Download - In Vitro Properties of Neurons in the Rat Pretectal Nucleus of the Optic Tract

doi: 10.1152/jn.00039.200797:3574-3584, 2007. First published 7 March 2007;J Neurophysiol

N. Prochnow, P. Lee, W. C. Hall and M. SchmidtNucleus of the Optic TractIn Vitro Properties of Neurons in the Rat Pretectal

You might find this additional info useful...

58 articles, 10 of which you can access for free at: This article citeshttp://jn.physiology.org/content/97/5/3574.full#ref-list-1

2 other HighWire-hosted articles: This article has been cited by http://jn.physiology.org/content/97/5/3574#cited-by

including high resolution figures, can be found at: Updated information and serviceshttp://jn.physiology.org/content/97/5/3574.full

can be found at: Journal of Neurophysiology about Additional material and informationhttp://www.the-aps.org/publications/jn

This information is current as of August 6, 2016.

at http://www.the-aps.org/. Copyright © 2007 by the American Physiological Society. ISSN: 0022-3077, ESSN: 1522-1598. Visit our websitetimes a year (monthly) by the American Physiological Society, 9650 Rockville Pike, Bethesda MD 20814-3991.

publishes original articles on the function of the nervous system. It is published 12Journal of Neurophysiology

by guest on August 6, 2016

http://jn.physiology.org/D

ownloaded from

by guest on A

ugust 6, 2016http://jn.physiology.org/

Dow

nloaded from

http://jn.physiology.org/content/97/5/3574.full#ref-list-1

http://jn.physiology.org/content/97/5/3574#cited-by

http://http://jn.physiology.org/content/97/5/3574.full

http://www.the-aps.org/publications/jn

http://jn.physiology.org/


In Vitro Properties of Neurons in the Rat Pretectal Nucleus of the Optic Tract

N. Prochnow,1 P. Lee,2 W. C. Hall,2 and M. Schmidt1

1Allgemeine Zoologie and Neurobiologie, Ruhr-Universitat Bochum, Bochum, Germany; and 2Department of Neurobiology,Duke University Medical Center, Durham, North Carolina

Submitted 11 January 2007; accepted in final form 26 February 2007

Prochnow N, Lee P, Hall WC, Schmidt M. In vitro properties ofneurons in the rat pretectal nucleus of the optic tract. J Neuro-physiol 97: 3574 –3584, 2007. First published March 7, 2007;doi:10.1152/jn.00039.2007. The nucleus of the optic tract (NOT) hasbeen implicated in the initiation of the optokinetic reflex (OKR) andin the modulation of visual activity during saccades. The presentexperiments demonstrate that these two functions are served byseparate cell populations that can be distinguished by differences inboth their cellular physiology and their efferent projections. Wecompared the response properties of NOT cells in rats using target-directed whole cell patch-clamp recording in vitro. To identify thecells at the time of the recording experiments, they were prelabeled byretrograde axonal transport of WGA-apo-HRP-gold (15 nm), whichwas injected into their primary projection targets, either the ipsilateralsuperior colliculus (iSC), or the contralateral NOT (cNOT), or theipsilateral inferior olive (iIO). Retrograde labeling after injections insingle animals of either WGA-apo-HRP-gold with different particlesizes (10 and 20 nm) or two different fluorescent dyes distinguishedtwo NOT cell populations. One projects to both the iSC and cNOT.These cells are spontaneously active in vitro and respond to intracel-lular depolarizations with temporally regular tonic firing. The otherpopulation projects to the iIO and consists of cells that show nospontaneous activity, respond phasically to intracellular depolariza-tion, and show irregular firing patterns. We propose that the sponta-neously active pathway to iSC and cNOT is involved in modulatingthe level of visual activity during saccades and that the phasicallyactive pathway to iIO provides a short-latency relay from the retina topremotor mechanisms involved in reducing retinal slip.

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

The nucleus of the optic tract (NOT) consists of a small andinconspicuous population of neurons intercalated among optictract fibers passing through the pretectum. Considering itssmall size and apparent lack of distinct nuclear subdivisions,NOT has been associated with a surprising variety of functionsthat appear remarkably conserved across mammalian species.The most extensively studied function of NOT is its role in thecontrol of compensatory eye movements during the horizontaloptokinetic reflex (OKR) (marsupials: Ibbotson et al. 1994;Volchan et al. 1996; rat: Benassi et al. 1991; Cazin et al. 1984;rabbit: Collewijn 1975a; cat: Hoffmann and Fischer 2001;monkey: Kato et al. 1986; Mustari and Fuchs 1990; Schiff et al.1988, 1990). In addition, NOT cells can modulate thalamocor-tical activity (cat: Fischer et al. 1998; Sudkamp and Schmidt1995; monkey: Bickford et al. 2000; Wilson et al. 1995) andprovide inhibition to visuosensory neurons in the superiorcolliculus (Baldauf et al. 2003; Born and Schmidt 2004).

The efferent connections of NOT are correspondingly di-verse (reviewed in Gamlin 2005; Ibbotson and Dreher 2005;

Simpson et al. 1988). They include pathways to ipsilateralvisuosensory thalamic nuclei, including the dorsal lateralgeniculate nucleus, the lateral posterior thalamic nucleus, andthe pulvinar, to the ipsilateral superior colliculus (iSC), to theipsilateral inferior olive (iIO) and nucleus prepositus hypo-glossi (iNPH), and to the contralateral NOT (cNOT). Thepathway from the NOT to the thalamus seems to carry signalsrelated to ongoing saccades because NOT cells that project toeither the dorsal lateral geniculate nucleus or the lateralisposterior/pulvinar complex in vivo respond to saccade-inducedretinal image shifts 30 to 50 ms after saccade onset (cat:Fischer et al. 1998; Schmidt 1996; Sudkamp and Schmidt1995; monkey: Fuchs et al. 1992). In contrast, the type ofinformation relayed to iSC and to cNOT remains to be deter-mined, partly because cNOT-projecting cells seem unrespon-sive in anesthetized animals (Schmidt et al. 1995). Finally, thepreoculomotor structures iIO and iNPH receive NOT signalsthat are related to the control of slow-phase eye movementsduring the OKR (Buttner-Ennever et al. 1996; Magnin et al.1989; Schmidt et al. 1995). In particular, NOT cells that projectto iIO and iNPH generate directionally selective responses tolow-velocity global retinal image movements from temporal tonasal in the contralateral hemifield (for review, see Gamlin2005; Ibbotson and Dreher 2005; Simpson et al. 1988).

Although a common property of NOT cells is that theypreferentially respond to moving rather than to stationaryvisual stimuli, significant differences occur in their preferredstimulus velocity range. OKR-related neurons prefer stimulusmotion below 100°/s (marsupials: Ibbotson et al. 1994; Vol-chan et al. 1989; rat: Cazin et al. 1980; Schmidt et al. 1993;rabbit: Collewijn 1975b; ferret: Klauer et al. 1990; cat: Hoff-mann and Schoppmann 1981; monkey: Ilg and Hoffmann1996; Mustari and Fuchs 1988), whereas other neurons areresponsive only to much faster stimulus velocities (marsupials:Ibbotson et al. 1994; Price and Ibbotson 2001; cat: Missal et al.2002; Schweigart and Hoffmann 1992; Sudkamp and Schmidt1995; monkey: Inoue et al. 2000). In addition, in monkeyneurons have been described whose activity is completelyblocked by a saccadic eye movement (Mustari et al. 1997).

Thus the known response properties in vivo suggest that,even though NOT cells as a group prefer moving visualstimuli, some selectively respond to high-velocity movementsand others to slow movements. However, with currently avail-able in vivo methods, it has been difficult to determine withcertainty whether NOT makes a similar contribution to each ofits diverse targets or, instead, it consists of several distinct cellpopulations, each with its own function and pattern of connec-

Address for reprint requests and other correspondence: N. Prochnow, Allge-meine Zoologie and Neurobiologie, Ruhr-Universitat Bochum, ND 6/32,D-44780 Bochum, Germany (E-mail: [email protected]).

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.

J Neurophysiol 97: 3574–3584, 2007.First published March 7, 2007; doi:10.1152/jn.00039.2007.

3574 0022-3077/07 $8.00 Copyright © 2007 The American Physiological Society www.jn.org



ownloaded from


tions. In the present experiments, we addressed these questionsin rats by performing in vitro whole cell patch-clamp record-ings on NOT cells that were prelabeled by retrograde axonaltransport after injections of neuronal tracers in the destinationsof their efferent projections. This approach allowed directcomparisons of the intracellular properties of pretectal cellpopulations with their identified projection targets. The resultsdemonstrate that NOT includes several cell populations thatcan be distinguished both by their intrinsic membrane proper-ties and by their efferent connections.

M E T H O D S

Retrograde labeling

Experiments were performed on 27 Long–Evans hooded rats be-tween 16 and 32 days postnatal age. Experimental procedures werecarried out in accordance with the European Communities CouncilDirective of November 24, 1986 (86/609/EEC) and approved by theDuke University Institutional Animal Care and Use Committee. Forstereotaxic surgery, the animals were deeply anesthetized by anintraperitoneal injection of ketamine (100 mg/kg body weight) andxylazine hydrochloride (1 mg/kg) and the level of anesthesia wasmaintained by additional injections of ketamine. Using stereotaxiccoordinates, the terminal zones of NOT neurons were injected with0.5 �l of the retrograde axonal tracer wheat germ agglutinin (WGA)-apo-horseradish peroxidase (HRP)-gold (15-nm particle size; Sanbio,Beutelsbach, Germany, and E-Y Laboratories, San Mateo, CA). As acell label, the (WGA)-apo-HRP-gold has an advantage over fluores-cent labels of being resistant to fading, nontoxic to the neurons, andvisible under bright-field illumination (Lee et al. 2001). In someexperiments, two injections of different-size gold particles (10 and 20nm), each in a different site, were made in single animals to determinewhether one cell type projects to both sites. After surgery, animalswere allowed to recover and survive for a period of 5–7 days, whichis sufficient for retrograde axonal transport of the tracer (Lee et al.2001). Although the particle clusters that appear in labeled cells are ofmuch larger size than the gold particles, control tracer injections witha single particle size revealed that smaller particles lead to consider-ably smaller clusters than do the larger particles (Fig. 1). This allowsa clear distinction between labels from different particle sizes indouble-label experiments.

The results obtained with WGA-apo-HRP-gold were confirmedusing double injections of fluorescent tracers. We used red textmarkerdye (“Stabilo Red”; Stabilo, Heroldsberg, Germany) for cNOT andiIO injections and green textmarker dye (“Stabilo Green”) for injec-tions into the iSC.

Slice preparation for patch-clamp recording

After injection of an overdose of ketamine and xylazine hydrochlo-ride, animals were transcardially perfused with ice-cold artificialcerebrospinal fluid (ACSF) containing the following components (inmM): NaCl, 124; KCl, 5; NaH2PO4, 1.25; NaHCO3, 26; MgSO4, 2;CaCl2, 2; and glucose, 10. The ACSF was continuously gassed withcarbogen (95% O2-5% CO2). After removal of the brain from theskull, 350-�m-thick coronal slices were cut and kept at 35°C in ACSFfor �1 h to allow the tissue to recover from the slicing procedure. Forwhole cell patch-clamp recording they were transferred to a submer-sion-type recording chamber and superfused at 3 ml/min with room-temperature ACSF.

Whole cell patch-clamp recording

Whole cell patch-clamp recordings from WGA-apo-HRP-gold–labeled neurons in the NOT were performed on a fixed-stage micro-

scope (Olympus BX51WI) under visual control with a �40-longworking distance objective, using infrared differential interferencemicroscopy (Dodt and Zieglgansberger 1998). For recording, boro-silicate glass micropipettes (impedance 5–8 M�) were filled withinternal solution containing (in mM): potassium gluconate, 130;sodium gluconate, 2; HEPES, 20; MgCl2, 4; Na2ATP, 4; NaGTP, 0.4;and EGTA, 0.5. To confirm that the recordings were made from theprelabeled cell and also for later morphological characterization of thecell, 0.5% biocytin was added to the pipette solution. The biocytindiffused into the cell during the recording and was later reacted with3�3�-diaminobenzidine (DAB), without heavy metal intensification.Neuronal signals were amplified and filtered by an EPC 9 amplifier(HEKA, Lambrecht, Germany), digitized at 10 kHz and displayed,stored, and analyzed using Pulse/Pulsefit software (HEKA). Measuredmembrane potentials were corrected for the junction potential of�10 mV.

Anatomic analyses

Twelve additional animals were used to examine the efferentconnections of NOT using retrograde tracers. After a survival periodof 5–7 days for retrograde tracer uptake and axonal transport, animalswere injected with an overdose of ketamine hydrochloride and xyla-zine and perfused transcardially with ACSF mixed with 0.1% heparinat room temperature. ACSF was followed by 4.0% paraformaldehydein 0.1 M phosphate buffer and a rinse in 10% sucrose in 0.1 Mphosphate buffer. Brains were removed from the skull and then stored

FIG. 1. Distribution of gold particle cluster sizes after retrograde transportof wheat germ agglutinin (WGA)–apo-horseradish peroxidase (HRP), labeledwith 10-nm-sized gold particles (gray bars, arrowheads in A) and 20-nm-sizedgold particles (black bars, arrows in B). Although clusters are much larger thanthe original gold particles retrograde label from either particle size can beeasily distinguished. Scale bar in A: 20 �m.

3575PROPERTIES OF NOT NEURONS IN VITRO

J Neurophysiol • VOL 97 • MAY 2007 • www.jn.org



ownloaded from


in 30% sucrose buffer overnight at 4°C for cryoprotection. The brainswith injections of WGA-apo-HRP-gold were cut with a freezingmicrotome into 40-�m-thick coronal slices through both the pretec-tum in its entire anterior–posterior extent and the tracer injection sites.Sections were collected in 0.1 M phosphate-buffered saline, put ongelatin-coated slides, and stained with 0.03% cresyl violet. Afterinjections of fluorescent dyes, the same procedure was performed,except the sucrose rinses were omitted and 40-�m-thick coronal sliceswere cut using a vibratome. Then, the brain slices were coverslippedin 0.01 M phosphate-buffered saline containing 1% glycerol to avoiddrying. Injection sites and labeled somata were identified and photo-graphed with an oil-immersion objective lens (�60) on a lightmicroscope.

Soma sizes of labeled neurons as well as sizes of gold particleswere analyzed using a computer-aided reconstruction system (Neu-rolucida, MicroBrightField, Williston, VT). All chemicals used wereobtained from Sigma–Aldrich (Deisenhofen, Germany).

R E S U L T S

Two populations of NOT cells were distinguished in theseexperiments. One population projects to the iIO, whereas theother projects to both the iSC and the cNOT. Whole cellpatch-clamp recordings from cells that were identified by theirefferent projections confirmed that these are two distinct cell

populations; the cells that project to iSC and cNOT are toni-cally active in vitro, whereas the cells that project to iIOgenerate phasic responses.

Tonically active cells

All of the clamped cells that projected to the iSC weretonically active (n � 15). For example, Fig. 2 illustrates theresults from one cell that was prelabeled by an injection ofWGA-apo-HRP-gold into the iSC (0.5 �l, 15-nm gold parti-cles). The arrows in Fig. 2A indicate gold particles that weretransported to the cell soma from the injection site. Thehomogeneous background in the soma and primary dendrite(asterisk) was produced by the biocytin that diffused into thecell from the patch pipette during the experiment, confirmingthat the recordings were obtained from the prelabeled cell.Figure 2B shows, in a current-clamp recording, that actionpotentials evoked by current injections were sustained for theentire duration of the depolarizations (500 ms) and varied infrequency with the injected current or membrane potential.Figure 2C illustrates that action potential frequency increasedwith stepwise increases in the injected current. Figure 2D plotsthe relationship between the response frequency and the mem-

FIG. 2. Results from a representative nucleus of the optictract (NOT) neuron retrogradely labeled from ipsilateral supe-rior colliculus (iSC). A: clustered gold particles are present inthe soma and primary dendrite of the neuron (arrows). Homo-geneous background label (asterisk) produced by biocytin dif-fusion from the patch pipette during recording confirms that therecording was from a prelabeled cell. Inset: location of theneuron within the slice (m, medial; l, lateral; d, dorsal; v,ventral). B: depolarizing current injection under current-clampconditions (duration 500 ms) evokes action potentials in aregular firing pattern. Neuron’s firing rate varies with themembrane potential. C: depolarizing current injections (ad-justed to achieve �5 mV steps of the membrane potential)increase the firing rate of the neuron, which was spontaneouslyactive at resting potential (�62 mV). Interspike intervals (ISIs)remain regular at all membrane potentials. D: firing rate plottedas a function of membrane potential. Maximum firing rate (27spikes/s) is reached at �30 mV.

3576 N. PROCHNOW, P. LEE, W. C. HALL, AND M. SCHMIDT




ownloaded from


FIG. 3. Representative NOT neuron labeled af-ter an injection of WGA-apo-HRP-gold into thecontralateral NOT (cNOT). A: arrows point to theclustered gold particles in the soma and primarydendrite that prelabeled the cell. Homogeneousbackground of biocytin confirms that the recordingwas from the prelabeled cell. B: depolarizing cur-rent injections during current clamp (5 and 10 pA,duration 500 ms) increase the firing rate as a func-tion of the amplitude of the injected current. C: asustained nonadapting tonic firing of action poten-tials (maximum firing rate: 18 spikes/s) is generatedduring current-clamp recordings at membrane po-tentials of �60 and �40 mV.

FIG. 4. Double labeling with 2 retrograde trac-ers. A: WGA-apo-HRP-gold with 20-nm-sizedgold particles were injected into the superficiallayer of the iSC and WGA-apo-HRP-gold with10-nm-sized gold particles into the cNOT. B: dis-tribution of retrogradely labeled neurons in a rep-resentative midbrain section containing caudalNOT. Double-labeled cells are shown in blue;cells labeled only from iSC or cNOT are shown inred and green, respectively. C: in 40-�m-thickcoronal sections counterstained with cresyl violet,NOT neurons retrogradely labeled with WGA-apo-HRP-gold contain clusters with both particlesizes (inset). Clustered 20-nm-sized particles ap-pear as prominent dots, in contrast to the smallerclusters of 10-nm-sized particles. Coappearance of10- and 20-nm-sized particles within individualNOT neurons (arrows) indicates that the neuronsends axon collaterals to both the iSC and thecNOT.





ownloaded from


brane potential; the action potentials increased in frequencywith the amount of depolarizing current from fewer than fivespikes/s when the cell was near its resting potential (�61.3mV) to a peak of about 27 spikes/s when the membranepotential approached �25 mV. The mean input resistance ofthe tonically active cells was 323.7 M� (SD � 119.7); themean time constant �, which was conventionally calculatedfrom the onset of long-lasting hyperpolarizing voltage steps,was 0.93 ms (SD � 1.09); and the mean resting potential was�55.1 mV (SD � 3.4). The membrane potential threshold forthe generation of action potentials, defined as the potential atwhich the first spike appeared when the cell was continuouslydepolarized from �90 mV by increasing the amplitude of thecurrent injections, was �59.6 mV (SD � 3.8).

Figure 3 illustrates the results from an experiment on a cellprelabeled after an injection of WGA-apo-HRP-gold in cNOT(Fig. 3A). Similar to the NOT–iSC cell illustrated in Fig. 1, thecell in Fig. 2 generated sustained action potentials that variedin frequency with the amplitude of the injected current in acurrent-clamp recording (Fig. 3B). Figure 3C shows trains ofaction potentials (two and 18 spikes/s, respectively) that weregenerated when the cell was clamped at a membrane potentialof �60 mV (left) or �40 mV (right). Similar results wereobtained for all cells that projected to cNOT (n � 8). The meaninput resistance of NOT–cNOT cells (Rmem) was 429.5 M�(SD � 106.0); the mean time constant � was 1.81 ms (SD �2.96); the mean resting potential was �53.5 mV (SD � 3.6);and the mean spike threshold was �56.6 mV (SD � 2.4).These parameters were not significantly different from those ofthe cells that projected to the iSC (P � 0.84 for membraneresistances, P � 0.35 for time constants, and P � 0.45 forresting potentials; P � 0.15 for spike thresholds, t-test).

To determine whether the NOT cells with in vitro tonicfiring patterns constitute a single population that projects toboth iSC and cNOT, single animals (n � 3) were injected with20-nm-sized gold particles in iSC and 10-nm-sized particles incNOT (Fig. 4A). Although the size of gold particles themselvesare below the resolution limit of the light microscope, the twotracers can be distinguished by the size of the aggregates thatoccur in the labeled cells. Aggregates in iSC, where 20-nm-sized gold particles had been injected, had diameters �2.5 �m(mean size 3.0 �m, SD � 0.3), whereas aggregates of the10-nm particles in iIO were �2.5 �m in diameter (mean 1.8�m, SD � 0.2), which was significantly smaller (P � 0.0001).The distribution of retrogradely labeled cells within the NOT inone representative midbrain section is depicted in Fig. 4B. Ofthe cells labeled from iSC (n � 267), 76% were doublelabeled; of the cells labeled from cNOT (n � 226) 90% weredouble labeled. Thus the majority of labeled cells containedboth tracers, indicating that they projected to both targets.Figure 3C illustrates a cluster of five cells in rostral NOT, eachof which contained both sizes of gold particles, confirming thatsingle NOT cells project to both targets. The same conclusionwas reached in the experiments illustrated in Fig. 5. In thiscase, the retrograde fluorescent tracer Stabilo Red was injectedinto the cNOT and Stabilo Green was injected into the iSC andclusters of cells with both labels (arrows) were present in NOT.

We also compared soma sizes of single-labeled and double-labeled cells. The average soma area of NOT neurons labeledonly from iSC was 191.0 �m2 (SD � 44.1); that of cellslabeled only from cNOT was 212.3 �m2 (SD � 46.1). Double-

labeled cells had a mean soma area of 221.7 �m2 (SD � 48.3).These differences were not statistically significant.

Phasic cells

Neurons prelabeled with WGA-apo-HRP-gold after injec-tions in iIO (0.1 �l, 15-nm-sized particles) generated phasic orphasic–tonic, fast-adapting, or bursting firing patterns in re-sponse to intracellular current injection and never showedspontaneous activity at resting potential. An example is shownin Fig. 6. Figure 6A shows a prelabeled cell containing the goldparticles (arrows) and the homogeneously distributed biocytin(asterisk). Figure 6B is a current-clamp recording showing thatthe cell generated a phasic–tonic firing pattern in response todepolarizing current injections leading to irregular maintainedfiring when higher depolarizing currents were injected. Ingeneral, NOT neurons prelabeled from the iIO showed less-uniform responses to intracellular depolarizations than neuronsprelabeled from either iSC or cNOT. Figure 6C shows theirregular firing pattern of another prelabeled cell when injectedcurrents induced small depolarizing steps of the membranepotential. In Fig. 6D, the response frequency is plotted againstthe membrane potential; the average firing rate increased withthe amount of depolarizing current from fewer than fivespikes/s at �40 mV to a peak of roughly 30 spikes/s at �25mV. More positive membrane potentials resulted in a reducedfiring rate and no spikes occurred above �15 mV.

FIG. 5. Results from double retrograde tracing with fluorescent dyes in40-�m-thick coronal vibratome sections. A and B: NOT neurons labeled afteran injection of “Stabilo Green” into the SGS of the iSC and an injection of“Stabilo Red” into the cNOT. Superimposed images for green and redfluorescence display the coincidence of green and red fluorescing tracer inneurons of the iNOT. Presence of both labels in single cells indicates that thesecells project to both the iSC and the cNOT.





ownloaded from


The mean input resistance of the cells that projected to iIOneurons (n � 17) was 200.7 M� (SD � 59.4), the mean timeconstant � was 3.07 ms (SD � 2.28), the mean resting potentialwas �57.9 mV (SD � 4.9), and the spike threshold was �39.9mV (SD � 7.9). Compared to the tonic cells that projected toiSC and cNOT, the phasic iIO-projecting cells were character-ized by significantly lower input resistances (P � 0.00012;t-test), larger time constants (P � 0.0122), more negativeresting potentials (P � 0.015), and, in particular, significantlymore positive spike thresholds (P � 0.0001).

A striking difference between iSC/cNOT- and iIO-project-ing NOT neurons in vitro concerned the regularity of theirongoing firing. To analyze the temporal firing pattern, inter-spike intervals (ISIs) were calculated from the ongoing firingof spontaneously active cells and from activity of nonsponta-neously active cells during maintained intracellular depolariza-tion over a 10-s period. ISIs of NOT neurons that project to iSCand cNOT (Fig. 7, A, D, and G) form a narrow unimodalGaussian distributions with only a slight variation (Fig. 7, B, E,and H), indicating a high degree of equally sized ISIs. Thetemporal regularity of the spontaneous firing was easily detect-able in autocorrelograms of the activity where multiple peaksappeared at constant time intervals (Fig. 7, C, F, and I). Thisdistribution was in contrast to the ISI distributions of iIO-projecting NOT neurons (Fig. 7, J, M, and P). The interval

histograms showed a much less symmetrical distribution (Fig.7, K, N, and Q). Furthermore, the lack of regularity in firing canbe derived from the autocorrelograms that showed no distinctpeaks (Fig. 7, I, L, and R).

To further confirm that the phasic NOT neurons constitute apopulation that is distinct from the tonically active cells, 10-nmgold particles were injected in the iIO and 20-nm particles wereinjected in iSC (Fig. 8A). In three experiments, no cells werefound that contained both labels. This result is shown in Fig.8B where the distribution of retrogradely labeled cells is shownin a representative midbrain section. Although the labeled cellswere intermingled throughout NOT, none of the cells (n � 90and n � 60, labeled from iSC and iIO, respectively) wasdouble labeled. Similar results were obtained when StabiloGreen tracer was injected in iSC and Stabilo Red in iIO, that is,no cells contained both tracers (Fig. 9, A–D, n � 2).

Comparisons of the soma sizes of labeled cells revealed thatthe average soma area of NOT neurons labeled from iSC was227.4 �m2 (SD � 49.9), whereas that of cells labeled from iIOwas 154.3 �m2 (SD � 40.2). Thus iIO-projecting NOT cellshad significantly smaller soma areas than those of iSC-project-ing cells (P � 0.001). In a comparison of soma shapes (Figs.8B and 9) it appeared that iSC-projecting cells had sphericalcell bodies whereas somata of iIO-projecting cells were moreelongated, particularly along the mediolateral axis.

FIG. 6. NOT neuron prelabeled with WGA-apo-HRP-gold(particle size 15 nm) after an injection into the ipsilateralinferior olive (iIO). A: gold particles are clustered in the somaand the primary dendrite (arrows). To confirm that the record-ing was from the prelabeled cell it was filled with biocytin fromthe patch pipette during whole cell patch-clamp recording(as-terisk). B: in this cell, depolarization (5 mV above restingpotential, middle trace) generated a phasic–tonic response withirregular action potentials. Increasing the membrane potential(30 mV above resting potential, top trace) leads to an increaseof the number of the evoked action potentials; the firing pattern,however, remains irregular. C: action potentials generated bycurrent injections that induce small depolarization steps atvarious holding potentials in another cell. Firing pattern isalways irregular. D: firing rate as a function of membranepotential. Firing rate increases rapidly above spike thresholduntil maximum firing level. Further depolarization leads toreduced firing.





ownloaded from


D I S C U S S I O N

We examined the in vitro activity patterns of neurons in theNOT of the rat pretectal nuclear complex that were distin-guished by differences in the destinations of their axons.Specifically, we examined NOT neurons that project to one ormore of three structures: the iIO, the iSC, and the cNOT. TheseNOT cells could be distinguished by their responses to intra-cellular depolarization: one group responded with a transientfast-adapting pattern of activity or burst firing, whereas asecond group responded tonically with only minor adaptation.The latter neurons were also spontaneously active whenclamped at their resting potential, which was a unique featureof this population. These two response patterns were clearlyassociated with the connections of the cells; the neurons thatprojected to iIO responded transiently, whereas those thatprojected to iSC and cNOT were tonically active. This distinc-tion between NOT cells that project to iSC and cNOT andthose that project to iIO was confirmed by anatomical experi-ments. That is, injections of different retrograde axonal tracersinto cNOT and iSC in single animals produced NOT cells withboth labels, indicating that at least some NOT cells project toboth structures. In contrast, after similar injections into either

iIO and iSC or iIO and cNOT, individual NOT cells containedonly one of the two labels, indicating that the cells that projectto iIO are distinct from those that project to iSC and cNOT.These results demonstrate that NOT constitutes at least two cellpopulations and that these populations differ in both their firingproperties and their patterns of connections.

NOT cell populations

Data in the literature on NOT anatomy and physiologyobtained from a variety of mammalian species demonstrate thatcell properties are very similar across different mammals(Gamlin 2005; Ibbotson and Dreher 2005; Simpson et al.1988). Therefore it seems reasonable to generalize resultsbetween mammalian species.

Previous in vitro experiments in the rat demonstrated differ-ences in the responses of NOT cells to intracellular depolar-ization (Prochnow and Schmidt 2004). One population of cellswas characterized by tonic firing to intracellular depolariza-tions, an almost linear relationship between membrane poten-tial and firing rate, and the generation of spontaneous activity.The activity pattern is characterized by a high degree of

FIG. 7. Temporal analysis of the firing pattern of NOTneurons. A, D, G: NOT neurons retrogradely labeled fromiSC/cNOT are characterized by maintained regular firing. B, E,H: ISI histograms of these neurons show narrow Gaussiandistributions with only limited variations of ISIs. C, F, I:regularity of firing is confirmed in the autocorrelograms by theappearance of equally spaced peaks arising from the slightvariation in ISIs. J, M, P: NOT neurons that project to the IOexhibit irregular firing in vitro. K, N, Q: ISI histograms show thespread distribution of events. L, O, R: irregular firing of thispopulation is further illustrated by the irregular event distribu-tions in the autocorrelograms. Scale bars in A, D, G and J, M,P: 30 �m.





ownloaded from


regularity leading to periodic peaks in activity autocorrelo-grams. The association of these properties was confirmed in thepresent study. Although we did not specifically test for anautonomous generation of spontaneous activity, all cells thatgenerated spontaneous activity at their resting membrane po-tential also responded with tonic firing to intracellular depo-larizations. Furthermore, the spike frequency of the cells in thispopulation was a linear function of the membrane potential andthe spontaneous firing showed strong periodicity. Becausethese properties always appeared together in individual neu-rons, we conclude that they define a distinct population of NOTcells. The experiments with two retrograde tracers demon-strated that at least some of the cells in this population projectto both iSC and cNOT. In contrast, NOT cells that project toiIO must be regarded as a separate NOT cell population, as canbe derived from fundamental differences in their temporalactivity pattern, which lacks any periodicity. This conclusion isconsistent with earlier results from cat and rat showing NOTcells that project to iIO send collaterals to the iNPH but not tothe cNOT (Schmidt et al. 1995). Furthermore, NOT cells thatproject to iSC were previously characterized as beingGABAergic (rat: Baldauf et al. 2003; Born and Schmidt 2004;rabbit: Nunes Cardozo et al. 1994). In contrast, NOT cells thatproject to iIO are not GABAergic (cat: Horn and Hoffmann1987) and show immunoreactivity for glutamate (rat: Lewaldet al. 1994).

A comparison of electrophysiological properties reveals thatNOT cells that project to iSC and cNOT have higher inputresistances, more positive resting potentials, and more negative

spike thresholds than do cells that project to iIO. This constel-lation of properties is similar to the properties of spontaneouslyactive NOT cells in an earlier study (Prochnow and Schmidt2004), which also showed higher input resistance, more posi-tive resting potentials, and lower spike thresholds than didNOT cells with phasic firing characteristics. Because the dif-ferences in firing were still present when the cells were phar-macologically isolated from synaptic input, by a substitution ofcalcium in the extracellular solution with cobalt (Prochnow andSchmidt 2004), they must result from differences in intrinsicphysiological properties of the two cell populations.

Although both response properties and efferent connectionsdistinguish between these two populations, other commoncriteria for differentiating between cell groups, such as differ-ences in spatial distribution and morphology, do not appear toapply to NOT. That is, the cells labeled by transport from iIO,iSC, and cNOT were intermingled at similar topographicallocations within the NOT. Furthermore, the morphology of thecells revealed by the biocytin label after recording did notreliably distinguish the two populations on a single-cell anal-ysis. This result agrees with our earlier observation that spon-taneously active NOT cells with tonic response patterns showdendritic morphologies similar to that of cells with phasicresponse characteristics (Prochnow and Schmidt 2004).

Although morphological comparisons of the biocytin-filledcells revealed no obvious differences, our results from thetracing experiments indicate that iIO-projecting cells on aver-age have smaller somata than those of cells that project to iSCand cNOT. Even though this might seem contradictory to

FIG. 8. Retrograde labeling with 2 tracers dem-onstrates that phasic cells are anatomically distinctfrom tonically active cells. A: WGA-apo-HRP-goldwith 20-nm-sized particles was injected into thesuperficial layer of the iSC (left) and WGA-apo-HRP-gold with 10-nm-sized particles was injectedinto the iIO (right). B: distribution of retrogradelylabeled cells shown in a representative midbrainsection through the caudal NOT. Despite the spatialoverlap of the retrogradely labeled cells, neurons areeither labeled from iSC (shown in red) or from iIO(shown in green). C: light micrograph of the leftiNOT in a 40-�m-thick coronal section counter-stained with cresyl violet. Whereas large clusters ofthe 20-nm-sized particles are localized in triangular-to ovoid-shaped somata, clusters of the smaller 10-nm-sized particles are present only in flattened orspindle-shaped somata. No neurons contained bothlabels, confirming that the in vitro phasically activeNOT neurons that project to the iIO are distinct fromthe tonically active cells that project to SC andcNOT.





ownloaded from


results from the biocytin-filled cells, the soma sizes of therecorded cells may be biased toward large cells because theyare easier to detect and to patch-clamp than are small cells.Thus because there is overlap of the soma sizes between NOTcells that project to cNOT and iSC and cells that project to iIO,we probably have primarily recorded from large iIO-projectingcells. On average, NOT cells labeled from iSC and cNOT hadsoma sizes that are comparable to those of spontaneouslyactive cells recorded from in a previous study (Prochnow andSchmidt 2004).

NOT cells involved in the optokinetic reflex

A priori, one might predict that the spontaneously active cellswith tonic firing patterns are involved in the generation of slowcompensatory eye movements during the horizontal optokineticreflex. Indeed, in vivo the NOT cells that serve this functiongenerate tonic responses to large moving visual stimuli as dofunctionally related cells in the adjacent dorsal terminal nucleus(DTN) of the accessory optic system (Cazin et al. 1980; forreview, see Gamlin 2005; Ibbotson and Dreher 2005; Simpson etal. 1988). Because neurons in the left NOT–DTN generate eyemovements only to the left and neurons in the right NOT–DTNgenerate movements only to the right (Cazin et al. 1980; Col-lewijn 1975b; Hoffmann and Schoppmann 1981; Mustari andFuchs 1990; for review, see Gamlin 2005; Ibbotson and Dreher2005), spontaneous activity may stabilize the eyes by maintainingan activity balance between the right and left NOT–DTN in theabsence of appropriate visual stimuli. However, the NOT–DTNneurons involved in horizontal OKR generation project to the iIO

(Gamlin 2005; Ibbotson and Dreher 2005; Simpson et al. 1988)and in our in vitro experiments none of the cells that projected toiIO was spontaneously active. Instead, all NOT neurons thatprojected to iIO showed phasic responses to intracellular depolar-ization and we assume similar responses are characteristic of DTNcells.

How might this in vitro response property be related tothe known function of OKR–NOT neurons? In general,phasic responses allow rapid transmission of afferent spikeswith high temporal precision. For the NOT cells that projectto iIO, the main driving input arises from retinal ganglioncells (rat: Kato et al. 1992; rabbit: Pu and Amthor 1990;ferret: Klauer et al. 1990; cat: Hoffmann and Schoppmann1981; Koontz et al. 1985; monkey: Perry and Cowey 1984).In cat, retinal afferents were classified as directionallyselective ganglion cells that respond tonically to visualstimuli that move slowly across the visual field (Hoffmannand Stone 1985). Because the directionally selective gan-glion cells are tuned to stimulus velocity (rabbit: Oyster1968; Oyster et al. 1972) changing the firing rate at theretinal ganglion cell-to-NOT cell relay might confound theinformation necessary to achieve the required eye velocity.That is, a temporally precise transfer of changes in spikefrequency might be better achieved by NOT cells that havephasic response properties than by spontaneously activecells that add self-generated spikes to the firing patternsgenerated by the retinal input.

NOT cells that project to cNOT and iSC

To our surprise, NOT cells that project to the cNOT and theiSC were tonically active in vitro. We expected these neuronswould respond phasically because the only NOT neurons thatare characterized by tonic responses in vivo are the OKR-related directionally selective NOT cells that project to the iIOand iNPH (Gamlin 2005; Ibbotson and Dreher 2005). Inanesthetized animals, many NOT neurons respond to saccadiceye movements or shifts of the retinal image at saccadicvelocities with short, high-frequency bursts (for review, seeGamlin 2005; Ibbotson and Dreher 2005). Previous studiesdemonstrated that neurons with these phasic responses projectto either the ipsilateral dorsal lateral geniculate nucleus or tothe ipsilateral extrageniculate visual thalamic nuclei (Schmidt1996; Sudkamp and Schmidt 1995). The response properties ofthe NOT neurons that project to iSC and/or cNOT have notbeen characterized in vivo. Based on results in marsupials, itwas previously argued that the NOT cells that project to cNOTcontribute to the binocularity of the directionally selectiveNOT cells that project to IO (Ibbotson et al. 2002; Pereira et al.1995). Because the responses of these cells to monocularstimulation of the ipsilateral eye are tonic, the expectationwould be that the commissural NOT cells are also tonicallyactive. In higher mammals, the binocularity of directionallyselective NOT cells seems to depend on cortical input (ferret:Sengpiel et al. 1990; cat: Distler and Hoffmann 1993; monkey:Hoffmann et al. 1992), suggesting that the commissural NOTcells have a function different from conveying input from theipsilateral eye. Whether cortical input is also responsible forNOT cell binocularity in lower mammals (rat: Schmidt et al.1993; guinea pig: Lui et al. 1994) remains to be confirmed.

FIG. 9. Results with 2 retrograde fluorescence tracers. A and B: NOTneurons retrogradely labeled after an injection of “Stabilo Green” into thesuperficial layer of the iSC and “Stabilo Red” into the iIO. A and C: examplesfor neurons in the left iNOT retrogradely labeled from the iSC with greenfluorescent tracer. At the same location, images in B and D show NOT neuronslabeled from the iIO with the red fluorescent tracer. Comparison of A and B andC and D indicates no colocalization of the green and the red fluorescent tracersin single NOT neurons, confirming that the in vitro tonically active NOT cellsare anatomically distinct from the phasically active cells.





ownloaded from


Without more information concerning the in vivo responseproperties of the NOT cells that project to cNOT and iSC, wecan only speculate about their functions. The evidence thatNOT is not retinotopically organized (Hoffmann and Schopp-mann 1981) suggests that its influence is modulatory. Tonicactivity of NOT neurons in vivo has been reported not only forOKR-related cells, but also for a population of NOT neurons inmonkey that are effectively suppressed by saccadic eye move-ments in a nondirectionally selective fashion (Mustari et al.1997). These neurons, termed pretectal omnidirectional pauseneurons, could represent the spontaneously acitve cells re-ported here. We also know that SC-projecting NOT neurons, asmany others, are GABAergic and that they project almostexclusively to non-GABAergic projection neurons in the su-perficial, visuosensory layers of the superior colliculus(Baldauf et al. 2003; Born and Schmidt 2004; Nunes Cardozoet al. 1994). If iSC-projecting NOT cells are the pretectal pauseneurons, they could provide a tonic inhibitory input to collicu-lar cells that might help to suppress the execution of unwantedsaccades. Alternatively, iSC-projecting neurons could haveproperties in vivo different from those of the pause neurons andmight be activated by fast image movements, as many NOTcells are. Then, these GABAergic NOT cells could provideinhibition to collicular cells that increases during saccades andreduces the likelihood that the saccade-induced shifts in thevisual field trigger subsequent unwanted eye movements. Incontrast, the GABAergic projection from NOT to dorsal lateralgeniculate nucleus terminates exclusively on GABAergic in-terneurons and disinhibits relay cells in the dorsal lateralgeniculate nucleus during saccades (Cucchiaro et al. 1991,1993; Fischer et al. 1998; Wang et al. 2002). Thus NOT maycontribute both to preventing unwanted saccades by attenuat-ing visuomotor activity in the superior colliculus during sac-cades (Richmond and Wurtz 1980) and to maintaining the relayof visuosensory signals from the thalamus to the cortex duringand immediately after a saccade (Buttner and Fuchs 1973;Judge et al. 1980).

In conclusion, the organization of NOT has been a puzzle.This small and, by usual morphological criteria, homogeneousnucleus has been associated with several apparently diversefunctions ranging from modulation of the activity of visualthalamic relay cells during saccades to inhibition of the visuallayers of the SC and, through its pathway to iIO and iNPH, tomediation of the OKR. The question addressed by the presentexperiments is whether NOT is one nucleus that makes a singlecontribution to all of these diverse structures and functions orwhether, in contrast, it constitutes several distinct nuclei thatoverlap spatially but differ in terms of their connections,physiology, and functions. The results demonstrate that NOTconsists of several cell types that are anatomically and physi-ologically distinct. Tonic spontaneous activity is propagated byNOT neurons in a branched pathway to the iSC and the cNOT,whereas the NOT–DTN neuronal population that projects tothe iIO is characterized by phasic response properties in vitro.

G R A N T S

This work was supported by an International Graduate School for Neuro-science (Ruhr-Universitat Bochum) grant to N. Prochnow, Deutsche For-schungsgemeinschaft Grant SFB 509 (“Neurovision,” TP A8) to M. Schmidt,and National Eye Institute Grant EY-08233 to W. C. Hall.

R E F E R E N C E S

Baldauf ZB, Wang XP, Wang S, Bickford ME. Pretecto-tectal pathway: anultrastructural quantitative analysis in cats. J Comp Neurol 464: 141–158,2003.

Benassi C, Lui F, Biral G, Ferrari R, Corazza R. Correlation betweenamount of retinal afferents to the pretectal nucleus of the optic tract anddorsal terminal accessory optic nucleus and performance of horizontaloptokinetic reflex in rat. Behav Brain Res 45: 87–95, 1991.

Bickford ME, Ramcharan E, Godwin DW, Erisir A, Gnadt J, ShermanSM. Neurotransmitters contained in the subcortical extraretinal inputs to themonkey lateral geniculate nucleus. J Comp Neurol 424: 701–717, 2000.

Born G, Schmidt M. Inhibition of superior colliculus neurons by a GABAer-gic input from the pretectal nuclear complex in the rat. Eur J Neurosci 20:3404–3412, 2004.

Buttner U, Fuchs AF. Influence of saccadic eye movements on unit activityin simian lateral geniculate and pregeniculate nuclei. J Neurophysiol 36:127–141, 1973.

Buttner-Ennever JA, Cohen B, Horn AKE, Reisine H. Efferent pathways ofthe nucleus of the optic tract in monkey and their role in eye movements.J Comp Neurol 373: 90–107, 1996.

Cazin L, Precht W, Lannou J. Firing characteristics of neurons mediatingoptokinetic responses to rat’s vestibular neurons. Pfluegers Arch 386:221–230, 1980.

Collewijn H. Oculomotor areas in the rabbits midbrain and pretectum. J Neu-robiol 6: 3–22, 1975a.

Collewijn H. Direction-selective units in the rabbit’s nucleus of the optic tract.Brain Res 100: 489–508, 1975b.

Cucchiaro JB, Bickford ME, Sherman SM. A GABAergic projection fromthe pretectum to the dorsal lateral geniculate nucleus in the cat. Neuro-science 41: 213–26, 1991.

Cucchiaro JB, Uhlrich DJ, Sherman SM. Ultrastructure of synapses fromthe pretectum in the A-laminae of the cat’s lateral geniculate nucleus.J Comp Neurol 334: 618–630, 1993.

Distler C, Hoffmann KP. Visual receptive field properties in kitten pretectalnucleus of the optic tract and dorsal terminal nucleus of the accessory optictract. J Neurophysiol 70: 814–827, 1993.

Dodt HU, Zieglgansberger W. Visualization of neuronal form and function inbrain slices by infrared videomicroscopy. Histochem J 30: 141–152, 1998.

Fischer WH, Schmidt M, Hoffmann KP. Saccade-induced activity of dorsallateral geniculate nucleus X- and Y-cells during pharmacological inactiva-tion of the cat pretectum. Vis Neurosci 15: 197–210, 1998.

Fuchs AF, Mustari MJ, Robinson FR, Kaneko CR. Visual signals in thenucleus of the optic tract and their brain stem destinations. Ann NY Acad Sci.656: 266–276, 1992.

Gamlin PD. The pretectum: connections and oculomotor-related roles. ProgBrain Res 151: 379–405, 2005.

Hoffmann KP, Distler C, Ilg U. Callosal and superior temporal sulcuscontributions to receptive field properties in the macaque monkey’s nucleusof the optic tract and dorsal terminal nucleus of the accessory optic tract.J Comp Neurol 321: 150–162, 1992.

Hoffmann KP, Fischer WH. Directional effect of inactivation of the nucleusof the optic tract on optokinetic nystagmus in the cat. Vision Res 41:3389–3398, 2001.

Hoffmann KP, Schoppmann A. A quantitative analysis of the direction-specific response of neurons in the cat’s nucleus of the optic tract. Exp BrainRes 42: 146–157, 1981.

Hoffmann KP, Stone J. Retinal input to the nucleus of the optic tract of thecat assessed by antidromic activation of ganglion cells. Exp Brain Res 59:395–403, 1985.

Horn AK, Hoffmann KP. Combined GABA-immunocytochemistry andTMB-HRP histochemistry of pretectal nuclei projecting to the inferior olivein rats, cats and monkeys. Brain Res 409: 133–138, 1987.

Ibbotson MR, Dreher B. Visual functions of the retinorecipient nuclei in themidbrain, pretectum and ventral thalamus of primates. In: The PrimateVisual System: A Comparative Approach, edited by Kremmer J. Chichester,UK: Wiley, 2005, p. 213–266.

Ibbotson MR, Mark RF, Maddess TL. Spatiotemporal response properties ofdirection-selective neurons in the nucleus of the optic tract and dorsalterminal nucleus of the wallaby, Macropus eugenii. J Neurophysiol 72:2927–2943, 1994.

Ibbotson MR, Marotte LR, Mark RF. Investigations into the source ofbinocular input to the nucleus of the optic tract in an Australian marsupial,the wallaby Macropus eugenii. Exp Brain Res 147: 80–88, 2002.





ownloaded from


Ilg UJ, Hoffmann KP. Responses of neurons of the nucleus of the optic tractand the dorsal terminal nucleus of the accessory optic tract in the awakemonkey. Eur J Neurosci 8: 92–105, 1996.

Inoue Y, Takemura A, Kawano K, Mustari MJ. Role of the pretectalnucleus of the optic tract in short-latency ocular following responses inmonkeys. Exp Brain Res 131: 269–281, 2000.

Judge SJ, Wurtz RH, Richmond BJ. Vision during saccadic eye movements.I. Visual interactions in striate cortex. J Neurophysiol 43: 1133–1155, 1980.

Kato I, Harada K, Hasegawa T, Igarashi T, Koike Y, Kawasaki T. Role ofthe nucleus of the optic tract in monkeys in relation to optokinetic nystag-mus. Brain Res 364: 12–22, 1986.

Kato I, Okada T, Watanabe S, Sato S, Urushibata T, Takeyama I. Retinalganglion cells related to optokinetic nystagmus in the rat. Acta Otolaryngol112: 421–428, 1992.

Klauer S, Sengpiel F, Hoffmann KP. Visual response properties and afferents ofnucleus of the optic tract in the ferret. Exp Brain Res 83: 178–189, 1990.

Koontz MA, Rodieck RW, Farmer SG. The retinal projection to the catpretectum. J Comp Neurol 236: 42–59, 1985.

Lee PH, Schmidt M, Hall WC. Excitatory and inhibitory circuitry in thesuperficial gray layer of the superior colliculus. J Neurosci 21: 8145–8153,2001.

Lewald J, Petrasch-Parwez EW, Veh RW. GABA-like and glutamate-likeimmuno-reactivity in the pretecto-olivary pathway in the rat. J Hirnforsch35: 279–294, 1994.

Lui F, Giolli RA, Blanks RH, Tom EM. Pattern of striate cortical projectionsto the pretectal complex in the guinea pig. J Comp Neurol 344: 598–609,1994.

Magnin M, Kennedy H, Hoffmann KP. A double-labeling investigation ofthe pretectal visuo-vestibular pathways. Vis Neurosci 3: 53–58, 1989.

Missal M, Coimbra A, Lefevre P, Olivier E. A quantitative analysis of thecorrelations between eye movements and neural activity in the pretectum.Exp Brain Res 143: 373–382, 2002.

Mustari MJ, Fuchs AF. Discharge patterns of neurons in the pretectal nucleusof the optic tract (NOT) in the behaving primate. J Neurophysiol 64: 77–90,1990.

Mustari MJ, Fuchs AF, Pong M. Response properties of pretectal omnidi-rectional pause neurons in the behaving primate. J Neurophysiol 77: 116–125, 1997.

Nunes Cardozo B, Mize RR, van der Want JJL. GABAergic and non-GABAergic neurons in the nucleus of the optic tract project to the superiorcolliculus: an ultrastructural retrograde tracer and immunocytochemicalstudy in the rabbit. J Comp Neurol 350: 646–656, 1994.

Oyster CW. The analysis of image motion by the rabbit retina. J Physiol 199:613–635, 1968.

Oyster CW, Takahashi E, Collewijn H. Direction-selective retinal ganglioncells and control of optokinetic nystagmus in the rabbit. Vision Res 12:183–193, 1972.

Pereira A, Volchan E, Bernardes RF, Rocha-Miranda CE. Binocularity inthe nucleus of the optic tract of the opossum. Exp Brain Res 102: 327–338,1994.

Perry VH, Cowey A. Retinal ganglion cells that project to the superiorcolliculus and pretectum in the macaque monkey. Neuroscience 12: 1125–1137, 1984.

Price NS, Ibbotson MR. Pretectal neurons optimized for the detection ofsaccade-like movements of the visual image. J Neurophysiol 85: 1512–1521, 2001.

Prochnow N, Schmidt M. Spontaneous activity of rat pretectal nuclearcomplex neurons in vitro (Abstract). BMC Neurosci 5: 29, 2004.

Pu ML, Amthor ML. Dendritic morphologies of retinal ganglion cellsprojecting to the nucleus of the optic tract in the rabbit. J Comp Neurol 302:657–674, 1990.

Richmond BJ, Wurtz RH. Vision during saccadic eye movements. II. Acorollary discharge to monkey superior colliculus. J Neurophysiol 43:1156–1167, 1980.

Schiff D, Cohen B, Buttner-Ennever J, Matsuo V. Effects of lesions of thenucleus of the optic tract on optokinetic nystagmus and after-nystagmus inthe monkey. Exp Brain Res 79: 225–239, 1990.

Schiff D, Cohen B, Raphan T. Nystagmus induced by stimulation of thenucleus of the optic tract in the monkey. Exp Brain Res 70: 1–14, 1988.

Schmidt M. Neurons in the cat pretectum that project to the dorsal lateralgeniculate nucleus are activated during saccades. J Neurophysiol 76: 2907–2918, 1996.

Schmidt M, Schiff D, Bentivoglio M. Independent efferent populations in thenucleus of the optic tract: an anatomical and physiological study in rat andcat. J Comp Neurol 360: 271–285, 1995.

Schmidt M, Zhang HY, Hoffmann KP. OKN-related neurons in the ratnucleus of the optic tract and dorsal terminal nucleus of the accessory opticsystem receive a direct cortical input. J Comp Neurol 330: 147–157, 1993.

Schweigart G, Hoffmann KP. Pretectal jerk neuron activity during saccadiceye movements and visual stimulations in the cat. Exp Brain Res 91:273–283, 1992.

Sengpiel F, Klauer S, Hoffmann KP. Effects of early monocular deprivationon response properties and afferents of nucleus of the optic tract in the ferret.Exp Brain Res 83: 190–199, 1990.

Simpson JI, Giolli RA, Blanks RH. The pretectal nuclear complex and theaccessory optic system. Rev Oculomot Res 2: 335–364, 1988.

Sudkamp S, Schmidt M. Physiological characterization of pretectal neuronsprojecting to the lateral posterior-pulvinar complex in the cat. Eur J Neu-rosci 7: 881–888, 1995.

Volchan E, Rocha-Miranda CE, Picanco-Diniz CW, Zinsmeisser B, Ber-nardes RF, Franca JG. Visual response properties of pretectal units in thenucleus of the optic tract of the opossum. Exp Brain Res 78: 380–386, 1989.

Wang S, Eisenback M, Datskovskaia A, Boyce M, Bickford ME. GABAer-gic pretectal terminals contact GABAergic interneurons in the cat dorsallateral geniculate nucleus. Neurosci Lett 323: 141–145, 2002.

Wilson JR, Hendrickson AE, Sherk H, Tigges J. Sources of subcorticalafferents to the macaque’s dorsal lateral geniculate nucleus. Anat Rec 242:566–574, 1995.





ownloaded from
