o Se ille, 41012 Se ille, Spain; and Depa men o Physiology and Biophysics, New Yo k Uni e si y Medical Cen e , New Yo k, NY 10016 Communica ed by R. Llinds, Augus 25, 1993 ABSTRACT Two ypes o cen al ne ous sys em in eg a- o s a e c i ical o oculomo o pe o mance. The is in e- g a es eloci y commands o c ea e posi ion signals ha hold ixa ion o he eye. The second s o es ela i e eloci y o he head and isual su ound o s abilize gaze bo h du ing and a e he occu ence o con inuous sel and wo ld mo ion. We ha e used eco dings om single neu ons o es ablish ha he "posi ion" and " eloci y" in eg a o s o ho izon al eye mo e- men occupy adjacen , bu nono e lapping, loca ions in he gold ish medulla. Lidocaine inac i a ion o each in eg a o esul s in he eye mo emen de ici s expec ed i ho izon al eye posi ion and eloci y signals a e p ocessed sepa a ely. These obse a ions also indica e ha each b ains em compa men gene a es and s o es hese signals. Consequen ly, each in eg a- o exhibi s unc ional au onomy. The e o e, we p opose ha he in insic elec ophysiological p ope ies o he cons i uen neu ons in each b ains em subnucleus may be su icien o p oducing in eg a o hy hmici y. Main aining gaze on a ge s ha mo e ela i e o he ob- se e equi es ha es ibula and isual e lexes de ec ing bo h sel and wo ld mo ion espond wi h app op ia e eye and/o head mo emen s. Neu al in eg a o s ac ing as com- pu a ional elemen s a e essen ial building blocks wi hin his senso y-mo o ans o ma ion (1-3). The concep o a "neu- onal" in eg a o whose inpu -ou pu cha ac e is ics could be desc ibed by a ma hema ical in eg al was i s en isioned a e obse ing he ac i i y o abducens mo oneu ons du ing changes in ho izon al eye posi ion (4). Cu en hypo heses equi e neu al ne wo ks o cascading, e e be a ing colla - e als wi h posi i e and/o nega i e eedback; howe e , pu- a i e neu onal mechanisms, le alone causal chains o syn- ap ic connec ions, a e s ill no ional in he ope a ion o in eg a o s o ei he spinal o b ains em ci cui s (5-7). He e we desc ibe he p ecise loca ion o hindb ain neu ons ha a e necessa y and su icien o accomplishing wo dis inc ypes o in eg a ion in he oculomo o sys em. In mammals, he p eposi us hypoglossi nucleus was ini- ially p oposed o ul ill he c i e ia pos ula ed as necessa y o he " eloci y o posi ion" ans o ma ion (8-11). Recen pha macological and lesion s udies implica e bo h he es i- bula and p eposi us egions o he b ains em (12-14). Join ly hese da a sugges ha all ho izon al eye mo emen - ela ed subsys ems (e.g., saccadic, es ibula , and isual) p obably sha e a common posi ion in eg a o (15, 16). Un o una ely, i has no been possible o causally place any iden i ied es ibula o p eposi us neu on wi hin his ci cui (5). The obse a ions ha es ibula and isual senso y signals ela ed o sel o ex e nally gene a ed mo emen a e la gely encoded in a eloci y domain (17, 18) led o iden i ica ion o a second ype o in eg a o ha could accumula e eye eloc- The publica ion cos s o his a icle we e de ayed in pa by page cha ge paymen . This a icle mus he e o e be he eby ma ked "ad e isemen " in acco dance wi h 18 U.S.C. §1734 solely o indica e his ac . i y in he ho izon al plane (19, 20). In his case, neu al in o ma ion p opo ional o he eloci y o he head and/o isual su ound is compiled and p ese ed cen ally (i.e., s o ed up). By discha ging he s o ed eloci y a e cessa ion o he ini ia ing senso y s imulus, his in eg a o ex ends he ime o oculomo o compensa ion o ei he head o isual wo ld mo ion. The loca ion o his in eg a o , le alone i s ope a ion, has emained qui e elusi e, al hough mos e i- dence sugges s a c ucial ole o he es ibula nuclei (14, 21). In his pape , he posi ion and eloci y in eg a o s will be shown o occupy spa ially sepa a e, non es ibula , loci in he gold ish medulla ha can be accu a ely iden i ied, eco ded om, and e e sibly inac i a ed. MATERIALS AND METHODS Gold ish (Ca assius au a us) we e p epa ed unde gene al anes hesia ( icaine me hanesul ona e; 1:20,000, w / ol) o eye mo emen and neu onal eco ding. Su ge y consis ed o implan ing an ac ylic pedes al o head s abiliza ion and ephining a hole in he occipi al bone o eco ding b ains em neu al ac i i y (22). Eye posi ion was moni o ed wi h scle al sea ch coils. A e a eco e y pe iod o se e al days, ully ale animals we e es ed in a e ical-axis op okine ic and es ibula s imula o . This sys em consis ed o a se o- con olled plane a ium and o a ing able in e aced by a wa e o m gene a o ha p oduced any desi ed in e ac ion in phase and/o ela i e eloci y be ween he wo s imuli (23). Compensa o y eye eloci y is nea ly o equal ampli ude, bu opposi ely di ec ed, o head eloci y and exhibi s nea ly 1800 phase shi wi h espec o he s imulus [e.g., es ibuloocula e lex (VOR) gain o 1.0 in Fig. 2 B and D]. Neu onal ac i i y was eco ded du ing spon aneous and/o isual- es ibula -induced e lex eye mo emen s (see Fig. 2). Ins an aneous i ing a e his og ams we e con- s uc ed as he ecip ocal o he in e spike in e al ( e . 22; see Fig. 2 A and B, FR). To p oduce local anes hesia, mic opipe es illed wi h 4% lidocaine and be eled o a ip size o 10 ,Am we e si ua ed a he physiological cen e o each iden i ied a ea by eco ding he su ounding ex acellula ac i i y. On a e age, 1 nl o lidocaine was injec ed wi h 5- o 10-ms, 200-kPa ai p essu e pulses o e 30 s (see Figs. 3 and 4). Fo bila e al inac i a ion, wo sepa a e injec ions we e comple ed wi hin 2 min. Biocy in dissol ed in 0.1 M phospha e bu e was ei he p essu e injec ed o ion opho esed in o he caudal lobe o he ce ebellum o label b ains em nuclei (see Fig. 1 B-E). A e pe usion wi h eleos saline and aldehyde ixa i e, 50-,um sec ions we e eac ed wi h a idin-bio in-pe oxidase com- plex (Vec o Labo a o ies) ollowed by diaminobenzidine his ochemical analysis ( e . 22; see Fig. 1 C-F). Abb e ia ion: VOR, es ibuloocula e lex. 807 P oc. Na l. Acad. Sci. USA 91 (1994) RESULTS Fou hindb ain nuclei dis ibu ed pe iodically a -500-,um in e als be ween he obex and he abducens nucleus we e designa ed om caudal o os al as a eas I-IV (Fig. 1 A and B). Each subg oup con ained a o al o 25-40 neu ons ha we e in ol ed in ei he posi ion (a eas I and III) o eloci y (a eas II and IV) in eg a ion o oculomo o (a eas I and II) and pos u al (a eas III and IV) con ol. Biocy in injec ions in he es ibuloce ebella lobe selec i i y labeled neu ons in he in e io oli e and a ea II (Fig. 1 C-F). Elec ical s imula ion o he ce ebellum (Fig. 1B, CL) p oduced an id omic ac i- a ion o a ea II eye eloci y-only neu ons desc ibed in Fig. 2 (32). A eas I and II we e si ua ed di ec ly abo e he in e io oli e as shown in bo h sagi al (Fig. 1 C and F) and co onal (Fig. 1 D and E) iews o he b ains em. The longi udinal dis ibu ion o he ou hindb ain nuclei could be ecognized in ela ionship o a se o epea ing ascula b anches ha was con i med by neu ophysiological co ela es in e e y animal. Neu ons we e iden i ied in each o he ho izon al eye mo emen - ela ed a eas I-IV wi h ex acellula single uni eco dings (Fig. 2). Discha ge p ope ies we e highly co e- la ed in he wo pos e io a eas (I and II) wi h eye mo emen pa ame e s, while ela ionships we e less ob ious in he wo os al a eas (III and IV). In he mos caudal subnucleus (a ea I), neu onal i ing a e du ing spon aneous eye mo emen s inc eased in p opo ion o he angle o eye de ia ion owa d he ipsila e al side o eco ding. Du ing di e gen saccades, _ F P... o p.X_S .. .. w. .. s C IV .. .............. a ea I neu ons encoded he posi ion o he ipsila e al eye (Fig. 2A; dashed lines). The majo i y o a ea I neu ons (70%) demons a ed a sensi i i y p opo ional o he eloci y o he eyes du ing saccades (Fig. 2A, a ow). Some o hese neu- ons we e pu ely posi ion ela ed (30%) since he eloci y sensi i i y du ing apid eye mo emen s was negligible (Fig. 2B, a ows). Du ing he slow phase o ei he es ibula (Fig. 2B) o op okine ic nys agmus, i ing a e o neu ons in a ea I was p opo ional o eye posi ion, exhibi ing li le eye eloci y sensi i i y. A ea II was loca ed immedia ely os al o a ea I. Neu ons in his a ea ei he emained silen o showed a s eady i ing a e du ing spon aneous saccadic eye mo emen (Fig. 2C). A ea II neu ons modula ed in phase wi h eye eloci y di- ec ed owa d he ipsila e al side o eco ding du ing he slow componen o ei he op okine ic (Fig. 2D) o es ibula nys agmus. The wo main ypes o neu onal esponses de- ec ed in a ea II we e eye eloci y only (51%) and eye combined wi h a head eloci y sensi i i y measu ed du ing isual supp ession o he VOR (49%). Some o hese neu ons also paused o all as phases (eye eloci y pause, 25%). Du ing ei he sinusoidal op okine ic o es ibula s imula- ion, he i ing a e o neu ons in a ea I peaked in phase wi h eye posi ion (Fig. 2B, do ed line), while neu ons in a ea II modula ed in phase wi h eye eloci y (Fig. 2D, do ed line). Thus, he main physiological di e ence be ween a eas I and II was an exquisi e seg ega ion o eye posi ion and eye eloci y signals. Ne e heless, neu ons in bo h a eas we e conside ed p emo o since hey i ed 15-20 ms p io o he FIG. 1. O ganiza ion o hind- :: :b ain nuclei in he gold ish. Whole + b ain (A) and sagi al Nissl sec- ions (B) showing he os ocaudal and en al loca ion o he majo $- FL V: j es ibula subdi isions (24) and '. ' .i3-..a eas I-IV. (C-F) Neu ons, ax- ons, and p ocesses labeled a e biocy in injec ion in o he es ib- uloce ebellum (CL). (C) Pa asag- i al sec ion 300 Pm om he mid- line showing he os al in e io oli e (TO) decussa ion and neu- *g _ s ons in a ea II. Mossy ibe pa h- ways (a) and g anule cell (g) clus- e s a e ma ked. (F) High magni- S # :: ica ion o a ea II (s a in C) illus a ing soma and dend i es o pu a i e eye eloci y in eg a o neu ons. (D) Co onal sec ion showing axons (a) o labeled con- ala e al IO neu ons c ossing he midline and he ipsila e al en o- medial loca ion o pu a i e a ea IT eye eloci y neu ons. (E) Clus- e ed somas (s) and dend i es (d) o a ea II neu ons a e shown a highe magni ica ion (s a in D). A, D, P, M, and T, an e io , de- scending, pos e io , magnoceliu- la , and angen ial subdi isions o es ibula nuclei; VC, CL, CC, and CB, al ula, caudal lobe, ce - -4i* d s ¢ ¢ * < ebella c es , and co pus o he ce ebellum; ABD, abducens; OC and T, oculomo o and ochlea nuclei; FL, acial lobe; I-IV, a eas I-IV o he hindb ain; OT, op ic ec um; SC, spinal co d; TE, elencephalon; VIIIn, oc a ola - e alis ne e; Xn, agal ne e; VL, agal lobe. 808 Neu obiology: Pas o e al. P oc. Na l. Acad. Sci. USA 91 (1994) 809 C I D 3000ms # ~~~~~~~~~~~~~ L Z . *. 100~~~~~~~~~~~~~~~~~oSp/s. .~0 FIG. 2. Discha ge cha ac e is- ics o a ea I (A and B) and a ea II (C and D) neu ons. (A) Fi ing a e (FR) o an a ea I neu on du ing spon aneous eye mo emen s. Dis- cha ge was co ela ed o le eye posi ion (LE) du ing ixa ion (dashed lines) and some neu ons also exhibi ed saccade sensi i i y (a ow). (B) FR o a pu ely posi- ion- ela ed a ea I neu on (a ows) du ing sinusoidal head o a ion ( i) in he da k. Head eloci y is in e ed in B and D o acili a e compa ison o eye eloci y. FR was associa ed wi h ho izon al eye posi ion (LE) bu no eye e- loci y (do ed line). (C) A ea II neu ons we e no modula ed du - ing spon aneous eye mo emen s. (D) Du ing head o a ion a ea II neu ons modula ed in phase wi h eye eloci y (1:E, do ed line) bu no eye posi ion (LE). Calib a- ions a e indica ed. occu ence o saccades, as phases, o eye esponses o sudden s ep o a ions o he head (da a no shown). Du ing spon aneous eye mo emen s, mos neu ons in a ea III (75%) exhibi ed an i egula i ing a e poo ly co ela ed ( < 0.7) wi h ipsila e al eye posi ion. A weak eye eloci y sensi i i y was de ec ed du ing es ibula and op okine ic e lexes. The second ype (25%) exhibi ed a simila p o ile bu paused o saccades and as phases in all di ec ions. A ea IV (100%) was composed o eye eloci y-like neu ons ha inc eased i ing du ing con ala e al eye mo emen . Th ee sepa a e clus e s o medially loca ed bu s ing neu ons we e si ua ed be ween a eas I and IV. E e y bu s neu on be ween a ea I and II i ed 30-40 ms be o e he occu ence o all ipsila e ally di ec ed as phases and saccades. Bu s neu ons loca ed be ween a eas II and III exhibi ed a a iable la ency and co ela ion wi h saccades while hose be ween a eas III and IV we e bidi ec ional. Selec i e lidocaine inac i a ion o he ou medulla y a eas was used o assess he po en ial con ibu ion o cons i uen neu ons o he in eg a ion p ocess. The inac i a ion o ei he a ea I o II, bu no a ea III o IV, p oduced quali a i ely dis inc changes in eye mo emen pa ame e s (Figs. 3 and 4). Bila e al lidocaine injec ion o a ea I p oduced se e e gaze- holding de ici s o ho izon al eye mo emen s. By 2 min a e he injec ion, and du ing he 20 min o e ec i e inac i a ion, ho izon al saccades o bo h eyes and in bo h di ec ions we e ollowed by an exponen ial d i di ec ed asymp o ically owa d a null poin (Fig. 3A). The exponen ial ime cou se o he d i can be in e p e ed as he passi e ecen e ing o he eye due o o bi al iscoelas ic issue o ces. The ime con- s an o he exponen ial d i was p og essi ely sho ened owa d a minimum o 0.3 s o e he i s 10 min (Fig. 3B, 10 min). Then he ime cons an p og essi ely eco e ed o con ol o e he nex 20 min (Fig. 3B, 30 min). Thus, inac i a ion o he in eg a o appea s o emo e he posi ion signals a ailable o he mo oneu ons o main aining he ho izon al angle o gaze. Bila e al inac i a ion o a ea I also p oduced de as a ing e ec s on bo h he VOR and op okine ic e lex. Sinusoidal s imula ion a low equency (s0.25 Hz) g ea ly educed VOR gain om 1.0 o an a e age o 0.2 and eye eloci y led able eloci y by 70° (Fig. 3C, do ed lines). A highe o a ional equencies, gain and phase p og essi ely ap- p oached ypical alues (Fig. 3D, do ed line). Consequen ly, VOR measu emen s a w1 Hz exhibi ed a no mal senso y o mo o ans o ma ion in he absence o any cen al neu onal p ocessing o eloci y o posi ion signals. Eye mo emen de ici s induced a e unila e al lidocaine inac i a ion o a ea I we e ne e as ma ked as hose ollowing bila e al injec ion and he ime cons an o he pos saccadic d i was ne e lowe ed o <1.0 s. Compa ison o da a a e many e e sible uni- and/o bila e al inac i a ions demon- s a ed ha a single in ac a ea I could p ocess posi ion signals capable o d i ing bo h eyes. Independen in eg i y o he bila e ally loca ed posi ion in eg a o s implies ha he in e connec ions be ween he wo a e no c i ical o he eloci y o posi ion ans o ma ion. By con as , unila e al es ibula and/o p eposi us lesions in mammals we e e- po ed as su icien o p oduce a o al gaze-holding ailu e (21, 25). Lidocaine inac i a ion o a ea II also p oduced dis inc oculomo o abno mali ies. A single unila e al lidocaine in- jec ion in a ea II caused a nys agmus o he wo eyes wi h as phases always di ec ed owa d he side o he inac i a ion (Fig. 4A). The saw oo h mo emen o he eyes likely o igi- na ed om imbalancing he eloci y ou pu om he bila e al a ea II nuclei (Fig. 2). The esul ing eye eloci y bias was maximal in he da k, 16-22°/s du ing he slow phase o nys agmus; howe e , in he p esence o an illumina ed s a- iona y backg ound, he bias was educed o :4°/s, indica - ing isual supp ession h ough he in ac a ea II (Fig. 4A, dashed line). This amp-like eye mo emen (Fig. 4B) was clea ly dis inguishable om he exponen ial d i in eye posi ion ollowing inac i a ion o a ea I (Fig. 3B). Du ing sinusoidal head o a ion in he da k, he eloci y bias shi ed he eye eloci y by a cons an alue o =20O/s (Fig. 4C, a ows). All he as phases o nys agmus occu ed in he di ec ion opposi e he eloci y bias because eye eloci y could no c oss ze o alue (Fig. 4C, LE). A e bila e al lidocaine inac i a ion o a ea II, saccades and ixa ions exhibi ed li le sign o ei he bias o gaze- holding ailu e (Fig. 4D). Ne e heless, bila e al anes hesia o a ea II maximally comp omised bo h he VOR and he op okine ic e lex. Eye eloci y p ocessing was o ally abol- ished du ing ei he sinusoidal o s ep o a ion o he head. The no mal ime cons an o he eloci y s o age in eg a o was educed om 10-12 s (Fig. 4E, a ow) o 0.3 s (Fig. 4F, a ows). By con as , he posi ion in eg a o con inued o pe o m one-s ep in eg a ion o he eloci y- ela ed signals because bo h eyes ollowed wi h an eye posi ion s ep esem- bling he s imulus eloci y p o ile (Fig. 4F, RE and LE). B Neu obiology: Pas o e al. 50SP/ W%l i P oc. Na l. Acad. Sci. USA 91 (1994) B 1oo. 0- 100- 50 50 50] 16~ O- w ~~~~. 500ms, 6E FIG. 3. E ec s o bila e al lidocaine injec ion in a ea I. (A) Ho izon al posi ion o he le (LE) and igh (RE) eyes 12 min a e a ea I inac i a ion. A e saccades in bo h di ec ions he eyes d i ed cen ipe ally wi h an exponen ial p o ile exhibi ing a ime cons an o 0.3 s. (B) Time cou se in minu es illus a ing inac i a ion and eco e y a e bila e al a ea I lidocaine injec ion. Saccades o simila size di ec ed le wa d we e supe imposed a he poin o peak ampli ude o illus a e sho ening o he eloci y o posi ion in eg a- o ime cons an . (C and D) E ec s o a ea I inac i a ion on he VOR a low (0.125 Hz; C) and high (1 Hz; D) equency. Lidocaine injec ions o 1 nl in o ei he a ea III o a ea IV did no in luence eye mo emen s. La ge amoun s (10 nl) p oduced some d i and eloci y bias, likely due o sp ead in o adjacen nuclei and/o ibe ac s. Lidocaine injec ions o 1 nl in o he medial bu s neu ons be ween a ea I and II nea ly abolished saccades and as phases bu did no a ec in eg a o ope a ion. Inac i a ion o a ea II did no in e up he es ibula con- nec ion o he posi ion in eg a o . These esul s a e no consis en wi h models (26) o lesion s udies (21) ha place he si e o eloci y s o age in a sys em o ecip ocal com- missu al connec ions be ween he es ibula complex. DISCUSSION Ou da a demons a e ha he neu al mechanisms equi ed o bo h eloci y o posi ion and eloci y s o age in eg a ion eside in adjacen medulla y nuclei clea ly isola ed om, bu likely in ima ely connec ed o, he es ibula nuclei. The spa ial disc e eness o hese in eg a o s co esponds o con- s an mo phological landma ks in he gold ish medulla ha e lec he adul e en ion o a segmen al bluep in desc ibed in la al zeb a ish (27). We p opose ha hese hindb ain nuclei in gold ish a e analogous in unc ion o he mammalian p eposi us nuclei and ha homologous neu onal ypes can be e en ually ecognized. The e o e, he hindb ain opog aphy in eleos s (Fig. 1 A and B) is mo e ad an ageous o esol ing neu ophysiological co ela es as well as in es iga ing he de elopmen al and gene ic plans unde lying assembly o hindb ain ho izon al eye mo emen pa hways (28). In each disc e e nono e lapping compa men , he quan- i a i e analysis o discha ge cha ac e is ics and oculomo o de ici s a e lidocaine injec ions we e complemen a y. A ea I and a ea II neu ons we e co ela ed wi h eye posi ion and eloci y, espec i ely. Signal p ocessing appea ed o be au onomous o each nucleus. Since a ea I esponses we e co ela ed o saccades, VOR, and isually induced e lexes, he posi ion in eg a o is indeed common o all eye mo e- men s (15). Ne e heless, a ea I p obably ep esen s he summing junc ion o wo qui e di e en modali ies o eye eloci y signal because as phases and spon aneous saccadic eye mo emen s we e spa ed a e bila e al lidocaine inac i- a ion o a ea II (Fig. 4D). Thus, es ibula and op okine ic eye eloci y signals p ocessed in a ea II each a ea I by a pa hway dis inc om ha o bu s neu ons p oducing sac- cades. An equally likely supposi ion is ha a ea II also ep esen s a summing junc ion o es ibula and isual eye eloci y, because hese neu ons do no i e o ei he sac- cades o as phases (Fig. 2 C and D). Homogenei y o discha ge pa e n and inac i a ion wi h lidocaine a gues ha a ea II plays wo sequen ial oles in eye mo emen . We sugges ha he neu ons i s gene a e and hen s o e eye eloci y since lidocaine inac i a ion no only abolished eloci y s o age bu also educed he VOR ime cons an o alues less han he a e age es ibula a e en ime cons an (0.3 s s. 3-4 s, espec i ely). Inac i a ion o a ea II g ea ly impai s bo h es ibula and op okine ic e- lexes because all neu ons in a ea II exhibi a con inuous le el o onic ac i i y (Fig. 2C). Hence, he VOR ime cons an s a e lowe han ha expec ed om p edic ions based on a e aged p ima y a e en discha ges (29), and, in ac , many species o eleos exhibi ypical VOR ime cons an s a less han mean a e en alues (30). This lowe ed VOR ime cons an was accompanied by a ma ked comp omise in he VOR when es ed wi h lowe equency sinusoidal s imuli. By con as , he di ec , p esumed h ee-neu on, VOR pa h- way emained in ac as indica ed by he la ency and ampli- ude o he eye posi ion s ep ( e . 23; Fig. 4F). We sugges ha a ea II eloci y neu ons comp ise a b ains em loop ha suppo s he di ec VOR pa hway loca ed in he descending oc a al nucleus (Fig. 1 A-C). O e all, he mos s aigh o wa d a gumen is ha a ea I and a ea II a e independen eye posi ion and eloci y gen- e a o s. Assu edly, a ea I is he equisi e posi ion in eg a o . P incipally, a ea II con e s senso y signal eloci y ( es ib- ula and isual) o mo o (eye) commands as well as assuming he impo an ole o eye eloci y s o age. The e o e, he pa e ns o neu al ac i i y wi hin each in eg a o can be p oposed o a ise om in insic p ope ies o hei cons i uen neu ons as opposed o in e ac ions wi h ex insic synap ic ci cui y. One-s ep signal gene a ion and s o age is an a ac i e idea because he same neu ons used o c ea e can also s o e he neu al ac i i y co ela ed wi h ei he posi ion o eloci y. Such neu onal beha io is p edic ed based on he s uc u al uniqueness, homogenei y, and clea sepa a ion o he indi- idual a eas om each o he and he es ibula nuclei (Fig. 810 Neu obiology: Pas o e al. P oc. Na l. Acad. Sci. USA 91 (1994) 811 100] 0 D LE l A, 80 I RE ls 1). Midline sagi al lesions be ween hese a eas do no block ei he posi ion o eloci y s o age (da a no shown); howe e , es ibula pa hways a e highly p obable h ough mo e os al commissu es (Fig. 1B). The medio en al loca ion o hese hindb ain nuclei and hei sepa a e inac i a ion clea ly dem- ons a e ha nei he axonal pa hways no neu ons wi hin he es ibula complex we e a ec ed by lidocaine (Figs. 3 and 4). We belie e, he e o e, ha all o he s uc u al and unc ional cha ac e is ics expec ed o each in eg a o a e embodied by neu ons cen alized wi hin a single nucleus. Since hese unusual compa men s a e necessa y and p e- sumed su icien o p ocess he neu onal ope a ions elemen- a y o in eg a ion, we hypo hesize ha ans o ma ions wi hin each compa men con e au o hy hmic cha ac e is- ics ha a e la gely modula ed, no commanded, by ex e nal synap ic ci cui y. In p inciple, he inhe en elec o espon- si e p ope ies o he neu ons alone may be su icien o p oduce in eg a ion (31). The e o e, we p opose ha signal ans o ma ions may la gely a ise om he in insic elec o- physiological p ope ies o cons i uen neu ons (32). The di e si y o eye mo emen s and he a ied in eg a o ime cons an s ha exis be ween eleos ean species should pe mi hese s uc u al and neu ophysiological hypo heses o be sa is ac o ily se led. We hank D s. M. U. L. Benne , B. Cohen, and T. Raphan o discussion and commen s on his manusc ip . 1. Robinson, D. A. (1%8) Science 161, 1219-1224. 2. Raphan, T., Ma suo, V. & Cohen, B. (1979) Exp. B ain Res. 35, 229-248. 3. Raphan, T. & Cohen, B. (1981) in Models o Oculomo o Beha io and Con ol, ed. Zube , B. L. (CRC, Boca Ra on, FL), pp. 91-109. 4. Ska enski, A. A. & Robinson, D. A. (1973) J. Neu ophysiol. 36, 724-738. 5. Robinson, D. A. (1989) Annu. Re . Neu osci. 12, 33-45. 6. Cannon, S. C., Robinson, D. A. & Shamma, S. (1983) Biol. Cybe n. 49, 127-136. 7. Anas asio, T. J. & Robinson, D. A. (1989) Biol. Cybe n. 61, 79-89. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 10°- O- 4s . J FIG. 4. E ec s o unila e al (A-C) I and bila e al (D-F) lidocaine inac i- a ion o a ea II. (A) Unila e al inac- i a ion o he le a ea II p oduced nys agmus o he le in he da k, bu slow phases we e a enua ed in he ligh (dashed line). (B) Supe imposi- ion o as phases demons a es J amp-like eye mo emen s a e inac- 1 11 I i a ion. (C) Du ing head o a ion in 4s . he da k eloci y bias added o he sinusoidal nys agmus in bo h eyes. l (D) Bila e al lidocaine injec ion p o- 8 duced nei he d i no eloci y bias du ing spon aneous eye mo emen s. (E) Pe - o a o y nys agmus (a ow) du ing cons an head eloci y exhib- 1 i ed a no mal ime cons an o 10-12 s. H 320/s (F) Bila e al lesion o a ea II i ually ] elimina ed eloci y s o age in bo h di ec ions (VOR ime cons an shown l /; by a ows) bu eloci y o posi ion in eg a ion was una ec ed [le eye (LE) and igh eye (RE)]. L6pez-Ba neo, J., Da lo , C., Be hoz, A. & Bake , R. (1982) J. Neu ophysiol. 47, 329-352. Bake , R. & Be hoz, A. (1975) B ain Res. 86, 121-127. Delgado-Ga c a, J. M., Vidal, P. P., G6mez, C. & Be hoz, A. (1989) Neu oscience 29, 291-307. Bake , R. (1977) in Eye Mo emen s, eds. B ooks, B. A. & Bajandas, F. J. (Plenum, New Yo k), pp. 145-178. Che on, G., Godaux, E., Laune, J. M. & Vande kelen, B. (1986) J. Physiol. (London) 372, 75-94. Cannon, S. C. & Robinson, D. A. (1987) J. Neu ophysiol. 57, 1383-1409. Kaneko, C. R. S. (1992) Ann. N. Y. Acad. Sci. 656, 408-427. Robinson, D. A. (1975) in Basic Mechanisms o Ocula Mo ili y and Thei Clinical Implica ions, eds. Lenne s and, G. & Bach-y-Ri a, P. (Pe gamon, New Yo k), pp. 337-374. Tomlinson, R. D. & Robinson, D. A. (1984) J. Neu ophysiol. 51, 1121-1136. Simpson, J. I. (1984) Annu. Re . Neu osci. 7, 13-41. Goldbe g, J. M. & Fe nandez, C. (1975) Annu. Re . Physiol. 37, 129-162. Collewijn, H. (1972) B ain Res. 36, 71-88. Cohen, B., Ma suo, V. & Raphan, T. (1977) J. Physiol. (Lon- don) 270, 321-344. S aube, A., Ku zan, R. & Bu ne , U. (1991) Exp. B ain Res. 86, 347-358. Pas o , A. M., To es, B., Delgado-Ga c a, J. M. & Bake , R. (1991) J. Neu ophysiol. 66, 2125-2140. Pas o , A. M., de la C uz, R. R. & Bake , R. (1992) J. Neu- ophysiol. 68, 2003-2015. Highs ein, S. M., Ki ch, R., Ca ey, J. & Bake , R. (1992) J. Comp. Neu ol. 319, 501-518. Che on, G. & Godaux, E. (1987) J. Physiol. (London) 394, 267-290. Galiana, H. L. & Ou e b idge, J. S. (1984) J. Neu ophysiol. 51, 210-241. Kimmel, C. B., Me cal e, W. K. & Schab ach, E. (1985) J. Comp. Neu ol. 233, 365-376. T e a ow, B., Ma ks, D. L. & Kimmel, C. B. (1990) Neu on 4, 669-679. Boyle, R. & Highs ein, S. M. (1990) J. Neu osci. 10, 1570- 1582. Die inge , N., Reichenbe ge , I. & G a , W. (1992) B ain Beha . E ol. 39, 289-304. Llinis, R. R. (1988) Science 242, 1654-1664. De la C uz, R. R., Pas o , A. M. & Bake , R. (1993) Soc. Neu osci. Abs . 19, 858. Neu obiology: Pas o e al. L-------