Cytochrome Oxidase,Acetylcholinesterase, and NADPH ... · studied in the supratemporal and insular cortex of 5 human brains (cases 1–5). Additionally, the acetylcho-linesterase

Cytochrome Oxidase, Acetylcholinesterase, and NADPH-DiaphoraseStaining in Human Supratemporal and Insular Cortex: Evidence

for Multiple Auditory Areas

Francois Rivier* and Stephanie Clarke*,†*Institut de Physiologie and †Division de Neuropsychologie, Universite de Lausanne, Lausanne Switzerland

Received May 19, 1997

The pattern of cytochrome oxidase, acetylcholinester-ase, and NADPH-diaphorase activity was studied inthe supratemporal plane, the posterior part of thesuperior temporal gyrus, and the insula of normalhuman brains. Five dark cytochrome oxidase regionswere found: (i) on Heschl’s gyrus (area TC of vonEconomo and Koskinas); (ii) on the planum polare(area TC/TG); (iii) posterior to Heschl’s gyrus (withinarea TA); (iv) on the posterior convexity of the superiortemporal gyrus (within area TA); and (v) on the postero-superior insula (area IB). More lightly stained cortexseparated these regions (areas IA, TD, and part of TB).The laminar distribution of cytochrome oxidase activ-ity varied in different areas. Acetylcholinesterase-positive fibers predominated in area TC and pyramidalneurons in areas TA and IA and in parts of TB; amixture of fiber and neuronal staining was found inTC/TG, TD, and IB. NADPH-diaphorase positive pro-files included large darkly stained nonpyramidal neu-rons, mostly in infragranular layers and in subcorticalwhite matter, small faintly stained cells, and a densearray of fibers. The NADPH-diaphorase staining pat-tern did not vary between areas. The present resultssuggest that the supratemporal plane, the posteriorpart of the superior temporal gyrus, and the insulacontain at least eight putative cortical areas. Compari-son with activation studies by others suggest that,apart from the primary auditory area, six other puta-tive areas may be auditory whereas one putative area,on posterior insula, may be vestibular. r 1997 Academic

Press

Key Words: temporal cortex; insula; auditory associa-tion cortex; Wernicke; human

INTRODUCTION

Human cerebral cortex found to respond to auditorystimuli comprises the supratemporal plane, the poste-rior two-thirds of the superior temporal gyrus, andparts of the supramarginal and angular gyri and of the

fronto-parietal operculum (Celesia, 1976). The firststages of auditory cortical processing occur in thesupratemporal plane and in adjacent parts of thesuperior temporal gyrus; often these regions are re-ferred to as the auditory cortex, whereas the supramar-ginal and angular gyri and the parietal operculum onthe left side are believed to be involved in the morespecific analysis of language.

The human auditory cortex is currently subdividedinto several architectonically defined areas. The num-ber of these areas as well as their extent and exactposition vary between authors. Using cytoarchitectoniccriteria, Brodmann (1909) distinguished three areas,called 41, 42, and 22; and von Economo and Koskinas(1925) four, TC, TD, TB, and TA. Using the pattern oflipofuscin staining, Braak (1978) defined four areas(areae temporalis granularis, paragranularis, magnopy-ramidalis centralis, magnopyramidalis medialis). Com-bining cyto- and myeloarchitecture, Galaburda andSanides (1980) identified in the same region eight areas(KAm, KAlt, PaAi, PaAc/d, PaAe, PaAr ProA, Tpt).Rademacher et al. (1996) identified six anteroposteriorbelts of three areas each on the supratemporal planeand the superior temporal gyrus and within the supe-rior temporal sulcus. Brodmann’s area 41 (1909), whichcorresponds to TC of von Economo and Koskinas (1925)and to KAm plus KAlt of Galaburda and Sanides(1980), is commonly accepted to be the primary audi-tory area (A1) and can be distinguished from thesurrounding cortex by its cyto- (see above) and myeloar-chitecture (Vogt and Vogt, 1919). Peroperatory record-ings demonstrated that evoked potentials to clicks hadhigher amplitudes and shorter latencies in A1 than inthe surrounding cortex (Celesia, 1976; Liegeois-Chau-vel et al., 1991). Retrograde degeneration studies car-ried out in cases with circumscribed cortical lesionsindicated that human A1 receives input from theparvocellular subdivision of the medial geniculatenucleus (Van Buren and Borke, 1972).

The cortex around the primary auditory area con-tains probably several other auditory areas as indi-

NEUROIMAGE 6, 288–304 (1997)ARTICLE NO. NI970304

2881053-8119/97 $25.00Copyright r 1997 by Academic PressAll rights of reproduction in any form reserved.

cated by changes in cortical architecture (Galaburdaand Sanides, 1980). These changes are, however, notclear enough to serve for unequivocal identification ofareas and different authors subdivide the supratempo-ral plane in different ways. Additional evidence formultiple auditory areas comes from electrophysiologi-cal studies. Peroperative recordings of evoked poten-tials in the auditory cortex show differences in responsecharacteristics in A1 and in the surrounding cortex(Celesia, 1976), and the existence of excitatory connec-tions from the primary auditory area to the surround-ing cortex (Liegeois-Chauvel et al., 1991).

We have analyzed the cortical architecture of thehuman supratemporal plane and insula in cytochromeoxidase, acetylcholinesterase, and NADPH-diaphorasestaining and found evidence for the presence of eightdistinct cortical areas. Preliminary results were pub-lished in abstracts (Rivier and Clarke, 1996; Clarkeand Rivier, 1997).

MATERIAL AND METHODS

The pattern of cytochrome oxidase staining wasstudied in the supratemporal and insular cortex of 5human brains (cases 1–5). Additionally, the acetylcho-linesterase and NADPH-diaphorase staining patternswere analyzed in two of them (cases 1 and 2). Allsubjects died from acute cardiac problems (Table 1) andhad no known neurological diseases or hearing com-plaints. Their ages were between 70 and 90 years.

Time between death and fixation varied between 8and 12 hours (Table 1). Each brain was taken from theskull and perfused simultaneously through the left andright internal carotid and the basilary arteries. First,the blood was washed out as much as possible withheparin in 0.9% NaCl solution. Then followed perfusionwith 4% paraformaldehyde in 0.1 M phosphate buffer(pH 7.4) for 45 to 60 min. During the whole perfusionprocedure, the brain was held in a skull-shaped con-

tainer, made of perforated plastic, to avoid distortionsof the unfixed tissue. After perfusion the brains werephotographed in standard views and each hemispherein medial view. Brains of cases 1, 2, and 5 were cut into20-mm-thick slices; the slices were photographed in acoordinate frame, so as to record anteroposteriorlycorresponding points. The slices were postfixed for 15 to36 h in 4% paraformaldehyde in 0.1 M phosphate buffer(pH 7.4) and kept in phosphate buffer (pH 7.4) withincreasing concentrations of saccharose (10–30%) for4–13 days. The whole temporal, insular, and occipitallobes of both hemispheres were cut frozen in sets ofcoronal sections. Each set comprised five 80-µm-thicksections (three used for cytochrome oxidase and 2 forNADPH-diaphorase staining) and fifteen 40-µm-thicksections (used for Nissl and acetylcholinesterase stain-ing). During cutting, photographs of the frozen blockwere taken between each set (i.e., every 1 mm); thisallowed us to correct later for distortions due to histo-logical procedures. Brains of cases 3 and 4 were used forflattening of the superior part of the temporal lobe.Immediately after the perfusion, the superior temporalgyrus, Heschl’s gyrus, the plana temporale and polare,and the insula were dissected from both hemispheres.Cuts were introduced into the white matter under thegyri and the cortex was gently spread and flattenedbetween two parallel glass slides. The flattened pieceswere postfixed for 15 h and kept in phosphate bufferwith increasing concentrations of saccharose (10–30%)for 4 to 14 days. Frozen 80-µm-thick sections were cuttangentially to the flattened cortical surface; photo-graphs of the frozen block were taken between eachsection. All sections were stained for cytochrome oxi-dase.

For cytochrome oxidase histochemistry, 80-µm-thickcoronal or tangential serial sections were processedfloating (modified from Seligman et al., 1968). In cases1 and 2, additional sections, adjacent to the floatingones, were processed mounted on glass slides. For bothprocedures, sections were incubated in a solution of0.03% cytochrome C (type III, Sigma), 0.05% diamino-benzidine, and 4% saccharose in 0.1 M phosphatebuffer (pH 7.4) at 37°C for 16 to 24 h, rinsed in the samephosphate buffer, mounted, air-dried, dehydrated, andcoverslipped. Selectivity of the cytochrome oxidasereaction was assessed as described in a previous report(Clarke, 1994). For acetylcholinesterase histochemistry(Emre et al., 1993), 40-µm-thick sections were mountedon glass slides and air-dried. They were rinsed sixtimes in 0.1 M acetate buffer (pH 5.5), incubated for 5 to6 h in a solution of 0.0072% ethopromazine, 0.075%glycine, 0.05% copper sulfate, 0.24% acethylthiocholineiodide, and 0.68% sodium acetate (pH 5.5; on a rockingtable, at room temperature), and rinsed six times in 0.1M acetate buffer (pH 5.5). The sections were developedin a solution of 3.8% sodium sulfide in 0.1 N HCl (pH

TABLE 1

Summary of Cases

Sex

Age atdeath(years)

Cause ofdeath

Delaydeath-fixation(hours)

Plane ofsection

Case 1 F 90 Cardiac arrest 12 CoronalCase 2 M 80 Heart failure 11 CoronalCase 3 M 71 Cardiac arrest 12 Tangential to flat-

tened cortexCase 4 M 70 Myocardial

infarction8 Tangential to flat-

tened cortexCase 5 F 71 Heart failure 10 Coronal

Note. All patients died from causes unrelated to the brain and hadno brain lesions. In all cases both hemispheres were analyzed.

289AUDITORY AREAS IN MAN

FIG. 1. Photomicrographs showing cytoarchitectonic criteria used to identify areas TC (A), TB (B), TA (C), TD (D), TG (E), IB (F), and IA(G). Within TC, a more granular part can be identified (5TC1; B). Calibration bar, inserted in the bottom right corner of G, is 200 µm.

290 RIVIER AND CLARKE

7.8), rinsed six times in 0.1 M acetate buffer (pH 5.5),and intensified in a 1% of silver nitrate solution. Theywere then rinsed in H2O, dehydrated in graded alco-hols, cleared in xylol, and coverslipped. For NADPH-diaphorase histochemistry, 80-µm-thick coronal sec-tions were processed floating (modified from DeFelipe,1993). They were stored overnight in 50 mM Tris–HClbuffer (pH 7.4) and then incubated for 5 h at 37°C in asolution of 1 mM b-NADPH, 0.5 mM nitrobluetetra-zolium, and 0.1% Triton X-100 in 50 mM Tris–HClbuffer (pH 8.0). They were then rinsed first in 50 mMTris–HCl buffer (pH 7.4) followed by 0.1 M phosphatebuffer (pH 7.4). They were mounted, dehydrated, andcoverslipped. Specificity of the reaction was tested byomitting either b-NADPH or nitrobluetetrazolium fromthe incubating medium; in either case no staining wasobtained. Adjacent 40-µm-thick sections were stainedwith cresyl violet.

Analysis of Staining Patterns

The pattern of cytochrome oxidase staining wasstudied in two ways. (i) Images of sections were cap-tured with either a XRS Omnimedia scanner linked to aSun Classic or with a Canon monochrome video TVcamera (M420) mounted on a Wild Makroscop (Ci-20PM). The files were then imported on a PowerMacintosh 7100 and analyzed by digital image process-ing (NIH Image 1.59ppc). The patterns of staining werestudied in pseudocolor images and in densitometricprofiles either in tangential or radial strips of cortex.The densitometry was calibrated within the individualsection (see Fig. 7). The density was indicated inarbitrary units along the y-axis; values left to layer Icorresponded to the optical density of the glass slidesand mounting medium, the peak to the maximumwithin the region measured.

(ii) Sections were observed under bright-field illumi-nation or using differential interference contrast micros-copy (according to Nomarski), and photographs weretaken at different locations. Cortical architecture wasstudied in adjacent Nissl-stained sections. The distribu-

tion of NADPH-diaphorase-positive neurons wascharted in selected regions of the cortex using either acamera lucida or low magnification photomicrographs.The distribution of acetylcholinesterase-positive neu-rons and nerve fibers was studied in selected regions ofthe cortex and documented by photomicrographs.

Talairach Coordinates

Individual human brains vary considerably in sizeand shape. This represents a serious problem forcomparison of data collected in different individuals.Talairach and Tournoux (1988) introduced a propor-tional grid and a coordinate system anchored in theforebrain commissures. Three standard planes aredefined: two coronal planes through the anterior andposterior commissures and a horizontal plane throughboth. The hemispheres are subdivided into ‘‘cuboids’’(orthogonal parallelepipeds) that result from the inter-sections of 12 horizontal slices (4 below and 8 above thebicommissural plane), 11 coronal slices (4 anterior tothe anterior commissure, 3 between the anterior andposterior commissures, 4 posterior to the posteriorcommissure), and 8 sagittal slices (4 for each hemi-sphere). We have determined the position of the corre-sponding planes and cuboids in our experimental brains.From the position of a point within a given cuboid in theexperimental brain we calculated the equivalent pointin the corresponding cuboid in the reference brain bylinear mapping.

For each putative cortical area described below theTalairach and Tournoux (1988) coordinates of the cen-ter of this area were determined (Table 5). In a firststep, each area represented in the flat reconstructionsof cases 1 and 2 (Figs. 3 and 4) was approximated by apolygon and its center of gravity was determined. Thecenter of gravity was then reported on the correspond-ing histological section and hence on the photograph ofthis section during cutting. The x, y, and z coordinateswere determined by applying the proportional stereo-taxic grid and linear mapping between correspondingcuboids.

TABLE 2

Cytoarchitectonic Characteristics That Were Used for the Identifications of Areas on the Supratemporal Planeand the Insula (Presence of Large or Middle-Sized Pyramids; Particular Cell Arrangements)

TA TB TC TD TG IA IB

Generalarrangement

Cell columns inIII–IV

Radial striae(rain-shower)II–VI

Irregular radialand laminararrangement

Cell poor; noradial stria-tion

Cell columns inIII–IV

Layer III Middle-sizedpyramids

Large pyramidsin IIIc

Middle-sizedpyramids

Middle-sizedpyramids

Layer IV Thick ThinLayer V Middle-sized

pyramidsLight Middle-sized

pyramids

Note. All these criteria remain reliable in partially obliquely cut cortex. For illustrations, see Figure 1.


RESULTS

Cytoarchitectonic areas, as described by von Economoand Koskinas (1925) and von Economo and Horn (1930)were delineated on the supratemporal plane and on theinsula of coronally cut brains (cases 1, 2, 5). Thedetailed descriptions by von Economo and his collabora-tors were based on sections that were essentiallyperpendicular to the pial surface. When working withserially cut material, relevant parts of cortex may becut obliquely, and certain cytoarchitectonic criteria, such asthickness of layers, can no longer be used. Other criteria,such as presence of large neurons in certain layers orspecial cellular arrangements, remain reliable and wereused by us (Fig. 1; Table 2). The location of distinctarchitectonic areas was as described by von Economo andHorn (1930). Area TC was found on Heschl’s gyrus in allhemispheres (see also Rademacher et al., 1993), sur-rounded posterolaterally by area TB, anteromedially byarea TG, and posteromedially by area TD. Area TA sur-rounded the posterolateral part of area TB and the antero-lateral part of area TC. Area IB was found in theposterior part of the insula, area IA in the anterior part.

Cytochrome Oxidase Pattern

Cytochrome oxidase was revealed in individual corti-cal neurons and in the neuropil (Fig. 2B). The intensityof staining varied between layers. A dense band wasfound in midcortex, corresponding to layer IV in someregions and to layers III and IV in others; its upper andlower limits were gradual (Fig. 3). Layers I, II, V, andVI were always lighter than the midcortical band.

The overall intensity of the staining was darker insome regions. The boundaries between dark and lightregions were not sharp, the changes occurring gradu-ally over 0.5 to 2 mm. Five darkly stained regions wereprominent in all hemispheres. The first dark region,relatively large, was found on Heschl’s gyrus; it wascoextensive with area TC and corresponded to theprimary auditory cortex (A1). The second dark region,rather small, was on the planum polare anterior toHeschl’s gyrus within area TC/TG. The third darkregion, relatively small, was posterior to Heschl’s gyruswithin the posterior part of area TA. The fourth darkregion was on posterior convexity of superior temporalgyrus within the lateral part of area TA. The fifth darkregion, elongated in shape, was on posterior insula,more or less coextensive with area IB. Three lightlystained regions were found between or next to the darkregions. The first light region was large and lay lateralto Heschl’s gyrus comprising large parts of area TB.

FIG. 2. Photomicrographs showing NADPH-diaphorase (A), cyto-chrome oxidase (B), and acetylcholinesterase (C) staining in humantemporal cortex (area TB from the left (A) and right (B, C) hemi-spheres of case 1). Calibration bar, 30 µm.


The second light region was medial to Heschl’s gyruswithin areas TD. The third light region was on theanteroinferior part of the insula, corresponding to areaIA. These dark and light cytochrome oxidase regionswere present in all hemispheres analyzed and in simi-lar positions. Figures 4 and 5 show the pattern ofcytochrome oxidase staining within the superior part ofthe temporal lobe and the insula in the left and righthemispheres of cases 1 and 2 and relate the dark and lightregions to the cytoarchitectonic map. A similar dispositionof dark and light cytochrome oxidase regions was observedin tangential sections through the left and right hemi-spheres of cases 3 and 4 (partially shown in Fig. 6).

The dark and light regions may correspond to dis-tinct cortical areas. This was confirmed by differences

in cytoarchitecture (see above), the different radialdistribution of the cytochrome oxidase, and the differ-ent acetylcholinesterase patterns (see below) betweensome of these areas. We propose to refer to theseputative areas outside A1 as AA (anterior area, corre-sponding to the second cytochrome oxidase dark regionwithin TC/TG); PA (posterior area, corresponding to thethird dark region within the posterior part of TA); STA(superior temporal area, corresponding to the fourthdark region within the lateral part of TA); PIA (poste-rior insular area, corresponding to the fifth dark regionin IB); LA (lateral area, corresponding to the first lightregion within TB); MA (medial area, corresponding to thesecond light region within TD); and AIA (anterior insularrea, corresponding to the third light region within IA).

FIG. 3. Photomicrographs showing darkly stained regions that may correspond to individual auditory areas. A1 is on the Heschl’s gyrus;area AA, anterior to Heschl’s gyrus; area STA, on posterior convexity of superior temporal gyrus; area PIA, on posterosuperior insula. Undereach photomicrograph is shown the densitometric profile through a 14-µm-wide strip of cortex within the darker midcortical band (arbitraryunits of density in abscissa). Asterisks mark regions that were considered boundaries. Note that changes in cytochrome oxidase staining atregion boundaries occurred gradually. All examples are from case 2; A1 and areas AA and PIA are from the right hemisphere, areas STA andA1-LA from the left. Note that staining within A1 is not uniform; a peak is present in mid-A1, corresponding to the more granular subpart TC1.


The radial profiles of the cytochrome oxidase stain-ing density were analyzed within the cortical areas.Figure 7 shows representative samples from five areas.Multiple samplings within a given area showed thatthe same type was found in its whole extent. Further-more, the same type of profile was associated with thesame area in the two hemispheres and in differentcases. In A1, the midcortical dark band was centered onlayer IV; two other bands were found in layers II andVI. The cortex above and below the midcortical banddid not differ much in staining density. Areas AA andLA had very similar profiles; the midcortical band wascentered on the boundary between layers III and IV

and the cortex above and below it had roughly the samedensity. In area STA the midcortical dark band wasrelatively discrete and centered on the boundary be-tween layers III and IV; the cortex above it was darkerthan the cortex below it. In PIA the mid-cortical darkband was centered on IV; the cortex above and belowhad similar densities. Systematic analysis of samplesfrom the supratemporal and insular cortices uncoveredthree basic types of radial cytochrome oxidase profiles.Type alpha was characterized by a prominent darkband centered on layer IV; it was found in A1 and inarea PIA. Type beta had a less prominent dark bandwhich was centered on the boundary between layers III

FIG. 4. Flat reconstruction of the superior part of the temporal lobe and of the insula of the right hemisphere in cases 1 and 2 showing thedistribution of dark cytochrome oxidase regions (top parts) and their relationship to cytoarchitectonic subdivisions (bottom parts). Dark andlight cytochrome oxidase regions were delimited in coronal sections at 16 (case 1) and 15 (case 2) rostrocaudal levels (3 adjacent cytochromeoxidase sections were analyzed per level) and projected onto a line running halfway through layer IV (black boxes, dark regions; open boxes,medium light regions). Individual sections, represented here by full lines, were aligned along the inferior circular sulcus (cs-i). Thin linesindicate the superior circular sulcus (cs-s), the inferior lip of the sylvian fissure (sf-i), the superior lip of the superior temporal sulcus (sts-s),and the fundus of the superior temporal sulcus (sts-f). Heschl’s gyrus is delimited by sulci temporalis transversus primus (medially) andsecundus (laterally); on the reconstruction it is found between cs-i and sf-i. In the bottom parts, dark cytochrome oxidase regions (black boxesin the upper part) are represented by different shading according to the auditory area they belong to (boundaries of auditory areas in thinlines, abbreviation of their names white on black). Cytoarchitectonic areas according to von Economo and Koskinas (delimited by thick lines)are indicated in italics.


and IV; it was found in areas AA, PA, LA, MA, and AIA.Type gamma had a relatively faint band centered onthe boundary between layers III and IV and the cortexabove it was darker than the cortex below it; it wasfound only in area STA.

The sizes of areas were measured in cases 1 and 2(both hemispheres; Table 3). The smallest measure forthe smallest area was 0.4 and the largest measure forthe largest area was 4.5 cm2. The size range was 2.3 to3.1 cm2 for A1, 0.6 to 1.1 cm2 for AA, 0.4 to 1.1 cm2 forPA, 2.7 to 3.1 cm2 for LA, 2.1 to 2.7 cm2 for MA, 1.0 to 1.5cm2 for STA, 0.6 to 3.8 cm2 for PIA, and 2.8 to 4.5 cm2 forAIA. The size of A1 was similar between subjects; incase 1 it had a similar size on either side, whereas itwas slightly larger on the right side in case 2. For AIAsmall side differences were found, but the larger sidewas on the left in case 1 and on the right in case 2. ForAA and PA, side differences greater than 30% wereobserved in case 2 but not in case 1. STA and PIA were

larger on the right side by more than 30% (i.e., R-L/R . 0.3) in both cases.

The relative sizes of the above areas were comparedin two ways. First, they were compared to the sum ofthe surface areas of the identified areas. A1 occupied 14to 19%, AA 4 to 7%, PA 3 to 6%, LA 15 to 22%, MA 12 to20%, STA 7 to 16%, PIA 4 to 19%, and AIA 16 to 23%.Second, the size of a given area was scaled to the size ofA1 in the corresponding hemisphere. AA and PA wereeach 0.2 to 0.4 times A1, LA 0.9 to 1.3, MA 0.8 to 1.2,STA 0.6 to 1.2, PIA 0.4 to 1.2, and AIA 1.2 to 1.6.

Acetylcholinesterase Pattern

The pattern of acetylcholinesterase staining wasstudied in serial coronal sections through the supratem-poral plane and the insula from both hemispheres ofcases 1 and 2. The superior temporal and the insularcortices contained a relatively high density of acetylcho-

FIG. 5. Flat reconstruction of the superior part of the temporal lobe and of the insula of the left hemispheres in cases 1 and 2 showing thedistribution of dark cytochrome oxidase regions (top parts) and their relationship to cytoarchitectonic subdivisions and putative auditoryareas (bottom parts). Same conventions as in Fig. 3.


linesterase positive neurons and fibers (Fig. 2C). Thepattern of staining varied; the variations appeared tobe related to cytoarchitectonic areas and to corticalareas defined by the cytochrome oxidase pattern. Fig-ure 8 shows samples from A1 and from five other areas.

Area TC (corresponding to A1) contained a rather densefiber network, concentrated in lower layer I and upperlayer IV. Only a few neurons were stained in layers IIIand V. The transitional area TC/TG (corresponding toAA) was similar to A1, but the fiber staining was lessprominent and a small number of lightly stained pyra-midal neurons were found in layer III. Area TB (corre-sponding to area LA) was rather poor in fibers, butcontained darkly and lightly stained pyramidal neu-rons in layer III. Area TD (corresponding to area MA)contained a fiber network at the layer I/II boundary andin layer VI as well as some rather lightly stainedpyramidal neurons in layer III. The lateral part of areaTA (corresponding to area STA) was characterized by alarge number of darkly stained neurons in layers IIIand V and relatively light fiber staining. Area IB(corresponding to area PIA) contained few stainedfibers, notably in layer I, and a relatively small numberof stained neurons in layers III and V. Area IA (corre-sponding to area AIA) was poor in fiber staining, butcontained numerous darkly and lightly stained neu-rons in layers III, V, and VI. The different stainingpattern that we observed could be classified accordingto the predominance of fiber versus neuronal somatastaining (Table 4). Fiber staining predominated in areaTC (corresponding to A1) and the transitional areaTC/TG (corresponding to the area AA). Neuronal so-mata staining predominated in area TB (correspondingto area LA), in the lateral part of area TA (correspond-ing to area STA), and in area IA (corresponding to areaAIA). Fiber and neuronal somata staining of comparableimportance was found in the posterior part of area TA(corresponding to area PA), area TD (corresponding to areaMA), and in area IB (corresponding to area PIA).

NADPH-Diaphorase Pattern

The distribution of NADPH-diaphorase-positive ele-ments was studied in serial coronal sections of thetemporal lobe and of the insula in both hemispheres ofcases 1 and 2. Four types of NADPH-diaphorase stain-ing were observed: (i) large darkly stained neurons,visible in both gray and white matter, with theirdendritic and axonal arbors (Fig. 2A); (ii) small faintlystained cells, probably small neurons or glial cells; (iii)axons, including terminal arborizations; and (iv) smallarterioles (but not capillaries). The distribution of thelarge NADPH-diaphorase-positive neurons was chartedin the cortical areas as defined above. In all areasanalyzed, almost no large stained neurons were foundin layers I and IV (Fig. 9) and only a few large neuronsin layers II and III. Layers V and VI had a relativelyhigh density of large neurons in area TC (correspond-ing to A1) and in the lateral part of area TA (correspond-ing to area STA), and low density in areas TC/TG, TB,TD, IA, IB, and the posterior part of area TA(correspond-ing to areas AA, LA, MA, AIA, PIA, and PA).

FIG. 6. Photomicrograph of 80-µm-thick cytochrome oxidase-stained sections through the flattened superior temporal cortex of theleft hemisphere of case 3. Anterior is down, medial left and lateralright. Areas A1 and AA are cut through layer IV in A and roughlythrough layer III in B. Calibration bar, 5 mm. Note that A1 and AAstained darkly.


DISCUSSION

The human auditory cortex is the entry to the mostpowerful communication system and yet relativelylittle is known about its functional organization. Mostcurrent beliefs are derived from work on nonhuman

primates or extrapolated from work on the humanvisual cortex. It is often assumed that the auditorycortex outside A1 contains several, perhaps function-ally specialized areas. This has been shown to be thecase in nonhuman primates. Macaque auditory cortex,defined as responding to auditory stimuli, comprises

FIG. 7. Densitometric profiles of cytochrome oxidase staining density along a radial scan (14 µm wide) through layers I–VI of putativeauditory areas. For each profile, layer I is to the left, the white matter to the right; the five short lines along the x-axis indicate the pial surface,the limits between layers I/II, III/IV, IV/V and the limit between layer VI and the white matter, respectively. The density is indicated inarbitrary units along the y-axis; values left to layer I correspond to the optical density of the glass slides and mounting medium, the peak to themaximum within the region measured. Note that in A1 and PIA the peak of the dark midcortex band was within layer IV, whereas in AA, LA,and STA it was shifted toward the layer III/IV boundary. The layers above and below the midcortex band had roughly the same density in A1,PIA, and AA, whereas in STA the supragranular layers were clearly darker than the infragranular ones. Bar, 0.5 mm.


the supratemporal plane and the adjacent parts of thesuperior temporal gyrus. Electrophysiological studiesdemonstrated the existence of five auditory fields ofwhich three, A1 and the adjacent areas RL and L, areorganized tonotopically (Merzenich and Brugge, 1973).Areas lateral to A1 were shown to be selective forcomplex acoustic stimuli and were proposed to be partof a hierarchical sequence in which complex stimuli areprocessed (Rauschecker et al., 1995). Some of theelectrophysiologically defined auditory fields have adistinct architecture. A1 can be delineated on thegrounds of its cyto- and myeloarchitecture; A1 and thefield rostral to it stain very heavily for acetyl cholines-terase and cytochrome oxidase (Morel et al., 1993).Several other areas were identified by cyto- and myelo-architectonic criteria and were proposed to form threeanteroposterior stripes (called root, core, belt; Gala-burda and Pandya, 1983).

Human Auditory Areas

We have used the cytochrome oxidase and acetylcho-linesterase staining patterns, combined with cytoarchi-tectonic, to identify cortical areas on the human supra-temporal and insular cortex. The following criteriawere taken into account for defining a cortical area: (i)cytoarchitectonic differences, as described by vonEconomo and Koskinas (1925); (ii) differences in theoverall intensity of cytochrome oxidase staining; (iii)differences in radial profiles of cytochrome oxidase

staining; and (iv) differences in acetylcholinesterasestaining pattern. By these criteria, eight putative corti-cal areas were defined (Table 4); five were on thesupratemporal plane (A1, and areas AA, PA, LA, MA),one on the posterior part of the superior temporal gyrus(area STA), and two on the insula (areas PIA, AIA). Thepattern of NADPH-diaphorase staining showed littlevariations between areas and was not used as criterion.

Cortical areas identified in this study varied in sizebetween 0.4 and 4.5 cm2 (Table 3). The size of theprimary auditory area as identified by us was between2.3 and 3.1 cm2, which is within the range given by vonEconomo and Horn (1930) for area TC. Auditory areassurrounding A1 measured between 0.4 and 4.5 cm2; i.e.,0.2 to 1.6 times the size of A1. Three of them (AA, PA,STA) were on average smaller than A1, and three (LA,MA, AIA) larger. According to our estimates, there wasmore variability in the size of auditory associationareas than in the size of A1. Comparison with the visualcortex reveals that A1 is much smaller than V1, butthat auditory association areas are about the same sizeas extrastriate visual areas. Recent estimates fromflatmounted cytochrome oxidase-stained sections indi-cate that human V1 is as large as 11 3 4 cm (Tootell andTaylor, 1995). Human extrastriate visual areas tend tobe much smaller than V1; V5/MT, an area believed to bespecialized in perception of visual motion, is ca 1 3 2 cm(Clarke and Miklossy, 1990; Clarke, 1994; Tootell andTaylor, 1995), which is roughly 0.05 times the size ofV1.

Recent data from electrophysiological studies sug-gested that there are several tonotopically organizedareas on the human supratemporal plane. Pantev et al.(1995) reported two, Liegeois-Chauvel et al. (1995)three tonotopic maps, and Talavage et al. (1997) severaltonotopic gradients. One of these areas corresponded toA1 because of its location on Heschl gyrus. It remains,however, uncertain which other areas correspond to theother tonotopically organized areas.

The functional specialization of putative corticalareas described in this study can be only derived fromindirect comparison with activation studies by others.To make individual and interstudy comparisons pos-sible, we calculated Talairach coordinates of the corti-cal areas. Working with postmortem material offersseveral sources of imprecision. Gross distortion mayoccur when brain tissue is fixed outside the skull.1 Toavoid this, we have placed the brain immediately afterremoval from skull into a skull-shaped container madeof perforated plastic and perfused it in this position.Lack of appropriate measures can compromise preciselocalization of a given section within the brain. Weopted for detailed photographical recordings at all

1 Fixation within the skull may pose problems of penetration (forimmersion procedure) or of resistence to flow (for fixation procedure)which may compromise even and rapid fixation.

TABLE 3

The Sizes of Putative Cortical Areas in the Superior Partof the Temporal Lobe and in the Insula as Determinedin Cytochrome Oxidase and Nissl Stains

Case 1 left Case 1 right Case 2 left Case 2 right

cm2 % cm2 % cm2 % cm2 %

A1 2.8 15 1 2.7 14 1 2.3 17 1 3.1 19 1AA 1.1 6 0.4 1.1 5 0.4 1.0 7 0.4 0.6 4 0.2PA 1.1 6 0.4 0.9 4 0.3 0.4 3 0.2 0.8 5 0.3LA 3.1 16 1.1 3.0 15 1.1 3.1 22 1.3 2.7 17 0.9MA 2.3 12 0.8 2.1 11 0.8 2.7 20 1.2 2.5 15 0.8STA 1.7 9 0.6 3.1 16 1.2 1.0 7 0.4 1.5 9 0.5PIAa 2.6 13 0.9 3.8 19 1.4 0.6 4 0.3 1.2 7 0.4AIA 4.5 23 1.6 3.2 16 1.2 2.8 20 1.2 3.8 24 1.2Total 19.2 100 6.8 19.9 100 7.4 13.9 100 6 16.2 100 5.3

Note. Several of these areas are most likely auditory (A1, AA, PA,LA, MA, STA) and others auditory or vestibular (PIA, AIA) areas. Theareal surface is as measured along layer IV in flat reconstructions.Boundaries of areas are as defined in the legends to Figs. 2, 3, and 4.For each hemisphere measured, the area surface is given in cm2 (leftcolumn), in percentage of the total surface occupied by the identifiedareas (middle column), and in proportion to A1 (right column;A1 5 1).

a PIA may be coextensive with the cytoarchitectonic area IB andcontain 2 compartments: a cytochrome oxidase dark one in itsposterosuperior part (the size of which is given here) and a light onein its anteroinferior part.


FIG. 8. Photomicrographs of acetylcholinesterase stained sections through the superior part of the temporal lobe and the insula (righthemisphere of case 1). Pia is up, white matter down. Note the differential distribution of stained neurons and axonal plexuses within differentareas.


steps of the processing. In particular, we photographed(i) standard views of the whole brain; (ii) medial viewsof each hemisphere; (iii) anterior and posterior views ofthe coronal slices, taken in a reference frame thatrecorded reference points for anteroposterior align-ment; and (iv) the frozen blocks during cutting. Seriallycut sections were collected in separate containers,allowing for an anteroposterior spatial resolution of160 µm or less. Photographs (i)–(iii) were used forcalculating the position of the proportional grid in eachhemisphere, (iv) for correction of distortions that oc-curred in histological sections during the staining andmounting. Thus, combination of (i)–(iv) allowed us todetermine Talairach coordinates of any point within ahistological section. Despite these precautions, we can-not exclude that our coordinates of the cortical areasare not distorted with respect to the living brain andour coordinates have to be taken with caution. How-ever, it has to be also considered that activation studies,especially when analyzed in terms of groups of subjects,are subject to imprecisions.

The Talairach coordinates of eight temporoinsularareas in the left and right hemispheres of cases 1 and 2are listed in Table 5. We observed certain interhemi-spheric and interindividual variability in the positionof what we believed to be corresponding areas. Thisvariability may be due to different distortions occurringin different hemispheres or may reflect genuine inter-hemispheric and interindividual variations. Penhuneet al. (1996) analyzed MRI from 20 normal subjects andfound that rightA1 (identified by its position on Heschl’sgyrus) was on average 5 to 8 mm more rostral then leftA1. We found the same asymmetry in postmortemmaterial.

When comparing the coordinates of cortical areaswith those of activation foci reported by others, itappears that six of eight cortical areas are most likelyactivated by auditory stimuli (Figs. 10 and 11). Focifalling within or being very near areas A1, AA, PA, andLA were activated in tasks that involve listening toenvironmental sounds (Engelien et al., 1995) or tomusic (Sergent et al., 1992), making pitch judgment or

using musical imagery (Zatorre et al., 1996) when thecontrol task is rest or visual. When words or speechwere used as stimuli, as compared to tones or noise,activation foci coincided with left or right area LA orwere anterior to it on the left side (Demonet et al., 1992;Zatorre et al., 1992; Fiez et al., 1996). Moving sound, ascompared to stationary sound, activated selectively theright insula, most probably corresponding to area AIA(Griffiths et al., 1994). Area AIA was, however, alsoactivated by nonauditory stimuli, such as contralateraltactile vibration or pain stimuli (Coghill et al., 1994).

The functional role of area PIA remains unclear. Itseems to be involved primarily in vestibular and notauditory functions. Vestibular stimulation by cold wa-ter activated contralateral PIA(Fig. 11), but also contra-lateral LA and A1 (Bottini et al., 1994). The latter couldbe ascribed to the noise produced by water circulatingin the external acoustic canal.

Hierarchical and Parallel Processing within HumanAuditory Areas?

Comparison between activation studies and our stud-ies suggests that areas A1, AA, PA, LA, MA, and AIAare involved in auditory processing, and area PIA investibular processing. As in the visual system (forreview see Felleman and Van Essen, 1991), auditoryareas may be organized in a hierarchical system.Although we do not have data on layers of origin andtermination of corticocortical connections of humanauditory areas, we may be able to derive some informa-tion on a putative hierarchical order from the cyto-chrome oxidase and acetylcholinesterase patterns. Pri-mary sensory areas receive their main input fromthalamic afferents and tend to have a distinctly densecytochrome oxidase staining in layer IV (for V1 see, e.g.,Wong-Riley et al., 1993). Conversely, high level associa-tion areas rely more on input from hierarchically lowerlevel areas (Van Essen, 1985) and may therefore dis-play a relatively high cytochrome oxidase activity insupragranular layers. A1 and area PIA had a cyto-chrome oxidase profile that was compatible with a

TABLE 4

Criteria Used for Defining Putative Cortical Areas on the Human Supratemporal Plane and Insula

A1 AA LA PA STA MA AIA PIA

Cytoarchitecture TC TC/TG TB TA TA TD IA IBCOX: overall intensity Dark Dark Light Dark Dark Light Light DarkCOX: radial profile Alpha Beta Beta Beta Gamma Beta Beta AlphaAChE F F N FN N FN N FN

Note. COX, cytochrome oxidase staining pattern (type alpha, dark band in layer IV; beta, less prominent dark band in centered on layers IIIand IV; gamma, no prominent dark band, supragranular layers darker than infragranular ones); AChE, acetylcholinesterase staining pattern(type F, predominantly fiber staining; N, predominantly neuronal somata staining; FN, fiber and neuronal somata staining of comparableimportance).


primary sensory area: dark band prominent withinlayer IV, and similar staining in supra- and infragranu-lar layers. Area STA had a very different profile, with amuch less prominent band in layers III and IV and withdarker staining in supra- than in infragranular layers;this profile may be compatible with a high orderassociation area. Areas LA, PA, MA, AA, and AIA hadan intermediate profile and may correspond to earlyauditory areas of an intermediate level between A1 andarea STA.

Previous studies showed that the pattern of acetylcho-linesterase staining varies considerably within thehuman neocortex and can be used to distinguish pri-mary sensory from association cortices. However, nocomparison has been done with cytochrome oxidasestaining pattern. Mesulam and Geula (1994) found thatin acetylcholinesterase-stained material, the primarysensory cortices, including A1, have a moderate densityof axons and a low density of pyramidal neurons. The‘‘upstream’’ auditory association cortex (defined as uni-modal association cortex receiving direct input from theprimary sensory cortex) contained slightly less axonsand more pyramidal neurons than A1. The ‘‘down-stream’’ auditory association cortex (defined as unimo-dal association cortex more than one synapse from theprimary sensory cortex) was relatively rich in axons.Hutsler and Gazzaniga (1996) analyzed acetylcholines-terase staining in human auditory and language corti-ces. They found that A1 contained a dense mesh offibers, but almost no neurons, whereas adjacent corti-cal regions contained less axons and more neurons.Posterior auditory regions contained a relatively highdensity of pyramidal neurons and axons. We havefound a similar gradient in acetylcholinesterase stain-ing. Areas A1 and PIA displayed characteristics compat-

FIG. 9. Layer distribution of large NADPH-diaphorase positiveneurons in 5 putative areas. Lines indicate pial surface, boundariesbetween areas I/II, III/IV, IV/V, and VI/white matter. All examples arefrom case 1; A1, AA, and PIA from the right, LA and STA from the lefthemisphere. Note the higher proportion of stained neurons in infra-than in supragranular layers.

TABLE 5

Talairach and Tournoux (1988) Coordinatesof Putative Auditory Areas

Areas

Lefthemisphere,

case 1

Righthemisphere,

case 1

Lefthemisphere,

case 2

Righthemisphere,

case 2

x y z x y z x y z x y z

A1 243 220 8 44 215 9 246 221 16 44 215 16AA 251 25 7 50 25 7 250 1 13 55 21 12PA 242 228 15 34 226 18 240 233 9 42 228 18LA 257 222 10 58 223 15 259 229 12 60 219 12MA 246 214 8 43 215 0 241 221 5 45 212 9STA 260 220 4 64 223 15 266 234 4 65 225 14PIA 237 220 17 34 215 25 240 221 25 31 214 22AIA 237 217 7 40 215 10 240 21 0 43 26 4

Note. Areas as defined in the present study are listed in column 1.The coordinates (x, mediolateral; left hemisphere, negative; y, antero-posterior; z, inferosuperior; mm) of the center of each area are listedin columns 2 to 5.


ible with primary sensory areas. Areas AA, PA, MA,and AIA were rather similar and compatible with‘‘upstream’’ association areas. Areas LA and STA dif-fered from the latter and had a profile compatible witha ‘‘downstream’’ association area.

The hierarchical organization within the humanauditory cortex proposed by previous studies (Mesulamand Geula, 1994; Hutsler and Gazzaniga, 1996) wasconfirmed here. However, our results suggest a struc-ture within which a certain degree of parallel process-ing could take place. The intermediate level (or ‘‘up-stream’’ association cortex) may comprise as much asfour areas (PA, MA, AA, AIA) which may belong tofunctionally distinct processing streams. Anterior in-sula, corresponding to our area AIA, was shown to beselectively activated by auditory motion (Griffiths et al.,1994). Conversely, a lesion including the insula as wellas large parts of the parietal cortex was shown toimpair auditory motion perception (Griffiths et al.,

1996). Such a specialization speaks in favor of parallelprocessing, such as described in the extrastriate visualcortex. Furthermore, parallel processing seems to playan important role in human verbal and nonverbalauditory recognition, as apparent from neuropsychologi-cal studies (Clarke et al., 1996).

Differences in NADPH-Diaphorase Staining Patternbetween Cortical Regions

NADPH-diaphorase staining pattern did not differsignificantly between the putative cortical areas on thesupratemporal plane and the insula. The darkly stainedcells were nonpyramidal neurons, predominantly ininfragranular layers. However, this distribution is dif-ferent from that found in other parts of the humancerebral cortex. In human V1, darker stained cells werealso nonpyramidal neurons, but they were found mainlyin supragranular layers (Hadjikhani and Clarke, 1995).

FIG. 10. The position of A1 and five putative auditory association areas on the supratemporal plane. Colored dots indicate the Talairachcoordinates of centers of the corresponding auditory areas as determined in cases 1 and 2. The extent of the areas indicated corresponds to theaverage extent of the area (determined as the region were these areas overlapped in both cases plus the adjacent half of the region occupied bythe area in a single case). Symbols indicate activation foci to auditory stimuli from quoted studies. The same brain outlines were used here forthe left and right hemispheres. Note, however, that right A1 was more anterior than left A1, confirming thus in vivo observations (Penhune etal., 1996).


In human primary motor cortex, darkly stained neu-rons included Betz cells (Wallace et al., 1995; ownunpublished observations).

ACKNOWLEDGMENTS

The brains were obtained from bodies legated to the Institute ofAnatomy, University of Lausanne; our gratitude goes to Professors J.Dorfl and G. M. Innocenti for this opportunity. We are very grateful toMrs. S. Tollardo Naegele for excellent histological assistance, to Mr.C. Verdan for photography, and to Professor F. de Ribaupierre and Dr.P. G. H. Clarke for comments on the manuscript. This work has beensupported by the Swiss National Science Foundation Grants 31-36297.92 and 3231-41607.94 (‘‘Score A’’ fellowship) to S. Clarke. Thisarticle is submitted by F. Rivier to the Medical Faculty of theUniversity of Lausanne in partial fulfillment of the requirements forthe degree of doctor of medicine.

REFERENCES

Bottini, G., Sterzi, R., Paulesu, E., Vallar, G., Cappa, S. F., Erminio,F., Passingham, R. E., Frith, C. D., and Frackowiak, R. S. J. 1994.Identification of the central vestibular projections in man: apositron emission tomography activation study. Exp. Brain Res.99:164–169.

Braak, H. 1978. The pigment architecture of the human temporallobe. Anat. Embryol. 154:213–240.

Brodmann, K. 1909. Vergleichende Localisationslehre der Grosshirn-rinde in ihren Prinzipien dargestellt auf Grund des Zellenbaues.Johann Ambrosius Barth, Leipzig.

Celesia, G. G. 1976. Organization of auditory cortical areas in man.Brain 99:403–414.

Clarke, S. 1994. Modular organisation of human extrastriate visualcortex: Evidence from cytochrome oxidase pattern in normal andmacular degeneration cases. Eur. J. Neurosci. 6:725–736.

Clarke, S., and Miklossy, J. 1990. Occipital cortex in man: Organiza-tion of callosal connections, related myelo- and cytoarchicture, andputative boundaries of functional visual areas. J. Comp. Neurol.298:188–214.

Clarke, S., and Rivier, F. 1997. Multiple areas within human auditorycortex: Evidence from cytochrome oxidase, NADPH-diaphoraseand acetylcholinesterase patterns. Neuroimage 5:S7 (Abstract).

Clarke, S., Bellmann, A., de Ribaupierre, F., and Assal, G. 1996.Non-verbal auditory recognition in normal subjects and brain-damaged patients: Evidence for parallel processing. Neuropsycholo-gia 34:587–603.

Coghill, R. C., Talbot, J. D., Evans, A. C., Meyer, E., Gjedde, A.,Bushnell, M. C., and Duncan, G. H. 1994. Distributed processing ofpain and vibration by the human brain. J. Neurosci. 14:4095–4108.

DeFelipe, J. 1993. A study of NADPH diaphorase-positive axonalplexuses in the human temporal cortex. Brain Res. 615:342–346.

Demonet, J. F., Chollet, F., Ramsay, S., Cardebat, D., Nespoulous, J.L., Wise, R., Rascol, A., and Frackowiak, R. 1992. The anatomy ofphonological and semantic processing in normal subjects. Brain115:1753–1768.

Emre, M., Heckers, S., Mash, D. C., Geula, C., and Mesulam, M.-M.1993. Cholinergic innervation of the amygdaloid complex in thehuman brain and its alterations in old age and Alzheimer’s disease.J. Comp. Neurol. 336:117–134.

Engelien, A., Silbersweig, D., Stern, E., Huber, W., Doring, W., Frith,C., and Frackowiak, R. S. J. 1995. The fonctional anatomy ofrecovery from auditory agnosia. Brain 118:1395–1409.

Felleman, D. J., and Van Essen, D. C. 1991. Distributed hierarchicalprocessing in the primate cerebral cortex. Cereb. Cortex 1:1–47.

Fiez, J. A., Raichle, M. E., Balota, D. A., Tallal, P., and Petersen, S. E.1996. PET activation of posterior temporal regions during auditoryword presentation and verb generation. Cereb. Cortex 6:1–10.

Galaburda, A. M., and Pandya, D. N. 1983. The intrinsic architectonicand connectional organization of the superior temporal region ofthe rhesus monkey. J. Comp. Neurol. 221:169–184.

Galaburda, A. M., and Sanides, F. 1980. Cytoarchitectonic organiza-tion of the human auditory cortex. J. Comp. Neurol. 190:597–610.

Griffiths, T. D., Bench, C. J., and Frackowiak, R. S. J. 1994. Humancortical areas selectively activated by apparent sound movement.Curr. Biol. 4:892–895.

Griffiths, T. D., Rees, A., Witton, C., Sahkir, R. S., Heening, G. B., andGreen, G. G. T. 1996. Evidence for a sound movement area in thehuman cerebral cortex. Nature 383:425–427.

Hadjikhani, N., and Clarke, S. 1995. NADPH-diaphorase staining inhuman primary visual cortex. Experientia 51:A82 (Abstract).

Hustler, J. J., and Gazzaniga, M. S. 1996. Acetylcholinesterasestaining in human auditory and language cortices: Regional varia-tion of structural features. Cereb. Cortex 6:260–270.

Liegeois-Chauvel, C., Musolino, A., and Chauvel, P. 1991. Localiza-tion of the primary auditory area in man. Brain 114:139–153.

Liegeois-Chauvel, C., Laguitton, V., Badier, J. M., Schwartz, D., andChauvel, P. 1995. Etudes des mecanismes corticaux de la percep-tion auditive chez l’homme: Apport des potentiels intracerebraux etdes champs magnetiques evoques par les stimulations auditives.Rev. Neurol. 151:495–504.

Merzenich, M. M., and Brugge, J. F. 1973. Representation of thecochlear partition on the superior temporal plane of the macaquemonkey. Brain Res. 50:275–296.

Mesulam, M.-M., and Geula, C. 1994. Chemoarchitectonics of axonaland perikaryal acetylcholinesterase along information processingsystems of the human cerebral cortex. Brain Res. Bull. 33:137–153.

Morel, A., Garraghty, P. E., and Kaas, J. H. 1993. Tonotopic organiza-tion, architectonic fields, and connections of auditory cortex inmacaque monkeys. J. Comp. Neurol. 335:437–459.

Pantev, C., Bertrand, O., Eulitz, C., Verkindt, C., Hampson, S.,Schuierer, G., and Elbert, T. 1995. Specific tonotopic organizationsof different areas of the human auditory cortex revealed bysimultaneous magnetic and electric recordings. Electroencepha-logr. Clin. Neurophysiol. 94:26–40.

Penhune, V. B., Zatorre, R. J., MacDonald, J. D., and Evans, A. C.

FIG. 11. The position of cortical areas AIA and PIA on the insula.Same conventions as in Fig. 10.


1996. Interhemispheric anatomical differences in humabn primaryauditory cortex: probabilistic mapping and volume measurementfrom magnetic resonance scans. Cereb. Cortex 6:661–672.

Rademacher, J., Caviness, V. S. Jr., Steinmetz, H., and Galaburda, A.M. 1993. Topographical variation of the human primary cortices:Implications for neuroimaging, brain mapping and neurobiology.Cer. Cortex 3:313–329.

Rademacher, J., Werner, C., Morosan, P., Schleicher, A., Zilles, K.,and Steinmetz, H. 1996. Localization and variability of cytoarchitec-tonic areas in the human superior temporal cortex. Neuroimage3:S456 (Abstract).

Rauschecker, J. P., Tian, B., and Hauser, M. 1995. Processing ofcomplex sounds in the macaque nonprimary auditory cortex.Science 268:111–114.

Rivier, F., and Clarke, S. 1997. Multiple auditory areas in man:Evidence from cytochrome oxidase pattern. Eur. J. Neurosci.Suppl. 9:90 (Abstract).

Seligman, A. M., Karnovsky, M. J., Wasserkrug, H. L., and Hanker, J.S. 1968. Nondroplet ultrastructural demonstration of cytochromeoxidase activity with a polymerizing osmiophilic reagent, diamino-benzidine (DAB). J. Cell. Biol. 38:1–14.

Sergent, J., Zuck, E., Terriah, S., and MacDonald, B. 1992. Distrib-uted neural network underlying musical sight-reading and key-board performance. Science 257:106–108.

Talairach, J., and Tournoux, P. 1988. Co-Planar Stereotaxic Atlas ofthe Human Brain. Thieme Medical Publishers, New York.

Talavage, T. M., Ledden, P. J., Sereno, M. I., Rosen, B. R., and Dale,A. M. 1997. Multiphase encoded tonotopic maps in human auditorycortex. Neuroimage 5(S8):17 (Abstract).

Tootell, R. B. H., and Taylor, R. T. 1995. Anatomical evidence for MT

and additional cortical visual areas in humans. Cereb. Cortex5:39–55.

Van Buren, J. M., and Borke, C. 1972. Variations and Connections ofthe Human Thalamus. Springer Verlag, Berlin Heidelberg, NewYork.

Van Essen, D. C. 1985. Functional organization of primate visualcortex. In Cerebral Cortex (Peters, A., and Jones, E. G., Eds.), Vol. 3,pp. 259–329. Plenum Press, New York/London.

Vogt, C., and Vogt, O. 1919. Allgemeinere Ergebnisse unserer Hirnfor-schung. J. Psychol. Neurol. 25:279–461.

von Economo, C., and Horn, L. 1930. Uber Windungsrelief, Masseund Rindenarchitektonik der Supratemporalflache, ihre indivi-duellen und ihre Seitenunterschiede. Z. f. d. ges. Neur. u. Psych.130:678–757.

von Economo, C., and Koskinas, G. N. 1925. Die Cytoarchitectonik derHirnrinde des erwachsenen Menschen. Julius Springer, Berlin.

Wallace, M. N., Brown, I. E., Cox, A. T., and Harper, M. S. 1995.Pyramidal neurons in human precentral gyrus contain nitric oxidesynthase. Neuroreport 6:2532–2536.

Wong-Riley, M. T. T., Hevner, R. F., Cutlan, R., Earnest, M., Egan, R.,Frost, J., and Nguyen, T. 1993. Cytochrome oxidase in the humanvisual cortex: Distribution in the developing and the adult brain.Visual Neurosci. 10:41–58.

Zatorre, R. J., Evans, A. C., Meyer, E., and Gjedde, A. 1992.Lateralization of phonetic and pitch discrimination in speechprocessing. Science 256:846–848.

Zatorre, R. J., Halpern, A. R., Perry, D. W., Meyer, E., and Evans, A. C.1996. Hearing in the mind’s ear: A PET investigation of musicalimagery and perception. J. Cogn. Neurosc. 8:29–46.


Documents

Cytochrome Oxidase,Acetylcholinesterase, and NADPH ... · studied in the supratemporal and insular cortex of 5 human brains (cases 1–5). Additionally, the acetylcho-linesterase