areas pmls and 21 a of cat visual cortex: two functionally distinct areas

Areas PMLS and 21a of Cat Visual Cortex:Two Functionally Distinct Areas

B. Dreher,1 C. Wang,' K. J. Turlejski,12 R. L. Djavadian12 and W.Burke1-3

'Department of Anatomy and Histology and 'Department ofPhysiology, Institute for Biomedical Research, The University ofSydney, NSW 2006, Australia2Current address: Department of Neurophysiology, NenckiInstitute of Experimental Biology, 3 Pasteur Street, Warsaw02-093, Poland

We have compared the receptive field properties of neurons recordedfrom visuotopically corresponding regions of area 21a and theposteromedial lateral suprasylvian area (PMLS) of cat visualcortex. In both areas, the great majority of neurons wereorientation-selective and binocular, and their responses to movingcontours were modulated by simultaneous in-phase or anti-phasemotion of large textured background stimuli ('visual noise').However, despite the great hodological similarity between the twoareas, PMLS neurons had on average significantly higher peakdischarge rates, exhibited substantially greater direction selectivityindices, and preferred substantially higher stimulus velocities thanarea 21a neurons. Furthermore, the majority of binocular neurons inthe PMLS area and in area 21a were dominated respectively by thecontralateral and the ipsilateral eyes. Finally, while 46% of PMLSneurons were excited by movement of visual noise per se. only 25%of area 21a neurons could be excited by such stimuli. We argue thatthe PMLS area, like its presumed primate homologue themiddle-temporal (MT) area, is mainly involved in motion analysis. Bycontrast area 21a appears to be involved in pattern analysis ratherthan motion analysis. It is likely that phylogenetically area 21aderives from the PMLS area.

IntroductionThe visual cortices of virtually all mammals appear to containa number of distinct areas distinguished on a number ofhodological and functional criteria (for reviews, see Kaas andKrubitzer, 1991; Sereno and Allman, 1991; Spear, 1991). Inmammals with well-developed vision, such as virtually allprimates and some carnivores (e.g. the domestic cat), thenumber of distinct visual cortical areas is quite large (15-30areas; for reviews, see Rosenquist, 1985; Dreher, 1986; Fellemanand Van Essen, 1991; Kaas and Krubitzer, 1991; Sereno andAllman, 1991; Spear, 1991; Payne, 1993). The establishment ofhomologies, or even analogies, between particular cortical areasin different species of the same order or between the species ofdifferent orders could provide some crucial insights concerningthe steps involved in the evolution of the mammalian visualcortex. The task is, however, very difficult (cf. Sereno andAllman, 1991; Kaas and Krubitzer, 1991; Krubitzer and Kaas,1993; Payne, 1993), not least because the criteria for designatinga particular region of the visual cortex as a distinct area are farfrom uniform. Thus, only some areas, such as area VI (primaryvisual cortex, area 17, striate cortex) in all mammals, areas 18and 19 in cat visual cortex, and the middle temporal area (MT) ofOld World and New World monkeys, are designated as distinctareas on the basis of a whole battery of criteria such as: (i)distinct cytoarchitectonics and/or myeloarchitectonics; (ii)distinct hodology (especially in relation to the principal thalamicvisual relay nucleus—the dorsal lateral geniculate nucleus—andassociational and commissural cortico-cortical connections); (iii)distinct topographic (visuotopic) organization; and (iv) distinctreceptive-field properties of their neurons.

By contrast, the designation of virtually all other corticalregions as distinct cortical areas, even in such widely studiedorders as carnivores, primates or rodents, is based on morelimited criteria. One of the examples of a region of the visualcortex being subdivided into a number of distinct areas on thebasis of rather limited criteria is the lateral suprasylvian region ofthe cat visual cortex. One of the extreme cases in this examplerelates to the designation of area 21a and the posteromediallateral suprasylvian area (PMLS—in the middle suprasylviangyrus, in the vicinity of the middle and posterior suprasylviansulci) as two distinct visual cortical areas on the basis of only twocriteria: (i) their fairly distinct visuotopic organization and (ii)rather limited information, available at the time, concerningtheir hodology (Heath and Jones, 1971; Palmer etal, 1978; Tusaand Palmer, 1980; Tusa et al, 1981). Not surprisingly, thedesignation of area 21a and the PMLS area as two separatecortical areas has been frequently challenged. Thus, there areremarkable similarities between the two areas in the pattern oftheir thalamic interconnections. In particular, the principalthalamic input to each area originates in the lateral striate-recipient (or cortico-recipient) part of the lateral posterior/pulvinar complex (for reviews, see Rosenquist, 1985; Dreher,1986; Sherk, 1986; Shipp and Grant, 1991). Furthermore, it hasbeen pointed out that the interpretation of the visuotopicorganization of the region as indicating the existence of twodistinct areas is somewhat arbitrary, that the two postulatedareas are cytoarchitectonically very similar, and that both areasreceive their principal associational inputs from the supra-granular laminae of area 17 and to a lesser extent thesupragranular laminae of areas 18 and 19 (cf. Montero, 1981;Sherk, 1986; Grant and Shipp, 1991; Sherk and Mulligan, 1993).On the basis of the hodological similarities and the overallpattern of the visuotopic organization of the visual areas locatedaround the lateral suprasylvian sulcus, a number of authors (cf.Sherk, 1986; Shipp and Grant, 1991) consider the PMLS area andat least part of the area designated as area 21a as parts of a largersingle visual cortical area originally described by Marshall et al(1943) and further delineated in 1954 by Clare and Bishop (cf.Hubel and Wiesel, 1969; Wright, 1969; Spear and Baumann,1975; Turlejski and Michalski, 1975; Djavadian andHarutiunian-Kozak, 1983; Zumbroich et al, 1986; Grant andShipp, 1991; Sherk and Mulligan, 1993).

The argument in favour of the idea that area 21a and the PMLSarea constitute separate visual areas is strongly supported bystudies reporting major differences in the receptive fieldproperties of cells in these two areas (cf. Dreher, 1986; Mizobeet al, 1988; Spear, 1991; Dreher et al, 1993; Toyama et al,1994). However, the analyses of the receptive field properties ofarea 21a neurons and those of PMLS neurons were, with only oneexception (cf. Mizobe et al, 1988; Toyama et al, 1994),conducted in different laboratories. Naturally, in such a situation

Cerebral Cortex July/Aug 1996:6:585-599; 1047-3211/96/$4.00

by guest on February 12, 2013http://cercor.oxfordjournals.org/

Dow

nloaded from

http://cercor.oxfordjournals.org/

substantial functional similarities (as well as substantialadditional differences) between the two areas might have beeneasily missed or misinterpreted. Secondly, at least some of thereported differences might be related to the fact that thecomparisons between the two areas are largely based on samplesof cells with receptive fields at non-corresponding retinaleccentricities. In particular, while the majority of area 21a cellsanalysed quantitatively had their receptive fields located within5° of the area centralis (cf. Wimborne and Henry, 1992; Dreheret al, 1993), the majority of PMLS neurons whose responseswere analysed quantitatively had receptive fields located moreperipherally (Spear and Baumann, 1975; Turlejski, 1975;Camarda and Rizzolatti, 1976; Harutiunian-Kozak et al, 1984;Blakemore and Zumbroich, 1987; Rauschecker et al, 1987a;Zumbroich and Blakemore, 1987; Gizzi et al, 1990a,b). Thirdly,while virtually all area 21a neurons are reported to exhibitgenuine orientation selectivity, there is a substantial controversyin the literature concerning the orientation selectivity versusaxis of motion selectivity of most PMLS neurons (for review, seeSpear, 1991). Fourthly, although there is considerable supportfor the idea that the PMLS area is involved in motion rather thanform analysis (Turlejski, 1975; von Grunau and Frost, 1983;Hamada* 1987; von Grunau et al, 1987), very little informationis available concerning the responsiveness of area 21a neurons tothe motion of two-dimensional random texture patterns (socalled 2D visual noise) per se (cf. Toyama et al, 1994) and 2Dvisual noise/elongated contours interactions in area 21a and thePMLS area.

In the present study, in order to clarify some of these issuesand establish a more solid basis for functional comparisonsbetween the two areas, as well as the basis for comparisonswith extrastriate areas of other species, we have analysedquantitatively a number of receptive field properties of area 21aand PMLS area neurons with the receptive fields located 5-20°from the area centralis. Secondly, we have examined to whatextent neurons in area 21a and the PMLS area are sensitive to thebackground motion of 2D visual noise per se and whether theresponses of area 21a neurons, like those of PMLS neurons andareas 17 and 18, which provide the principal associational inputsto area 21a and the PMLS area, are influenced by relative motionbetween the contoured objects and the textured background.Preliminary communications, describing some of the findings,have been published (Wang et al, 1992a,b).

Materials and Methods

Animal PreparationThe general experimental procedures were similar to those described ina series of recent papers from our laboratory (Burke et al, 1992; Dreheret al, 1992, 1993). The initial surgery, which included intravenouscannulation, tracheotomy, bilateral cervical sympathectomy andcraniotomy, was carried out on cats anesthetized with 1-1.5% halothanein an N2O/O2 (67%/33%) gaseous mixture. During the recording sessionsthe animals were paralysed with gallamine triethiodide at the rate of 7.5mg/kg/h and artificially ventilated. Anaesthesia was maintained withN2O/O2 (67%/33%) gaseous mixture and i.v. infusion of 1 mg/kg/h ofsodium pentobarbitone in sodium lactate (Hartmann's solution) or 5%dextrose solution, the solutions being alternated every 12 h. Expired CO2was maintained at 3.7-4.0% by adjusting the stroke volume of thepulmonary pump. Body temperature was automatically maintained at-37.5°C. Heart rate and electroencephalogram (EEG) were monitored,and with the above regime the EEG always showed a slow wave record.Antibiotic (amoxycillin trihydrate, 125 mg; Amoxil, Beecham),dexamethasone phosphate (4 mg) and atropine sulphate (0.3 mg) weregiven i.m. on a daily basis.

The corneas were protected with zero-power, air-permeable plastic

contact lenses. Pupils were dilated and accommodation paralysed with1% atropine sulphate solution. The nictitating membranes were retractedwith 0.128% phenylephrine hydrochloride (Isopto-Frin, Alcon). Artificialpupils (3 mm in diameter) were placed in front of the contact lenses. Slitretinoscopy was used to determine the value of correcting lensesappropriate to focus the eyes on a tangent screen located 114 cm away.The locations of the optic discs and the areae centrales were plotted atleast twice daily using a back-projecting fundus camera.

RecordingCraniotomy for recording single neuron activities from areas 17 and 18 ofthe primary visual cortex was made at Horsley-Clarke coordinatesanterior 2 to posterior 6, lateral -2 to 8 (cf. Dreher et al, 1980, 1992;Burke et al, 1992). Craniotomy for recording single neuron activitiesfrom area 21a and the PMLS area was made at Horsley-Clarke coordinatesanterior 5 to posterior 7, lateral 9-17 (cf. Grant and Shipp, 1991; Dreheret al, 1993). A smaller opening in the dura was made above theappropriate region of the cortex and a perspex cylinder -15 mm high wasmounted around the opening and glued to the skull with dental acrylic toform a well around the craniotomy. A platinum/iridium glass-coatedmicroelectrode was lowered under visual control until its tip waspositioned at the cortical surface. The cylinder was then filled with 4%warm agar in physiological saline. To reduce brain pulsations the agar wascovered by warm liquid bone wax, which when cooled formed a firmclosure of the cylinder. The microelectrode was advanced into the cortexwith a hydraulic micromanipulator. In the case of recordings from area 17or 18 the shaft of the electrode was vertical. For recordings from area 21athe electrode shaft was angled in the coronal plane, so that the tippointed medially -30° to the vertical. This resulted in the electrode at thelevel of entry into the cortex being approximately orthogonal to thecortical surface. In the case of recordings from the PMLS area theelectrode shaft was angled (in the coronal plane) 30-40° to the vertical;this resulted in the electrode advancing roughly parallel to the surface ofthe medial bank of the lateral suprasylvian sulcus. The recordings fromthe PMLS area were restricted to coronal planes between anterior 3 andposterior 2. The action potentials of single neurons were recordedextracellularly and amplified conventionally; triggered standard pulseswere then fed to a PDP-11 computer.

Visual StimulationThe receptive fields of recorded units were plotted with a hand-held lightslit projector. The size of the receptive fields was determined as the sizeof the discharge field using moving bars that were lighter or darker thanthe background. Consistent with previous reports in the literature, wehave found that in virtually all cortical areas (cf. Dreher etal, 1993), thewidth of a discharge field, i.e., the dimension of the field perpendicular tothe axis of preferred orientation of a neuron, could be determined morereliably than its length, i.e., the dimension of the field along the axis of thepreferred orientation. Therefore, we have compared only the widthsrather than the size of discharge fields. Computer-controlled light slitsfrom a slide-projector, used for studying the properties of the receptivefields, usually had a luminance of 15 cd/m2 against the backgroundluminance of 0.9 cd/m2. A number of receptive field properties, e.g.,orientation tuning profile, velocity and direction selectivity, wereexamined quantitatively. The width of the orientation tuning curve wasmeasured at the half-height of the profile. The direction selectivity index(DO of a cortical cell was calculated by the following formula:

DI = [(/?, - R^)/Rr] * 100%

where R? and Rnp are the peak responses at the preferred andnon-preferred direction respectively (cf. Guedes et al, 1983; Burke etal,1992; Dreher etal, 1992, 1993).

To study the responses to 2D visual noise, one of the two texturepatterns routinely used by us (cf. Hammond and MacKay, 1977, 1981;Mason 1981; von Grunau and Frost, 1983; von Grunau etal, 1987; Crook1990a,b) was projected onto the tangent screen. One texture pattern(designated pattern 1) contained elementary rectangles of 0.28-0.33°. Inthe second texture pattern (designated pattern 2) elementary rectangleswere much smaller (0.05-0.07°). In each case the whole patternsubtended 32° x 22° on the screen. The responses elicited by the motionof pattern 1 and pattern 2 per se were examined separately. To study the

586 Areas PMLS and 2 la of Cat Visual Cortex • Dreher etal.


Dow

nloaded from


noise/contour interaction properties of the cortical cells, twocomputer-controlled slide-projectors were used, one for projecting animage of an oriented contour stimulus, usually a light slit, and the otherfor the contourless 2D visual noise (background stimulus).

The effect of a moving noise pattern on the response to a moving lightslit was determined in two test conditions: a slit and noise patternmoving: (i) in the same direction (in-phase) and (ii) in the opposite(anti-phase) direction. The responses were compared with a controlcondition when a light slit moved on the background of the stationarynoise pattern. Both the slit and noise pattern were superimposed at thecentre of the receptive field and moved, at optimal velocity, along the axisorthogonal to the orientation optimal for the slit. The amplitudes ofexcursions of the stimuli were such that when the slit travelled across thereceptive field, the noise pattern always covered the entire receptive field.

The net peak firing rate was used for assessment of the influence ofthe noise pattern on the response elicited by a moving slit. For the cellswhich did not respond to the noise pattern perse, net peak response wasobtained by subtracting the mean spontaneous discharge rates from thepeak responses. If the cell responded to the moving noise per se witheither distinct peaks or an increase of mean discharge rate, the noiseresponse locked in time with the peak responses obtained in thepresence of both moving slit and noise were subtracted to get the netpeak response. This was done under the assumption that the excitatoryprocesses elicited by either the slit or noise alone were processedadditively by cortical neurons. The noise effect index (NEI) was calculatedusing the following formula:

NEI = (R^JR^ " !) x 1°°%

where Rp and Rpc are the net peak responses at test and controlconditions respectively.

StatisticsThe sign of the effect, plus or minus, indicates the type of effect(facilitatory or inhibitory), while the absolute values give the relativestrength of the effect. In most cases, non-parametric tests—the Mann-Whitney {/-test, the chi-squared test, Fisher's exact probability test, theKruskal-Wallis test and Wilcoxon's matched-pairs signed-ranks test-wereused (Siegel, 1956). The parametric West was applied only for the paireddata when there was no significant difference in their variance. Statisticaldifferences were assessed at a level of P < 0.05. In all tests a two-tailedcriterion was used.

•Localization of Recording SitesLocalization of recording sites in area 21a and the PMLS area was basedon: (i) their relationship to the lateral suprasylvian and lateral sulci; (ii)visuotopic maps of area 21 a and the PMLS area (Palmer et al, 1978; Tusaand Palmer, 1980; Djavadian and Harutiunian-Kozak, 1983; Zumbroich etal, 1986; Grant and Shipp, 1991; Sherk and Mulligan, 1993); (iii)reconstruction of electrode penetrations; and (iv) different receptivefield-properties of neurons recorded in the reconstructed penetrationsthrough the part of area 19 which abuts area 21a (e.g., smaller receptivefields, high proportion of end-stopped neurons; for reviews, see Orban,1984; Dreher, 1986). Area 19 differs cytoarchitectonically from area 21a(cf. Sanides and Hoffmann, 1969; Heath and Jones, 1971). The recon-structions of electrode penetrations was based on small electrolyticlesions (5-7 uA for 10 s) in first and last electrode penetrations at -1.0mm separations. At the end of the recording session (lasting usually 3-5days) the animal was deeply anaesthetized with an i.v. injection of 120 mgof sodium pentobarbitone and perfused transcardially with warmHaitmann's solution followed by a 4% solution of paraformaldehyde in0.2 M phosphate buffer (pH 7.4). The brains were stereotactically blockedand sectioned coronally at 50 urn on a freezing microtome, mounted onslides and counterstained with cresyl violet.

Results

Receptive Field Locations and SizeThe receptive field centers of PMLS neurons were located 5-20°from the area centralis within a horizontal strip +3° to -10° fromthe horizontal meridian. The receptive field centers of area 21a

B

Figure 1 . W Side view of the left cerebral hemisphere of the cat with outlines of areas17,18.19. 21a. 21b. 20a and 20b, as well as lateral suprasylvian areas (anteromedial,anterolateral, posteromedial and posterolateral, dorsal and ventral; AMLS, PMLS, ALLS,PLLS, DLS and VLS, respectively), posterior suprasylvian (PS) and ectosylvian visual area(EVA). M. medial; C, caudal (after Tusa ef.a/., 1981; Mucke ef al.. 1982; Rosenquist,1985; Updyke. 1986; Olson and Graybiel, 1987). Black dots within areas 21a and PMLSindicate the locations of our electrode penetrations. (B) Outlines of two coronal sectionsthrough the left cerebral hemisphere with locations of electrode penetrations throughareas PMLS (1) and 21a (2).

neurons were located 5-20° from the area centralis within ahorizontal strip of+10° to -5° from the horizontal meridian (Fig.it; cf. Tusa and Palmer, 1980; Dreher etal.,\993).

For our sample of area 21a neurons the width of dischargefields varied from 1.0° to 7.2°, while for the sample of PMLSneurons it varied from 2.2° to 14.4°. Although the width ofdischarge fields varied considerably at all eccentricities, in bothareas there was a weak positive correlation between width ofdischarge fields and distance of the receptive field center fromthe area centralis. At a given eccentricity, however, the dischargefields of PMLS neurons tended to be wider than those of area 21aneurons. Thus, while for area 21a the mean width of dischargefields of receptive fields located 5-10° from the area centraliswas 365 ± 1.0° (n = 64), that for PMLS neurons was 5.6± 3.2° (n= 18). Similarly, while the mean width of dicharge fields of area21a cells with receptive fields located 10-20° from the areacentralis was 4.7 ± 0.9° (« = 32), that for PMLS neurons was 7.45± 4.7° (« = 40). The differences are statistically significant (P <0.05; Mann-Whitney C/test).

Receptive Field ClassesAlmost 80% (58/74) of area 21a cells and >75% (34/45) of PMLSneurons gave detectable, often quite strong, responses toappropriately oriented stationary flashing elongated (14° x 0.6°)light slits positioned in their receptive fields. The great majoritiesof cells recorded in either area either gave transient responses atboth onset and offset of flashing stationary stimuli positionedanywhere within their discharge fields (ON/OFF discharge

Cerebral CortexJuly/Aug 1996, V 6 N 4 587


Dow

nloaded from


Area 21a PMLS area

1J« 40

1 30S 20

I 1°

-o— moving bar g- • • • flashing bar 601

unit 14.720

—o— moving bar-••- fla<htng bar

Area 2 la i')3 cells)

OS -120 -80 -40 40 80 120 -120 -SO -40

Stimulus orientation (deg)40 80 120

D Area 21a (62 cells)• PMLS area (42 cells)

0-30 30-60 60-90 90-120 120-150

Orientation tuning width (deg)

Figure 2. Orientation tuning in areas 21a and PMLS. \AB) Examples of orientationtuning curves illustrating the effect of moving versus flashing bars. (C) Histogram of thedistribution of orientation tuning widths (width at half-height) for moving light bars forour samples of area 21 a and PMLS area neurons. Nor, non-oriented cells.

fields), and/or their discharge fields plotted with moving stimulidarker than the background overlapped completely with thoserevealed with moving stimuli brighter than the background. Onthat basis, almost 95% (70/74) of cells recorded from area 21aand >93% (42/45) of cells recorded from the PMLS area wereidentified as C-cells (see C-cells and B-cells of Henry, 1977; cf.complex cells of Hubel and Wiesel, 1962, 1965). The remainingfew cells were identified as subtypes of S-cells since they hadonly the ON or only the OFF subregions in their discharge fields(see S-cells and A-cells of Henry, 1977; cf. simple cells of Hubeland Wiesel, 1962).

Orientation SelectivityFor neurons in area 21a or the PMLS area, the orientation tuningcurve obtained using a stationary flashing light slit was alwaysnarrower than that obtained a using moving light slit (Fig. 24 JJ*).Indeed, the mean orientation tuning width (at half-height)derived from the responses of 20 cells, recorded in area 21a, to astationary elongated (10° x 0.6°) flashing light slit was 42.4 ±17.8°, i.e., significantly narrower (P < 0.05; Wilcoxon's test) thanthat (53.6 ± 25.6°) derived from the responses of the same 20cells to elongated moving light slits. Similarly, for 16 PMLSneurons for which we have obtained the complete orientationtuning curves using both moving and stationary flashingelongated slits, the mean width of orientation tuning curves,derived from the responses to stationary flashing stimuli, was43.4 ± 18.2° while that derived from the responses to movingstimuli was 55.5 ± 24.5°. The difference was statisticallysignificant (P < 0.05; Wilcoxon's test).

Generally, consistent with previous findings (Wimborne andHenry, 1992; Dreher etaL,\993), for a given 21a cell, the widthof the orientation tuning curve was very similar irrespective

2 3 4

Eye dominance class

-? B

w 50 i D Area 21a (77 cells)I PMLS area (53 cells)

40-

0-20 20-40 40-60 60-80 >80

Peak discharge rate (spikes/s)

50-

40-

30-

20-

10-

n -

a•

J H

Area 2 la (79 cells)PMLS area (53 cells)

_ JOE0-20 2O40 40-60 60-80 80-100

Direction selectivity index (%)

Figure 3. Eye dominance, peak discharge rates and direction selectivity indices. (4)Histograms of the proportion of neurons in each of five eye dominance classes for cellsin the PMLS area and area 21a. Monocular neurons which could be photically excitedonly through the contralateral or through the ipsilateral eye were classified, respectively,as class 1 or class 5 neurons. Binocular neurons in which the photic stimuli presentedvia the contralateral eye evoked the excitatory response stronger than, equal to orweaker than that evoked by photic stimuli presented via the ipsilateral eye wereclassified respectively as class 2, 3 and 4 neurons (cf. Noda era/.. 1971; Burke era/.,1992; Dreher eta/., 1992,1993). (fi) Histograms of the proportion of neurons in each offive groups covering the range of peak discharge rates for cells in area 21 a and the PMLSarea. (C) Histogram form of the direction selectivity indices for neurons in area 21a andthe PMLS area. Note the significantly higher indices for PMLS neurons.

(within limits) of the length of elongated slits or bars used. Bycontrast, consistent with numerous previous reports (cf. Spearand Baumann, 1975; Camarda and Rizzolatti, 1976; Blakemoreand Zumbroich, 1987; Toyama etaL, 1994), PMLS neurons werebroadly tuned when stimulated with moving light spots. Indeed,the mean widths of the tuning curves derived from the responsesof 13 cells to a moving light spot (2.5° in diameter) at 92.4 ±38.1° was significantly broader (/> < 0.05; Wilcoxon's test) thanthat (555 ± 119°) derived from the responses of the same cellsto a moving elongated (16° x 0.6°) light slit.

Although, unlike in the case of our sample of area 21aneurons, some PMLS cells did not exhibit orientation selectivityto long (16° x o.6°) moving light slits, and very few of themexhibited sharp orientation selectivity of <30° (Fig. 2Q, themean width, at half-height, of orientation tuning curves tomoving long light slits of PMLS neurons at 57.5 ± 24.9° is notsignificantly different (P = 0.35; Mann-Whitney U-tesi) from that(54.9 ± 27.8°) for the sample of area 21a neurons.

588 Areas PMLS and 21a of Cat Visual Cortex • Drcher et aL


Dow

nloaded from


Area 21a neurons PMLS neurons

100D

100 1000

In"35

1.inSo

tesp

40

30

20

10

n

B

Af/Ay / •• /

—o— unit 9.9- • — unit 14.14

/ \ - - • - • unil 14.5

Sv. >

100

80

60

40

20

0

•o— unit 903.21— unit 903.16••-• unit 903.11

100 1000

unit 11.5- unit 10.8

100 1000

- unit 903.8-unit 903.18

100 1000

Velocity (deg/s)

Figure 4. Examples of velocity response curves for neurons in area 21a and the PMLSarea. (4) Low-pass cells; \&£) low or moderate velocity-tuned cells; (C/) broad-bandcells; (D) high-pass cells. Few PMLS neurons were low-pass and no 21a neurons werehigh-pass.

Ocular DominanceFor 93 neurons recorded from area 21a and 44 neurons recordedfrom the PMLS area we compared the strength of the responsesevoked by photic stimulation of each eye and classified themaccording to the five-group scale of eye dominance modifiedfrom the seven-group scale of Hubel and Wiesel (1962). Asindicated in Figure 3-4, the great majorities of cells in either area(93% of cells recorded from the PMLS area and 85% of thoserecorded from area 21a) could be activated by photic stimulationvia either eye (binocular neurons). However, it is clear fromFigure 3A that PMLS neurons are dominated by the contralateraleye (eye dominance classes 1 and 2), whereas area 21a neuronsare dominated by the ipsilateral eye (eye dominance classes 4and 5). Indeed, the differences between the two areas in theproportion of cells in different eye dominance groups weresignificant (P = 0.0005; chi-squared test).

Peak Discharge RatesAs indicated in Figure 3B in both the PMLS area and area 21a thepeak discharge rates for optimally oriented stimuli moving atpreferred velocities across the receptive fields varied quiteconsiderably. The range of peak discharge rates for neurons inboth areas was fairly similar (12-158 spikes/s for PMLS neuronsand 6-127 spikes/s for area 21a neurons). However, the meanpeak discharge rate for our sample of PMLS neurons wassubstantially higher than that for our sample of area 21a neurons(58.3 ± 35.1 spikes/s for the PMLS area; 34.6 ± 24.7 spikes/s for

area 21a). The difference between the two areas is highlysignificant (P < 0.00006; Mann-Whitney (/-test).

Direction SelectivityFor all cells (whether in the PMLS area or in area 21a) exhibitingstrong directional selectivities (direction-selectivity indices>70%) the preferred direction of stimulus motion along the axisorthogonal to the optimal orientation was the same whether thestimulus was brighter or darker than the background. On theother hand, most cells exhibiting strong direction selectivities atoptimal velocities were less or not at all direction selective whenstimuli moved at slow velocities. Quite often direction selectivityindices also decreased at velocities substantially higher than theoptimal. Therefore, for each cell the DI was calculated foroptimally oriented light slits moving at optimal velocities. It isapparent from Figure 3C that many PMLS cells, but few 21a cells,were strongly direction selective. Not surprisingly, the meandirection selectivity index for our sample of PMLS neurons at72.3 ± 21.4% is much higher than that (36.7 ± 23.3%) for oursample of area 21a neurons. The difference between the twopopulations was highly significant (P < 0.00006; Mann-Whitney£/-test).

Velocity PreferencesNeurons in area 21a and the PMLS area differ substantially intheir velocity tuning profiles. Thus, applying the terminology ofOrban and his colleagues (for review, see Orban, 1984), -48.5%(35/72) of area 21a neurons were either low velocity-pass (Figs4A and 5A) or low velocity-tuned (Figs 4B and 5A).

About 41.5% (30/72) of area 21a cells were tuned formoderate velocities (Figs 4B and 5A) and only a small proportion(9-5%; 7/72) of area 21a neurons responded more or less equallywell to a wide range of stimulus velocities (broad-band cells; Figs4C and 5/1).

By contrast, relatively few PMLS neurons could be classified aslow velocity-tuned (Fig. 4E) or low-pass cells (21%; 10/74; Fig.48). Moderate velocity-tuned (Fig. 4E) and broad-band neurons(Fig. 4F) together constituted a large majority (72.5%; 34/47; Fig.5B) of PMLS neurons. Finally, unlike in area 21a, a distinct groupof PMLS neurons (6.0%; Figs 4D and 5A) were high-pass cellswhich responded only at moderate and high stimulus velocities.

Apart from the differences in the velocity tuning profiles, area21a and the PMLS area differed substantially in distribution of thepreferred stimulus velocities. Thus, as indicated in Figure 5B,while almost 45% (34/78) of area 21 neurons respondedoptimally at velocities not exceeding 5°/s, the proportion ofsuch cells among PMLS neurons was very low (3/53). Bycontrast, none of area 21a neurons responded optimally atstimulus velocities exceeding 50°/s, while among PMLS neurons-22.5% (12/53) of cells responded optimally at such velocities.The mean preferred velocity for our sample of PMLS neurons at69.9 ± 1537s was significantly higher (P < 0.00006; Mann-Whitney C/test) than that (12.8 ± 12.5°/s) for the present sampleof area 21a neurons.

Excitatory Responses to Visual Noise Per Se

Area 21a and the PMLS areaForty-six per cent (12/26) of neurons in the PMLS area gaveexcitatory responses to noise motion per se, a figure substantiallyhigher than that (25%; 9/36) in area 21a. The difference wassignificant (P < 0.05; chi-squared test). However, <10% of cells ineach area (3/36, 8.3% of cells in area 21a; 2/26, 7.7% of cells in

Cerebral CortexJuly/Augl996, V6N4 589


Dow

nloaded from


the PMLS area) responded tonically to the noise motion. By tonicexcitatory response we mean that the response lasted for thesame time as the noise motion, although, due to the responselatencies, the onset and offset of the response did not coincidewith the onset and cessation of noise motion (Fig. 6A,C). Thistype of response to 2D noise is strongly reminiscent of theso-called 'field' responses of some collicular or striate cortexneurons (Mason, 1979; Gulyas etaL, 1987)- Field responses werepresent irrespective of the size of the pixels in the noise patternused. Thus, if, in a given cell, pattern 1 (with its large pixels)evoked a field response, pattern 2 (with its much smaller pixels)also evoked a 'field' response (Fig. 6Q.

However, in both areas, the majority of cells responding to themotion of the visual noise perse (6/9 cells in area 21a and 10/12cells in the PMLS area) appeared to respond to distinct featuresin the noise pattern rather than the motion of the whole noise. Inparticular, the responses of these cells to noise moving throughtheir receptive fields did not last for the whole duration ofmotion but consisted of one or several discharge peaks similar tothose evoked by moving slits of light (Fig. 6BJJ). Theseresponses are strongly reminiscent of the so-called 'grain'responses characterizing the responses to the motion of visualnoise of virtually all geniculate neurons (Mason, 1976; Gulyas etaL, 1987), and those of many neurons in the superior colliculusand areas 17 and 18 (Mason, 1979; Gulyas et aL, 1987).Consistent with the grain character of these responses, in somecells studied by us these responses were abolished when noisepattern 1, with its large pixels, was substituted by noise pattern2, with its smaller pixels. The proportion of cells responding todistinct features of the noise pattern was significantly lower inthe case of area 21a neurons than in the case of PMLS neurons(16.5% of the sample of area 21a neurons versus 38.5% of thesample of PMLS neurons).

Whatever the type of response to the visual noise, the peakdischarge rate in the excitatory response evoked by the moving2D noise per se was in both areas usually, but not always, lowerthan that in the excitatory response evoked by the moving lightslit (Fig. 6). For all cells responding to the visual noise per se thepreferred direction and velocity for the visual noise were similarto those for the light slits (Fig. 6). In both areas, the directionselectivity indices for the noise moving along the axis optimal forthe light slits were similar to those for the light slits. In particular,for area 21a neurons, the mean direction selectivity index of theresponses evoked by the motion of 2D noise at 41.3 ± 19-4% (n =9) was not significantly different from that (47.7 ± 20.6%; n = 9)of the responses evoked in the same cells by moving light slits (P> 0.10; Wilcoxon's test). Similarly, in the PMLS area the meandirection selectivity index of responses evoked by the motion of2D noise at 58.4 ± 28.5% (» = 12) was not significantly differentfrom that (77.7 ± 17.5%; n = 12) of the responses evoked in thesame cells by moving light slits (P > 0.10; Wilcoxon's test).

Although in area 21a all neurons giving excitatory responsesto the visual noise per se had C-type receptive field organization,in the PMLS area, some neurons (all of them exhibiting a 'grain'type of response) had S-type receptive field organization.

Areas 17 and 18Responses to visual noise in areas 17 and 18 were examined forpurposes of comparison. Consistent with previous reports (cf.Gulyas et aL, 1987, 1990; Crook, 1990a), a majority of cells inthe present sample of area 17 and area 18 neurons was sensitiveto moving noise per se. In particular, 75% (15/20) of cellsrecorded from area 17 and 83.5% (15/18) of cells recorded from

D Area 21a (72 cells)• PMLS area (47 cells)

B

Low-Pass LowVT Medium VT ^,re"J High-PassBand

Velocity Selectivity Profiles

Area 2 la (78 cells)PMLS area (S3 cells)

5-10 10-20 20-50 50-100

Preferred velocity (deg/s)

>100

Figure 5. Distribution of velocity selectivity profiles (4) and preferred velocities (6) forneurons in area 21a and the PMLS area. See text for further explanations.

area 18 responded to 2D noise moving along the axis orthogonalto the optimal orientation. Of these a majority (10/15, 66.5% ofarea 17 cells and 12/15, 80% of area 18 cells) exhibited a 'grain'type of response. Consistent with the original reports ofHammond and MacKay (1975,1977), all area 17 and area 18 cellswhich exhibited field type excitatory responses to the motion of2D visual noise had C-type receptive field organization (cf.,however, Skottun et aL, 1988; Casanova et aL, 1995). On theother hand, in both areas 17 and 18 a substantial proportion ofcells exhibiting strong grain type responses were S-type cells(Gulyas etaL, 1987, 1990; Crook, 1990a; Casanova etaL, 1995).

In our sample of area 17 cells, the mean DI of the responsesevoked by motion of 2D visual noise was 49.8 ± 24.0% (n = 15)and not significantly different (P > 0.10; Wilcoxon's test) fromthat of the responses evoked by light slits moving along the sameaxis (56.1 ± 33.8%; n = 15). Similarly, in our sample of area 18cells, the mean DI of the responses evoked by motion of 2Dvisual noise at 51.8 ± 21.0% (n = 16) was not significantlydifferent (P> 0.10; Wilcoxon's test) from that (52.6 ± 318%; n =16) of the responses evoked by light slits moving along the sameaxis.

Effect of Movement of Visual Noise on the Responses to aMoving SlitThe responses of the great majority (89%, 32/36) of area 21a cellsto a moving slit were modulated by simultaneously moving 2Dvisual noise. Furthermore, most cells (67%, 24/36) showed asuppressive effect irrespective of the relative motion phasebetween the slit and the noise. Figure 74 shows an example ofthe responses of a typical area 21a neuron. Moving noise pattern

590 Areas PMLS and 21a of Cat Visual Cortex • Drchcrctal.


Dow

nloaded from


Bar

Noise

2.05s

2.05s

D

3.3s

20 spikcs/s

unit 903.13

3.0s

Noise

"I 3.0s 1.3s

Figure 6. Peristimulus-time histograms of the responses of neurons in area 21 a and the PMLS area to moving light bars and moving 2D noise. The small icons show the angle of thebar and the direction of movement. On the left side of each histogram the noise or bar moves in one direction across the receptive field and on the right side moves back across thereceptive field. The stimulus moves only during the time corresponding to the hatched bars beneath the abscissae. At other times the stimulus is stationary. The reverse trajectorycommences at the arrowheads. Note that the response to the noise pattern can be either maintained ('field response'; AC) or transient ('grain response'; BJ)).

unil 10.4 (C-lype)

moving bar+ stationary noise

moving bar4- in-phase moving noise

moving bar+ anti-phase moving noise

B

Ia

O cells not responding lo noise per se• cells responding 10 noise per te

-100 -80 -60 -40 -20 0 20 40

Effect of in-phase movement of noise

Figure 7. Interactions between moving bar and noise in area 21a. (A) Peristimulus histograms for four conditions. The bar moves in one direction across the receptive field in the leftside of each histogram; at the arrowhead it begins to move back across the receptive field, (fl) Graph of the relative effects of in-phase and anti-phase noise. Negative values on theordinates indicate reduction of the response to the bar. The dashed line indicates an equal effect of in-phase and anti-phase noise. The solid line is the linear regression line for all points.Note that the in-phase noise has a greater inhibitory effect than the anti-phase noise.

per se did not produce an excitatory response. On the otherhand, 2D noise moving in-phase or anti-phase with the light slitconsiderably reduced the response to a slit. Furthermore, for thiscell the reduction during in-phase motion of the noise wasgreater than that during anti-phase motion.

Generally, as indicated in Figure IB, the in-phase suppression(45.4 ± 23.6%) was substantially stronger (P = 0.0001; Mest) thanthe anti-phase suppression (28.4 ± 21.9%). Strong facilitatoryeffects of motion of 2D noise on the response to a movingcontour were very rare (one cell, anti-phase motion; Figs IB and8/J).

We found a clear correlation (r = 0.54, P < 0.02) betweenthe strength of the in-phase and anti-phase noise/contourinteractions. However, for both in-phase and anti-phasenoise/contour interactions the strength of suppressive effect wassimilar for cells exhibiting excitatory responses to noise per seand those which did not (P > 0.2; Mann-Whitney U-test; cf. Fig.IB). The strength of noise/contour interactions was notcorrelated (r < 0.32, P > 0.05) with other receptive fieldproperties, such as width of the receptive field, directionselectivity index or optimal velocity.

The relative motion of noise and contour stimuli appears also

Cerebral CortexJuly/Aug 1996, V 6 N 4 591


Dow

nloaded from


uron

s

=

opor

tion

oi

80-

60-

40-

20-

0-

80-

60 '

40-

20-

0-

A

n<-60

B

EL

Area 21aD in-phase (mean -45.4%)

^ ^ • anti-phase (mean-28.5%)

H 36 cells

Ii,n

20--60 -20-+20 +20-460 >+60

Area 17Q in-phase (mean -32.2%)• anti-phase (mean -23.5%;

_ 20 cells

l l l n-20--60 -20-+20 +20-+60 >+60

Area 18D in-phase (mean-42.0%)• anti-phase (mean -32.7%)

18 cells

<-60 -20--60 -2O-+2O +20-+60 >+60suppression facilitation

± 24% (n = T) the mean DIs for the same cells when the light slitsmoved in-phase or anti-phase with moving 2D noise wererespectively 70.3 ± 16.9% and 69.6 ± 27.6%.

Areas 17 and 18The distributions of 2D noise motion effects on the magnitude ofthe responses to moving elongated contours in cells recordedfrom areas 17 and 18 are shown in Figure SB and C respectively.Consistent with previous reports (cf. Gulyas et al., 1987, 1990;Crook, 1990a), the major effect of noise/contour interactions inthe present sample of neurons in areas 17 and 18 was thereduction of the amplitude of the responses to elongatedcontours. Furthermore, statistical analysis indicates that the cellsin each area (21a, PMLS, 17 and 18) are subject to a similar levels(P > 0.40; Kruskal-Wallis test) of modulatory effects of 2D noisemotion.

It has been shown in the past that about a third ofnon-direction-selective cells in area 17 became more direction-selective when the 2D noise and bar moved at the same speed,either in-phase or in anti-phase (Orban et al., 1987). In thepresent samples of area 17 and 18 neurons the mean DI when anoptimally oriented light slit moved against a background ofstationary 2D noise was 62.2 ± 8.0% (« = 20) for area 17 and 58.6± 25.6% in = 17) for area 18. During in-phase movement of thebackground the DIs increased to 68.6 ± 23.8 and 71.3 ± 24.9% forareas 17 and 18 respectively. The differences are significant (P <0.04 and P < 0.02 for areas 17 and 18 respectively; Wilcoxon'stest). By contrast, however, DIs during anti-phase movement of2D noise for area 17 (59.1 ± 293%) and area 18 (59.1 ± 28.2%)were not significantly different from those when the light slitsmoved on a background of stationary noise (P > 0.10; Wilcoxon'stest).

Figure 8. Distribution of effects of movement of 2D noise on the responses to themoving bars in areas 21a, 17 and 18. The dominant effect in all areas is reduction of theresponse to the moving bar, the in-phase motion generally having a stronger effect.Facilitatory effects were rare in all areas.

to modulate the direction selectivity of some cells (see Fig. 7A).Indeed, the mean DI of 29.5 ± 20.5% (n = 36) when a single barmoved on a stationary noise background increased to 36.8 ±24.2% and to 43-6 ± 26.8% respectively during in-phasemovement of the background and during anti-phase movementof the background. However, the changes in direction selectivityindices were not significant (P > 0.1; Wilcoxon's test).

For all seven PMLS neurons tested, the responses to a movinglight slit were modulated by simultaneously moving 2D visualnoise. All but one of them showed a suppressive effectirrespective of the relative motion phase between the slit and thenoise. Unlike in area 21a, the mean suppressive effect of thein-phase motion (at 61.7 ± 26.9%) was not significantly differentfrom that (75.4 ± 459%) for the anti-phase motion (P > 0.20;Mann-Whitney {/-test). In one cell simultaneous motion of 2Dnoise and light slit substantially increased the cell's firing rate.The facilitatory effect was restricted to the anti-phase motion,the in-phase motion of the noise and the slit resulting in aresponse of the same magnitude as that evoked by the slit movingagainst a background of stationary 2D noise.

The direction selectivity indices of PMLS neurons were notsignificantly affected by the noise/contour interactions. Inparticular, whereas the mean DI for the optimally oriented lightslits moving on the background of the stationary 2D noise was 72

Discussion

Area 21a versus the PMLS Area: Differences andSimilaritiesAs mentioned in the Introduction there are strong similaritiesbetween area 21a and the PMLS area in the composition of theirprincipal thalamic and cortical associational afferents. Consistentwith this, we have found that the two areas contain almostidentical proportions of C-cells and S-cells and their orientationtuning widths are not significantly different (cf. Table 1).However, in almost all other aspects of their receptive fieldorganization, the two areas differed significantly from each other(Table 1).

One could argue that since our comparisons of functionalproperties of neurons in the two areas were restricted to cellswith receptive fields located 5-20° from the area centralis, ourconclusions are valid only in relation to those eccentricities.However, several lines of evidence suggest to us that ourconclusions about the significant differences in the receptivefield properties between area 21a and the PMLS area are alsovalid for cells with centrally located receptive fields. Inparticular, the receptive field properties of area 21a neuronswith receptive fields located within 5° of the area centralis are inall aspects of receptive field organization (except receptive fieldsize and sharpness of the orientation tuning) indistinguishablefrom those with receptive fields located more peripherally (cf.Dreheretai, 1993). Furthermore, although to our knowledge nosystematic quantitative comparison is available, most PMLS cellswith centrally located receptive fields like those withperipherally located fields are binocular, dominated by the

592 Areas PMLS and 21a of Cat Visual Cortnc • Dreher et aL


Dow

nloaded from


Table 1Comparison of RF properties of neurons in the PMLS area and area 21a

Property PMLS area Area 21a

% of C (complex) cellsOrientation tuning width

(deg, mean ± SO)% ocular dominance

Class 2Class 4

Peak discharge rate(spikes/sec, mean ± SD)

Direction selectivity index|%. mean ± SD)

Optimal velocity(deg/s. mean (range))

% of cells excited by 'visual noise'

93.0

57.5 ± 24.9

48.028.0

58.3 ± 35.1

72.3 ± 21.4

69.9 (5-990)

46.0

95.054.9 ± 27.8

22.546.0

34.6 ± 24.7

36.7 ± 23.3

12.8(2-50)

25.0

>0.6.NS"

0.35. NS6

<0.00053

<0.00006"

<0.00006b

<0.00006b

<0.05"

"Chi- squared test two-tailed criterion.bMann-Whitney (/-test two-tailed criterion.

contralateral eye and strongly direction selective (cf. Hubel andWiesel, 1969; Camarda and Rizzolatti, 1976; Rauschecker et al.,1987a; Gizzi et al., 199Ob).

Receptive-Field Classes.The high proportions of C-type and low proportions of S-typecells in the present sample of area 21a neurons are in goodagreement with our own data in the previous sample of area 21aneurons (Dreher et al, 1993; cf. Dreher, 1986) as well as withthe data of Wimborne and Henry (1992). Athough the greatmajority of PMLS cells responding to stationary flashing stimuligive transient ON/OFF responses and would thereforecorrespond to our C-cells, a proportion of PMLS cells could beclassified as subtypes of S-cells since they contained either onlyOFF or only ON discharge regions in their receptive fields(Turlejski, 1975; Harutiunian-Kozak et al, 1984; Blakemore andZumbroich, 1987; for review, see Spear, 1991). The pre-dominance of C-type cells in the PMLS area is in good agreementwith previous reports (Hubel and Wiesel 1969; Wright, 1969;Morrone et al, 1986; Blakemore and Zumbroich, 1987;Zumbroich and Blakemore, 1987; Gizzi etaL, 1990b).

Orientation SelectivityWhen moving elongated light bars were used, area 21a and thePMLS area do not differ significantly with regard to the widths oforientation tuning curves of their respective neurons. However,while cells in area 21a give similar orientation tuning curves,regardless of the length of the bar used (Wimborne and Henry,1992; Dreher et al., 1993), cells in the PMLS area wereconsidered to exhibit axial motion selectivity rather thanorientation selectivity since they are very broadly tuned fororientation when the short bars or spots are used (Spear andBaumann, 1975; Camarda and Rizzolatti, 1976; Blakemore andZumbroich, 1987; Toyama et al, 1994; cf. present study). On theother hand, the present findings (cf. also Danilov et al, 1995)that the great majority of PMLS cells are clearly orientationselective when long (14° or more) stationary flashing light slitsare used, combined with the previous findings (Blakemore andZumbroich 1987; Hamada 1987; Gizzi et al, 1990a) that PMLSneurons respond in an orientation-selective manner whenmoving or stationary flashing gratings are used, indicate that thegreat majority of PMLS neurons is indeed orientation selectiverather than exhibiting axial motion selectivity (cf. Henry et aL,1974a,b, 1994). This conclusion is consistent with the original

conclusions of Hubel and Wiesel (1969) and Wright (1969)based on the responses of Clare-Bishop (PMLS) cells to movingsingle light slits and dark bars.

Ocular DominanceThe proportions of binocular neurons in area 21a and in thePMLS area, as well as the overall distributions of eye dominanceclasses in these areas, are significantly different from those in oursamples of area 17 and area 18 cells (Burke etaL, 1992; Dreher etal, 1992; cf. Sherk, 1989). In the striate- (or cortico-) recipientzone of the lateral posterior-pulvinar complex (LP1) whichreceives its principal input from areas 17 and 18 and provides theprincipal thalamic input to area 21a and the PMLS area, 75-90%of neurons are binocular (Rauschecker etaL, 1987b; Casanova etal, 1989; Chalupa and Abramson, 1989; Casanova and Savard,1995). As in the PMLS area, but unlike in area 21a, only a smallproportion of binocular neurons in the cortico- recipient zone ofLP1 is dominated by the ipsilateral eye and the majority ofmonocular neurons could be activated via the contralateral eye(Rauschecker et aL, 1987b; Casanova et al., 1989; Chalupa andAbramson, 1989; Casanova and Savard, 1995).

The effect of inactivation of areas 17 and 18 on oculardominance of area 21a neurons is not known. However, chronicor acute, irreversible bilateral or unilateral (ipsilateral) removalsof areas 17 and 18 result in a dramatic reduction in theproportion of PMLS neurons that can be activated via theipsilateral eye—the large majority of neurons become monocularand can be activated only via the contralateral eye (Spear andBaumann, 1979; cf, however, Guedes et al, 1983).

Peak Discharge RatesThe peak discharge rates of PMLS neurons are significantlyhigher than those in our sample of area 17 neurons (cf. Burke etal., 1992) but not significantly different from those in our sampleof area 18 neurons (cf. Dreher et al., 1992). By contrast, the peakdischarge rates of area 21a neurons are substantially lower thanthe peak discharge rates of the present sample of PMLS neurons,those of our sample of area 18 neurons (Dreher et aL, 1992), aswell as slightly but significantly lower than those of our sampleof area 17 neurons (Burke etaL, 1992). Thus, the peak dischargerate of area 21a neurons seems to be mainly determined by theinput from area 17, whereas the discharge rate of PMLS neuronsappears to reflect the input from area 18.

Direction SelectivityThe mean direction selectivity indices for the present samples ofarea 21a neurons are not only significantly lower than those ofPMLS neurons (PMLS, 72.3% versus area 21a, 36.7%) but alsosignificantly lower than those of area 17 and area 18 neurons,while the direction selectivity indices of PMLS neurons aresignificantly higher than those of area 17 and area 18 neurons(cf. Burke et al., 1992; Dreher et al, 1992). Despite the strongdirection selectivity indices of PMLS neurons, removal of thePMLS area (and the PLLS area) does not produce a detectabledeficit in the animal's ability to discriminate directions ofmoving gratings (Pasternak et al., 1989). Thus, it appears that thedirection discrimination is based on signals coming from visualareas outside the lateral suprasylvian cortex. Chronic or acute,bilateral or unilateral (ipsilateral) removal of areas 17 and 18results in a dramatic reduction in the proportion of direction-selective PMLS neurons (Spear and Baumann, 1979; cf., however,Guedes et al., 1983), but tends to increase the directionselectivity of area 21a neurons (Michalski etaL, 1993). However,

Cerebral Cortex July/Aug 1996, V 6 N 4 593


Dow

nloaded from


since such ablations strongly reduce the responsiveness of bothsets of neurons to photic stimuli (cf. Spear and Baumann, 1979;Michalski et aL, 1993, 1994) this result is difficult to interpret.

The contribution of neurons in LP1 (which constitute theprincipal thalamic input to both area 21a and the PMLS area) tothe direction selectivities of area 21a and PMLS neurons is notknown. However, direction selectivity indices of LP1 neuronsappear to be much greater than those of area 21a neurons butcomparable to those of areas 17 and 18 or even to those of PMLSneurons (Rauschecker et aL, 1987b; Casanova et aL, 1989;Chalupa and Abramson, 1989; Casanova and Savard, 1995; seealso Mason, 1981).

Velocity PreferencesAs in the case of our previous sample of area 21a neurons(Dreher et aL, 1993), preferred velocities of the present samplesof area 21a were not significantly different from those of area 17neurons (Burke et aL, 1992; for review, see Orban, 1984).Furthermore, as in the case of area 17 neurons (for review, seeOrban, 1984), almost all area 21a neurons in the present samplewere either velocity low-pass or velocity-tuned with a very smallproportion of broad-band cells. By contrast, preferred velocitiesof area 21a neurons in the present sample were significantlylower than those in our sample of area 18 neurons (Dreher et aL,1992; for a review, see Orban, 1984).

In the present sample of PMLS neurons, the distribution ofpreferred velocities and the range of velocities to whichindividual neurons respond is very similar to those in severalearlier samples of PMLS neurons (Rauschecker et aL, 1987a; vonGriinau et aL, 1987; Zumbroich and Blakemore, 1987; cf.Turlejski, 1975; Spear and Baumann, 1975; Camarda andRizzolatti, 1976). The preferred velocities of the present sampleof PMLS neurons are significantly higher than those of area 21aneurons and those of area 17 neurons. By contrast, the meanpreferred velocity of our sample of PMLS neurons is substantiallylower than that of area 18 neurons (Dreher et al., 1992).

Both areas 21a and PMLS receive >50% of their associationalinput from laminae 2 and 3 of area 17 (Dreher, 1986; cf.Rosenquist, 1985). Furthermore, both areas receive 15-20% oftheir associational input from supragranular layers of area 18 and-10% of their associational afferents originate in thesupragranular layers of area 19 (Dreher, 1986; cf. Rosenquist,1985). Despite these striking similarities in the associationalinputs, the velocity profiles of the neurons in the two areas arequite different. The velocity profile of area 21a neurons is verysimilar to that of area 17 neurons, while the velocity profile ofPMLS neurons appears to reflect the properties of itsassociational inputs from both area 17 and 18 neurons (cf. BurkeetaL, 1992; Dreher etaL, 1992).

There are, however, significant—albeit numerically small-differences in the composition of afferents to the PMLS area andarea 21a which are likely to underlie the differences in thevelocity preferences of cells in the two areas. Y-type input to thePMLS area presumably underlies the preference of >20% of PMLSneurons in our sample for stimulus velocities >50°/s (cf. DreheretaL, 1980). Indeed, almost all PMLS neurons receive excitatoryY-type input (Berson, 1985; Rauschecker et aL, 1987b). Botharea 21a and the PMLS area receive direct input from thepretecto-recipient zone of the lateral posterior-pulvinarcomplex (cf. Symonds et aL, 1981; Raczkowski and Rosenquist,1983; for reviews, see Rosenquist, 1985; Dreher, 1986). In turn,many pretectal neurons that project to the LP-pulvinar complexreceive a Y-input and respond optimally to rapid motion

(>1000c/s; Sudkamp and Schmidt, 1995; cf. Schoppmann andHoffmann, 1979). These cells presumably project to the PMLSarea but not to area 21a. In addition, some of the Y-type input tothe PMLS area probably originates directly in the retino-recipientparts of the thalamus (especially the medial interlaminarnucleus) and/or is relayed to the PMLS area via area 18 (Berson,1985; Dreher, 1986; Lee etaL, 1986).

By contrast, excitatory Y-type input to area 21a, althoughpresent, appears to be rather minor (Dreher et aL, 1993)- Ourretrograde double-labelling experiments indicate that topo-graphically corresponding parts of area 21a and the PMLS areareceive their thalamic and associational inputs from almostcompletely non-overlapping populations of neurons (Turlejski etaL, 1992; Dreher et aL, 1996). Thus, it is possible that Y-typeneurons, which are quite numerous in the medial interlaminarnucleus (for review, see Garey et aL, 1991), project to the PMLSarea but not to area 21a.

Similarly, it is likely that most area 18 neurons which projectto area 21a prefer low stimulus velocities. By contrast, projectionfrom area 18 to the PMLS area is likely to include a substantialproportion of neurons which respond optimally or exclusively atstimulus velocities >50c/s (broad-band and high-pass cells).Furthermore, input from area 18 to area 21a (but not that to thePMLS area) might actively suppress the responses of area 21aneurons to high-velocity stimuli. Such a possibility is suggestedby the fact that in -40% of cells located in layer 5 of area 17,reversible inactivation of area 18 reveals responses tohigh-velocity stimuli to which the cells were previously hardlyresponsive (Alonso etaL, 1993). Finally, it is important to pointout that many cells in the LP1, which provides the principalthalamic input to area 21a and the PMLS area, are broad-bandcells that respond over a wide range of stimulus velocities, with>75% of cells responding reliably at stimulus velocities >200°/s(Chalupa and Abramson, 1989; cf. Rauschecker et aL, 1987a,b).Again it appears that populations of neurons projecting from theLP1 to area 21a and those projecting to the PMLS area virtually donot overlap (Turlejski et aL, 1992; Dreher et al., 1996). It is likelythat the LP1 neurons preferring fast velocities do not project toarea 21a. Reversible bilateral inactivation of areas 17 and 18reduces the response of some area 21a neurons to slowly movingstimuli (<10°/s), while in others the reduction of the responsesto faster-moving stimuli (15-20°/s) is apparent (Michalski et aL,1993).

Excitatory Responses to Visual Noise Per SeWhile in area 21a only 25% of cells exhibited excitatoryresponses to the motion of visual noise per se, in the PMLS areathe proportion of such cells was significantly larger (46%; Table1). Less than 10% of cells in each area responded tonically to thenoise motion ('field' responses). In both areas, the majority ofcells responding to the motion of the visual noise per seresponded to distinct features of the noise pattern rather thanthe motion of the whole noise ('grain' responses).

The proportions of area 17 and 18 neurons responding to themotion of 2D visual noise were significantly higher than those ofarea 21a and PMLS area neurons (75% of our sample of area 17cells and 83.5% of our sample of area 18 cells). Again, as in thecase of area 21a and the PMLS area, only minorities of those cells(25% in area 17 and 16.5% in area 18) exhibited a field type ofresponse. Furthermore, in our samples, the proportions ofneurons in areas 17 and 18 which exhibited the field type andthe grain type of response were very similar to those in the

594 Areas PMLS and 21a of Cat Visual Cortex • Dreher el al.


Dow

nloaded from


earlier samples of cells recorded from those areas (Gulyas et aL,1987, 1990; Crook, 1990a).

Although it has been reported that some cells in areas 17 and18 of the cat exhibit dissimilar preferred directions for noise andfor elongated contour stimuli (Hammond and MacKay, 1977;Hammond, 1978; Crook, 1990a), several lines of evidencesuggest that the low proportions of cells responding to themotion of 2D visual noise in our samples of area 21a and PMLSneurons do not represent a serious underestimate due to therestriction of our testing procedure to the axis of movementoptimal for the moving slit. Thus, none of four area 21a cells andnone of three PMLS cells which did not respond to the noisemotion per se at the velocities and axis of movement optimal forthe moving slit showed a response when the noise was presentedat various orientations and velocities. Secondly, as mentionedabove, the proportions of area 17 and 18 neurons responding tothe motion of 2D noise per se were much higher than those inarea 21a and the PMLS area and very similar to the proportions ofsuch neurons in the samples of area 17 and area 18 neurons inwhich the testing procedures were not restricted to the motionsof the noise along the axis optimal for the moving slits and/orbars (Gulyas et aL, 1987, 1990; Crook, 1990a). Thirdly, layers 2and 3 of area 17—the principal source of associational input toarea 21a and the PMLS area (cf. Rosenquist, 1985; Dreher,1986)—contain very few cells that are strongly sensitive to themotion of 2D visual noise (Edelstyn and Hammond, 1988;Casanova et aL, 1995). Such cells are located mainly in layer 5(Hammond and MacKay, 1975, 1977; Edelstyn and Hammond,1988), which contributes very little to the associationalprojections to area 21a and the PMLS area (cf. for reviewsRosenquist, 1985; Dreher, 1986).

The low proportion of area 21a and PMLS area cells exhibitinga field type of response to motion of 2D visual noise contrastswith a high proportion (almost 40%) of such cells in the striate-(corticd-) recipient zone of the LP-pulvinar complex (Casanovaand Savard, 1995; cf. Mason, 1981). It is likely, therefore, thatarea 21a and the PMLS area receive their principal thalamic inputfrom subpopulations of cells in the cortico-recipient zone whichrespond poorly to the motion of 2D visual noise.

Noise/Contour InteractionsIt has been demonstrated in the present study that the responsesof a majority (86%) of cells in area 21a to moving contour stimuliare modulated by the background motion of 2D noise. With oneexception, the effect of motion of 2D noise on responses evokedby the moving contour stimuli was inhibitory. The proportion ofarea 21a neurons whose responses to the moving contours areaffected by motion of 2D noise is fairly similar to the proportionsof such cells in the present samples of area 17 (70%) and area 18(89%) neurons (cf. also earlier studies of areas 17 and 18 byHammond and MacKay, 1981; Gulyas et aL, 1987,1990).

In all cortical areas studied here the suppression was equallystrong in cells not responding to noise motion per se and thoseresponding to it (cf. Gulyas et aL, 1987, for area 17). In area 21a,in-phase suppression was significantly stronger than anti-phasesuppression, but in areas 17, 18 and the PMLS area there was nosignificant difference. In the present study a moving noisebackground did not affect significantly the direction selectivityof neurons in areas 21a, PMLS, 17 or 18 (for similar results forneurons in areas 17 and 18, see Orban et aL, 1987; Gulyas et aL,1990).

The suppressive effect of moving background is alreadypresent in the geniculate neurons and the suppression observed

in cortical areas 17, 18 and 21a, based on the simplestassumption of linear additivity, appears to be similar to theeffects observed in their geniculate afferents (Gulyas et aL, 1987,1990; the present study). Similarity of modulatory effects ofmoving noise on neurons in areas 17, 18 and 21a suggests thatneurons in area 21a reflect, to a large extent, the properties oftheir associational afferents from areas 17 and 18. On the otherhand, the responses of some direction-selective cells in thelateral suprasylvian areas (LSSAs), including the PMLS area, tendto be suppressed during in-phase background motion and muchless suppressed or even strongly facilitated during anti-phasebackground motion (von Griinau and Frost, 1983). Indeed, boththe excitatory receptive fields and suppressive surrounds ofsome direction-selective cells in LSSAs areas are directionselective but have opposite preferred directions (von Griinauand Frost, 1983). This may suggest that receptive field propertiesof directionally specific cells in LSSAs (including the PMLS area),unlike those of area 21a, rather than simply reflecting theproperties of their afferents, are largely elaborated by the localcircuitry.

'Division and Sharing of Labor' by the PMLS Area and Area21aThe PMLS area, with its responsiveness over a wide range ofvelocities, high direction selectivity indices and broadorientation selectivity for short contours, appears to be part of amotion-processing stream; area 21a, with its preference for lowvelocities, low direction selectivity indices and sharp orientationselectivity to both long and short contours, appears to be part ofa pattern-processing stream. Not surprisingly, the selectiveablations of the lateral suprasylvian cortex (including the PMLSarea, but not area 21a) do not seem to affect significantlydiscriminations of stationary patterns (Sprague et aL, 1977;Spear et aL, 1983). By contrast, chronic removals of the PMLSarea (as well as the PLLS area) without involvement of area 21aresult in permanent deficits in a cat's ability to discriminatedifferences in speed of moving (especially low-contrast) stimuli(Pasternak et al., 1989).

PMLS neurons appear to be clearly involved in discriminationof object-induced and self-induced motion (Rauschecker et aL,1987a; von Griinau et aL, 1987). Indeed, Kriiger et al. (1993b)reported that the bilateral removal of the lateral suprasylvianareas (including the PMLS area) results in a specific deficit in theanimal's ability to discriminate the patterns moving in anti-phase(or even in-phase as long as there is a relative movement betweenpatterns and background) with the background. Although theeffect of ablations of area 21a on the cat's ability to discriminatepatterns moving relatively to the background is not known, webelieve that the strong involvement of this area in discriminationof object-induced and self-induced motion is unlikely. Inparticular, in area 21a, unlike in the PMLS area, there is a paucityof neurons that are influenced by the 'efference copy signals' ofeye movements (corollary oculomotor signals: Kennedy andMagnin, 1977; Vanni-Mercier and Magnin, 1982; Komatsu et aL,1983; Yin and Greenwood, 1992b).

Despite the fact that the great majority of cells in areas 21a andthe PMLS area could be activated through either eye, and bothareas appear to play a role in depth perception, each area seemsto play a different role in this process. Area 21a appears to beinvolved in processing static binocular positional disparity(Wang and Dreher, 1996), while the PMLS area appears to beinvolved mainly in dynamic or movement-in-depth binocular

Cerebral Cortex July/Aug 1996, V6N 4 595


Dow

nloaded from


perception (cf. Turlejski, 1975; Toyama et al, 1985; Kriiger etaL, 1993a).

The idea of the 'division of labor' between the PMLS area andarea 21a is also supported by distinct patterns of efferentprojections from the two areas. Thus, while the projection ofarea 21a to the ipsilateral superior colliculus (SC) terminatesexclusively in the superficial layers, PMLS projections terminatein both the superficial SC layers and in the dorsal part of theintermediate gray layer (for reviews, see Dreher, 1986; Halting etal., 1992). Indeed, like the deep collicular laminae, thecorticotectal projection from the lateral suprasylvian cortex(including the PMLS area) is clearly implicated in visually guidedbehavior and plays an important role in determining directionand velocity selectivities as well as the binocularity and spatialorganization of receptive field of the collicular neurons locatedin the deep laminae (for review, see Stein and Meredith, 1991).Secondly, the PMLS area, but not area 21a, is stronglyinterconnected with the anteromedial lateral suprasylvian area(AMLS), the ventral lateral suprasylvian area (VLS) and, albeit toa lesser extent, the ectosylvian visual area (EVA), i.e., the areaswhich appear to constitute some of the 'higher order' areas inthe motion processing stream (cf. Tusa et al., 1981; Mucke etal,1982; Rosenquist, 1985; Dreher, 1986; Olson and Graybiel,1987; Spear, 1991). By contrast, area 21a, unlike the PMLS area,is directly interconnected with area 20b, which together witharea 20a has been implicated in the learning of discrimination ofstationary patterns (Sprague etal, \911).

On the other hand, there is clearly a substantial sharing ofsome visual tasks by area 21a and the PMLS area. For example,both areas, unlike areas 17, 18 and 19, appear to make asignificant contribution to the interhemispheric transfer of visuallearning (cf. Marzi et al, 1982; Ptito and Lepore, 1983).Furthermore, both areas project directly to the accessory optictract nuclei (Berson and Graybiel, 1980; Marcotte and Updyke,1982) which play a major role in horizontal optokineticnystagmus (for reviews, see Hoffmann, 1986; Grasse andCynader, 1991). Indeed, both area 21a and the PMLS area mightplay some role in the control of horizontal optokineticnystagmus since in both areas there is a significant increase inglucose utilization when animals are viewing vertically orientedstripes moving horizontally (optokinetic nystagmus drum) ascompared to viewing the stationary nystagmus drum (Herdmanet al, 1989). Finally, both areas appear to contribute toforeground/background interactions. Thus, in both areas themotion of 2D visual noise has a clear effect on the responses tomoving contour stimuli (for area 21a, see the present study; forthe PMLS area, see von Griinau and Frost 1983; von Griinau et al,1987; the present study).

One Area, Two Areas or Spatially Segregated DistinctFunctional Modules?There is very strong evidence (a review of which is beyond thescope of the present paper) indicating that the PMLS area ishomologous with the middle temporal (MT or V 5) 'motion area'area of Old World and New World monkeys (cf. Zeki, 1974;Payne, 1993; cf., however, Gizzi et aL, 1991a). The extensivehodological similarities between area 21a and the PMLS area,combined with the fact that the PMLS area contains mainly therepresentation of the lower visual field and the horizontalmeridian (cf. Hubel and Wiesel, 1969; Turlejski, 1975; Palmer etaL, 1978; Djavadian and Harutiunian-Kozak, 1983; Zumbroich etaL, 1986; Grant and Shipp, 1991; Sherk and Mulligan, 1993)while area 21a contains mainly the representation of the central

part of the horizontal streak (±5° on either side of the horizontalmeridian) plus a small representation of the central part of theupper visual field (cf. Tusa and Palmer, 1980; Dreher et al.,1993), are consistent with the suggestion that area 21a and thePMLS area constitute distinct parts of a single cortical area(Clare-Bishop or LS; Sherk, 1986; Grant and Shipp, 1991). Onecan try to accommodate the existence of clear functionaldifferences between the two regions (reaffirmed, or revealed forthe first time, in the present study) in the concept of a single areaby regarding the PMLS area and area 21a as distinct modulescomparable to hodologically and functionally distinct modulesdescribed in areas VI and V2 of primate visual cortex (see forreviews Livingstone and Hubel, 1988; Zeki and Shipp, 1988,Kaas and Krubitzer, 1991; Casagrande and Kaas, 1994).However, the distinct modules in areas VI and V2 of primatevisual cortex, unlike the PMLS area and area 21a, are inter-mingled and neighbouring modules contain the representationof the same parts of the visual field. It has been postulated thatthe initial modulation of a given cortical field and eventualsegregation of distinct modules into distinct new areas might berelated to slight differences in composition of thalamic and/orcortical afferents to different regions of a given area (Krubitzer,1995). The small differences in composition of thalamic andcortical afferents (for reviews, see Rosenquist, 1985; Dreher,1986; Turlejski et al., 1992; Dreher et al., 1996) might indeedhave played an important role in the spatial segregation of thePMLS area and area 21a.

NotesWe are grateful to Dr B. G. Cleland for making his laboratory available tous. This study was supported by grants from the National Health andMedical Research Council of Australia and the Australian ResearchCouncil.

Address correspondence to Dr B. Dreher, Department of Anatomy andHistology, Institute for Biomedical Research F13, The University ofSydney, NSW 2006, Australia.

ReferencesAlonso JM, Cudeiro J, Perez R, Gonzalez F, Acufia C (1993) Influence of

layer V of area 18 of the cat visual cortex on responses of cells in layerV of area 17 to stimuli of high velocity. Exp Brain Res 93:363-366.

Berson DM (1985) Cat lateral suprasylvian cortex: Y-cell inputs andcorticotectal projection. J Neurophysiol 53:544-555.

Berson DM, Graybiel AM (1980) Some cortical and subcortical fiberprojections to the accessory optic nuclei in the cat. Neuroscience5:2203-2217.

Blakemore C, Zumbroich TJ (1987) Stimulus selectivity and functionalorganization in the lateral suprasylvian visual cortex of the cat. JPhysiol 389:569-603.

Burke W, Dreher B, Michalski A, Cleland BG, Rowe MH (1992) Effects ofselective pressure block of Y-type optic nerve fibres on thereceptive-field properties of cells in the striate cortex of the cat. VisNeurosci 9:47-64.

Camarda R, Rizzolatti G (1976) Visual receptive fields in the lateralsuprasylvian area (Clare-Bishop area) of the cat. Brain Res101:427-443.

Casagrande VA, Kaas JH (1994) The afferent, intrinsic, and efferentconnections of primary visual cortex in primates. In: Cerebral cortex,Vol 10, Primary visual cortex in primates (Peters A, Rock land KS, eds),pp 201-259. New York: Plenum.

Casanova C, Savard T (1995) Responses to moving texture patterns ofcells in the striate-recipient zone of the cat's lateral posterior-pulvinarcomplex. Neurosci (in press).

Casanova C, Freeman RD, Nordmann JP (1989) Monocular and binocularresponse properties of cells in the striate-recipient zone of the cat'slateral posterior-pulvinar complex. J Neurophysiol 62:544-557.

Casanova C, Savard T, Nordmann JP, Molochnikofi S, MinviUe K (1995)Comparison of the responses to moving texture patterns of simpleand complex cells in the cat's area 17. J Neurophysiol 74:1271-1286.

Chalupa LM, Abramson BP (1989) Visual receptive fields in the striate-

596 Areas PMLS and 21a of Cat Visual Cortex • Dreher etal.


Dow

nloaded from


recipient zone of the lateral posterior-puMnar complex. J Neurosci9:347-357.

Clare MH, Bishop GH (1954) Responses from an association areasecondarily activated from optic cortex. J Neurophysiol 17:271-277.

Crook JM (1990a) Directional tuning of cells in area 18 of the feline visualcortex for visual noise, bar and spot stimuli: a comparison with area17. Exp Brain Res 80:545-561.

Crook JM (1990b) Modulatory influences of a moving visual noisebackground on bar-evoked responses of cells in area 18 of the felinevisual cortex. Exp Brain Res 80:562-576.

Danilov Y, Moore RJ, King VR, Spear PD (1995) Are neurons in catposteromedial lateral suprasylvian visual cortex orientation sensitive?Tests with bars and gratings. Vis Neurosci 12:141-151.

Djavadian RL, Harutiunian-Kozak BA (1983) Retinotopic organization ofthe lateral suprasytvian area of the cat. Acta Neurobiol Exp 43:251-262.

Dreher B (1986) Thalamocortical and corticocortical interconnections inthe cat visual system: relation to the mechanisms of informationprocessing. In: Visual neuroscience (Pettigrew JD, Sanderson KJ,Levick WR, eds), pp 290-314. Cambridge, Cambridge UniversityPress.

Dreher B, Leventhal AG, Hale PT (1980) Geniculate input to cat visualcortex: a comparison of area 19 with areas 17 and 18. J Neurophysiol44:804-826.

Dreher B, Michalski A, Cleland BG, Burke W (1992) Effects of selectivepressure block of Y-type optic nerve fibres on the receptive-fieldproperties of cells in area 18 of the cat. Vis Neurosci 9:65-78.

Dreher B, Michalski A, Ho RHT, Lee CWF, Burke W (1993) Processing ofform and motion in area 21a of cat visual cortex. Vis Neurosci10:93-115.

Dreher B, Djavadian RL, Turlejski KJ, Wang C (1996) Areas PMLS and 2 laof cat visual cortex are not only functionally but also hodologicallydistinct. Prog Brain Res (in press).

Edelstyn NMJ, Hammond P (1988) Relationship between corticallamination and texture sensitivity in complex neurones of the striatecortex in cats. J Comp Neurol 278:397-404.

Felleman DJ, Van Essen DC (1991) Distributed hierarchical processing inthe primate cerebral cortex. Cereb Cortex 1 :l-47.

Garey LJ, Dreher B, Robinson SR (1991) The organization of the visualthalamus. In: Neuroanatomy of the visual pathways and theirdevelopment. (Dreher B, Robinson SR, eds), Vision and visualdysfunction, Vol 3 (Cronly-Dilllon J, ed), pp 176-234. Basingstoke:Macmillan.

Gizzi MS, Katz E, Schumer RA, Movshon JA (1990a) Selectivity fororientation and direction of motion of single neurons in cat striate andextrastriate visual cortex. J Neurophysiol 63:1529-1543-

Gizzi MS, Katz E, Movshon JA (1990b) Spatial and temporal analysis byneurons in the representation of the central visual field in the cat'slateral suprasylvian visual cortex. Vis Neurosci 5:463-468.

Grant S, Shipp S (1991) Visuotopic organization of the lateral suprasylvianarea and of an adjacent area of the ectosylvian gyrus of cat cortex: aphysiological and connectional study. Vis Neurosci 6:315-338.

Grasse KL, Cynader MS (1991) The accessory optic system in frontal-eyedanimals. In: The neural basis of visual function (Leventhal AG, ed),Vision and visual dysfunction, Vol 4 (Cronly-Dilllon J, ed), pp111-139. Basingstoke: Macmillan.

Guedes R, Watanabe S, Creutzfeldt OD (1983) Functional role ofassociation fibres for a visual association area: the posteriorsuprasylvian sulcus of the cat. Exp Brain Res 49:13-27.

Gulyas B, Orban GA, Duysens J, Maes H (1987) The suppressive influenceof moving textured backgrounds on responses of cat striate neuronsto moving bars. J Neurophysiol 57:1767-1791.

Gulyas B, Spileers W, Orban GA (1990) Modulation by a moving texture ofcat area 18 neuron responses to moving bars. J Neurophysiol63:404-423.

Hamada T (1987) Neural response to the motion of textures in the lateralsuprasylvian area of cats. Behav Brain Res 25:175-185.

Hammond P (1978) Directional tuning of complex cells in area 17 of thefeline visual cortex. J Physiol 285: 479-491.

Hammond P, MacKay DM (1975) Differential responses of cat visualcortical cells to textured stimuli. Exp Brain Res 22:427-430.

Hammond P, MacKay DM (1977) Differential responsiveness of simpleand complex cells in cat striate cortex to visual texture. Exp Brain Res30:275-296.

Hammond P, MacKay DM (1981) Modulatory influences of moving

textured backgrounds on responsiveness of simple cells in felinestriate cortex. J Physiol 319:431- 442.

Hailing JK, Updyke BV, Van Lieshout DP (1992) Corticotectal projectionsin the cat: anterograde transport studies of twenty-five cortical areas.J Comp Neurol 324:379-414.

Harutiunian-Kozak BA, Djavadian RL, Melkumian AV (1984) Responses ofneurons in cat's lateral suprasylvian area to moving light and darkstimuli. Vis Res 24:189-195.

Heath CJ, Jones EG (1971) The anatomical organization of thesuprasylvian gyrus of the cat. Erg Anat 45:1-61.

Henry GH (1977) Receptive field classes of cells in the striate cortex ofthe cat. Brain Res 133:1-28.

Henry GH, Bishop PO, Dreher B (1974a) Orientation, axis and directionas stimulus parameters for striate cells. Vis Res 14:767-777.

Henry GH, Dreher B, Bishop PO (1974b) Orientation specificity of cellsin cat striate cortex. J Neurophysiol 37:1394-1409.

Henry GH, Michalski A, Wimborne BM, McCart RJ (1994) The nature andorigin of orientation specificity in neurons of the visual pathways.Prog Neurobiol 43:381-437.

Herdman SJ, Tusa RJ, Smith CB (1989) Cortical areas involved inhorizontal OKN in cats: metabolic activity. J Neurosci 9:1150-1162.

Hoffmann K-P (1986) Visual inputs relevant for the optokineticnystagmus in mammals. In: Progress in brain research, Vol 64 (FreundHJ, Biittner U, Cohen B, Noth J, eds) pp 75-83. Amsterdam: FJsevier.

Hubel DH, Wiesel TN (1962) Receptive fields, binocular interaction, andfunctional architecture in the cat's visual cortex. J Physiol160:106-154.

Hubel DH, Wiesel TN (1965) Receptive fields and functional architecturein two non-striate visual areas. J Neurophysiol 28:229-286.

Hubel DH, Wiesel TN (1969) Visual area of the lateral suprasylvian gyrus(Clare-Bishop area) of the cat. J Physiol 202:251-260.

Kaas JH (1980) A comparative survey of visual cortex organization inmammals. In: Comparative neurology of the telencephalon (EbbessonSOE, ed), pp 483-503. New York: Plenum.

Kaas JH, Krubitzer LA (1991) The organization of extrastriate visualcortex. In: Neuroanatomy of the visual pathways and theirdevelopment (Dreher B, Robinson SR, eds), Vision and visualdysfunction, Vol 3 (Cronly-Dilllon J, ed), pp 302-323. Basingstoke:Macmillan.

Kennedy H, Magnin M (1977) Saccadic influences on single neuronactivity in the medial bank of the cat's suprasylvian sulcus(Clare-Bishop area) Exp Brain Res 27:315-317.

Komatsu Y, Shibuki K, Toyarna K (1983) Eye movement-related activitiesin cells of the lateral suprasylvian cortex of the cat. Neurosci Lett41:271-276.

Krubitzer L (1995) The organization of neocortex in mammals: arespecies differences really so different. Trends Neurosci 18:408-417.

Krubitzer LA, Kaas JH (1993) The dorsomedial visual area of owlmonkeys: Connections, myeloarchitecture, and homologies in otherprimates. J Comp Neurol 334:497-528.

Kruger K, Kiefer W, Groh A (1993a) Lesion of the suprasylvian corteximpairs depth perception of cats. NeuroReport 4:883-886.

Kruger K, Kiefer W, Groh A, Dinse HR, von Seelen W (1993b) The role ofthe lateral suprasylvian visual cortex of the cat in object-backgroundinteractions: permanent deficits following lesions. Exp Brain Res97:40- 60.

Lee C, Malpeli JG, Weyand TG (1986) Simple cells of cat area 18 mediateinput from layer A of the lateral geniculate nucleus to theClare-Bishop area. Neurosci Abstr 12:584.

Livingstone M, Hubel D (1988) Segregation of form, color, movement,and depth: anatomy, physiology, and perception. Science240:740-749.

Marcotte RR, Updyke BV (1982) Cortical visual areas of the cat projectdifferentially onto the nuclei of the accessory optic system. Brain Res242:205-217.

Marshall WH, Talbot SA, Ades HW (1943) Cortical response of theanesthetized cat to gross photic and electrical afferent stimulation. JNeurophysiol 6:1-5.

Marzi CA, Antonini A, Di Stefano M, Legg CR (1982) The contribution ofthe corpus callosum to receptive fields in the lateral suprasylvian areasof the cat. Behav Brain Res 4:155-176.

Mason R (1976) Responses of cells in the dorsal lateral geniculatecomplex of the cat to textured visual stimuli. Exp Brain Res25:323-326.

Cerebral CortexJuly/Aug 1996, V6N 4 597


Dow

nloaded from


Mason R (1979) Responsiveness of cells in the cat's superior coUiculus totextured visual stimuli. Exp Brain Res 37:231-240.

Mason R (1981) Differential responsiveness of cells in the visual zones ofthe cat's LP-pulvinar complex to visual stimuli. Exp Brain Res43:25-33.

Michalski A, Wimborne BM, Henry GH (1993) The effect of reversiblecooling of cat's primary visual cortex onthe responses of area 21aneurons. J Physiol 466:133-156.

Michalski A, Wimborne BM, Henry GH (1994) The role of ipsilateral andcontralateral inputs from primary cortex in responses of area 21aneurons in cats. Vis Neurosci 11:839-849.

Mizobe K, Itoi M, Kaihara T, Toyama K (1988) Neuronal responsiveness inarea 21a of the cat. Brain Res 438:307-310.

Montero VM (1981) Topography of the cortico-cortical connections fromthe striate cortex of the cat. Brain Behav Evol 18:194-218.

Morrone MC, Di Stefano M, Burr DC (1986) Spatial and temporalproperties of neurons of the lateral suprasylvian cortex of the cat.JNeurophysiol 56:969-986.

Mucke L, Norita M, Benedek C, Creutzfeldt O (1982) Physiologic andanatomic investigation of a visual cortical area situated in the ventralbank of the anterior ectosylvian sulcus of the cat. Exp Brain Res46:1-11.

Noda H, Freeman RB Jr, Gies B, Creutzfeldt OD (1971) Neuronalresponses in the visual cortex of awake cats to stationary and movingtargets. Exp Brain Res 12:389-405.

Olson CR, Graybiel AM (1987) Ectosylvian visual area of the cat: location,retinotopic organization, and connections. J Comp Neurol261:277-294.

Orban GA (1984) Neuronal operations in the visual cortex. Berlin:Springer-Verlag.

Orban GA, Gulyas B, Vogels R (1987) Influence of a moving texturebackground on direction selectivity of cat striate neurons. JNeurophysiol 57:1792-1812.

Palmer LA, Rosenquist AC, Tusa RJ (1978) The retinotopic organization oflateral suprasylvian visual areas in the cat. J Comp Neurol177:237-256.

Pasternak T, Horn KM, Maunsell JHR (1989) Deficits in speeddiscrimination following lesions of the lateral suprasylvian cortex inthe cat. Vis Neurosci 3:365-375.

Payne BR (1993) Evidence for visual cortical area homologs in cat andmonkey. Cereb Cortex 3:1-25.

Ptito M, Lepore F (1983) Effects of unilateral and bilateral lesions of thelateral suprasylvian area on learning and interhemispheric transfer ofpattern discimination in the cat. Behav Brain Res 7:211-227.

Raczkowski D, Rosenquist AC (1983) Connections of the multiple visualcortical areas with the lateral posterior-pulvinar complex and adjacentthalamic nuclei in the cat. J Neurosci 3:1912-1942.

Rauschecker JP, von Grunau MW, Poulin C (1987a) Centrifugalorganization of direction preferences in the cat's lateral suprasylvianvisual cortex and its relation to flow field processing. J Neurosci7:943-958.

Rauschecker JP, von Grunau MW, Poulin C (1987b) Thalamo-corticalconnections and their correlation with receptive field properties inthe cat's lateral suprasylvian visual cortex. Exp Brain Res 67:100-112.

Rosenquist AC (1985) Connections of visual cortical areas in the cat. In:Cerebral cortex, Vol 3, Visual cortex (Peters A, Jones EG, eds), pp81-116. New York: Plenum.

Sanides F, Hoffmann, J (1969) Cyto- and myeloarchitecture of the visualcortex of the cat and of the surrounding integration cortices. JHirnforsch 11:79-104.

Schoppmann A, Hoffmann K-P (1979) A comparison of visual responsesin two pretectal nuclei and in the superior coUiculus of the cat. ExpBrain Res 35:495-510.

Sereno MI, Allman JM (1991) Cortical visual areas in mammals. In: Theneural basis of visual function. (Leventhal AG, ed), Vision and visualdysfunction, Vol 4 (Cronly-Dilllon J, ed), pp 160-172. Basingstoke:Macmillan.

Sherk H (1986) Location and connections of visual cortical areas in thecat's suprasylvian sulcus. J Comp Neurol 247:1-31.

Sherk H (1989) Visual response properties of cortical inputs to anextrastriate cortical area in the cat. Vis Neurosci 3:249-265.

Sherk H, Mulligan K A (1993) A reassessment of the lower visual field mapin the striate-recipient lateral suprasylvian cortex. Vis Neurosci10:131-158.

Shipp S, Grant S (1991) Organization of reciprocal connections between

area 17 and the lateral suprasylvian area of cat visual cortex. VisNeurosci 6:339-355.

Siegel S (1956) Nonparametric statistics for the behavioral sciences. NewYork: McGraw-Hill.

Skottun BC, Grosof DH, De Valois RL (1988) Responses of simple andcomplex cells to random dot patterns: a quantitative comparison. JNeurophysiol 59:1719-1735.

Spear PD (1991) Functions of extrastriate visual cortex in non-primatespecies. In: The neural basis of visual function (Leventhal AG, ed),Vision and visual dysfunction, Vol 4 (Cronly-Dilllon J, ed), pp339-369. Basingstoke: Macmillan.

Spear PD, Baumann TP (1975) Receptive-field characteristics of singleneurons in lateral suprasylvian visual area of the cat. J Neurophysiol38:1403-1420.

Spear PD, Baumann TP (1979) Effects of visual cortex removal onreceptive-field properties of neurons in lateral suprasylvian visual areaof the cat. J Neurophysiol 42:31-56.

Spear PD, Miller S, Ohman L (1983) Effects of lateral suprasylvian visualcortex lesions on visual localization, discimination, and attention incats. Behav Brain Res 10:339-359.

Sprague JM, Levy J, DiBerardino A, Berlucchi G (1977) Visual corticalareas mediating form discrimination in the cat. J Comp Neurol172:441-488.

Stein BE, Meredith MA (1991) Functional organization of the superiorcoUiculus. In: The neural basis of visual function (Leventhal AG, ed),Vision and visual dysfunction, Vol 4 (Cronly-DUUon J, ed), pp 85-110.Basingstoke: Macmillan.

Sudkamp S, Schmidt M (1995) Physiological characterization of pretectalneurons projecting to the lateral posterior-pulvinar complex in thecat. EurJ Neurosci 7:881-888.

Symonds LL, Rosenquist AC, Edwards SB, Palmer LA (1981) Projections ofthe pulvinar-lateral posterior complex to visual cortical areas in thecat. Neuroscience 6:1995-2020.

Toyama K, Komasu Y, Kasai H, Fujii K, Umetani K (1985) Responsivenessof Clare-Bishop neurons to visual cues associated with motion of avisual stimulus in three-dimensional space. Vis. Res 25:407-414.

Toyama K, Mizobe K, Akase E, Kaihara T (1994) Neuronal responsivenessin areas 19 and 21a, and the posteromedial lateral suprasylvian cortexof the cat. Exp Brain Res 99:289-301.

Turlejski K (1975) Visual responses of neurons in the Clare-Bishop areaof the cat. Acta Neurobiol Exp 35:189-208.

Turlejski K, Michalski A (1975) Clare-Bishop area in the cat: location andretinotopical projection. Acta Neurobiol Exp 35:179-188.

Turlejski K, Djavadian RL, Dreher B (1992) Diencephalic and tel-encephalic inputs to areas 21a and PMLS of cat visual cortex: a doublelabel study. Proc Austr Physiol Pharmacol Soc 23:133P.

Tusa RJ, Palmer LA (1980) Retinotopic organization of areas 20 and 21 inthe cat. J Comp Neurol 193:147-164.

Tusa RJ, Palmer LA, Rosenquist AC (1981) Multiple cortical visual areas:Visual field topography in the cat. In: Cortical sensory organization,Vol 2, Multiple visual areas (Woolsey CN, ed), pp. 1-31. Clifton, NJ:Humana.

Updyke BV (1986) Retinotopic organization within the cat's posteriorsuprasylvian sulcus and gyrus. J Comp Neurol 246:265-280.

Vanni-Mercier C, Magnin M (1982) Retinotopic organization ofextra-retinal saccade-related input to the visual cortex in the cat. ExpBrain Res 46:368-376.

von Grunau M, Frost BJ (1983) Double-opponent-process mechanismunderlying RF-structure of directionaUy specific cells of cat lateralsuprasylvian visual area. Exp Brain Res 49:84-92.

von Grunau MW, Zumbroich TJ, Poulin C (1987) Visual receptive fieldproperties in the posterior suprasylvian cortex of the cat: acomparison between the areas PMLS and PLLS. Vis Res 27:343-356.

Wang C, Dreher B (1996) Binocular interactions and disparity coding inarea 21a of cat extrastriate visual cortex. Exp Brain Res 108:257-272.

Wang C, Dreher B, Qeland B (1992a) The modulatory effects of movingvisual noise background on bar-evoked responses of neurones in area21a of cat visual cortex. Proc Austr Neurosci Soc 3:90P.

Wang C, Dreher B, Turlejski K, Djavadian RL, Burke W (1992b) Receptivefield properties of neurones in areas 21a and PMLS of cat visual cortex:two functional areas. Proc Austr Physiol Pharmacol Soc 23:132P.

Wimborne BM, Henry GH (1992) Response characteristics of the cells ofcortical area 21a of the cat with special reference to orientationspecificity. J Physiol 449:457-478.

598 Areas PMLS and 21a of Cat Visual Cortex • Dreher et al.


Dow

nloaded from


Wright MJ (1969) Visual receptive fields of cells in a cortical area remotefrom the striate cortex in the cat. Nature 223:973-975.

Yin TC, Greenwood M (1992a) Visual response properties of neurons inthe middle and lateral suprasylvian cortices of the behaving cat. ExpBrain Res 88:1-14.

Yin TC, Greenwood M (1992b) Visuomotor interactions in responses ofneurons in the middle and lateral suprasylvian cortices of thebehaving cat. Exp Brain Res 88.15-32.

Zeki SM (1974) Functional organization of a visual area in the posterior

bank of the superior temporal sulcus of the rhesus monkey. J Physiol236:549-573.

Zeki SM, Shipp S (1988) The functional logic of cortical connections.Nature 335:311-317.

Zumbroich TJ, Blakemore C (1987) Spatial and temporal selectivity in thesuprasylvian visual cortex of the cat. J Neurosci 7:482-500.

Zumbroich TJ, von Griinau M, Poulin C, Blakemore C (1986) Differencesof visual field representation in the medial and lateral banks of thesuprasylvian cortex (PMLS/PLLS) of the cat. Exp Brain Res 64:77-93-

Cerebral Cortex July/Aug 1996, V 6 N 4 599


Dow

nloaded from


areas pmls and 21 a of cat visual cortex: two functionally distinct areas

Documents