p#18-Publications-Acta Neuropathol May 2006 (1)

8/13/2019 p#18-Publications-Acta Neuropathol May 2006 (1)

1/17

Abstract Autism is characterized by qualitativeabnormalities in behavior and higher order cognitivefunctions. Minicolumnar irregularities observed inautism provide a neurologically sound localization toobserved clinical and anatomical abnormalities. Thisstudy corroborates the initial reports of a minicol-umnopathy in autism within an independent sample.The patient population consisted of six age-matchedpairs of patients (DSM-IV-TR and ADI-R diagnosed)and controls. Digital micrographs were taken fromcortical areas S1, 4, 9, and 17. The image analysis

produced estimates of minicolumnar width (CW),mean interneuronal distance, variability in CW ( V CW ),cross section of Nissl-stained somata, boundary lengthof stained somata per unit area, and the planar con-vexity. On average CW was 27.2 l m in controls and25.7 l m in autistic patients ( P = 0.0234). Mean neuronand nucleolar cross sections were found to be smallerin autistic cases compared to controls, while neurondensity in autism exceeded the comparison group by23%. Analysis of inter- and intracluster distances of aDelaunay triangulation suggests that the increased celldensity is the result of a greater number of minicol-umns, otherwise the number of cells per minicolumnsappears normal. A reduction in both somatic andnucleolar cross sections could reect a bias towardsshorter connecting bers, which favors local computa-tion at the expense of inter-areal and callosal connec-tivity.

Keywords Autistic disorder/pathology Childdevelopment disorders Pervasive Neocortex Neuropil Pyramidal cells

Introduction

Minicolumns are basic architectonic and physiologicalelements identied in all regions of the neocortex [ 16]and in all mammalian species thus far evaluated [ 37].The minicolumnar circuit is an evolutionarily andontogenetically conserved template adapted in thevarious cortical areas according to their specic devel-opmental and functional requirements. The minicol-umnar core comprises radially oriented arrays of pyramidal projection neurons. At the core and periph-

M. F. Casanova ( & )Department of Psychiatry and Behavioral Sciences,University of Louisville, 500 South Preston St Bldg55A Ste 217, Louisville, KY 40292, USAe-mail: [email protected]

I. A. J. van Kooten H. van EngelandDepartment of Child and Adolescent Psychiatry,University Medical Center, Utrecht, The Netherlands

A. E. Switala J. Trippe J. StoneDepartment of Psychiatry and Behavioral Sciences,University of Louisville, Louisville, KY, USA

H. W. M. Steinbusch C. Schmitz I. A. J. van KootenDepartment of Psychiatry and Neuropsychology,Division of Cellular Neuroscience, Maastricht University,and European Graduate School of Neuroscience(EURON), Maastricht, The Netherlands

H. HeinsenMorphologic Brain Research Unit, University of Wu rzburg,Wu rzburg, Germany

P. R. Hof Department of Neuroscience,Mount Sinai School of Medicine, New York, NY, USA

Acta NeuropatholDOI 10.1007/s00401-006-0085-5

1 3

ORIGINAL PAPER

Minicolumnar abnormalities in autism

Manuel F. Casanova

Imke A. J. van Kooten

Andrew E. Switala

Herman van Engeland Helmut Heinsen Harry W. M. Steinbusch Patrick R. Hof Juan Trippe Janet Stone Christoph Schmitz

Received: 28 March 2006 / Revised: 9 May 2006 / Accepted: 10 May 2006 Springer-Verlag 2006


2/17

ery of the minicolumn, combinations of GABAergicinterneurons provide for a diversity of signaling prop-erties that serve to dynamically modulate pyramidalcell inputs and outputs that perform area and task-specic information processing needs [ 22, 30, 38, 51].

Available neuropathological and structural imagingdata suggest that autism is the result of a develop-mental lesion capable of affecting brain growth. Onepossible explanation for this is the recent nding of minicolumnar abnormalities in autism (i.e., minicol-umns of reduced size and increased numbers) [ 20]. Inthis initial study measures of minicolumnar mor-phometry were obtained relative to pyramidal cell ar-rays in nine autistic cases and in an equal number of controls. The feature extraction properties of thealgorithms were corrected for minicolumnar frag-ments, curvature of the tissue section, and 3D pro-portions (stereological modeling) [ 24]. Later on, thesame patient population was used to conrm thepresence of cortical radial abnormalities in a studyusing the gray level index (GLI), i.e., proportional areacovered by Nissl-stained to unstained elements inpostmortem samples [ 21]. Other studies have providedevidence that the minicolumnar alterations in autismare not a nonspecic effect of mental retardation.Investigators have found that minicolumnar width inDown syndrome patients reaches adult proportionsearlier than normal, possibly as a result of acceleratedaging [16, 17]. In these studies minicolumnar size wasreported to be normal despite the small brain size of Down syndrome patients.

An increase in the number of minicolumns isthought to underlie the neocortical expansion accom-panying human encephalization, i.e., the process bywhich the brain has increased in size to a degreegreater than expected when taking body size into ac-count [ 88]. Empirical evidence and theoretical modelsindicate that local circuit neurons increase in number,complexity, and proportion relative to projectionneurons during primate encephalization [ 45, 86]. Thesetrends reect the emergence in primates of a distinctpopulation of dorsal telencephalic-derived inhibitoryinterneurons [ 59] modulating activity of minicolumnarpyramidal cells. Furthermore, isocortical areas such asdorsolateral prefrontal cortex, lacking direct homologsin non-primates, contain a well-dened granule celllayer of excitatory interneurons and increased numbersof supragranular local projection neurons [ 80]. Multi-ple polymorphisms associated with autism may be aconsequence of phylogenetically recent changes ingenetic programs guiding the development of species-specic cytoarchitectonic features. The morphometricanalysis of such features complements genetic analysis.

This study therefore investigates minicolumnopathy inan independent sample of autistic patients. It also ex-pands on previous ndings by studying cortical cell sizeand density as related to pyramidal cell arrays. Thechanges in these parameters, early during develop-ment, would provide for basic alterations in cortico-cortical connections and information processing.

Materials and methods

Clinical dataset

The diagnosis for each autistic patient was establishedpostmortem by the Autism Tissue Program (ATP). Acertied rater and trainer arranged for a postmortemvisit with the family to obtain, with written consent,medical and clinical information via a questionnairethat included the Autism Diagnostic Interview-Revised(ADI-R; [63]).

The Harvard Brain Tissue Resource Center(HBTRC) questionnaire was modied to include aut-ism-specic questions for ATP use. The informationobtained included: donor and respondent identifyinginformation; ethnicity, handedness and known exposureto hazardous materials; diagnostic information includingdates and physician; genetic tests; pre- and postnatalmedical history; immunization, medication, and hospi-talization information; family history and additionalinformation about donor participation in any training orresearch studies such as imaging, medication trials, and/or genetic studies. The supporting documents such asautopsy reports, death certicates, medical, clinical, and/or educational records were obtained at the time of thehome visit or by sending written requests, signed by thelegal next-of-kin, to named providers.

Brain specimens

Postmortem brains (one hemisphere per case) from sixautistic cases (mean interval between death and au-topsy 20.0 2.9 h) and from six age-matched controls(mean interval between death and autopsy24.0 11.1 h) were analyzed (Table 1). Brains wereobtained from several brain banks in the USA andGermany (see Acknowledgments). All autistic patientsmet the DSM-IV-TR [ 2] and ADI-R [ 62] criteria forautism. None of them exhibited any chromosomalabnormalities. In all of the cases, autopsy was per-formed after informed consent was obtained from arelative. The use of these autopsy cases was approvedby the relevant Institutional Review Boards. Clinicalrecords were available for all cases (see Appendix ).

Acta Neuropathol

1 3
http://-/?-http://-/?-


3/17

T a

b l e 1

C l i n i c a l c h a r a c t e r i s t i c s o f t h e c a s e s i n c l u d e d i n t h i s s t u d y

P a t i e n t S e x

H e m i s p h e r e A g e ( y ) C a u s e

o f d e a t h

C l i n i c a l

h i s t o r y

H i s t o r y o f

s e i z u r e s

M

e d i c a t i o n

h i s t o r y

B r a i n

w e i g h t ( g )

P M I

( h )

S e c t i o n

t h i c k n e s s

( l m )

A 1

M

L

4

D r o w n i n g

A s t h m a / b r o n c h i t i s

N o

D

a i l y m e d i c a t i o n

( n o t s p e c i e d ) f o r

a s t h m a / b r o n c h i t i s

1 , 1 6 0

3 0

2 0 0

C 1

M

L

4

M y o c a r d i a l

i n f a r c t T a k a y a s u

a r t e r i t i s

N H

N H

N

H

1 , 3 8 0

5

5 0 0

A 2

F

L

5

C a r a c c i d e n t

E a r i n f e c t i o n s

N o

A

n t i b i o t i c s f o r e a r i n f e c t i o n s

1 , 3 9 0

1 3

2 0 0

C 2

F

R

4

L y m p h o c y t i c

m y o c a r d i t i s

N H

N H

N

H

1 , 2 2 2

2 1

2 0 0

A 3

M

R

8

S a r c o m a

S y n d a c t y l y o f t h e n g e r s

a n d f e e t ; c o l i t i s ; h i g h

f e v e r ; n e u t r o p e n i a ;

m e t a s t a t i c a l v e o l a r

r h a b d o m y o s a r c o m a ;

l a r g e p a r a v e r t e b r a l

m a s s e x t e n d i n g

f r o m c h e s t

c a v i t y t o a b d o m e n

A b n o r m a l E E G ;

n o t d i a g n o s e d

w i t h s e i z u r e

d i s o r d e r

D

e p a k o t e ( 1 y e a r a f t e r E E G ) ;

c h e m o t h e r a p y ; P e r i d e x ;

N y s t a t i n ; G C S F ; B e n a d r y l ;

P h e e r g a n ; D e x a m e t h a s o n e ;

M o r p h i n e

1 , 5 7 0

2 2

2 0 0

C 3

F

R

7

S t a t u s a s t h m a t i c u s

N H

N H

N

H

1 , 3 5 0

7 8

5 0 0

A 4

M

L

1 3

S e i z u r e s

S e v e r e h y p o t o n i a ;

k e t o g e n i c d i e t

f o r 1 . 5 y e a r s

Y e s

D

i l a t i n ( s e i z u r e s ) ;

A n t i c o n v u l s a n t s ;

T r i l e p t a l ( s e i z u r e s ) ;

T r a z a d o n e ( s l e e p )

1 , 4 2 0

2 6

2 0 0

C 4

M

R

1 4

E l e c t r o c u t i o n

N H

N H

N

H

1 , 6 0 0

2 0

2 0 0

A 5

F

R

2 0

O b s t r u c t i v e

p u l m o n a r y

d i s e a s e

A D H D ; m i c r o c e p h a l y ;

e p i l e p s y ; s c h i z o p h r e n i a

Y e s ( 3 t i m e s )

V

a r i o u s p s y c h o t r o p i c

m e d i c a t i o n s i n c l u d i n g H a l d o l ,

R i t a l i n

, a n d C o n g e n t i n ;

D e p o P r o v e r a ( b i r t h c o n t r o l ) ;

M e l l a r i l ( s l e e p ) ; Z o l o f t

1 , 1 0 8

1 5

2 0 0

C 5

M

R

2 3

R u p t u r e d s p l e e n

N H

N H

N

H

1 , 5 2 0

6

2 0 0

A 6

M

R

2 4

D r o w n i n g

P n e u m o n i a ; b r o n c h i t i s ;

b e h a v i o r a l p r o b l e m s

F i r s t s e i z u r e

p r i o r t o d e a t h

Q

u e t i a p i n e ( 2 0 0 m g B I D ) ;

P r o p a n o l o l ( 4 0 0 m g B I D ) ;

T h i o r i d a z i n e ( 5 0 m g H S )

1 , 6 1 0

1 4

2 0 0

C 6

M

R

2 5

C a r d i a c

t a m p o n a d e

N H

N H

N

H

1 , 3 8 8

1 4

5 0 0

A a u t i s m , C

c o n t r o l , M m a l e , F f e m a l e , L l e f t

, R r i g h t , N H n o h i s t o r y , B W b r a i n w e i g h t , P M I p o s t m o r t e m i n t e r v a l , y y e a r s , h h o u r s

Acta Neuropathol

1 3


4/17

Tissue processing

After immersion-xation in 10% formalin for at least3 months all hemispheres were mounted with celloidinand cut into entire series of 200- l m-thick coronal sec-tions as described in detail elsewhere [ 43]. Threehemispheres were cut at 500- l m thickness. These dif-ferences did not inuence the results of this study, sinceimaging of the tissue was done at high magnication,with a depth of eld much narrower than 200 l m (seebelow, Image capture ). Every third section (in onehemisphere: every second) was stained with gallocya-nin.

Brain regions

Gallocyanin-stained sections were used by three of us(van Kooten, Heinsen, and Schmitz) to identify cor-tical areas M1, V1, frontal association cortex, and S1(areas 4, 17, and 9 of Brodmann [ 13] and area 3b of Vogt and Vogt [ 101], respectively) according to ana-tomical landmarks and cytoarchitectural criteria(Fig. 1). Gross anatomical landmarks for M1 includethe anterior wall of the central sulcus and adjacentportions of the precentral gyrus. Cytoarchitectonically,the region is clearly demarcated by its giant Betz cellsand minimization of layer IV [ 89]. Area 17 (V1) islocated along the walls of the calcarine sulcus in theoccipital lobe and adjacent portions of the cuneus andlingual gyrus [ 19]. It is dened histologically by abroad lamina IV divided into three sublayers with

numerous very small pyramidal cells in layers II andIII. It is noted for the dense line of Gennari in myelinstains. Area 9 lies in the superior and middle frontalgyrus. Rajkowska and Goldman-Rakic [ 83, 84] foundthat it was located in the middle third of the superiorfrontal gyrus in all the cases they examined. It cov-ered both dorsolateral and dorsomedial surfaces of the gyrus and extended, in some cases, to the depth of the superior frontal gyrus and portions of the middlefrontal gyrus.

Image capture

Regions of interest were delineated with a stereologyworkstation, consisting of a modied BX50 lightmicroscope with UPlanApo objectives (Olympus, To-kyo, Japan), motorized specimen stage for automaticsampling (Ludl Electronics, Hawthorne, NY, USA),HV-C20AMP CCD color video camera (Hitachi, To-kyo, Japan), and StereoInvestigator software (Micro-BrightField, Williston, VT, USA). Delineations wereperformed with a 10 objective (NA = 0.40). Digitalmicrographs each measuring about 200 l m by 150 l mwere produced using the stereology workstation de-scribed above and a 40 oil objective (NA = 1.0). Afew hundred such images were captured per region of interest to cover the entire cortical thickness. Theseimages were assembled into one mosaic using theVirtual Slice module of the StereoInvestigator soft-ware. Only slight adjustments of contrast and bright-ness were made, without altering the data of theoriginal materials.

Depth of eld dtot of the mosaic component imagescan be computed according to the formula

d tot kNA

eM nNA ;

where k is the wavelength of the illumination, M is themagnication, e is the resolution of the CCD (twice thedistance between detectors), and n is the refractiveindex of the medium. These last two were e = 16.2 l mfor the HV-C20AMP camera and n = 1.5 for oil.Taking k on the order of 1 l m, d tot was equal to about2.1 l m, so images represent only a small virtual slice of the 200- or 500- l m-thick sections, and differences insection thickness were thus not considered as a con-found.

Computerized image analysis

Multiple techniques reduced the images to sets of descriptive parameters. The minicolumn fragment

Fig. 1 Primary sensory cortex of an 8-year-old autistic male.Automatic segmentation has classied pixels as background,shown in black , and neurons, in gray scale . Clumps of neuronshave been further separated into individual objects using themorphological watershed transform

Acta Neuropathol

1 3
http://-/?-http://-/?-


5/17

method [ 24] produced estimates of the mean width(CW) of minicolumns in a region of interest, the rela-tive deviation ( V CW ) of minicolumnar widths about themean, and the mean distance (MCS) between neigh-boring neurons within the same minicolumn fragment(the CW and MCS notation being used for con-sistency with earlier publications [ 20]). The parametersmean grain size A ; mean grain perimeter U ; andintensity ( k) of a Boolean spatial model [ 99] were t toeach image. The GLI method [ 91, 92] reported themean ( D ) and standard deviation (sd D ) of distancesbetween ridges of high stain intensity, mean ( W ) andstandard deviation (sd W ) of the width of these ridges,and their average height ( A) above background.Lastly, the distribution of distances between neigh-boring neurons within each image was modeled as amixture of two distributions with means ( mnear andm far ) and standard deviations ( snear and sfar , respec-tively) with mnear < mfar . Details of each of the fourmethods are outlined below.

Computerized image analysis of minicolumnarstructure in laminae II through VI was performed withalgorithms described in the literature [ 24]. The featureextraction properties of the program were correctedfor minicolumnar fragments, curvature of the tissuesection, and have been adapted to 3D proportions(stereological modeling). The resulting estimates of minicolumnar width have also been validated againstphysiological measurements using intrinsic optical sig-naling and against anatomical measurements usingmyelinated ber bundles [ 24].

Color mosaics were converted to intensity imagesand adaptively thresholded using a scale space ap-proach [ 58]. Overlapping objects in the thresholdedimages were further separated using the watershedtransformation. Each connected region in the resultingimage was further classied according to size. Objectssmaller than 10 l m2, assumed to be glia, neuron frag-ments, or noise, were discarded altogether. Of theremaining objects, those with areas above the 15th

Fig. 2 Minicolumns inBrodmann area 4, lamina III,in an autistic patient ( bottom )and an age-matched controlcase ( top ). Insets highlight thecores of minicolumnfragments identied by oursoftware, illustrating thereduction in minicolumnarwidth (CW). Scale barsmeasure 200 l m on left and50 l m on right

Acta Neuropathol

1 3


6/17

percentile were used for minicolumn fragment detec-tion, while the smaller objects were classied as inter-neurons based on the fraction of all neurons estimatedto be inhibitory by Braitenberg and Schu z [12].

Local neuron density was computed as the sum of equal contributions from large objects above the 15thpercentile in the cross section. Each large neuronproduced a density hump, peaking at the objectscentroid, with elliptical contours oriented so that themajor axes were parallel to the local radial direction,i.e. along the axis of the minicolumn, at each point inan image. The ridges in the neuronal density indicatedthe cores of minicolumn fragments, and objectsincluding those classied as interneurons were parcel-lated into clusters according to the nearest fragmentcore. The average distance between neighboring frag-ment cores was addressed as the minicolumnar widthCW (Fig. 2). Standard deviation of the logarithm of these distances was the scale-independent measurevariability in minicolumn width ( V CW ). Interneuronaldistance (MCS) was the average distance betweencentroids of neighboring neurons within the samefragment. Only those fragments with more than tenneurons were considered when computing the meansand standard deviations.

From the adaptively thresholded images, comput-erized image analysis produced estimates of the rela-tive amount of area occupied by Nissl-stained tissue( A A ), the total boundary length of stained tissue perunit area ( L A ), and the planar convexity ( N A+ ); thesequantities are of no interest in themselves but sufce to

compute the parameters of a Boolean model withconvex primary grains using the method of moments[99]. Each realization of the Boolean model is theunion of a number of convex random closed sets(grains)independently and identically distributedwith mean area A and mean perimeter U located, orcentered, at the points of a Poisson process withintensity k. These three parameters, which completelydetermine the model, were obtained algebraically fromthe following equations [ 99], substituting for A A , L A ,and N A+ their respective estimates:

A A 1 e k A ;

N A k 1 A A ;

L A N A U :

Considering that the foreground pixels of a thres-holded image correspond to locations in the originalimage overlapped by one or more stained cell soma, Ais the average area of a neuronal cross section and k istheir density, i.e. number of neuronal cross sections perunit area. This technique was used to obtain thesevalues without any need to segment individual neu-rons. The method of minicolumn fragments, on theother hand, does attempt to segment individual cellsomata but uses the size of segmented objects only tosort them into size classes as described above.

Thresholded images were also analyzed according tothe GLI method [ 91, 92], slightly modied so that theGLI would vary smoothly across a region (Figs. 3, 4).

Fig. 3 Left : A 0.5 mm 0.5 mm eld from normal humanprimary motor cortex, lamina III. Right : Local GLI in thevicinity of each point in the eld. Values are shown in gray scalefrom 1% ( black ) to 70% ( white ). For the signicance of points X and Y , see Fig. 4. The gray level prole in this gure differs

qualitatively from the example in Schleicher and Zilles [ 91]because we estimated GLI with a smooth, bell-shaped kernel anddid not subsample the image. This allows for better localizationof gray level peaks and troughs

Acta Neuropathol

1 3


7/17

Here, the inverse thresholded image, with foregroundset equal to one and background set equal to zero, wasconvolved with a kernel measuring 11 l m by 110 l m athalf-maximum. The long axis of the kernel was parallelto the axes of the minicolumns, assumed to be theimage Y -axis. The result was an estimate of the GLI, orstaining intensity, in the neighborhood of each pixel(Fig. 3). Each GLI image so produced was measured inproles in the X direction to obtain the followingparameters: (1) mean ( D ) and (2) standard deviation(sd D ) in the distance between local maxima of the GLI,(3) amplitude ( A) of the local maxima, or differencebetween peak GLI and the lowest GLI between thepeak in question and the next peak in the prole, and(4) mean ( W ) and (v) standard deviation (sd W ) in thewidth, i.e. full width to half maximum, of the peaks(Fig. 4).

A follow-up study considered the distance betweenneighboring objects. The objects were those classiedas large neurons for the purposes of minicolumnaranalysis (see above). Neurons were considered neigh-bors if (a) they were endpoints of an edge of theDelaunay triangulation of object centroids (Fig. 5),and (b) they were not further apart than the distancefrom either of them to the edge of the region of interest. Criterion (a) is equivalent to the geometricstatement that three neurons are neighbors if the circledrawn through their centroids does not enclose thecentroid of any other neuron. The criterion (b) isnecessary to correct for a boundary effect where dis-tant neurons are incorrectly labeled as neighbors be-cause the true neighbors of one or more of them lieoutside the region of interest. Now given that objects in

a single lamina were clustered, as cells in minicolumns,an objects neighbors may include the members of thesame cluster or the members of a nearby cluster(Fig. 6). Accordingly, the distribution of distances be-tween neighbors ( d) would then be a mixture of in-tracluster or near distances, and intercluster or fardistances. This was modeled as a two-component log-normal mixture distribution; that is to say, the log dwere assumed to be drawn from either of two Gaussiandistributions of near and far distances:

log d aN l near ; r2near 1 a N l far ; r 2far ; 0 a 1:

Nucleolar size

Nucleoli were identied visually in digitized images.

Twelve random locations were selected, per mosaic, bycomputer. The user identied the nucleolus nearest toeach random point, excepting those points that did notfall within laminae II through VI. Nucleoli were seg-mented from the surrounding cytoplasm by thres-holding according to Otsus [ 73] method. If thethresholded nucleolus appeared to touch another object, the user manually removed pixels from that objectuntil the nucleolus was separate. Nucleolar cross sec-tions were measured by counting pixels in the thres-holded nucleoli.

Fig. 5 Delaunay triangulation of larger neurons ( gray scale ) inthe primary sensory cortex of an 8-year-old autistic male. Givena point set in the plane, neuron centroids in this case, theDelaunay triangulation is dened such that three points form thevertices of a triangle if and only if the circle through those pointshas no point from the set in its interior

Fig. 4 GLI parameters are illustrated on a prole from points X to Y in Fig. 3: amplitude A, the difference between peak GLIand average trough GLI, distance D between peaks, and fullwidth at half amplitude W

Acta Neuropathol

1 3


8/17

Statistical analysis

Autistic and control cases were sorted into age-mat-ched pairs for the purposes of statistical analysis.Wilcoxon signed-rank tests were used to verify theabsence of pairwise differences in brain weight orpostmortem interval. Both tests were insignicantwith P = 0.84. Otherwise, statistical analysis for allmeasurements employed repeated measures analysisof variance. Case pair (16) was the random factor,and xed factors included diagnosis, sex, cortical area,and hemisphere. Interactions of the main effectsinvolving both sex and cerebral hemisphere were ex-cluded from the model due to multicolinearity. Forthe minicolumn fragment data only, a follow-up testwas performed using a second model comprising theoriginal model plus a factor for cortical lamina: III orV/VI. The model tting and inferential statistics wereperformed with Matlab (The MathWorks, Natick,MA, USA).

Results

Minicolumnar width (CW) was greater in controls thanin autistic persons by 1.46 l m or 5.54% of the mean[Student t = 2.466 with 19 degrees of freedom ( df );P = 0.0234] (Table 2). CW also varied with corticalarea ( F = 17.50 with 3 numerator df and 19 denomi-nator df ; P < 0.0001), but no other effects were sta-tistically signicant. MCS exhibited a sex dependencein that the difference between control and autisticcases was greater in females than in males ( t = 2.227with 19 df ; P = 0.0382); mean MCS was 18.3 and18.9 l m in autistic females and males, respectively, and19.7 and 18.6 l m in normal females and males,respectively. Again the only other signicant effect wascortical area ( F = 16.09 with 3 numerator df and 19denominator df ; P < 0.0001) (Table 2). There were nosignicant ndings in V CW . The follow-up analysis onthose minicolumn fragments lying outside of laminaIV, using the supplemental model with lamina includedas a factor, found no signicant dependence of anymeasurement on lamina or lamina by diagnosis inter-action.

Simultaneous multivariate analysis of neuron densityand neuron prole area and perimeter (Table 3) re-vealed signicant diagnosis dependence ( F = 5.47 with3 numerator df and 17 denominator df ; P = 0.0081) anddiagnosis by cortical area interaction ( F = 3.65 with 9numerator df and 41.5 denominator df ; P = 0.0020).Mean particle cross section was 30.5 l m2 less in autisticcases compared to controls ( t = 3.804 with 19 df ,P = 0.0012), while particle density was greater in aut-ism: 5.15 103 mm 2 versus 4.19 103 mm -2 in controls(t = -2.723 with 19 df ; P = 0.0135).

Each GLI parameter (Table 4) was reduced in aut-ism: D by 1.93 l m ( t = 2.8465 with 19 df ; P = 0.0103),sdD by 0.96 l m (t = 2.1402 with 19 df ; P = 0.0455), A

Fig. 6 a Owing to the clustering of neurons within minicolumns,the triangulation will include edges between objects in the samecluster ( solid lines ) and edges between objects in other clusters(dotted lines ). The distribution of edge lengths over the wholegraph will be a mixture of these two subgraphs distributions. bWe have estimated the locations and scales of distances withinand between clusters by modeling the logarithm of edge lengths

as a two-component Gaussian mixture (see text for details)

Table 2 Average minicolumnar morphometry, broken down bydiagnosis and cortical area

Area CW ( l m) MCS ( l m) V CW (%)

Autistic Control Autistic Control Autistic Control3 25.8 27.0 18.4 18.7 15.1 14.84 27.5 28.2 19.9 20.0 14.2 14.69 26.5 29.4 19.1 20.7 15.0 14.917 23.0 24.1 17.0 17.3 14.4 14.7sd 1.2 0.8 0.9

Scale-independent variation in minicolumnar width V CW is ex-pressed as percentage of the meanCW minicolumnar width; MCS mean cell spacing (interneuronaldistance); V CW variability in minicolumnar distance; sd within-group standard deviation

Acta Neuropathol

1 3


9/17


10/17

objective, the use of thicker sections allows for a sub-stantial overlap of neighboring minicolumns (see Imagecapture ). The result is a relative diminution in minicol-umnar width for both patients and controls when com-paring the present study to previous ones [ 20].

Our initial nding, now corroborated in this study, isthat the brains of the autistic patients have minicol-umns of reduced width and consequently of increasednumbers per linear length of distance [ 20]. Minicol-umnar width varied with cortical area but no othereffects were statistically signicant. Of the four sam-pled areas (areas 3, 4, 9, and 17), the dorsolateralprefrontal cortex showed the greatest reduction inminicolumnar width when comparing autistics andcontrols (Table 2). Since the lateral or granular pre-frontal cortex is apparently unique to primates [ 82], thending could have important implications regardingputative animal models for autism, especially in ro-dents. Further topographical inferences regarding aminicolumnopathy in autism would require a largerpatient population and a greater number of regions of interest.

The presence of supernumerary minicolumns is saidto account for the process of cortical expansion duringencephalization [ 88]. Cortical expansion necessitatesan increased number of neurons, but not proportion-ally [26, 47]. Although the human brain, largely theneocortex, is three times the size of the chimpanzeebrain, there is only a 25% increase in the total numberof neurons [ 47]. In our study evidence suggestive of anincreased number of minicolumns per linear length inthe brains of autistic individuals led us to examineneuronal density. A point process model indicated in-creased neuronal density but failed to discern the basisfor the same, e.g., the presence of supernumerary mi-nicolumns, an increase in the total number of cells perminicolumn, or both. A subsequent analysis based on aDelaunay triangulation addressed the aforementionedpossibilities. The new algorithm indicated signicantdifferences in the edges of intercluster distancesbut not within intracluster distances. Minicolumns

appeared to be packed closer together in autism(reduction in far distances) but the total number of cells per minicolumn (near distances) appeared nor-mal. Finally, neurons within the minicolumns of theautistic patients were smaller in size and had a reduc-tion in the size of their nucleoli. In light of the smallsample size, the fact that all the various measurementsrevealed statistically signicant differences is surpris-ing. There is a strong possibility, then, that the trueeffect size associated with some of these measurementsis greater than we have observed. Further studies witha larger sample may put tighter bounds on the mag-nitude of these differences between autistic patientsand controls. The following paragraphs discuss thepreviously mentioned results from the perspective of both neuropathology and possible clinical correlates.

Minicolumns (development and numbers)

Cell arrays are the rst radially aligned structureappreciable during development. These modulescomprise layer V pyramidal cells (whose dendritesform a bundle extending through layer II), clusters of layer II and III cells, as well as associated interneurons[78]. They have been investigated in a number of species, including barrel eld cortex of the mouse[103], barrel eld cortex of the rat [ 75], area 17 of cat[77], visual cortex of non-human primates [ 14, 76, 78,79], and various human cortical areas [ 11, 94, 95].Pyramidal cell arrays have been the focus of investi-gation in the studies of columnar development in aseries of fetal and adult postmortem tissues. Krmpotic -Nemanic et al. [54] described the development of cellarrays in human fetal and postnatal auditory cortex in aseries from celloidin-embedded postmortem materialwith dates ranging from 9 weeks gestational to3 months postnatal. They identied continuity in thedevelopment of ontogenetic columns into pyramidalarrays with regional differences in the elaboration of the basic columnar structure. Pyramidal cell arraystherefore derive from the ontogenetic cell column andprovide the matrix around which growing axonal anddendritic processes are organized. In both monkeysand humans, most of the founder cell divisions thataccount for the number of cortical columns occur be-fore embryonic day 40 [ 85, 87]. The genesis for an in-crease in the total number of minicolumns in autismwould therefore transpire at an early stage of gestation.

Minicolumnar size

The possible signicance of smaller minicolumns inautistic patients can be gleaned from the results ob-

Table 6 Mean nucleolar equivalent radius, i.e. the radius of acircle with area equal to the nucleolar cross section, by diagnosisand cortical area

Area r equiv ( l m)

Autistic Control

3 2.27 3.19

4 2.24 3.089 2.56 3.1917 2.03 2.46

Within-group standard deviation is 0.31 l m

Acta Neuropathol

1 3
http://-/?-http://-/?-http://-/?-http://-/?-


11/17

tained in the comparative studies of columns in corticalarea 17 (V1) [ 31, 7479]. In primates, the primary visualcortex contains small minicolumns when compared tomany other mammals. This is the case even though thebrain size of primates is many times greater than thecomparison species. Studies using uniformed sectionthickness in rhesus monkeys have reported minicol-umns to be about 2331 l m on the basis of apicaldendrite bundles. By contrast they range from 50 to60 l m in other small-brained mammals [ 15, 18].Researchers have concluded that the narrow minicol-umns reect the increased processing complexity of primate vision and interpreted the smaller minicol-umns as being a more complex, interconnected system.In effect, reduced minicolumnar spacing may providefor an increased overlap of their dendritic trees makingthe function of neighboring minicolumns more inter-dependent [ 96, 97].

Cell size and number

Studies estimating cell counts and/or describing cyto-architectural features in autism have been few innumber and have revealed no consistent ndings.Aarkrog [ 1] described some cell increase in a frontallobe biopsy. Coleman et al. [ 28] found signicant dif-ferences in six out of 42 comparisons when studying thebrain of a 21-year-old autistic female and two controls.Kemper and Bauman [ 52] used qualitative means todescribe disturbed lamination in the anterior cingulategyrus in ve out of six autistic patients. No abnormali-ties in cell counts were reported by Bailey et al . [4] intheir six autistic cases. Previous results by Casanovaet al. [20] indicate that the brains of autistic patientshave smaller and more abundant minicolumns. Thetighter packing of cortical modules suggested increasedcellular density. A more recent case study examinedthree cortical areas in nine autistic patients and 11controls [ 21]. The overall mean GLI in this series was19.4% for the control group and 18.7% for the autisticgroup ( P = 0.724) with diagnosis-dependent effects inD . The latter authors concluded that in autism a normalGLI and an increased number of modules indicate areduction in the total number of cells per minicolumn.

In the present study the overall GLI did not differsignicantly between autistic patients and controls. Thedifferences in gray level inhomogeneity as described byD and W (Fig. 4) corroborate previous ndings thatminicolumnar width is reduced in autism [ 20, 21].Signicant differences in sd D and sd W , together withthe lack of signicant differences in scale-independentvariability in minicolumnar width, imply a direct pro-

portionality between the mean and standard deviationof minicolumnar widths within a cortical area.

Instead of cell loss the present study proposes analternate explanation to the preservation of GLI andincreased number of minicolumns: a reduction inneuronal size. Our results indicate the presence of diminished neuronal cell size and increased density inthe brains of autistic patients. These ndings variedaccording to brain region (diagnosis by cortical areainteraction). The results counter the general notionthat, with the exception of striate cortex, cell density issimilar across different cortical areas and even amongspecies [ 90]. The impression of cellular uniformity incortical columns has been refuted by using modernunbiased techniques [ 7, 8].

The neurons in our study sample had well-denednucleoli that were easily distinguishable from othercomponents of the karyoplasm. Histological studieshave characterized the nucleolus as an RNA organellewhose function is to regulate protein synthesis. As suchthe volume of the nucleolus is a constitutional factoradjusted to match the basal metabolic requirements of a cell [66]. The exception to this rule appears to be fastspiking neurons where increased metabolism appearscoupled to smaller neuronal size and nucleoli [ 3].Pathological conditions causing either increased ordecreased cellular activity have an impact on pyrami-dal cells nucleolar size in accordance with their proteinsynthesis requirements [ 66]. Studies of this and similarmorphometric indices have illustrated their utility inconditions such as Alzheimers disease [ 6466] andschizophrenia [ 29, 61]. In autism a signicant diminu-tion in nucleolar size, after thresholding interneurons,suggests a corresponding reduction in the neuronalmetabolic activity/efciency.

Clinical correlates

Brain growth causes the isolation of non-adjacentneurons by expanding their intervening distance. Withlonger distances the presence of smaller neurons im-poses a metabolic constraint on connectivity. Longerconnections necessitate larger and more active neu-rons, where each cell is networked into a dynamicallycontrolled energy system [ 27, 56]. A cortex biased to-wards smaller neurons would facilitate signal delaysand metabolic inefciencies when linking disparatebrain areas. The result would be reduced or impairedfunctional connectivity between distant cortical regions[9, 48, 50]. In such a system both sensory coding andmotor output, the endpoints of networked chaintranslating sensory information into behavioral actions,

Acta Neuropathol

1 3


12/17

are normal. It is the intervening step of informationanalysis that is abnormal.

The presence of smaller neurons in autism may helpexplain the fact that autism spectrum disorder (ASD)patients normally activate the fusiform gyrus whenviewing faces as compared to non-face stimuli [ 39]. Thedata indicate that the face-processing decits encoun-tered in ASD are not due to dysfunction of an indi-vidual area, in this case the fusiform gyrus, but due tomore complex anomalies in the distributed network of interconnected brain regions [ 10, 39, 81]. In effect, thelevel of synchronization during activation tasksinvolving distant brain regions suggests impaired con-nectivity in autism [ 25, 48, 50, 53, 104]. It may be, asMinshew et al. [ 69] have suggested, that autism ismanifested as abnormalities in high-level tasks when-ever an elevated demand is placed on informationprocessing [ 33, 70].

Just et al. [ 50] have subsumed the evidence for alower degree of information integration in autism un-der the rubric of the underconnectivity theory.However, the term may prove to be a misnomer whenapplied to shorter intra-areal connections (arcuate or ubers). In autism, an increase in the total number of minicolumns requires a scale increase (roughly a 3/2power law) in white matter to maintain modular in-terconnectivity [ 46]. This additional white matter takesthe form of short-range connections which makes upthe bulk of intracortical connections [ 23]. Recentstructural imaging studies suggest that short associationbers are overrepresented in autism [ 44]. This fact mayhelp explain the superior abilities of the autistic pa-tients when performing tasks that require local infor-mation processing [ 41]. A diminution in neuronal cellsize and a concomitant increase in the total number of minicolumns biases cortical connectivity in favor of local rather than global information processing [ 6]. Theresult is a hyper-specic brain [ 36], where segmentsof perception are retained at the expense of losing thebig picture [ 98]. In autism, a computational per-spective validates this framework and sees hyperspec-icity as a possible framework of neural codes incharge of elaborating concepts [ 67].

It is noteworthy that an adaptive strategy forincreasing the metabolic efciency of a system drivenby smaller neurons is to increase their total cell num-bers while reducing the activity of each neuron [ 60].This approach, called sparse distributed coding,achieves high representational capacity by distributingsmall amounts of activity over a large population of neurons [ 57]. In sparse coding, neurons have the po-tential to respond strongly to focal features of sensoryinputs [ 32, 40]. Since neurons rarely encounter their

feature, they will re in short bursts, sparingly distrib-uted in time [ 57]. Sparse-distributed coding is charac-teristic of the visual system [ 100, 102]. A brain whoseneuronal population is biased towards small neuronsand corresponding metabolic exigencies would there-fore create and execute strategies that emphasizeselective convergence of information among closelyadjacent modules, as e.g. in the visual system.

Large system operations can be subdivided intotask-specic modules where information processingproceeds along hierarchical stages [ 49]. Metabolicconstraints facilitate the connectivity of closely adja-cent cortical areas [ 35, 68] which represent similarfeatures of perception [ 93]. The analogy is to an operonwhere related genes (e.g., coding for enzymes in thesame metabolic pathway) cluster together so that out-side inuences can provide for simultaneous negativeor positive control to all of its units. This type of assembly is efcient because perceptions commonlyoverlap with one another, sharing parts which con-tinue unchanged from one moment to another [ 5]. Byway of contrast information at higher echelonsemphasizes complex conjunctions of perceptual attri-butes.

When comparing the hierarchy of perceptual net-works, primary association cortices, modules becomelarger and representation more categorical [ 34, 42]. Asa consequence, information at lower echelons of theperceptual hierarchy is more localized and lesions re-sult in concrete decits. At higher levels, information ismore dependent on distributed networks. One goodexample of a physiological characteristic dependent ona distributed network is face recognition, i.e., nodes intemporal, fusiform, and prepiriform cortex [ 34]. An-other example of greater relevance to autism is jointattention. More so than face processing, multiplestudies relate joint attention to diagnosis and outcomein autism [ 72]. Development of joint attention involvesthe prefrontal cortical areas 8 and 9 and anterior cin-gulate area 24. These areas serve to integrate self-monitored information about social intentions withinformation about goal-related behavior in other peo-ple [55] processed in the parietal and temporal lobe.This so-called social executive process [ 71] may beespecially prone to disturbance wherever an increase innumber and proportion of small neurons facilitate theintegration of information within a given region whileattenuating distal connectivity and coherence.

Acknowledgments We are grateful to the following institutionsfor the provision of the specimens: Mount Sinai School of Medicine (New York, NY, USA), University of Wu rzburgMorphological Brain Research Unit (Wu rzburg, Germany),University of Maryland Brain and Tissue Bank for Develop-

Acta Neuropathol

1 3


13/17

mental Disorders (Baltimore, MD, USA), Harvard Brain TissueResearch Center (Belmont, MA, USA), New York State Insti-tute for Basic Research in Developmental Disabilities (StatenIsland, NY, USA) and the US Autism Tissue Program (Prince-ton, NJ, USA). We thank E.K. Broschk and A. Bahrkel for theirtechnical support. This article is based upon work supported bythe Stanley Medical Research Foundation (H.H., C.S., P.R.H.and M.F.C.), the Korczak Foundation (H.v.E.), the NationalAlliance for Autism Research (C.S. P.R.H. and M.F.C.), theMcDonnell Foundation (P.R.H.) and NIMH grants MH61606(M.F.C.), MH62654 (M.F.C.), MH69991 (M.F.C), NIHMH66392 (P.R.H).

Appendix: Cognitive/functional level of the autisticpatients in this study

A1

Regression at 24 months of age Normal or early attainment of all developmental

milestones before 2 years of age Articulation of 23 word sentences with difculty,

echoing No spontaneous use of pointing Inconsistency in responding to his own name Inability to socially greet someone No sensitivity to noise Stereotypic rather than creative/imaginative play Poor eye contact No unusual preoccupations or rituals, other than

spinning wheels on transportation toys Anxiety when routines were changed No aggression toward others or himself Tantrums Sometimes walking on his toes; no spinal problems;

very agile Age-appropriate growth prole No idiosyncratic hand or nger mannerisms No neurocutaneous stigmata or musculoskeletal

abnormalities Immature pencil grip when attempting graphomo-

tor tasks

A2

Motor milestones met within normal limits; nevertoilet trained

Somewhat delayed ne and gross motor skills Isolated and withdrawn, in addition to engaging in

repetitive behaviors, at 18 months of age Language delays Lack of speech and low frustration tolerance

Speech therapy, physical therapy, and occupationaltherapy

No consistent use of any words at the age of 5 No imitation of others actions or direction of oth-

ers attention to things of interest to her Qualitative impairments in reciprocal social inter-

action Reduced eye contact and difculty regulating eye

contact in social situations Inappropriate facial expressions, such as laughing

and crying for no apparent reason Occasional engagement in imaginative play by her-

self; no engagement in imaginative play with others Very little interest in other children Repetitive play Repetitive body movements, such as nger icking

and hand apping

A3

Motor milestones were met within normal limits;fully toilet trained

Use of single words at 18 months of age No development of phrase speech by 3 or 4 years of

age Echoing at 3 years of age, occurring only occa-

sionally by the age of 8 Poor eye contact between 4 and 5 years of age, with

improvement during development Speech therapy, physical therapy, and occupational

therapy Ability to speak in simple sentences, using verbsand other grammatical markings, by 8 years of age

Difculty engaging in reciprocal conversation ontopics the patient himself introduced

Difculty answering direct questions Difculty pronouncing certain words Regular use of stereotyped words and phrases Difculty spontaneously imitating the action of

others Pointing to make requests Inattentive to those talking to him

Well-developed receptive language skills Qualitative impairments in reciprocal social inter-action

Inability to express or explain his own pain Typical range of facial expressions, but occasionally

inappropriate, such as laughing for no apparentreason

Engagement in some pretend play on his own at theage of 8

Acta Neuropathol

1 3


14/17

Difculty engaging in reciprocal social play Repetitive play Stereotyped whole body movements, such as

jumping up and down on his tiptoes Extremely affectionate, loving, and kind-hearted Well-developed visuospatial skills Excellent memory

A4

Severe hypotonia at 6 months of age Ability to sit up without support at 7 months of age No walking without support until 4 years of age Never toilet trained Developmental delays at 29 months of age Delays in both ne and gross motor skills Physical therapy, occupational therapy, and inten-

sive speech therapy

Qualitative impairments in communication Language delays evident at 15 months of age, usinga few single words inconsistently

Ability to sign a few words learned at about18 months of age but lost by the age of 4

Rare spontaneous use of conventional or instru-mental gestures between 4 and 5 years of age, onlyoccasional looking up when someone would enterthe room without calling patients name

Vocalization in the form of jargon and consonantvowel sounds that were not consistently directed atanyone, at 13 years of age

Ability to follow only simple directions at 13 yearsof age Rare spontaneous imitation of another persons

actions Frequent screaming Hypersensitivity to certain noises Qualitative impairments in reciprocal social inter-

action Ability to make brief make eye contact with

familiar adults Vocalizing, jumping up and down, and apping his

arms and hands to express excitement

Struggle to understand emotional experience of others Limited range of facial expressions between 4 and

5 years of age, but sufcient by the age of 13 for thepatients parents to understand the major emotionsthat he experienced, such as happiness, anger, andfrustration

Inappropriate facial expressions, such as laughingfor no apparent reason, worsening with age

No engagement in imaginative play by himself orwith others between the ages of 4 and 5, playingnext to other children

Rare interaction with siblings and unresponsivenessto their social approaches or the approaches of lessfamiliar children

Repetitive play Anxiety caused by minor changes in his routine and

changes to his immediate environment Difculty processing new environments Many unusual sensory interests History of aggression towards family members,

especially between the ages of 10 and 11, sometimesat school

Occasional self-injurious behavior, such as headbanging

Affectionate, good natured, and sweet

A5

Delayed motor milestones Delayed language development Use of single words, such as dog and hi, at

approximately 3 years of age Ability to speak in short phrases at age 5; no progress

beyond spontaneous use of two word combinations;no ability to speak three word phrases spontaneously

Language regression at age 5 following a grand malseizure

Occupational and physical therapy, music and art

therapy as well as speech and language, social andlife skills training No interest in her peers between the ages of 4 and 5 Qualitative impairments in communication Some difculty coordinating her gaze Occasional imitation of noises she had heard; no

spontaneous imitation of the actions of others No engagement in any form of pretend play Qualitative impairments in social interaction Difculty making and maintaining eye contact Exceptional visuospatial skill Greater interest in certain parts of toys rather than

using the toy as it was intended No unusual hand or nger mannerisms Bouncing up and down while bent over at the waist

with arms pulled in tightly to chest to expressexcitement

No unusual sensory interests Sensitivity to bright lights and excessive noise Tactile defensiveness, responding negatively to

being touched by others

Acta Neuropathol

1 3


15/17

A6

Speech delay, rocking, hyperactivity to the point of breaking several playpens, and periodic cryingspells at 26 months of age

Poor eye contact, failure to attend to the breast

while feeding, and crying spells as early as 7 monthsof age (noted retrospectively) Independent walking not delayed, noted at

14 months No word use until 9 years of age; no phrase use Gradual loss of interest of toileting self-care skills Disrupted sleeping habits Poor eye contact Absence of social smiling and limited facial

expression No imaginative play with others, nor interest in

other children

Absence of shared attention skills Lack of empathy Use of others body to communicate Laughing for no apparent reason at 45 years of age No use of any form of gesture to communicate,

including pointing and head shaking, and no imi-tation of others

Repetitive behaviors including circumscribedinterest in listening to music all day, repetitive toyplay, insistence on carrying out certain rituals,smelling everything (including non-food items),hand and nger mannerisms, and rocking

Extreme anxiety with any change in his routine(leading to self-hitting) Sound sensitivity, hyperactivity, and aggression to-

ward others (punching)

References

1. Aarkrog T (1968) Organic factors in infantile psychoses andborderline psychoses: retrospective study of 45 cases sub- jected to pneumoencephalography. Dan Med Bull 15:283288

2. American Psychiatric Association (2000) Diagnostic andstatistical manual of mental disorders, 4th edn. Text rev.APA, Washington

3. Bacci A, Huguenard JR (2006) Enhancement of spike-timing precision by autaptic transmission in neocorticalinhibitory interneurons. Neuron 49:119130

4. Bailey A, Luthert P, Dean A,Harding B, Janota I, Monto-gomery M, Rutter M, Lantos P (1998) A clinicopathologicalstudy of autism. Brain 121:889905

5. Barlow HB (1972) Single units and sensation: a neurondoctrine for perceptual psychology? Perception 1:371394

6. Baron-Cohen S (2004) The cognitive neuroscience of aut-ism. J Neurol Neurosurg Psychiatr 75:945948

7. Beaulieu C, Colonnier M (1989) Number and size of neu-rons and synapses in the motor cortex of cats raised indifferent environmental complexities. J Comp Neurol289:178181

8. Beaulieu C (1993) Numerical data on neocortical neuronsin adult rat, with special references to the GABA popula-tion. Brain Res 609:284292

9. Belmonte MK, Allen G, Belckel-Mitchener A, BoulangerIM, Carper RA, Webb S (2004) Autism and abnormaldevelopment of brain connectivity. J Neurosci 24:92289231

10. Boddaert N, Zilbovicius M (2002) Functional neuroimagingand childhood autism. Pediatr Radiol 32:17

11. Bonin G von, Mehler WR (1971) On columnar arrange-ment of nerve cells in cerebral cortex. Brain Res 27:110

12. Braitenberg V, Schu z A (1998) Cortex: statistics andgeometry of neuronal connectivity. Springer, Berlin Hei-delberg New York

13. Brodmann K (1909) Vergleichende Lokalisationslehre derGrohhirnrinde. Barth, Leipzig

14. Buxhoeveden DP, Casanova MF (2000) Comparative lat-eralisation patterns in the language area of human, chim-panzee, and rhesus monkey brains. Laterality 5:315330

15. Buxhoeveden D, Casanova MF (2002) The minicolumn andevolution of the brain: a review. Brain Behav Evol 60:125151

16. Buxhoeveden D, Fobbs A, Casanova MF (2002) Quantita-tive comparison of radial cell columns in developingDowns syndrome and normal cortex. J Intellect DisabilRes 46:7681

17. Buxhoeveden D, Casanova MF (2004) Accelerated matu-ration in brains of patients with Down syndrome. J IntellectDisabil Res 48:704705

18. Buxhoeveden D, Casanova MF (2005) The cell column incomparative anatomy. In: Casanova MF (ed) Neocorticalmodularity and the cell minicolumn. Nova Biomedical, NewYork, pp 93116

19. Carpenter M (1985) Core text of neuroanatomy, 3rd edn.Williams and Wilkins, Baltimore

20. Casanova MF, Buxhoeveden D, Switala A, Roy E (2002)Minicolumnar pathology in autism. Neurology 58:428432

21. Casanova MF, Buxhoeveden D, Switala A, Roy E (2002)Neuronal density and architecture (gray level index) in thebrains of autistic patients. J Child Neurol 17:515521

22. Casanova MF, Buxhoeveden D, Gomez J (2003) Disruptionin the inhibitory architecture of the cell minicolumn:implications for autism. Neuroscientist 9:496507

23. Casanova MF (2004) White matter volume increase andminicolumns in autism. Ann Neurol 56:453

24. Casanova MF, Switala AE (2005) Minicolumnar mor-phometry: computerized image analysis. In: Casanova MF(ed) Neocortical modularity and the cell minicolumn. Nova

Biomedical, New York, pp 16118025. Castelli F, Frith C, Happe F, Frith U (2002) Autism, As-perger syndrome and brain mechanisms for the attributionof mental states to animated shapes. Brain 125:18391849

26. Changizi MA (2001) Principles underlying mammalianneocortical scaling. Biol Cybern 84:207215

27. Chklovskii DB, Koulakov AA (2004) Maps in the brain:what can we learn from them? Annu Rev Neurosci 27:369392

28. Coleman PD, Romano J, Lapham L, Simon W (1985) Cellcounts in cerebral cortex of an autistic patient. J AutismDev Disord 15:245255

Acta Neuropathol

1 3


16/17

29. Colon EJ (1972) Quantitative cytoarchitectonics of thehuman cerebral cortex in schizophrenic dementia. ActaNeuropathol 20:110

30. DeFelipe J (1997) Types of neurons, synaptic connectionsand chemical characteristics of cells immunoreactive forcalbindin-D28K, parvalbumin and calretinin in the neo-cortex. J Chem Neuroanat 14:119

31. Feldman ML, Peters A (1974) A study of barrels andpyramidal dendritic clusters in the cerebral cortex. BrainRes 77:5576

32. Field DJ (1994) What is the goal of sensory coding? NeuralComput 6:559601

33. Frith C (2004) Is autism a disconnection disorder? LancetNeurol 3:577

34. Fuster JM (2003) Cortex and mind: unifying cognition.Oxford University Press, Oxford

35. Goodhill GJ (1997) Stimulating issues in cortical mapdevelopment. Trends Neurosci 20:375376

36. Grandin T (2005) Animals in translation: using the mys-teries of autism to decode animal behavior. Scribner, NewYork

37. Gressens P, Evrard P (1993) The glial fascicle: an ontogenicand phylogenic unit guiding, supplying and distributingmammalian cortical neurons. Brain Res Dev Brain Res76:272277

38. Gupta A, Wang Y, Markram H (2000) Organizing princi-ples for a diversity of GABAergic interneurons and syn-apses in the neocortex. Science 287:273278

39. Hadjikhani N, Joseph RM, Snyder J, Chabris CF, Clark J,Steele S, McGrath L, Vangel M, Aharon I, Feczko E, HarrisGJ, Tager-Flusberg H (2004) Activation of the fusiformgyrus when individuals with autism spectrum disorder viewfaces. Neuroimage 22:11411150

40. Hahnloser R, Kozhevnikov A, Fee M (2002) An ultrasparsecode underlies the generation of neural sequences in asongbird. Nature 419:6570

41. Happe F (1999) Autism: cognitive decit or cognitive style?Trends Cogn Sci 3:216222

42. Hawkins J (2004) On intelligence. Times Books, New York43. Heinsen H, Heinsen YL (1991) Serial thick, frozen, gallo-

cyanin stained sections of human central nervous system. JHistotechnol 14:167173

44. Herbert MR, Ziegler DA, Makris N, Filipek PA, KemperTL, Normandin JJ, Sanders HA, Kennedy DN, Caviness VSJr (2004) Localization of white matter volume increase inautism and developmental language disorder. Ann Neurol55:530540

45. Hofman MA (1985) Neuronal correlates of corticalizationin mammals: a theory. J Theor Biol 112:7795

46. Hofman MA (2001) Brain evolution in hominids: are we atthe end of the road? In: Falk D, Gibson KR (eds) Evolu-tionary anatomy of the primate cerebral cortex. CambridgeUniversity Press, Cambridge, pp 113127

47. Holloway RL (1968) The evolution of the primatebrain: some aspects of quantitative relations. Brain Res7:121172

48. Horwitz B, Rumsey JM, Grady CL, Rapoport SI (1988) Thecerebral metabolic landscape in autism: intercorrelations of regional glucose utilization. Arch Neurol 45:749755

49. Hubel DH, Wiesel TN (1962) Receptive elds, binocularinteraction and functional architecture of the cats visualcortex. J Physiol 160:106154

50. Just MA, Cherkassky VL, Keller TA, Minshew NJ (2004)Cortical activation and synchronization during sentencecomprehension in high functioning autism: evidence of un-derconnectivity. Brain 127:18111821

51. Kawaguchi Y, Kubota Y (1997) GABAergic cell subtypesand their synaptic connections in rat frontal cortex. CerebCortex 7:476486

52. Kemper TL, Bauman ML (1993) The contribution of neu-ropathologic studies to the understanding of autism. NeurolClin 11:175187

53. Koshino H, Carpenter P, Minshew N, Cherkassky V, KellerT, Just M (2005) Functional connectivity in an fMRIworking memory task in high-functioning autism. Neuro-image 24:810821

54. Krmpotic -Nemanic J, Kostovic I, Nemanic (1984) Pre-natal and perinatal development of radial cell columns inthe human auditory cortex. Acta Otolaryngol 97:489495

55. Lau HC, Rogers RD, Haggard P, Passingham RE (2004)Attention to intention. Science 303:12081210

56. Laughlin SB, Sejnowski TJ (2003) Communication in neu-ronal networks. Science 301:18701874

57. Laughlin SB (2004) The implications of metabolic energyrequirements for the representation of information in neu-rons. In: Gazzaniga MS (ed) The cognitive neurosciences,3rd edn. MIT Press, Cambridge, pp 187196

58. Lega E, Scholl H, Alimi J-M, Bijaoui A, Bury P (1995) Aparallel algorithm for structure detection based on waveletand segmentation analysis. Parallel Comput 21:265285

59. Letinic K, Zoncu R, Rakic P (2002) Origin of GABAergicneurons in the human neocortex. Nature 417:645649

60. Levy WB, Baxter RA (1996) Energy-efcient neural codes.Neural Comput 8:531543

61. Lohr JB, Jeste DV (1986) Cerebellar pathology in schizo-phrenia? a neuronometric study. Biol Psychiatr 21:865875

62. Lord C, Rutter M, Le Couteur A (1994) Autism diagnosticinterviewrevised: a revised version of a diagnostic inter-view for caregivers of individuals with possible pervasivedevelopmental disorders. J Autism Dev Disord 24:659685

63. Lord C, Pickles A, McLennan J, Rutter M, Bregman J,Folstein S, Fombonne E, Leboyer M, Minshew N (1997)Diagnosing autism: analyses of data from the autism diag-nostic interview [Journal Article. Multicenter Study]. JAutism Dev Dis 27(5):501517

64. Mann DMA, Yates PO, Barton CM (1977) Cytophoto-metric mapping of neuronal changes in senile dementia. JNeurol Neurosurg Psychiatry 40:299302

65. Mann DMA, Sinclair KGA (1978) The quantitativeassessment of lipofuscin pigment, cytoplasmic RNA andnucleolar volume in senile dementia. Neuropathol ApplNeurobiol 4:129135

66. Mann DMA (1982) Nerve cell protein metabolism anddegenerative disease.Neuropathol Appl Neurobiol 8:161176

67. McClelland JL (2000) The basis of hyperspecicity in aut-ism: a preliminary suggestion based on properties of neuralnets. J Autism Dev Disord 30:497502

68. Miller KD (1994) A model for the development of simplecell receptive elds and the ordered arrangement of orien-

tation columns trough activity-dependent competition be-tween ON- and OFF-center inputs. J Neurosci 14:41041669. Minshew NJ, Goldstein G, Maurer RG, Bauman ML,

Goldman-Rakic PS (1989) The neurobiology of autism: anintegrated theory of the clinical and anatomic decits. J ClinExp Neuropsychol 11:38

70. Mundy P, Neal AR (2001) Neural plasticity, joint attention,and a transactional social-orienting model of autism. IntRev Res Ment Retard 23:139168

71. Mundy P (2003) The neural basis of social impairments inautism: the role of the dorsal medial-frontal cortex andanterior cingulate system. J Child Psychol Psychiatr 44:793809

Acta Neuropathol

1 3


17/17

72. Mundy P, Burnette C (2005) Joint attention and neurode-velopmental models of autism. In: Volkmar FR, Paul R, KlinA, Cohen D (eds) Handbook of autism and pervasive devel-opmental disorders, 3rd edn. Wiley, New York, pp 650681

73. Otsu N (1979) A threshold selection method from grey-level histograms. IEEE Trans Syst Man Cybern 9:377393

74. Peters A, Walsh TM (1972) A study of the organization of apical dendrites in the somatic sensory cortex of the rat. JComp Neurol 144:253268

75. Peters A, Kara DA (1987) The neuronal composition of area 17 of rat visual cortex, IV: the organization of pyra-midal cells. J Comp Neurol 260:573590

76. Peters A, Sethares C (1991) Organization of pyramidalneurons in area 17 of monkey visual cortex. J Comp Neurol306:123

77. Peters A, Yilmaz E (1993) Neuronal organization in area 17of cat visual cortex. Cereb Cortex 3:4968

78. Peters A, Sethares C (1996) Myelinated axons and thepyramidal cell modules in monkey primary visual cortex. JComp Neurol 365:232255

79. Peters A, Sethares C (1997) The organization of doublebouquet cells in monkey striate cortex. J Neurocytol26:779797

80. Petrides M, Pandya DN (2002) Comparative cytoarchitec-tonic analysis of the human and the macaque ventrolateralprefrontal cortex and corticocortical connection patterns inthe monkey. Eur J Neurosci 16:291310

81. Pierce K, Haist F, Sedaghat F, Courchesne E (2004) Thebrain response to personally familiar faces in autism: nd-ings of fusiform activity and beyond. Brain 127:27032716

82. Preuss TM (1995) Do rats have a prefrontal cortex? theRose Woolsey-Akert program reconsidered. J Cogn Neu-rosci 7:124

83. Rajkowska G, Goldman-Rakic PS (1995) Cytoarchitectonicdenition of prefrontal areas in the normal human cortex, I:remapping of areas 9 and 46 using quantitative criteria.Cereb Cortex 5:307322

84. Rajkowska G, Goldman-Rakic PS (1995) Cytoarchitectonicdenition of prefrontal areas in the normal human cortex, II:variability in locations of areas 9 and 46 and relationship tothe Talairach Coordinate System. Cereb Cortex 5:323337

85. Rakic P (1974) Neurons in rhesus monkey visual cortex:systematic relation between time of origin and eventualdisposition. Science 183:425427

86. Rakic P (1975) Local circuit neurons. Neurosci Res ProgBull 13:295416

87. Rakic P (1985) Limits of neurogenesis in primates. Science227:10541056

88. Rakic P (1995) A small step for the cell, a giant leap formankind: a hypothesis of neocortical expansion duringevolution. Trends Neurosci 18:383388

89. Rivara CB, Sherwood CC, Bouras C, Hof PR (2003) Ste-reologic characterization and spatial distribution patterns of Betz cells in the human primary motor cortex. Anat Rec A270:137151

90. Rockel AJ, Hiorns RW, Powell TPS (1980) The basic uni-formity in structure of the neocortex. Brain 103:221244

91. Schleicher A, Zilles K (1990) A quantitative approach tocytoarchitectonics: analysis of structural inhomogeneities innervous tissue using an image analyzer. J Microsc 157:367381

92. Schleicher A, Palomero-Gallagher N, Morosan P, Eickhoff SB, Kowalski T, Vos K de, Amunts K, Zilles K (2005)Quantitative architectural analysis: a new approach to cor-tical mapping. Anat Embryol 210:373386

93. Schwartz EL (1980) Computational anatomy and functionalarchitecture of striate cortex: spatial mapping approach toperceptual coding. Vision Res 20:643669

94. Seldon HL (1981) Structure of human auditory cortex, I:cytoarchitectonics and dendritic distributions. Brain Res229:277294

95. Seldon HL (1981) Structure of human auditory cortex, II:axon distributions and morphological correlates of speechperception. Brain Res 229:295310

96. Seldon HL (1982) Structure of human auditory cortex, III:statistical analysis of dendritic trees. Brain Res 249:211221

97. Seldon HL (1985) The anatomy of speech perception: hu-man auditory cortex. In: Jones EG, Peters A (eds) Associ-ation and auditory cortices. Plenum, New York, pp 273327

98. Shakow D (1946) The nature of deterioration in schizo-phrenic conditions. Coolidge Foundation, New York

99. Stoyan D, Kendall WS, Mecke J (1995) Stochastic geometryand its applications, 2nd edn. Wiley, Chichester

100. Vinje WE, Gallant JL (2000) Sparse coding and decorre-lation in primary visual cortex during natural vision. Science287:12731276

101. Vogt C, Vogt O (1919) Allgemeinere Ergebnisse unsererHirnforschung, dritte Mitteilung: die architektonische Rin-denfelderung im Lichte unserer neuesten Forschungen. JPsychol Neurol 25: 361376

102. Weliky M, Fiser J, Hunt RH, Wagner DN (2003) Coding of natural scenes in primary visual cortex. Neuron 37:703718

103. White EL, Peters A (1993) Cortical modules in the pos-teromedial barrel subeld (Sm1) of the mouse. J CompNeurol 334:8696

104. Zilbovicius M, Garreau B, Samson Y, Remy P, Barthele-my C, Syrota A, Lelord G (1995) Delayed maturation of the frontal cortex in childhood autism. Am J Psychiatr152:248252

Acta Neuropathol

1 3

Date post:	04-Jun-2018
Category:	Documents
Upload:	aldo-javier-barbero
View:	213 times
Download:	0 times

p#18-Publications-Acta Neuropathol May 2006 (1)

Documents