DISC1 Immunoreactivity at the Lightand Ultrastructural Level in the Human

Neocortex

BRIAN KIRKPATRICK,1* LEYAN XU,1 NICOLA CASCELLA,2 YUJI OZEKI,2

AKIRA SAWA,2–4AND ROSALINDA C. ROBERTS1

1Maryland Psychiatric Research Center, Department of Psychiatry, University ofMaryland School of Medicine, Baltimore, Maryland 21228

2Department of Psychiatry and Behavioral Sciences, Johns Hopkins University School ofMedicine, Baltimore, Maryland

3Department of Neuroscience, Johns Hopkins University School of Medicine,Baltimore, Maryland

4Program in Cellular and Molecular Medicine, Johns Hopkins University School ofMedicine, Baltimore, Maryland

ABSTRACTDisrupted-In-Schizophrenia 1 (DISC1) is one of two genes that straddle the chromosome 1

breakpoint of a translocation associated with an increased risk of schizophrenia. DISC1 has beenidentified in the brain of various mammalian species, but no previous immunocytochemicalstudies have been conducted in human neocortex. We examined DISC1 immunoreactivity infrontal and parietal cortex (BA 4, 9, 39, and 46) in normal human brain. At the light microscopiclevel, immunolabeling was prominent in the neuropil, in multiple populations of cells, and in thewhite matter. At the ultrastructural level, staining was prominent in structures associated withsynaptic function. Immunolabeled axon terminals comprised 8% of all terminals and formed bothasymmetric and symmetric synapses. Labeled axon terminals formed synapses with labeledspines and dendrites; in some, only the postsynaptic density (PSD) of the postsynaptic structurewas labeled. The most common configuration, however, was an unlabeled axon terminal formingan asymmetric synapse with a spine that had immunoreactivity deposited on the PSD andthroughout the spine. The presence of DISC1 in multiple types of synapses suggests the involve-ment of DISC1 in corticocortical as well as thalamocortical connections. Staining was also presentin ribosomes, parts of the chromatin, in dendritic shafts, and on some microtubules. Labeling wasabsent from the Golgi apparatus and multivesicular bodies, which are associated with proteinexcretion. These anatomical localization data suggest that DISC1 participates in synaptic activ-ity and microtubule function, and are consistent with the limited data on its adult function. J.Comp. Neurol. 497:436–450, 2006. © 2006 Wiley-Liss, Inc.

Indexing terms: development; schizophrenia; synapses; postmortem; plasticity

Schizophrenia is a common, chronic disorder character-ized by psychosis, cognitive impairments and deficit symp-toms in a subset of patients. Studies of environmental riskfactors point to gestation, suggesting that abnormalitiesin early brain development underlie the disorder (Lewisand Levitt, 2002). Genetic factors also have a role in therisk of schizophrenia (Harrison and Owen, 2003; Shirtsand Nimgaonkar, 2004). In addition to increased risk fordeveloping schizophrenia in relatives of those with thedisease, numerous studies have found genetic mutationslinked to the disease. For example, a (1/11) (q42.1;q14.3)balanced chromosomal translocation was found in a Scot-tish family (St. Clair et al., 1990); that is, in affectedsubjects, a segment of chromosome 1 was located on chro-mosome 11, and vice versa. Two genes straddle the break-point on chromosome 1. One transcript with an open read-

Grant sponsor: Stanley Medical Research Institute (to B.K., R.C.R., andA.S.); Grant sponsor: National Alliance for Research on Schizophrenia andDepression (to A.S.); Grant number: MH60744 (to R.C.R.); Grant number:MH66123 (to R.C.R.); Grant number: MH69853 (to A.S.); Grant sponsor:S-R Foundation (to A.S.).

Dr. Kirkpatrick’s current address is Department of Psychiatry, MedicalCollege of Georgia, 1515 Pope Avenue, Augusta, GA 30912.

*Correspondence to: Rosalinda C. Roberts, Maryland Psychiatric Re-search Center, PO Box 21247, Baltimore, MD 21228. E-mail: rroberts@mprc.umaryland.edu

Received 11 April 2005; Revised 11 August 2005; Accepted 24 February2005

DOI 10.1002/cne.21007Published online in Wiley InterScience (www.interscience.wiley.com).

THE JOURNAL OF COMPARATIVE NEUROLOGY 497:436–450 (2006)

© 2006 WILEY-LISS, INC.

ing frame, Disrupted-In-Schizophrenia (DISC1), is ex-pressed as protein; the other transcript, DISC2, is anti-sense to DISC1, and appears not to be expressed as aprotein product (Millar et al., 2000). When a psychiatricevaluation of family members was undertaken, this chro-mosomal abnormality was associated with an increasedrisk of psychiatric disorders. The strength of the associa-tion between the chromosomal translocation and psychi-atric illness was greatest for a broad phenotype that in-cluded recurrent major depression, bipolar disorder, andschizophrenia (Blackwood et al., 2001; Blackwood andMuir, 2004; Millar et al., 2000, 2001). This association hasnow been extended to the general Scottish population(Thomson et al., 2005b).

Other studies confirm that this region has a gene of riskfor schizophrenia in populations outside of Scotland. Re-cently, a frameshift mutation in DISC1 has been reportedin American probands with schizophrenia (Sachs et al.,2005). The highest lod score in a linkage analysis ofschizophrenia in a Finnish population was associated withan intragenic marker in the DISC1 gene (Ekelund et al.,2001), a finding confirmed with haplotype transmissionanalysis (Hennah et al., 2003). A polymorphism in thissame region on chromosome 1 has also been associatedwith an increased risk of either schizophrenia or bipolardisorder in some (Detera-Wadleigh et al., 1999; Ekelundet al., 2001) but not all studies (Devon et al., 2001; Zhanget al., 2005). Thus, this area on the genome is one ofseveral that appear to be determinants of risk for bothschizophrenia and bipolar disorder (Badner and Gershon,2002; Berrettini, 2001; Hodgkinson et al., 2004).

Molecular studies in rodents have found that DISC1interacts with multiple cytoskeleton and centrosome pro-teins, and have suggested a role in neurite extension,early brain development, interaction with proteins thatlocalize receptors to cell membranes, and protein trans-ducing signals (Miyoshi et al., 2003; Morris et al., 2003;Ozeki et al., 2003). In adult monkeys, DISC1 is highlylocalized in many brain regions, and is particularly robustin the limbic system (Austin et al., 2003). In humans,DISC1 has been localized by immunohistochemical tech-niques in the hippocampus (James et al., 2004) and anexonic single nucleotide polymorphism (SNP) in DISC1 isassociated with normal cognitive changes in the agingprocess (Thompson et al., 2005a). In the current study, wehave examined the light microscopic distribution and sub-cellular localization of DISC1 in the normal human neo-cortex. To our knowledge, this is the first immunocyto-chemical study of DISC1 at both the light andultrastructural level. Information on the localization ofDISC1 should help elucidate its role in the normal adultbrain and its role in the pathophysiology of schizophreniaand affective disorders. This work has been presented inpreliminary form (Kirkpatrick et al., 2002, 2003; Xu et al.,2003).

METHODS

All tissue came from the Maryland Brain Collection,where subjects had a gross neuropathological examinationin the Office of the Chief Medical Examiner of Marylandthat was judged to be normal. On the basis of informationfrom relatives, treating health professionals, autopsy re-ports, and medical records, the subjects were also deter-mined not to have a major neuropsychiatric disorder

(other than alcohol abuse) or a history of neurologicaldisease. DSM-IV diagnosis was confirmed by two Mary-land Brain Collection psychiatrists. Informed consent fortissue donation was obtained from the next of kin. Allprocedures were approved by the University of MarylandInstitutional Review Board.

A total of seven adult subjects (Table 1) were studied atthe light microscopic level; a subset of three subjects wasalso studied at the electron microscopic level. For lightmicroscopy, coronal blocks (1 cm thick) of tissue fromBrodmann area (BA) BA39, BA4, and BA9 were dissectedand placed in 0.1 M phosphate-buffered saline, pH 7.2–7.4, containing 4% paraformaldehyde for a period of atleast 1 week (4°C). Tissue used in electron microscopy(EM) was placed in 0.1 M phosphate buffer (pH 7.2–7.4)containing 1.0% glutaraldehyde plus 4% paraformalde-hyde. All sections were cut on a vibratome at a thicknessof 60 microns. Vibratomes have been shown to distorthistology less than do methods that involve freezing, andallow more sensitive (albeit specific) detection of immuno-reactivity (Kung et al., 1998; Roberts and Knickman,2002; Roberts et al., 2005).

The pH of all of the brains but one was determined fromarchived frozen tissue. In one case (#18), only fixed tissuesamples were available so it was impossible to obtain thepH. For the other cases, the pH of a sample of cerebellumwas determined according to the techniques of Harrison etal. (1995) and Johnston et al. (1997). A sample piece oftissue (a 1.5- to 2.0-cm chunk) was dissected with anautopsy saw from the frozen brain. The tissue was homog-enized for approximately 1 minute so as not to increasethe temperature of the sample, then the pH probe fromOmni’s PCR Tissue Homogenizing Kit was inserted tomeasure the pH.

The polyclonal antibody (C1) used for immunohisto-chemistry was raised in rabbits in the laboratory of Dr.Sawa. The epitope for this antibody consisted of 250 res-idues adjacent to the translocation break point (aminoacids 347–600). The antibody was affinity purified accord-ing to standard techniques used previously in Dr. Sawa’slaboratory (Ozeki et al., 2003, Sawamura et al., 2005). Theantibody’s specificity was confirmed by using two meth-ods: 1) Western blot analysis including preadsorption ofhuman embryonic kidney (HEK) 293 cells and humanbrain tissue; and 2) preadsorption with antigen for immu-nocytochemistry. Samples for the Western blot preadsorp-

TABLE 1. Subject Characteristics

ID ARG PMI pH BA Cause of death Toxicology

1 36CM1 4 6.70 9 MVA Ethanol4 43AF 6 7.18 9 ASCVD Negative5 37CF1 5 7.03 9 ASCVD Ethanol, THC10 48CM 12 6.56 4, 39 Ruptured aortic aneurysm Pseudoephedrine13 40AF 24 6.26 4 PE, deep vein thrombosis Morphine18 21CM1 6 na 46 MVA Ethanol19 52CF 12 6.70 39, 46 ASCVD Negative

Notes come from the Office of the Chief Medical Examiner of Maryland, and from thepsychiatric diagnoses of the Maryland Brain Collection. All cases were examined at thelight microscopic level; cases 1, 4, and 5 were also examined at the electron microscopiclevel.Abbreviations: ARG, age, race, gender; PMI, postmortem interval (hours); BA, Brod-mann areas; A, African-American; C, Caucasian; M, male; F, female; MVA, motorvehicle accident; ASCVD, atherosclerotic cardiovascular disease; THC, tetrahydrocan-nabinol; PE, pulmonary embolism.1Alcohol abuse.The toxicology report indicates any positive findings. Case #13 was given morphinepostoperatively. The pH of case #18 was not available (na).

The Journal of Comparative Neurology. DOI 10.1002/cne

437DISC1 ANATOMY IN HUMAN CORTEX

tion experiments were taken from archived frozen tissuefrom two cases (BA10 from case #1, and BA321 from case#19). Approximately 1 cubic centimeter was dissectedfrom each area using an autopsy bone saw. Extracts fromnormal human brain were prepared and analyzed byWestern blot as described in our previous study(Sawamura et al., 2005). For the immunocytochemicalcontrols, the primary antibody was preadsorbed for 24hours with antigen at a concentration of 25 �g/ml, thenthe tissue was stained using the same immunohistochem-ical protocol as for other sections.

All immunohistochemical incubations were done atroom temperature unless otherwise specified, and exceptas noted all solutions were made in 0.9% phosphate-buffered saline (PBS). All PBS washes lasted 5 minutes.Free-floating sections were processed through a series ofincubations, with PBS washes between the steps: 1% so-dium borohydride solution for 15 minutes; 1.5% hydrogenperoxide for 5 minutes; then incubation in 10% rabbitserum for 15 minutes followed by incubation in the pri-mary antibody solution with 1.5% horse serum at 4°C for3 days. On the third day the tissue was washed four timesin 1.5% horse serum, followed by incubation in biotinyl-ated anti-IgG (1:200) in 0.5% horse serum for 1 hour, then0.5% horse serum and reagents from an ABC kit for 30minutes (Elite ABC kit PK6100, Vector, Burlingame, CA,with recommended dilutions and preparation). Sectionswere then developed with nickel diaminobenzidine (DAB;kit #34065 from Pierce, Rockford, IL), using 1 part DABsubstrate to 9 parts stable peroxide substrate buffer. Wetested a dilution series with an antibody concentration of1:1,000, 2,000, 4,000, 10,000, 12,000, 15,000, 20,000,25,000, and 30,000. This dilution series produced a gradi-ent of staining: the darkest labeling was at high antibodyconcentrations and successively became lighter until allspecific staining disappeared. The results presented werein tissue stained with the antibody at a concentration of1:25,000.

We have previously shown that it is possible to do EMalone and combined with immunocytochemistry on hu-man postmortem brain tissue (Kung et al., 1998; Robertsand Knickman, 2002; Roberts et al., 2005). Tissue to beused for EM was immunolabeled as described above, ex-cept that none of the buffers contained Triton. After theimmunohistochemical staining, tissue for EM was reactedin 1% osmium tetroxide for 1 hour, stained en bloc with 1%uranyl acetate for 2 hours, dehydrated in alcohols, andembedded flat in resins as previously described (Kung etal., 1998; Roberts and Knickman, 2002). Samples fromcortical gray matter were removed, glued to epon beemcapsules, and serially thin sectioned at 90 nm.

The following strategy was used to characterize theultrastructural localization of DISC1 immunoreactivity incell bodies and in the neuropil. Care was taken to includein the thin section the tissue near, but not at the edge,where labeling is optimal. Moreover, the edge was ori-ented perpendicular to the pial surface so that immuno-reactivity would be comparable through the various corti-cal layers. For this analysis, at least two samples weresectioned from BA9 per case, one from the superficiallayers (I–III) and one from the deep layers (IV–VI). Boththe superficial and deep layers were examined and photo-graphed from each case. Because the orientation of thesections was maintained (small edge of the trapezoid wassuperficial and the long edge of the trapezoid was deep), it

was possible to move through the sections from superficialto deep while taking photographs along the way.

For the cell body analysis, at least 44 pictures weretaken per area per case. Labeled cell bodies in the regionof optimal immunoreactivity were photographed at lowpower (3.5–5,000�) and portions of the soma were photo-graphed at higher magnification (10–15,000�). For thesynaptic counts, tissue exhibiting optimal immunoreactiv-ity was randomly photographed at a magnification of 10–15,000, through all of the cortical layers present in theindividual section. This analysis yielded neuropil contain-ing both labeled and unlabeled elements, and a total of328 synapses were categorized. Synapses were classifiedby morphological criteria depending on the symmetry(asymmetric or symmetric) of the synapse and the identityof the postsynaptic structure (spine or dendrite). Thus, thesynapses were divided into asymmetric (Asym) and sym-metric (Sym) subgroups. Asym and Sym synapses werefurther divided by postsynaptic target into asymmetricaxospinous (AS) or asymmetric axodendritic (AD), andsymmetric axospinous (SS) or symmetric axodendritic(SD) synapses. All synapses combined as well as the sub-types were also divided into unlabeled synaptic complexes(no immunolabeling in the axon terminal [AT], postsyn-aptic density [PSD], or postsynaptic structure) and labeledsynaptic complexes (at least one of the structures labeled:AT, PSD, or postsynaptic structure). Labeled synapticcomplexes were further broken down into those with la-beling in the presynaptic AT or the postsynaptic structure(PSD, spine, or dendrite). Synapses were counted usingpreviously described techniques (Kung et al., 1998).

RESULTS

Antibody specificity

The DISC1 antibody used in the present study (C1) hasbeen used previously and is well characterized (discussedin detail in Ozeki et al., 2003). Briefly, in human tissue,four major signals at 210, 130, 105, and 65 kDa wereobtained (Fig. 1). There is almost exclusively one majorsignal from cell extracts of HEK293 cells overexpressingthe exogenous DISC1—hemagglutinin (HA). The size ofthe signal from the HEK293 cells is almost identical tothat of the 105-kDa signal from human brain tissue,which is preabsorbed with the original antigen (DISC1amino acids 347–600) (Fig. 1). The preadsorption per-formed using 0.02 �g/ml was enough to almost totallyremove the three upper bands and to largely decrease thelower-molecular-weight band. A second preadsorption wasperformed with a very high concentration of protein (30�g/ml) which produced a total elimination of the bands.The 105-kDa signal, corresponding to the full lengthDISC1, was abolished in HEK293 cells (not shown).DISC1 immunolabeling was also eliminated in preadsorp-tion control tissue in the immunocytochemical prepara-tions (Fig. 2).

Light microscopy

In the gray matter, DISC1 had prominent staining inneuronal cell bodies located throughout most of the graymatter layers. Labeling was widespread and included bothpyramidal and nonpyramidal neurons (Figs. 2–5). Thispattern of labeling was similar in the three cortical re-gions that were examined among all the cases. Cellular

The Journal of Comparative Neurology. DOI 10.1002/cne

438 B. KIRKPATRICK ET AL.

staining was most notable in the somata and apical den-drites of pyramidal cells in layers 3, 5, and 6 (Figs. 2–5).Little cellular labeling was noted in Layer IV. Diffuseneuropil staining, as well as more concentrated particu-late staining usually not associated with cell bodies, wasalso seen in both the gray matter and the subcorticalwhite matter (Figs. 3–5). Immunoreactive cells, morpho-logically similar to neurons (Kirkpatrick et al., 1999), werescattered in the subcortical white matter (Figs. 3, 5). Glialstaining was also present, especially in superficial layers,although many glial cells were not labeled (Fig. 4).

Electron microscopy

The labeling at the EM level was similar among thethree cases that we examined. In the neuropil, denseimmunolabeling was found in several structures associ-ated with the synapse, namely, ATs, postsynaptic densi-ties (PSD), and dendritic spines, although not all suchstructures were labeled. Several combinations of labeledand unlabeled pre- and postsynaptic structures are shownin Figures 6 and 7 and quantified in Table 2. The mostcommon configuration was an unlabeled AT forming anasymmetric synapse with a spine that had immunoreac-tivity deposited on the PSD and throughout the spine.Unlabeled ATs also formed synapses with unlabeledspines and dendrites (neither PSD nor any other contents

of the postsynaptic structure was labeled). Labeled ATsformed synapses with labeled spines and dendrites; insome of these, only the PSD of the postsynaptic structurewas labeled. Immunolabeled ATs formed asymmetric andsymmetric synapses with equal frequency. Some dendriticshafts had immunolabeling on microtubules, but this wasnot common; in contrast, elements that appear to be ribo-somes in dendritic shafts were frequently labeled (Figs.7, 8).

In neuronal cell bodies (Fig. 9), some areas of the chro-matin were labeled, as well as areas on the outer nuclearmembrane (probably in ribosomes). Ribosomes in the cy-toplasm and the rough endoplasmic reticulum were alsofrequently labeled. Labeling was consistently absent fromthe Golgi apparatus, mitochondria, lysosomes, and mul-tivesicular bodies. There did not appear to be a differencein the staining pattern of cell bodies based on the mor-phology of the neuron (pyramidal versus nonpyramidal).In glial somata, the labeling was present on similar struc-tures as in neuronal somata (Fig. 10). Labeling in smallglial processes was not often apparent.

A breakdown of labeled synapses is shown in Table 2.Out of 328 total synapses, 43% of the synapses had at leastone labeled structure: AT, PSD, and/or another postsyn-aptic structure. By far the majority of the labeled synapseshad a labeled postsynaptic structure (113/140, or 81%),whereas 19% (27/140) only had a labeled AT. There was astriking difference in the labeling pattern between sym-metric and asymmetric synapses. For labeled asymmetricsynapses, 89% (109/123) had a labeled PSD, whereas 11%(14/123) had only a labeled AT. In contrast, 24% (4/17) ofsymmetric synapses had a labeled PSD, but 76% (13/17) ofthe ATs were labeled. Of the 328 ATs forming synapses,only 27 (8%) were labeled. These labeled terminals formedasymmetric (14/27) and symmetric (13/27) synapses withapproximately equal frequency. Of the 283 dendriticspines forming synapses, 43% (121/283) were labeled.

DISCUSSION

DISC1 immunoreactivity was prominent and wide-spread at the light microscopic level in the four Brodmannareas of the human neocortex that were studied. No obvi-ous differences in labeling pattern were seen across thecortical areas, suggesting a similar cellular function forDISC in these regions. In different cortical regions, cellbodies (especially pyramidal neurons), apical dendrites,and neuropil were prominently stained. To our knowledge,the present article is the first report on DISC1 localizationat the electron microscopic level. At the ultrastructurallevel in the prefrontal cortex, synaptic structures werefrequently, but not always, labeled. Many, but not allribosomes and rough endoplasmic reticulum were immu-noreactive in the cell body, whereas some microtubulesand structures that are probably ribosomes were labeledin dendrites. In contrast, the machinery of protein excre-tion and mitochondria were not immunoreactive. Many ofour findings are consistent with those in the literature,although the lack of mitochondrial labeling and the pres-ence of DISC1 at the PSD are unique to our study.

The specificity of our staining was confirmed with bothWestern blot and preadsorption experiments, and wassupported by the selective anatomical distribution wefound at the ultrastructural level. We and others haveshown multiple bands on Western blots using different

Fig. 1. Western blot results with DISC1 antibody. The publishedantibody against human DISC1 (amino acids 347–600) was used inthe present study (Ozeki et al., 2003). There is almost exclusively onemajor signal from cell extracts of human embryonic kidney (HEK) 293cells overexpressing the exogenous full length DISC1 with the hem-agglutinin (HA) tag. There are four major signals (210, 130, 105, and65 kDa) in human brain extracts from normal controls. The 105-kDasignal in the human brain is equivalent to the signal from the exog-enous DISC1 in HEK293 cells, which is preadsorbed with the originalDISC1 antigen. The signal (lanes marked by �) was abolished in thepreadsorption study (lanes marked by �). This latter preadsorptionwas exposed longer than the rest in order to ensure that the bandshad totally disappeared. The immunopurified DISC1 antibody at1:500 dilution was used.

The Journal of Comparative Neurology. DOI 10.1002/cne

439DISC1 ANATOMY IN HUMAN CORTEX

DISC1 antibodies (Miyoshi et al., 2003; Ozeki et al., 2003;Schurov et al., 2004). For example, antibody D27 raisedagainst residues 734–753 of the mouse sequence (Schurovet al., 2004), and the antibody in the present study raisedin rabbit against residues 347–600 (Ozeki et al., 2003)both produce four bands. In rat brain, the antibody used inMiyoshi et al. (2003) raised in rabbit against the C termi-nal produced two bands and antibody R1211 producedthree bands (James et al., 2004). The isoform at 105 kDacorresponds to the full-length DISC1, and has been repro-

ducibly detected in the four aforementioned studies. The130-kDa band detected by our antibody may be a brain-specific phosphorylated form of the protein at the 105-kDasignal, because preliminary data from our laboratory sug-gest that treatment of brain extracts with alkaline phos-phatase weakens the signal at 130 kDa. The identity ofthis band needs further clarification. The intensity of thesignal at 210 kDa can be reduced by the addition of anexcess amount of the reducing agent, such as�-mercaptoethanol, suggesting that this signal may reflect

Fig. 2. Preadsorption test of DISC1 immunoreactivity in Brodmann Area 46. A: Immunolabeling ofDISC1 without counterstain. Prominent pyramidal cell labeling (arrows) contrasts with the absence oflabeling with the preadsorption control (B). Both tissue samples were from the same subject (11). Scalebar � 250 �m

The Journal of Comparative Neurology. DOI 10.1002/cne

440 B. KIRKPATRICK ET AL.

dimmers of the DISC1 isoform at 105 kDa. The identity ofthe signal at 65 kDa is unknown, but may represent theproduct of protease activity and/or autolysis processes.Even when a protease inhibitor cocktail is used whilepreparing samples for Western blot, protein degradationin human tissue may occur because of the postmorteminterval and tissue dissection of frozen archived sampleswith a hot bone saw. As well, long-term storage of samplesdoes not totally prevent the occurrence of protease activity(Rouy et al., 2005).

To date there are only a few studies that have examinedthe in vivo localization of DISC1 in the mammalian brain.In the mouse, DISC1 mRNA is prominently expressed inthe dentate gyrus, and is also found in other areas of thehippocampus, cerebellum, cerebral cortex, and olfactorybulb (Austin et al., 2004; Ma et al., 2002; Miyoshi et al.,2003). These results were confirmed and extended in animmunocytochemical study in mouse, which also showedthat DISC1 was expressed in both excitatory and inhibi-tory cortical neurons (Schurov et al., 2004). In the green

Fig. 3. DISC1 immunoreactivity in Brodmann Area 46. A, B: Montage of the cortex immunolabeledfor DISC and counterstained with cresyl violet; layers (I–VI) indicated. Layer V overlaps at the star atthe bottom of A, and the top of B. Immunolabeled neurons are present throughout the gray matter andin the subcortical white matter (SCWM). Scale bar � 250 �m.

The Journal of Comparative Neurology. DOI 10.1002/cne

441DISC1 ANATOMY IN HUMAN CORTEX

monkey, Austin et al. (2003) found the expression ofDISC1 in the same regions, as well as in other regions,indicating a more extensive distribution in the monkeythan in the mouse. In the monkey study, expression wasmore prominent in the dentate gyrus and lateral septumthan in cerebral cortex, amygdala, hypothalamus, cerebel-lum, and the interpeduncular and subthalamic nuclei. Inhuman hippocampus, James et al. (2004) also describedwidespread DISC1 localization in multiple populations ofneurons. The mouse, monkey, and human cortex showed a

pattern consistent with our findings, in that immunoreac-tivity or mRNA was detected across all the cortical layerswith relatively little difference in expression across thelayers. In the aforementioned studies, most of the expres-sion was associated with neurons, although in the mon-key, expression consistent with glial localization wasfound in the molecular layer of the cerebellum. In thepresent study, we detected DISC1 labeling in both neu-rons and glial cells in the cortex. The reason for thisdifference may be attributable to a species difference

Fig. 4. DISC1 immunoreactivity in layers I, II, and III in Brod-mann Area 46. Cortex immunolabeled for DISC and counterstainedwith cresyl violet. A: Layers I and II. Both neurons (red arrows) andglia (blue arrowhead) are labeled. B: Note cell body labeling, espe-

cially in pyramidal cells (green arrows), which display prominentapical dendrite staining (red arrowheads). Note rich neuropil stain-ing, and that many glial cells are unlabeled. Scale bar � 50 �m.

The Journal of Comparative Neurology. DOI 10.1002/cne

442 B. KIRKPATRICK ET AL.

(mouse versus human), unique location in different brainregions (human hippocampus versus human cortex), ormethodological issues (the resolution of in situ hybridiza-tion versus that of EM).

The results in the present article are the first to describeDISC1 localization at the electron microscopic level. Manyof our ultrastructural findings are consistent with molec-ular evidence that DISC1 interacts with a number of otherproteins, including centrosome and cytoskeletal proteins,

proteins that localize receptors to membranes, and signaltransduction proteins (Miyoshi et al., 2004; Morris et al.,2003; Ozeki et al., 2003). The nuclear localization ofDISC1 that we found is not surprising, as the DISC1 genecontains the nuclear localization signal in its open readingframe, as well as the leucine-zipper motifs that are fre-quently found in nuclear proteins, especially transcrip-tional factors (Taylor et al., 2003). Moreover, bothSawamura et al. (2005) and James et al. (2004) found

Fig. 5. DISC1 immunoreactivity in layers V and VI and whitematter in Brodmann Area 46. Cortex immunolabeled for DISC andcounterstained with cresyl violet. A: Layer V. Pyramidal cell labeling(green arrows) is prominent. B: Layer VI and subcortical white matter

(SCWM). Neuronal labeling (green arrows) is prominent in layer VI,whereas diffuse granular labeling is present in the white matter. Noterich neuropil staining, and that most glial cells are unlabeled. Scalebar � 50 �m.

The Journal of Comparative Neurology. DOI 10.1002/cne

443DISC1 ANATOMY IN HUMAN CORTEX

DISC1 immunoreactivity in nuclear fractions using sub-cellular fractionation and cell culture, respectively. Ourobservation of DISC1 labeling of some microtubules isconsistent with other reports showing DISC1 interactionswith cytoskeletal elements (Brandon et al., 2005). Therelationship of DISC1 with microtubules causes abnormalneurite extension in development and could cause prob-lems with proper cellular movement of mitochondria inboth development and adult.

Some of our ultrastructural findings are unique to thepresent study. For example, we did not find labeling inmitochondria, whereas others have done so. Brandon et al.(2005) found DISC1 labeling in mitochondria in culturedmouse cortical neurons, whereas James et al. (2004) im-munolabeled mitochondria in cultured human neuroblas-toma cells and in postmortem human hippocampus. Inter-estingly, James et al. (2004) pointed out that thedistribution of DISC1 in their study was not completelyconsistent among the different antibodies that were used,nor was the subcellular localization always consistent be-tween different cell types. The observations of Brandon etal. (2005) and some of James et al. (2004) were made in acultured system, which could provide different results toours. Our tissue was labeled with a high amount of glu-taraldehyde, which can reduce the immunocytochemicalsignal, and could have resulted in false negatives. How-ever, the very discrete staining found in other organellesmakes this possibility unlikely. Another novel finding inour study is the ultrastructural localization of DISC1 inmany, but not all, PSD. Others have not detected DISC1in this location, raising the possible that DISC labeling of

the PSD is an artifact. Although the PSD frequently givesfalse-positive results, we believe this explanation is un-likely because not all PSDs are labeled even in the samemicrographs (Figs. 6, 7). Moreover, we found reelin local-ized in many PSDs in an ultrastructural study of humancortex, strengthening the validity of the presence of devel-opmentally important proteins at the PSD in the adulthuman brain (Roberts et al., 2005).

The results of the present study indicate that in adulthumans DISC1 is prevalent throughout the cortical layersin multiple populations of neurons, ATs, and postsynaptictargets. Although prominent labeling was detected instructures related to synaptic function, such as ATs andPSDs, it was not present in all such structures. In someinstances, only the PSD was labeled within the spine ordendrite, whereas at other synapses there was also moreextensive labeling in the postsynaptic element. The func-tional significance of these differences remains to be de-termined, but it is intriguing that in instances where onlythe PSD was labeled, such labeling occurred only in asym-metric synapses, typical of excitatory inputs. The presenceof DISC1 on the pre- and/or postsynaptic side of asymmet-ric axospinous synapses suggests the involvement ofDISC1 in corticocortical as well as thalamocortical connec-tions, because cortical and thalamic inputs make asym-metric synapses (DeFelipe et al., 2002; Peters, 2002). Thepresence of DISC1 in symmetric synapses suggests itsinvolvement in inhibitory local circuit connections withinthe cortex, as cortical interneurons form symmetric syn-apses (DeFelipe et al., 2002; Peters, 2002). The location ofDISC1 at many synapses is consistent with its localization

Fig. 6. Electron micrograph of DISC1 immunoreactivity in the neuropil in Brodmann Area 9.Immunolabeled axon terminals, spines, and postsynaptic densities (PSD) are indicated. Unlabeledstructures are also indicated. Note that only a subset of spines, PSDs, and axon terminals are immu-noreactive. Scale bar � 1 �m.

The Journal of Comparative Neurology. DOI 10.1002/cne

444 B. KIRKPATRICK ET AL.

in many types of cortical neurons, and suggests thatDISC1 may have a role in synaptic function in the adultbrain. The anatomical specificity may relate to synapticdifferences in neurotransmitters, long-term potentiation,or some other factor.

DISC1 appears to change location and function betweenthe developing and the mature brain. Other examples ofproteins that do this include reelin (Guidotti et al., 2000;Roberts et al., 2005; Weeber et al., 2002), growth associ-ated protein (Benowitz and Routtenberg, 1997; DiFiglia et

Fig. 7. Electron micrographs of DISC1 immunoreactivity in theneuropil in Brodmann Area 9. A: An immunoreactive dendrite with anexample of an unlabeled microtubule is indicated. Both labeled andunlabeled axon terminals, spines, and postsynaptic densities are in-dicated in the surrounding neuropil. B: Another example of a largeimmunoreactive dendrite showing labeled and unlabeled structuresthat are likely ribosomes. Labeled rough endoplasmic reticulum (rER)

and an unlabeled mitochondrion is also indicated. Note that the rERis expanded, rather than the typical flattened shape that is present inwell-fixed animal tissue; this observation has been noted before inhuman postmortem cortex and striatum and appears to be a postmor-tem artifact (Hutcherson and Roberts, 2005; Roberts et al., 2005).Scale bars � 1 �m (A) and 0.5 �m (B).

The Journal of Comparative Neurology. DOI 10.1002/cne

445DISC1 ANATOMY IN HUMAN CORTEX

al., 1990), neural cell adhesion molecule (Uryu et al.,1999), and brain derived neurotrophic factor (Drake et al.,1999). In the mouse, DISC1 is expressed from embryonicday 10 through adult life (Austin et al., 2004; Schurov etal., 2004). In some brain regions such as the neocortex andlimbic regions, DISC1 mRNA was detected both duringdevelopment and in the adult. In other areas, such as thebed nucleus of the stria terminalis, and the reticular andreuniens thalamic nuclei, DISC1 labeling was present

during development, but disappeared in adulthood (Aus-tin et al., 2004; Schurov et al., 2004). Observations duringdevelopment suggest that DISC1 is involved in neuriteoutgrowth (Kamiya et al., 2005; Miyoshi et al., 2003; Oz-eki et al., 2003), and neuronal migration (Brandon et al.,2004; Kamiya et al., 2005) based on the interaction ofDISC1 with NUDEL (Brandon et al., 2005; Hayashi et al.,2005). Clearly the location of DISC1 changes over the lifeof the mouse within certain brain regions, and may changeits subcellular distribution pattern as well. It is not knownwhether DISC1 changes its cellular or subcellular local-ization in the human between development and adult-hood, because a developmental study of DISC1 in thehuman brain has yet to be performed. However, our re-sults from the adult suggest a role for DISC1 in synapticfunction.

Postmortem studies of schizophrenic brain show abnor-malities in several developmental markers and moleculesinvolved in early neuronal migration (Akbarian et al.,1996; Guidotti et al., 2000; Impagnatiello et al., 1998;Kirkpatrick et al., 1999), which are consistent with theneurodevelopmental hypothesis of schizophrenia (Lewisand Levitt, 2002; Woods, 1998). Abnormal function ofDISC1 seems to be a risk factor for both schizophrenia andaffective disorders (Ekelund et al., 2001; Hodgkinson etal., 2004; Millar et al., 2000; St. Clair et al., 1990). Recent

Fig. 8. Electron micrographs of DISC1 immunoreactivity in den-dritic shafts in Brodmann Area 9. A: In this dendritic shaft (outlinedwith dashed lines), there is immunoreactivity in structures that areprobably ribosomes, although not all ribosomes are labeled (see clus-

ter of unlabeled ribosomes). Microtubules in this dendrite are notimmunolabeled. B: In another process (outlined with dashed line),some microtubules exhibit immunoreactivity. Scale bars � 1 �m.

TABLE 2. Labeled Synaptic Complexes by Synaptic Subtype

Total Asym Sym AS AD SS SD

Unlabeled synapses 188 156 32 145 11 17 15Labeled synapses 140 123 17 113 10 8 9Synaptic structure labeled

Axon terminal 27 14 13 11 3 6 7PSD, spine, or dendrite 113 109 4 102 7 2 2PSD only 30 30 0 25 5 0 0PSD and spine or dendrite 83 79 4 77 2 2 2

For this analysis, a total of 328 synapses were classified from the EM cases. Totalsynapses were classified into asymmetric (Asym) and symmetric (Sym) synapses. Asymand Sym synapses were further divided by postsynaptic target into asymmetric axospi-nous (AS) or axodendritic (AD), and symmetric axospinous (SS) or axondendritic (SD).The total synapses and subtypes are divided into unlabeled synaptic complex (no labelin terminal, PSD, or postsynaptic structure) and labeled synaptic complex (any or all ofthe terminal, PSD, or postsynaptic structure labeled). Labeled synaptic complexes arefurther broken down into those with labeling in the axon terminal or the postsynapticside (�PSD and the spine or dendrite).

The Journal of Comparative Neurology. DOI 10.1002/cne

446 B. KIRKPATRICK ET AL.

studies on schizophrenia have found abnormalities inDISC1 in subjects other than the Scottish family wherethe abnormality was first discovered (Burdick et al., 2005;Callicott et al., 2005; Sachs et al., 2005; Sawamura et al.,2005; Thompson et al., 2005b). It has been reported thatthe subcellular distribution of one of the DISC isoforms isaltered in orbitofrontal cortex in postmortem tissue fromsubjects with schizophrenia (Sawamura et al., 2005). Aparticular allelic variation in DISC1 not only increases therisk for schizophrenia, but alters the structure and func-tion of the hippocampus, an important brain region impli-cated in the disease (Callicott et al., 2005). The DISC1genotype also appears to be related to certain neurocogni-tive deficits in schizophrenia (Burdick et al., 2005; Cannonet al., 2005). The mutant form of DISC1 (Brandon et al.,2004; Ozeki et al., 2003) fails to bind to NUDEL, a proteinessential for cortical development, neuronal migration,and axon growth (Niethammer et al., 2000; Sasaki et al.,2000; Sweeney et al., 2001), Thus, by this mechanism,DISC1 inhibits neurite outgrowth in vitro (Kamiya et al.,2005; Ozeki et al., 2003) and interferes with normal cor-tical development in vivo (Kamiya et al., 2005). Disruptionof normal development may contribute to the reducedneuropil volume found in postmortem cortex in schizo-phrenia (Selemon and Goldman-Rakic, 1999), and in fron-tal cortical gray matter volume in schizophrenic patientswith the DISC1 gene (Cannon et al., 2005). Other aspectsof the neurobiology of DISC1 are consistent with a role inschizophrenia. The amount of DISC1 peaks in the mousebrain during the time of embryonic neurogenesis and

again during puberty (Austin et al., 2004; Schurov et al.,2004), two critical time points implicated in the patho-physiology of schizophrenia (Lewis and Levitt, 2002).DISC1 is also present in animals in many of the brainregions known to be abnormal in schizophrenia (Harrison,1999; McCarley et al., 1999; Pearlson and Marsh, 1999),and we found it in the human prefrontal and parietalcortices in the present study. The location of DISC1 inmultiple brain areas that subserve the functions disturbedin patients with schizophrenia suggests that DISC1 mayhave a role in some of the many neuropsychiatric prob-lems that afflict patients with the disease (Green et al.,2003; Millar et al., 2004; Sawa and Snyder, 2005). Al-though DISC1 serves as a gene of risk in a small percent-age of patients with schizophrenia, it may prove to be auseful entry point for studying the pathophysiology of thiscomplex disease (Millar et al., 2004; Sawa and Kamiya,2003; Sawa and Snyder, 2005).

ACKNOWLEDGMENTS

The authors thank and acknowledge the technical as-sistance of Joy Knickman Roche for the electron micros-copy and Koko Ishizuka for the Western blots. We are verygrateful to Dr. Emma Perez-Costas for reading the manu-script and offering expert interpretation of the Westernblot experiments. We also acknowledge and thank themembers of the Maryland Brain Collection: Drs. RobertConley and Carol Tamminga for diagnoses, TerriU’Prichard for permissions and interviews, and Dr. Sami

Fig. 9. Electron micrograph of DISC1 immunoreactivity in thenucleus and cell body of a neuron in Brodmann Area 9. Immunoreac-tivity is present in some but not all of the following organelles: areasof the chromatin, ribosomes in the cytoplasm, and the rough endo-

plasmic reticulum (rER). Note that the rER is expanded as explainedin Figure 7b. Immunolabeling is also present on parts of the outernuclear membrane, probably in ribosomes. Labeling is absent fromthe Golgi apparatus and a mitochondrion. Scale bar � 1 �m.

The Journal of Comparative Neurology. DOI 10.1002/cne

447DISC1 ANATOMY IN HUMAN CORTEX

Daoud for dissections. We also thank the Chief MedicalExaminer Dr. David Fowler and the Deputy Chief MedicalExaminer Dr. Jack Titus at the Office of the Chief MedicalExaminer in Baltimore for their support and cooperationin collecting the cases.

LITERATURE CITED

Akbarian S, Kim JJ, Potkin SG, Hetrick WP, Bunney WE Jr, Jones EG.1996. Maldistribution of interstitial neurons in prefrontal white matterof the brains of schizophrenic patients. Arch Gen Psychiatry 53:425–436.

Austin CP, Ky B, Ma L, Morris JA, Shughrue PJ. 2004. Expression ofdisrupted-in-schizophrenia-1, a schizophrenia-associated gene, isprominent in the mouse hippocampus throughout brain development.Neuroscience 124:3–10.

Austin CP, Ma L, Ky B, Morris JA, Shughrue PJ. 2003. DISC1 (Disruptedin Schizophrenia-1) is expressed in limbic regions of the primate brain.Neuroreport 14:951–954.

Badner JA, Gershon ES. 2002. Meta-analysis of whole-genome linkagescans of bipolar disorder and schizophrenia. Mol Psychiatry 7:405–411.

Benowitz LI, Routtenberg A. 1997. GAP-43: an intrinsic determinant ofneuronal development and plasticity. Trends Neurosci 20:84–91.

Berrettini WH. 2001. The human genome: susceptibility loci. Am J Psy-chiatry 158:865.

Blackwood DH, Fordyce A, Walker MT, St Clair DM, Porteous DJ, MuirWJ. 2001. Schizophrenia and affective disorders—cosegregation with atranslocation at chromosome 1q42 that directly disrupts brain-expressed genes: clinical and P300 findings in a family. Am J HumGenet 69:428–433.

Blackwood DH, Muir WJ. 2004. Clinical phenotypes associated withDISC1, a candidate gene for schizophrenia. Neurotox Res 6:35–41.

Brandon NJ, Handford EJ, Schurov I, Rain JC, Pelling M, Duran-Jimeniz

B, Camargo LM, Oliver KR, Beher D, Shearman MS, Whiting PJ. 2004.Disrupted in Schizophrenia 1 and Nudel form a neurodevelopmentallyregulated protein complex: implications for schizophrenia and othermajor neurological disorders. Mol Cell Neurosci 25:42–55.

Brandon NJ, Schurov I, Camargo LM, Handford EJ, Duran-Jimeniz B,Hunt P, Millar JK, Porteous DJ, Shearman MS, Whiting PJ. 2005.Subcellular targeting of DISC1 is dependent on a domain independentfrom the Nudel binding site. Mol Cell Neurosci 28(4):613–24.

Burdick KE, Hodgkinson CA, Szeszko PR, Lencz T, Ekholm JM, Kane JM,Goldman D, Malhotra AK. 2005. DISC1 and neurocognitive function inschizophrenia. Neuroreport 16:1399–1402.

Callicott JH, Straub RE, Pezawas L, Egan MF, Mattay VS, Hariri AR,Verchinski BA, Meyer-Lindenberg A, Balkissoon R, Kolachana B, Gold-berg TE, Weinberger DR. 2005. Variation in DISC1 affects hippocam-pal structure and function and increases risk for schizophrenia. ProcNatl Acad Sci USA 102:8627–8632.

Cannon TD, Hennah W, van Erp TG, Thompson PM, Lonnqvist J, Hut-tunen M, Gasperoni T, Tuulio-Henriksson A, Pirkola T, Toga AW,Kaprio J, Mazziotta J, Peltonen L. 2005. Association of DISC1/TRAXhaplotypes with schizophrenia, reduced prefrontal gray matter, andimpaired short- and long-term memory. Arch Gen Psychiatry 11:205–13.

DeFelipe J, Alonso-Nanclares L, Arellano JI. 2002. Microstructure of theneocortex: comparative aspects. J Neurocytol 31:299–316.

Detera-Wadleigh SD, Badner JA, Berrettini WH, Yoshikawa T, Goldin LR,Turner G, Rollins DY, Moses T, Sanders AR, Karkera JD, Esterling LE,Zeng J, Ferraro TN, Guroff JJ, Kazuba D, Maxwell ME, Nurnberger JIJr, Gershon ES. 1999. A high-density genome scan detects evidence fora bipolar-disorder susceptibility locus on 13q32 and other potential locion 1q32 and 18p11.2. Proc Natl Acad Sci USA 96:5604–5609.

Devon RS, Anderson S, Teague PW, Burgess P, Kipari TM, Semple CA,Millar JK, Muir WJ, Murray V, Pelosi AJ, Blackwood DH, Porteous DJ.2001. Identification of polymorphisms within Disrupted in Schizophre-nia 1 and Disrupted in Schizophrenia 2, and an investigation of their

Fig. 10. Electron micrograph of DISC1 immunoreactivity in thenucleus and cell body in a glial cell in Brodmann Area 9. Immunore-activity is present in some areas of the chromatin (Labeled Chroma-tin) but not others (Unlabeled Chromatin). Immunolabeling is also

present on the nuclear membrane, as well as on ribosome-like struc-tures in the cytoplasm and the rough endoplasmic reticulum (rER).Scale bar � 1 �m.

The Journal of Comparative Neurology. DOI 10.1002/cne

448 B. KIRKPATRICK ET AL.

association with schizophrenia and bipolar affective disorder. Psychi-atr Genet 11:71–78.

DiFiglia M, Roberts RC, Benowitz LI. 1990. Immunoreactive GAP-43 in theneuropil of adult rat neostriatum: localization in unmyelinated fibers,axon terminals, and dendritic spines. J Comp Neurol 302:992–1001.

Drake CT, Milner TA, Patterson SL. 1999. Ultrastructural localization offull-length trkB immunoreactivity in rat hippocampus suggests multi-ple roles in modulating activity-dependent synaptic plasticity. J Neu-rosci 19:8009–8026.

Ekelund J, Hovatta I, Parker A, Paunio T, Varilo T, Martin R, Suhonen J,Ellonen P, Chan G, Sinsheimer JS, Sobel E, Juvonen H, Arajarvi R,Partonen T, Suvisaari J, Lonnqvist J, Meyer J, Peltonen L. 2001.Chromosome 1 loci in Finnish schizophrenia families. Hum Mol Genet10:1611–1617.

Green AI, Canuso CM, Brenner MJ, Wojcik JD. 2003. Detection andmanagement of comorbidity in patients with schizophrenia. PsychiatrClin North Am 26:115–139.

Guidotti A, Auta J, Davis JM, Di-Giorgi-Gerevini V, Dwivedi Y, GraysonDR, Impagnatiello F, Pandey G, Pesold C, Sharma R, Uzunov D, CostaE, DiGiorgi Gerevini V. 2000. Decrease in reelin and glutamic aciddecarboxylase67 (GAD67) expression in schizophrenia and bipolar dis-order: a postmortem brain study. Arch Gen Psychiatry 57:1061–1069.

Harrison PJ. 1999. The neuropathology of schizophrenia: a critical reviewof the data and their interpretation. Brain 122:593–624.

Harrison PJ, Heath PR, Eastwood SL, Burnet PWJ, McDonald B, PearsonRCA. 1995. The relative importance of premortem acidosis and post-mortem interval for human brain gene expression studies: selectivemRNA vulnerability and comparison with their encoded proteins. Neu-rosci Lett 200:151–154.

Harrison PJ, Owen MJ. 2003. Genes for schizophrenia? Recent findingsand their pathophysiological implications. Lancet 361:417–419.

Hayashi MA, Portaro FC, Bastos MF, Guerreiro JR, Oliveira V, Gorrao SS,Tambourgi DV, Sant’Anna OA, Whiting PJ, Camargo LM, Konno K,Brandon NJ, Camargo AC. 2005. Inhibition of NUDEL (nuclear distri-bution element-like)-oligopeptidase activity by disrupted-in-schizophrenia 1. Proc Natl Acad Sci USA 102(10):3828–3833.

Hennah W, Varilo T, Kestila M, Paunio T, Arajarvi R, Haukka J, Parker A,Martin R, Levitzky S, Partonen T, Meyer J, Lonnqvist J, Peltonen L,Ekelund J. 2003. Haplotype transmission analysis provides evidence ofassociation for DISC1 to schizophrenia and suggests sex-dependenteffects. Hum Mol Genet 12:3151–3159.

Hodgkinson CA, Goldman D, Jaeger J, Persaud S, Kane JM, Lipsky RH,Malhotra AK. 2004. Disrupted in schizophrenia 1 (DISC1): associationwith schizophrenia, schizoaffective disorder, and bipolar disorder.Am J Hum Genet 75:862–872.

Hutcherson L, Roberts RC. 2005. The immunocytochemical localization ofsubstance P in the human striatum: a postmortem ultrastructuralstudy. Synapse 57:191–201.

Impagnatiello F, Guidotti AR, Pesold C, Dwivedi Y, Caruncho H, Pisu MG,Uzunov DP, Smalheiser NR, Davis JM, Pandey GN, Pappas GD, Tu-eting P, Sharma RP, Costa E. 1998. A decrease of reelin expression asa putative vulnerability factor in schizophrenia. Proc Natl Acad SciUSA 95:15718–15723.

James R, Adams RR, Christie S, Buchanan SR, Porteous DJ, Millar JK.2004. Disrupted in Schizophrenia 1 (DISC1) is a multicompartmental-ized protein that predominantly localizes to mitochondria. Mol CellNeurosci 26:112–122.

Johnston NL, Cerevnak J, Shore AD, Torrey EF, Yolken RH. 1997. Mul-tivariate analysis of RNA levels from postmortem human brains asmeasured by three different methods of RT-PCR. Stanley Neuropathol-ogy Consortium. J Neurosci Methods 77:83–92.

Kamiya A, Kubo K, Tomoda T, Takaki M, Youn R, Ozeki Y, Sawamura N,Park U, Kudo C, Okawa M, Ross CA, Hatten ME, Nakajima K, Sawa A.2005. A schizophrenia-associated mutation of DISC1 perturbs cerebralcortex development. Nat Cell Biol 7:1067–1078.

Kirkpatrick B, Cascella N, Ozeki Y, Roberts R, Sawa A. 2003. DISC1-immunoreactive cells in the human brain. Schizophr Res 60:72.

Kirkpatrick B, Cascella N, Roberts R, Ozeki Y, Sawa A. 2002. DISC1-immunoreactive cells in the human neocortex. Am College Neuropsy-chopharmacol 41:347.

Kirkpatrick B, Conley RC, Kakoyannis A, Reep RL, Roberts RC. 1999.Interstitial cells of the white matter in the inferior parietal cortex inschizophrenia: an unbiased cell-counting study. Synapse 34:95–102.

Kung L, Force M, Chute DJ, Roberts RC. 1998. Immunocytochemicallocalization of tyrosine hydroxylase in the human striatum: a postmor-tem ultrastructural study. J Comp Neurol 390:52–62.

Lewis DA, Levitt P. 2002. Schizophrenia as a disorder of neurodevelop-ment. Annu Rev Neurosci 25:409–432.

Ma L, Liu Y, Ky B, Shughrue PJ, Austin CP, Morris JA. 2002. Cloning andcharacterization of Disc1, the mouse ortholog of DISC1 (Disrupted-in-Schizophrenia 1). Genomics 80:662–672.

McCarley RW, Wible CG, Frumin M, Hirayasu Y, Levitt JJ, Fischer IA,Shenton ME. 1999. MRI anatomy of schizophrenia. Biol Psychiatry45:1099–1119.

Millar JK, Christie S, Anderson S, Lawson D, Hsiao-Wei Loh D, Devon RS,Arveiler B, Muir WJ, Blackwood DH, Porteous DJ. 2001. Genomicstructure and localisation within a linkage hotspot of Disrupted InSchizophrenia 1, a gene disrupted by a translocation segregating withschizophrenia. Mol Psychiatry 6:173–178.

Millar JK, James R, Brandon NJ, Thomson PA. 2004. DISC1 and DISC2:discovering and dissecting molecular mechanisms underlying psychi-atric illness. Ann Med 36(5):367–378.

Millar JK, Wilson-Annan JC, Anderson S, Christie S, Taylor MS, SempleCA, Devon RS, Clair DM, Muir WJ, Blackwood DH, Porteous DJ. 2000.Disruption of two novel genes by a translocation co-segregating withschizophrenia. Hum Mol Genet 9:1415–1423.

Miyoshi K, Asanuma M, Miyazaki I, Diaz-Corrales FJ, Katayama T, To-hyama M, Ogawa N. 2004. DISC1 localizes to the centrosome by bind-ing to kendrin. Biochem Biophys Res Commun 317:1195–1199.

Miyoshi K, Honda A, Baba K, Taniguchi M, Oono K, Fujita T, Kuroda S,Katayama T, Tohyama M. 2003. Disrupted-In-Schizophrenia 1, a can-didate gene for schizophrenia, participates in neurite outgrowth. MolPsychiatry 8:685–694.

Morris JA, Kandpal G, Ma L, Austin CP. 2003. DISC1 (Disrupted-In-Schizophrenia 1) is a centrosome-associated protein that interacts withMAP1A, MIPT3, ATF4/5 and NUDEL: regulation and loss of interac-tion with mutation. Hum Mol Genet 12:1591–1608.

Niethammer M, Smith DS, Ayala R, Peng J, Ko J, Lee MS, Morabito M,Tsai LH. 2000. NUDEL is a novel Cdk5 substrate that associates withLIS1 and cytoplasmic dynein. Neuron 28:697–711.

Ozeki Y, Tomoda T, Kleiderlein J, Kamiya A, Bord L, Fujii K, Okawa M,Yamada N, Hatten ME, Snyder SH, Ross CA, Sawa A. 2003. Disrupted-in-Schizophrenia-1 (DISC-1): mutant truncation prevents binding toNudE-like (NUDEL) and inhibits neurite outgrowth. Proc Natl AcadSci USA 100:289–294.

Pearlson GD, Marsh L. 1999. Structural brain imaging in schizophrenia: aselective review. Biol Psychiatry 46:627–649.

Peters A. 2002. Examining neocortical circuits: some background and facts.J Neurocytol 31:183–193.

Roberts RC, Knickman J. 2002. The ultrastructural organization of thepatch matrix compartments in the human striatum. J Comp Neurol452:128–138.

Roberts RC, Xu L, Roche JK, Kirkpatrick B. 2005. The ultrastructurallocalization of reelin in the cortex in postmortem human brain. J CompNeurol 482:294–308.

Rouy D, Ernens I, Jeanty C, Wagner DR. 2005. Plasma storage at 08-degrees C does not protect matrix metalloproteinase from degradation.Anal Biochem 338:294–298.

Sachs NA, Sawa A, Holmes SE, Ross CA, Delisi LE, Margolis RL. 2005. Aframeshift mutation in Disrupted in Schizophrenia 1 in an Americanfamily with schizophrenia and schizoaffective disorder. Mol Psychiatry10:758–764.

Sasaki S, Shionoya A, Ishida M, Gambello MJ, Yingling J, Wynshaw-BorisA, Hirotsune S. 2000. A LIS1/NUDEL/cytoplasmic dynein heavy chaincomplex in the developing and adult nervous system. Neuron 28:681–696.

Sawa A, Kamiya A. 2003. Elucidating the pathogenesis of schizophrenia.BMJ 327:632–633.

Sawa A, Snyder SH. 2005. Two genes link two distinct psychoses. Science310:1128–1129.

Sawamura N, Sawamura-Yamamoto T, Ozeki Y, Ross CA, Sawa A. 2005. Aform of Disrupted-in-Schizophrenia-1 (DISC1) enriched in the nucleushas altered subcellular distribution in schizophrenia brains. Proc NatlAcad Sci USA 102:1187–1192.

Schurov IL, Handford EJ, Brandon NJ, Whiting PJ. 2004. Expression ofdisrupted in schizophrenia 1 (DISC1) protein in the adult and devel-oping mouse brain indicates its role in neurodevelopment. Mol Psychi-atry 9:1100–1110.

Selemon LD, Goldman-Rakic PS. 1999. The reduced neuropil hypothesis: acircuit based model of schizophrenia. Biol Psychiatry 45:17–25.

The Journal of Comparative Neurology. DOI 10.1002/cne

449DISC1 ANATOMY IN HUMAN CORTEX

Shirts BH, Nimgaonkar V. 2004. The genes for schizophrenia: finally abreakthrough? Curr Psychiatry Rep 6:303–312.

St Clair D, Blackwood D, Muir W, Carothers A, Walker M, Spowart G,Gosden C, Evans HJ. 1990. Association within a family of a balancedautosomal translocation with major mental-illness. Lancet 336:13–16.

Sweeney KJ, Prokscha A, Eichele G. 2001. NudE-L, a novel Lis1-interacting protein, belongs to a family of vertebrate coiled-coil pro-teins. Mech Dev 101:21–33.

Taylor MS, Devon RS, Millar JK, Porteous DJ. 2003. Evolutionary con-straints on the Disrupted in Schizophrenia locus. Genomics 81:67–77.

Thomson PA, Harris SE, Starr JM, Whalley LJ, Porteous DJ, Deary IJ.2005a. Association between genotype at an exonic SNP in DISC1 andnormal cognitive aging. Neurosci Lett 389:41–45.

Thomson PA, Wray NR, Millar JK, Evans KL, Hellard SL, Condie A, MuirWJ, Blackwood DH, Porteous DJ. 2005b. Association between theTRAX/DISC locus and both bipolar disorder and schizophrenia in theScottish population. Mol Psychiatry 10:657–668.

Uryu K, Butler AK, Chesselet MF. 1999. Synaptogenesis and ultrastruc-tural localization of the polysialylated neural cell adhesion molecule inthe developing striatum. J Comp Neurol 405:216–232.

Weeber EJ, Beffert U, Jones C, Christian JM, Forster E, Sweatt JD,Herz J. 2002. Reelin and ApoE receptors cooperate to enhancehippocampal synaptic plasticity and learning. J Biol Chem 277:39944 –39952.

Woods BT. 1998. Is schizophrenia a progressive neurodevelopmental dis-order? Toward a unitary pathogenetic mechanism. Am J Psychiatry155:1661–1670.

Xu L, Kirkpatrick B, Roberts R, Cascella N, Ozeki Y, Sawa A. 2003.DISC-1-Immunoreactive cells in the human brain. Soc Neurosci Abstr29:317.6.

Zhang X, Tochigi M, Ohashi J, Maeda K, Kato T, Okazaki Y, Kato N,Tokunaga K, Sawa A, Sasaki T. 2005. Association study of the DISC1/TRAX locus with schizophrenia in a Japanese population. SchizophrRes 79:175–180.

The Journal of Comparative Neurology. DOI 10.1002/cne

450 B. KIRKPATRICK ET AL.