OUP user menu

Differences Between the Pattern of Developmental Abnormalities in Autism Associated With Duplications 15q11.2-q13 and Idiopathic Autism

Jerzy Wegiel PhD, N. Carolyn Schanen MD, PhD, Edwin H. Cook MD, Marian Sigman MD, W. Ted Brown MD, PhD, Izabela Kuchna MD, PhD, Krzysztof Nowicki MD, Jarek Wegiel MSc, Humi Imaki PhD, Shuang Yong Ma MD, PhD, Elaine Marchi MSc, Teresa Wierzba-Bobrowicz MD, PhD, Abha Chauhan PhD, Ved Chauhan PhD, Ira L. Cohen PhD, Eric London MD, Michael Flory PhD, Boleslaw Lach MD, PhD, Thomas Wisniewski MD
DOI: http://dx.doi.org/10.1097/NEN.0b013e318251f537 382-397 First published online: 1 May 2012

Abstract

The purposes of this study were to identify differences in patterns of developmental abnormalities between the brains of individuals with autism of unknown etiology and those of individuals with duplications of chromosome 15q11.2-q13 (dup[15]) and autism and to identify alterations that may contribute to seizures and sudden death in the latter. Brains of 9 subjects with dup 15, 10 with idiopathic autism, and 7controls were examined. In the dup(15) cohort, 7 subjects (78%) had autism, 7 (78%) had seizures, and 6 (67%) had experienced sudden unexplained death. Subjects with dup(15) autism were microcephalic, with mean brain weights 300 g less (1,177 g) than those of subjects with idiopathic autism (1,477 g; p < 0.001). Heterotopias in the alveus, CA4, and dentate gyrus and dysplasia in the dentate gyrus were detected in 89% of dup(15) autism cases but in only 10% of idiopathic autism cases (p < 0.001). By contrast, cerebral cortex dysplasia was detected in 50% of subjects with idiopathic autism and in no dup(15) autism cases (p < 0.04). The different spectrum and higher prevalence of developmental neuropathologic findings in the dup(15) cohort than in cases with idiopathic autism may contribute to the high risk of early onset of seizures and sudden death.

Key Words
  • Autism
  • Chromosome 15q11.2-q13 duplication
  • Developmental brain alterations
  • Seizures
  • Sudden unexpected death

Introduction

Autism is the most severe form of autism spectrum disorder (ASD) and is characterized by qualitative impairments in reciprocal social interactions, qualitative impairments in verbal and nonverbal communication, restricted repetitive and stereotyped patterns of behavior, interests and activities, and onset before the age of 3 years (1). Autism is heterogeneous, both phenotypically and etiologically. In 44.6% of affected children, autism is associated with cognitive impairment as defined by intelligence quotient scores of less than 70 (2). Epilepsy is a comorbid complication diagnosed in up to 30% of individuals with autism (3). In 90% to 95% of cases, the etiology of autism is not known (id-iopathic or nonsyndromic autism) (4, 5). Twin and family studies have indicated both a strong and a moderate heritability for ASD (68).

Approximately 5% to 10% of individuals with an ASD have an identifiable genetic etiology corresponding to a known single gene disorder, for example, fragile X syndrome or chromosomal rearrangements, including maternal duplication of 15q11-q13. Supernumerary isodicentric chromosome 15 [idic(15)] (formerly designated as inverted duplication 15) is a relatively common genetic anomaly that most often leads to tetrasomy or mixed trisomy/tetrasomy of the involved segments and arises from a U-type crossover between a series of low copy repeats (LCRs) located on the proximal long arm. Small heterochromatic idic(15) chromosomes, which do not include the 15q11-q13 region, are often familial and are not associated with a clinical abnormality (9, 10). The symptoms in people with idic(15) markers correlate with the extent of duplication of the Prader-Willi syndrome/Angelman syndrome critical region (15q11-q13) (10, 11). Larger supernumerary idic(15) chromosomes, which include the imprinted chromosome 15q11-q13 region, are associated with a cluster of clinical features that include intellectual deficits (IDs), seizures, autistic behavior, hypotonia, hyperactivity, and irritability (12). Duplications of chromosome 15q11-q13 account for approximately 0.5% to 3% of ASD and may be the most prevalent cytogenetic aberration associated with autism in most studies. These duplications range from 4 to 12 Mb and may occur either through generation of supernumerary idic(15) chromosomes or as interstitial duplications and triplications. For interstitial duplications, maternal origin confers a higher risk for an abnormal phenotype (13, 14), and most of the reported chromosome 15 duplications (dup[15]) are maternally derived. A small number of subjects with duplications of paternal origin have been variously reported as being unaffected (13, 1517), affected but without ASD (16, 18), or affected with ASD (19). Interstitial triplications (int trp[15]) are relatively rare but have invariably been associated with a severe phenotype, including ID, ASD, or autistic features, and frequently with seizures. The parent-of-origin effect is not evident in the reported cases of int trp(15), with both maternal and paternal triplications associated with poor outcome.

Clinical studies indicate that most individuals diagnosed with dup(15) of maternal origin fulfill the criteria for the diagnosis of autism. In the first clinical reports, 24 individuals with idic(15) and autism were identified using standardized criteria of the Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised (13, 15, 2024). In several other studies of subjects with idic(15) chromosomes, autism was clinically diagnosed, although without the application of standardized measures of autism (2527). Application of the Gilliam Autism Rating Scale (28) to another idic(15) cohort confirmed a high prevalence of autism; 20 (69%) of 29 children and young adults with idic(15) had an ASD (29).

The link between the extent of genetic duplication and clinical phenotype has not yet been determined, but genomic and functional profiling provides insights into the direct and indirect effects of the copy number gains associated with chromosome 15 duplications. The γ-aminobutyric acid type A (GABAA) receptor subunit genes (α5, β, and γ3) that have been implicated in the etiology of autism (30) are located in the susceptibility segment of duplicated chromosome 15(3134). A map of parent-of-origin-specific epigenetic modifications suggests that this imprinted locus may have links not only with autism but also with other psychiatric phenotypes (35). Differential methylations in the 15q11-q13 region, including the GABAA gene (30, 3638), may contribute to epigenetic modifications and a broader clinical phenotypes in dup(15)/autism. Several other genes located in or near the 15q11-q13 region may contribute to a variable phenotype of autism, including a gene for juvenile epilepsy located near D15S165 (39) and a locus for agenesis of the corpus callosum (40).

We hypothesized that the neuropathology of autism with dup(15) differs from that of idiopathic autism and that it would provide an explanation for the high prevalence of seizures and associated sudden death in the dup(15) cohort. The aim of this comparative postmortem study of the brains of individuals diagnosed with idic(15) or int trp(15) (collectively referred to as dup[15]) and of individuals with id-iopathic autism was to identify common neuropathologic developmental defects for both cohorts and the patterns of changes distinguishing dup(15) from idiopathic autism. The dup(15) cohort examined exhibited a strikingly high prevalence of epilepsy, including intractable epilepsy, and a high rate of sudden unexpected death in childhood and early adulthood. Therefore, the second aim of the study was to identify patterns of neuropathologic changes that may contribute to epilepsy and sudden death in the dup(15) cohort.

Materials and Methods

The cohort of subjects diagnosed with dup(15) consisted of 9 subjects (range, 9–39 years), including 5 males (55%) and 4 females (45%). The cohort with autism consisted of 10 subjects (range, 2–52 years), including 9 males (90%) and 1 female (10%). The control cohort consisted of 7 subjects from 8 to 47 years, including 4 males (43%) and 3 females (57%). The brain of 1 subject diagnosed with dup(15) was excluded because of very severe autolytic changes, and the brain of 1 control subject was excluded because of lack of information about cause of death. The mean postmortem interval varied from 23.9 hours in the dup(15) cohort, to 19.6 hours in the idiopathic autism cohort, and 13.4 hours in the control group (Table 1).

View this table:
TABLE 1

Material Examined

GroupCase No.SexAge, yCause of DeathPMI, hHemBrain Weight, g
dup(15) autism1M9SUDEP13.6R1,130
2M10SUDEP17.7R1,070
3M11SUDEP10.5R1,540
4F15SUDEP (suspected)24.0LR1,141
5F15PneumoniaLR1,125
6M20Cardiopulmonary arrest (choking on food)28.1L1,190
7M24Pneumonia36.3L1,200
8F26SUDEP (suspected)28.6R1,310
9F39 Mean (SD)SUDEP32.8 23.9 (9.2)R890 1,177 (177)
Idiopathic autism1M2Asphyxia (drowning)4.0R1,328
2F5Asphyxia (drowning)33.0R1,360
3M5Asphyxia (drowning)25.5R1,560
4M8Asthma attack13.8R1,740
5M9SUDC3.7R1,690
6M11Asphyxia (drowning)R1,400
7M28Seizure-related43.0R1,580
8M32Brain tumor (glioblastoma multiforme)L1,260
9M51Heart failure22.2R1,530
10M52 Mean (SD)Heart failure11.5 19.6 (14.0)L1,324 1,477 (166)
Control1F8Rejection of cardiac transplant20.0R1,340
2M14Asphyxia (hanging)5.0R1,420
3M14Multiple traumatic injuries16.0R1,340
4M32Heart failure14.0R1,401
5F33Bronchopneumonia6.0L1,260
6F43Sepsis10.1L1,350
7M47 Mean (SD)Myocardial infarct23.0 13.4 (6.8)L1,450 1,366 (63)

PMI, postmortem interval; Hem, hemisphere; R, right; L, left; SUDEP, sudden unexpected death of subject with known epilepsy: SUDC, sudden unexpected death in childhood.

    Clinical and Genetic Characteristics

    Psychological, behavioral, neurologic, and psychiatric evaluation reports and reports by medical examiners and pathologists were the source of the medical records of the examined postmortem subjects. Medical records were obtained after consent for release of information from the subjects' parents. In most subjects diagnosed as being affected by dup(15) and idiopathic autism, the Autism Diagnostic Interview — Revised (ADI-R) was administered to the donor family after the subject's death as a standardized assessment tool to confirm an autism diagnosis (41). In addition, 6 of the subjects with dup(15) chromosomes were enrolled in a study of molecular contributors to the phenotype. The study was approved by the institutional review boards of the University of California, Los Angeles, and Nemours Biomedical Research. Before their deaths, 6 subjects (Cases 1–4, 7, and 8) had undergone behavioral and cognitive testing using the ADI-R, the Autism Diagnostic Observation Scale – Generic (ADOS-G) (42), and the Mullen Scales of Early Learning (43, 44) or the Stanford-Binet Intelligence Scales (45). Age at evaluation ranged from 45 to 251 months.

    Molecular genetic evaluation using antemortem peripheral blood samples and lymphoblast cell lines for 8 of the dup(15) cases included genotyping with 19 to 33 short tandem repeat polymorphisms from chromosome 15, Southern blot analysis of dosage with 5 to 12 probes, and measurement of the methylation state at SNRPN exon α, as described (46). In addition, array comparative genomic hybridization was performed, using a custom bacterial artificial chromosome array (47). Morphology of the duplication was confirmed by fluorescent in situ hybridization using 5 to 8 probes that detect sequences on chromosome 15q11-q14 (46).

    Tissue Preservation for Neuropathologic Study

    One brain hemisphere from each subject was fixed in 10% buffered formalin, dehydrated in a graded series of ethanol, infiltrated with polyethylene glycol 400 (no. 807 485; Merck, Whitehouse Station, NJ) and embedded in fresh polyethylene glycol 1000 (48) and stored at 4∘C. Tissue blocks were then cut at a temperature of 18∘C into 50-μm-thick serial sections identified with a number and stored in 70% ethyl alcohol at room temperature. Free-floating hemispheric sections were stained with cresyl violet and mounted with Acrytol. To increase the probability of detection of small focal developmental defects, on average, 140 hemispheric cresyl violet-stained sections at a distance of 1.2 mm were examined per case.

    Tissue acquisition for this project was based on individual tissue transfer agreements between the project's principal investigator and several tissue banks, including (i) the National Institute of Child Health and Human Development Brain and Tissue Bank for Developmental Disorders at the University of Maryland School of Medicine, Baltimore, MD; (ii) the Harvard Brain Tissue Resource Center, McLean Hospital, Belmont, MA; and (iii) the Brain and Tissue Bank for Developmental Disabilities and Aging of the New York State Institute for Basic Research in Developmental Disabilities. The institutional review board of the Institute for Basic Research in Developmental Disabilities approved the methods applied in this study.

    Statistical Analysis

    Differences among groups in the presence or absence of heterotopias and dysplasias were examined using the Fisher exact test. Differences in the numbers of abnormalities were examined using the Mann-Whitney U (Wilcoxon signed ranks) test or (for comparisons of all 3 groups) the Kruskal-Wallis one-way analysis of variance (an extension of the U test), with exact probabilities computed using version 12 of the SPSS statistical package (SPSS for Windows, Release 12.0, 2003; SPSS, Inc, Chicago, IL). Brain weights were compared in oneway analysis of variances controlled for age.

    Results

    Genetic Characteristics

    For 8 subjects (Cases 1–5 and 7–9) in the dup(15) cohort, duplication chromosomes was characterized using a combination of genotyping, fluorescent in situ hybridization, Southern blot, and array comparative genomic hybridization with lymphoblasts generated from antemortem blood samples (Table 2). All were maternally derived; 7 of these subjects were tetrasomic for the imprinted region between brake points (BP)2 and BP3, although the BP involved was variable. The idic(15) present in cells from Case 1 was generated by an exchange between copies of LCR3, causing tetrasomy that extended only to BP3. Four subjects (Cases 2, 4, 5, and 8) had the most common form of idic(15) chromosomes arising by nonallelic homologous recombination (NAHR) between BP4 and BP5, leading to tetrasomy of the region between the p-arm and BP4, with trisomy for the interval between BP4 and BP5. Another subject (Case 7) had a similar idic(15) chromosome but also carried a deletion between BP1 and BP2 on 1 homolog of chromosome 15 (Subject 00-03) (47). Case 3 had a complex tricentric supernumerary chromosome arising from NAHR between multiple copies of BP3, rendering him hexasomic for the region between the centromere and BP3 (46). One subject (Case 9) had an int trp(15) chromosome that led to tetrasomy between BP1 and BP4 and trisomy for the interval between the fourth and fifth LCRs, similar to the dosage seen in the BP4:BP5 idic(15) chromosomes.

    View this table:
    TABLE 2.

    Chromosome 15 Abnormalities in the dup(15) Cohort

    Case #Chromosomal AlterationsPrader-Willi Syndrome/Angelman Syndrome Critical Region (PWS/ASCR)Parental Origin of Abnormality
    147,XY,+idic(15)(q13;q13); Idic(15) arising from BP3:BP3 exchange.TetrasomyMaternal
    Subject 02–18, Wang et al (47)
    247,XY,+idic(15)(q13;q13); Idic(15) arising from BP4:BP5 exchange. Subject 01–22, Wang et al (47)TetrasomyMaternal
    347,XY,+der(15)(pter>q13::q13>cen>q13::q13>pter). Tricentric chromosome 15 arising from BP3:BP3 exchanges.HexasomyMaternal
    Case 2, Mann et al (46); Subject 00–29, Wang et al (47)
    447,XX,+idic(15)(q13;q13); Idic(15) arising from BP4:BP5 exchange. Subject 99-93, Wang et al (47)TetrasomyMaternal
    547,XX,+der(15)(q13;q13); Idic(15) arising from BP4:BP5 exchangeTetrasomyMaternal
    6IDIC15 (2 extra copies of region from beginning of array to BP4)TetrasomyNot determined
    747,XY,del(15)(q11.2)+idic(15)(q13;q13); Idic(15) arising from BP4:BP5 exchange; deletion of BP1:BP2 on 1 homolog of chromosome 15.TetrasomyMaternal
    Subject 00–03, Wang et al (47)
    847,XX,+idic(15)(q13;q13); Idic(15) arising from BP4:BP5 exchange.TetrasomyMaternal
    Subject 99-27, Wang et al (47)
    946,XX,trp(15)(q11.2q13). Subject 02–9, Wang et al (47)TetrasomyMaternal

    PWS/ASCR, Prader-Willi syndrome/Angelman syndrome critical region.

      The study of 5 subjects with idiopathic autism (Cases 1–4 and 9) revealed the absence of the relevant 15q11-q13 deletion or duplication between BP2 and BP3. In CAL105 normal karyotype was found. In 4 subjects (Cases 5, 7, 8, and 10), frozen tissue and genetic data were not available.

      Clinical Characteristics

      Of the 9 subjects with dup(15), 7 (78%) were diagnosed with autism (Table 3). In 6 cases, autism was diagnosed clinically and was confirmed with postmortem ADI-R. In a 9-year-old boy (Case 1), autism was diagnosed with ADOS-G (Table 3). This case was reported as Case 2 in Mann et al (46). An 11-year-old boy (Case 3) revealed impairments consistent with the diagnosis of pervasive developmental disorder — not otherwise specified.

      View this table:
      TABLE 3

      Autism Diagnostic Interview-RevisedYBased Diagnosis of Autism

      Communication
      GroupCase No.Reciprocal Social Interactions (10)Verbal (8)Nonverbal (7)Restricted, Repetitive, and Alterations Evident Stereotyped Behavior (3)Alterations Evident Before 36 mo (1)Diagnosis (Test)
      dup(15) autism17NA005Autism (ADOS-G)
      224NA1153Autism (ADI-R)
      318NA225PDD-NOS
      42622NA125Autism (ADI-R)
      5Autism (ADI-R) (score not available)
      62718NA45Autism (ADI-R)
      72322NA93Autism (ADI-R)
      82816NA95Autism (ADI-R)
      9Unknown
      Idiopathic autism114NA965Autism (ADI-R)
      2Autism (ADOS)
      322NA1465Autism (ADI-R)
      411NA824Atypical autism - ASD (ADI-R)
      526NA1254Autism (ADI-R)
      62520NA44Autism (ADI-R)
      72216NA53Autism (ADI-R)
      8Autism (ADI-R) (score not available)
      92719965Autism (ADI-R)
      10Atypical autism -ASD (ADI-R score not available)

      ADI-R, Autism Diagnostic Interview — Revised (cutoff scores); ADOS-G, Autism Diagnostic Observation Scale — Generic (42); NA, not applicable; PDD-NOS, Pervasive Developmental Disorder — Not Otherwise Specified.

        In the idiopathic autism group, all 10 subjects were diagnosed clinically as having an ASD. Postmortem ADI-R confirmed a classification of autistic disorder in 6 cases. Two subjects, an 8-year-old boy (Case 4) and a 52-year-old man (Case 10), were diagnosed with atypical autism or high-functioning autism. In a 5-year-old girl (Case 2), autism was diagnosed with ADOS-G. A 32-year-old man (Case 8) and a 52-year-old man (Case 10) were clinically diagnosed as having autism, but the ADI-R could not be conducted postmortem owing to the unavailability of a caregiver who could report on their behavior as a child.

        Among the 9 examined subjects with dup(15) chromosomes, 7 individuals were diagnosed with seizures (78%) (Table 4). In 6 cases, death was sudden and unexplained in patients with epilepsy (SUDEP, 6/9 [67%]). In the 10 subjects with autism, epilepsy was diagnosed in 3 (30%) and death was seizure related in 1 (10%; Table 1). In a previously described cohort of 13 subjects with autism who were subject to postmortem examination (49), seizures were reported in 6 (46%) and death was seizure related in 4 (31%).

        View this table:
        TABLE 4

        Behavioral, Neurologic, and Other Clinical Observations

        GroupCase No.Psychiatric Disorders and Neurologic SymptomsCognitive AssessmentSeizures (Age at Onset)
        dup(15) autism1Severe hypotonia. Regression in infancy. Abnormal response to pain and heat.Profound ID (DQ < 20)Infantile spasms. Intractable epilepsy (10 mo)
        2Hypotonia. Severe regression at age of 15 mo. Head banging.Profound ID (DQ = 22)Intractable epilepsy (8 mo)
        3Regression with infantile spasms. Severe hypotonia.Profound ID (DQ < 20)Infantile spasms. Intractable epilepsy (10 mo). Vagus nerve stimulator.
        4Delay of motor skills. Mild to moderate spastic quadriparesis. Abnormal response to pain, cold.Severe ID (DQ = 31)Seizures (11 y)
        5Hyperactive, verbal(—)No record
        6Sleep disorder. Abnormal response to pain, heat and cold.(—)No record
        7Abnormal gaitProfound ID (DQ < 20)Intractable epilepsy (7 y). Vagus nerve stimulator. Callosotomy.
        8Obsessive compulsive symptomsModerate ID (IQ = 36)Epilepsy (16 y)
        9Cerebral palsy. MicrocephalySevere ID.Intractable epilepsy (9 y). Vagus nerve stimulator.
        Aggressive behavior. Trichotillomania.(—)
        Idiopathic autism1Self-stimulatory behavior(—)No record
        2Hyperactivity, attention deficit. Sleep disorder. Enhanced sensitivity to light and sound.Cognitive delay (IQ = 65)No record
        3Sleep disorder(—)No record
        4Self-stimulatory behavior(—)Epilepsy (8 y)
        5HypotoniaModerate IDNo record
        6No recordModerate IDEpilepsy
        7Bipolar disorder(—)Epilepsy
        8No record(—)No record
        9Enhanced sensitivity to sound, heat, and cold. Low pain threshold.(—)One grand malseizure
        10Bipolar disorder, social anxiety(—)No record

        DQ, developmental quotients; ID, intellectual deficit; —, no formal assessment of ID available.

          Brain Weight

          The mean brain weight in dup(15) autism was 1,177 g, 300 g less than in idiopathic autism and 189 g less than in the control group (Table 1). Age-adjusted means for these 3 groups were 1,171, 1,474, and 1,378 g, respectively (F2 = 9.79, p < 0.001). Post hoc tests showed the difference between the idiopathic and dup(15) groups with autism to be significant (Scheffé-corrected p = 0.001). The difference between the dup(15) and control groups was not significant, although suggestive (p = 0.06). The difference between the idiopathic and control groups was not significant.

          Developmental Abnormalities in Autism Associated With dup(15) and Idiopathic Autism

          Three major types of developmental changes, including (i) heterotopias, (ii) dysplastic changes, and (iii) defects of proliferation resulting in subependymal nodular dysplasia, were detected in both the dup(15) (Table, Supplemental Digital Content 1, http://www.jnen.oxfordjournals.org/lookup/suppl/doi:10.1097/NEN.0b013e318251f537/-/DC1) and idiopathic autism (Table, Supplemental Digital Content 2, http://www.jnen.oxfordjournals.org/lookup/suppl/doi:10.1097/NEN.0b013e318251f537/-/DC1) cohorts. In the control group, 1 subject had small cerebellar subependymal heterotopias and 2 cases had dysplastic changes in the cerebellar flocculus and nodulus (Table, Supplemental Digital Content 3, http://www.jnen.oxfordjournals.org/lookup/suppl/doi:10.1097/NEN.0b013e318251f537/-/DC1). Although all 9 dup(15) subjects and all 10 subjects with autism had developmental abnormalities, there were significant differences between the dup(15) autism and idiopathic autism cohorts in the number of developmental defects and their distribution (Table 5).

          View this table:
          TABLE 5

          Topography and Type of Major Developmental Alterations

          HippocampusCerebral cortexCerebellum
          GroupCase No.Heterotopia (Alveus, CA4, DG)Dysplasia (DG)DysplasiaHeterotopiaDysplasia
          dup(15) autism1++++
          2++++
          3+++++
          4++
          5++++++
          6++++
          7+++++++++
          8+++++
          9+++++
          8 (89%)8 (89%)05 (56%)6/8 (75%)
          Idiopathic autism1+++
          2++
          3++
          4++++
          5+
          6+
          7++
          8+
          9
          10
          1 (10%)1 (10%)5 (50%)6 (60%)4/8 (50%)
          Control1
          2
          3
          4
          5
          6
          7
          +
          p00001 (14%)
          dup(15) vs idiopathic autism0.0010.0010.03nsns
          dup(15) vs control0.0010.001ns0.030.04
          Autism vs controlnsns0.040.03ns

          align="center">+, The number of types of developmental alterations and percentages of subjects with developmental defects are in parentheses; ns, not significant; —, missing structure; DG, dentate gyrus.

          align="center">Statistical analyses: Fisher exact test and Mann-Whitney U test.

            Heterotopias

            Migration developmental abnormalities were detected as heterotopias in the hippocampal alveus, the CA4 sector, the dentate gyrus (DG) molecular layer, and the cerebral and cerebellar white matter (Fig. 1). A description of each is given below.

            FIGURE 1

            Topography and morphology of heterotopias in the brains of individuals with duplications of chromosome 15q11.2-q13 (subjects with dup[15] and autism). (A, B) Heterotopia (A, arrowhead) in the alveus composed of bipolar, multipolar, and pyramidal-like neurons shown at higher magnification in B. (CE) Heterotopias composed of cells with the morphology of granule neurons in the CA4 sector (C, arrowhead) and in the molecular layer (arrowhead) of the dentate gyrus (D); higher magnification is shown in E. (F) Heterotopia (arrowhead) in the cerebral white matter. (GJ) Heterotopias in the cerebellum consist of mixed components of the cerebellar cortex (G, arrowhead), including granule and Purkinje cells (H); one is composed of one type of neuron (I, arrowhead), with the morphology of cells of cerebellar deep nuclei (J). (K) Multifocal heterotopias of both types in cerebellar white matter (arrowheads). (AC, F) dup(15), Case 2; (D, E) dup(15), Case 6; (GJ) dup(15), Case 5; (K) idiopathic autism, Case 5.

            Heterotopias in the Hippocampus

            A relatively large proportion of heterotopic cells in the alveus had the morphology of pyramidal neurons, although they were much smaller than neurons in the cornu Ammonis and were spatially disoriented. Heterotopias composed of neurons with the morphology of granule neurons of the DG were detected in the CA4 sector and in the molecular layer of the DG. Heterotopias in the alveus, CA4, and DG were found in 8 dup(15) subjects (89%), 1 subject with idiopathic autism (10%), and no control subjects. This difference between dup(15) autism and idiopathic autism cohorts (p < 0.001) and dup(15) and control subjects was highly significant (p < 0.001; Table 5). The difference between the idiopathic autism and control groups was not significant.

            Heterotopias in the Cerebellar White Matter

            The morphology of the cerebellar heterotopias reflected 2 types of migration defects. The presence of a mixture of granule and Purkinje cells suggests that clusters of cerebellar cortical neurons do not reach their destination site (Type 1). The second type of cerebellar heterotopia (Type 2) is composed of 1 type of cells with the morphology of cerebellar deep nuclei neurons. Both types of heterotopias were composed of neuronal nuclear marker-positive cells mixed with glial fibrillary acidic protein-positive astrocytes (not shown). In some cases, multiple Type 1 and 2 heterotopias were detected in the cer-ebellar white matter. In contrast to the significant difference in the prevalence of hippocampal heterotopias, there was no significant difference in the prevalence of heterotopias in the cer-ebellar white matter between the dup(15) (56%) and idiopathic autism (60%) groups (Table 5). However, the differences between the dup(15) autism and control groups (p < 0.04) and between the idiopathic autism and control groups (p < 0.04) were significant. Heterotopias in cerebral white matter were rare in both the dup(15) (1/9; 11%) and idiopathic autism (1/10; 10%) groups (Table, Supplemental Digital Content 1, http://www.jnen.oxfordjournals.org/lookup/suppl/doi:10.1097/NEN.0b013e318251f537/-/DC1, and Table, Supplemental Digital Content 2, http://www.jnen.oxfordjournals.org/lookup/suppl/doi:10.1097/NEN.0b013e318251f537/-/DC1).

            Dysplasias

            Dysplastic changes reflect focal microdysgenesis in the hippocampal DG and cornu Ammonis, the amygdala, and the cerebral and cerebellar cortices. Several types of dysplasia were detected in the DG, including hyperconvolution of the DG, duplication of the granular layer distorting the architecture of the molecular layer of the DG, irregular protrusions of the granular layer into the molecular layer, focal thinning and/or thickening of the granular layer, and fragmentation of the granular layer with the formation of isolated nests of granule cells (Fig. 2). The susceptibility of the DG to developmental abnormalities was several times more apparent in dup(15) syndrome than in idiopathic autism cases. They were detected in 8 (89%) of 9 subjects in the dup(15) group and in only 1 subject with autism (1/10, 10%; p < 0.001; Table 5). The number of different types of developmental abnormalities in the dup(15) group ranged from 2 per case in 4 subjects, to 3 per case in 3 subjects, and 5 types in 1 case. The total number of different types of dysplasia was 22 times greater in the dup(15) cohort than in the idiopathic autism cohort (p < 0.001). The difference between idiopathic autism (1 positive case) and the control group (no dysplastic changes in the DG) was not significant.

            FIGURE 2

            Six types of dysplastic changes in the dentate gyrus of subjects diagnosed with duplications of chromosome 15q11.2-q13 (dup(15]) syndrome. (AF) Hyperconvolution of the dentate gyrus within the hippocampal body (A), irregular large protrusions of the granular layer (B), duplication of the granular layer (C), focal thinning and discontinuity of granular layer (arrowhead), and thickening of the granular layer confirmed by examination of serial sections (D, 2 arrowheads), hippocampal malrotation and granular layer fragmentation into small clusters of cells of irregular shape and variable size (E, F). (A) dup(15), Case 8; (B) dup(15), Case 7; (C, E, F) dup(15), Case 3.

            The spectrum of dysplastic changes in the cornu Ammonis comprised abnormal convolution of the CA1 sector, focal deficits of pyramidal neurons, and distortion of the shape, size, and spatial orientation of pyramidal neurons, clustering of dysplastic neurons in the CA1 sector, and many foci of severe microdys-genesis in the CA4 sector, with clustering of immature neurons (Figs. 3A-E). Dysplastic changes in the amygdala resulted in multiple irregular nests of 20 to 40 cells composed of relatively few small immature neurons and numerous oval or bipolar hy-perchromatic neurons that were larger than normal amygdala neurons (Fig. 3F). Dysplastic changes in the cornu Ammonis were detected in 2 subjects with dup(15) syndrome and in 2 brains of subjects with autism (Table, Supplemental Digital Content 1, http://www.jnen.oxfordjournals.org/lookup/suppl/doi:10.1097/NEN.0b013e318251f537/-/DC1, and Table, Supplemental Digital Content 2, http://www.jnen.oxfordjournals.org/lookup/suppl/doi:10.1097/NEN.0b013e318251f537/-/DC1).

            FIGURE 3

            Multiple dysplastic changes in the cornu Ammonis (CA) and amygdala in an 11-year-old boy with hexasomy of chromosome 15q11.2q13 (Case 3). (AC) There is abnormal convolution of the CA1 sector (A) with focal microdysgenesis of the pyramidal layer (B, arrowhead) and clustering of disoriented polymorphic neurons (C, arrowheads). (D, E) Marked multifocal microdysgenesis in the CA4 sector (D, arrowheads), with clustering of a mixture of small and large polymorphic neurons (E). (F) Multifocal microdysgenesis (arrowheads) in the amygdala is composed of small immature neurons and neurons that are larger than normal amygdala neurons.

            The presence of dysplastic changes in the cerebral cortex of 5 of the 10 subjects with idiopathic autism (50%) was in striking contrast to the absence of these changes in the dup(15) (p < 0.03) and control subjects (p < 0.04; Table 5). Three types of cerebral cortex dysplasia were found in the idiopathic autism group: focal polymicrogyria, multifocal cortical dysplasia, and bottom-of-a-sulcus dysplasia (Supplemental Digital Content 2, http://www.jnen.oxfordjournals.org/lookup/suppl/doi:10.1097/NEN.0b013e318251f537/-/DC1).

            Focal polymicrogyria, which reflects a gyrification defect, was found in the frontal lobe in an 8-year-old boy with autism diagnosed with ASD (Case 4), which resulted in local abnormal folding of the cortex with formation of numerous small and irregular microgyri and distortion of the cortical thickness and vertical/horizontal cytoarchitecture (Fig. 4A). The most common developmental abnormality was cortical dysplasia with focal hypocellularity or acellularity and loss of cortical vertical and horizontal cytoarchitecture (Fig. 4B).

            FIGURE 4

            Cerebral cortex dysplasia autism spectrum disorder (ASD). (A, B) Polymicrogyria (A, arrowheads), focal hypocellularity or acellularity with lack of vertical and horizontal cortical organization (B, arrowhead) in an 8-year-old boy with ASD. (CF) Bottom-of-a-sulcus dysplasia in a 5-year-old girl with idiopathic autism (Case 2). Focal dysplasia was limited to deep cortical layers (C), affected all layers except the molecular layer (D), and caused cortex fragmentation (E) or disruption of cortical ribbon continuity (F, arrowheads).

            Another observed gyrification defect was multifocal bottom-of a-sulcus dysplasia with selective changes in the deepest layer, expansion of dysplastic changes to 2 to 3 deep layers of the cortex, or affecting the entire thickness of the cortex. This developmental abnormality was most often seen in the superior frontal and temporal gyrus, the Heschl gyrus, the middle temporal gyrus, the insula, and the parahippo-campal gyrus in a 5-year-old girl diagnosed with idiopathic autism (Case 2; Figs. 4C-F).

            Three types of dysplastic changes were found in the cerebellum of the dup(15) subjects and the subjects with autism. These included dysplasia in parts of the nodulus and flocculus, vermis dysplasia, and focal polymicrogyria (Table, Supplemental Digital Content 1, http://www.jnen.oxfordjournals.org/lookup/suppl/doi:10.1097/NEN.0b013e318251f537/-/DC1, and Table, Supplemental Digital Content 2, http://www.jnen.oxfordjournals.org/lookup/suppl/doi:10.1097/NEN.0b013e318251f537/-/DC1). In the nodulus and flocculus, dysplasia resulted in total spatial disorganization of the granular, molecular, and Purkinje cell layer; only a few small abnormally branched Purkinje cells were found to be dispersed among the granule cells in the affected areas. There were many interindividual differences observed in the nodulus or flocculus volume affected by dysplastic changes. Cerebellar dysplasia was commonly observed in both cohorts. Nodulus dysplasia was present in 7 (87%) of 8 dup(15) subjects and in 6 (75%) of 8 subjects with autism (Table, Supplemental Digital Content 1, http://www.jnen.oxfordjournals.org/lookup/suppl/doi:10.1097/NEN.0b013e318251f537/-/DC1, and Table, Supplemental Digital Content 2, http://www.jnen.oxfordjournals.org/lookup/suppl/doi:10.1097/NEN.0b013e318251f537/-/DC1). Flocculus dysplasia was detected in 6 (75%) of 8 dup(15) subjects, in 4 (50%) of 8 subjects with autism, and in 1 (14%) of 7 control subjects (Table 5). The difference between the groups with autism was not significant, but the difference between the dup(15) autism and control group was significant (p < 0.05).

            Subependymal Nodular Dysplasia

            Nodular dysplasia was found in the brain of a 15-year-old adolescent girl diagnosed with dup(15) (Case 5). This consisted of a single large nodule in the wall of the temporal horn of the lateral ventricle and numerous nodules in the wall of the lateral ventricle in the occipital lobe. Subependymal nodular dysplasia was also detected in the brain of a 39-year-old woman diagnosed with dup(15) (Case 9; Table, Supplemental Digital Content 1, http://www.jnen.oxfordjournals.org/lookup/suppl/doi:10.1097/NEN.0b013e318251f537/-/DC1). There were numerous subependymal nodules less than 1 to 3 mm in diameter in the wall of the occipital horn of the lateral ventricle in a 32-year-old subject with idiopathic autism (Case 8; Table, Supplemental Digital Content 2, http://www.jnen.oxfordjournals.org/lookup/suppl/doi:10.1097/NEN.0b013e318251f537/-/DC1). These were composed of dysplastic neurons with a partially modified morphology of pyramidal, multipolar, or bipolar neurons and oval medium and small cells. In all 3 cases, the nodules were free of oval or polygonal giant cells or ballooned glial cells. Examination of the thalamus, caudate, putamen, nucleus accumbens, and globus pallidus did not reveal developmental qualitative abnormalities in these cohorts.

            Differences Between the Global Pattern of Developmental Abnormalities in dup(15) and Autism

            Although all dup(15) and subjects with autism had developmental abnormalities, the number of different types of developmental alterations detected in the brains of the dup(15) group was, on average, 2.3 times greater (6.9 per case; n = 9) than in the subjects with autism (3 per case; n = 10). Analysis of developmental alterations in 13 subjects with autism previously reported (49) revealed developmental abnormalities in the brains of all subjects with autism and a similar prevalence of alterations.

            Other Neuropathologic Changes

            Selective and marked neuronal loss without gliosis was found in the pyramidal layer in the CA1 sector in the brain of a 10-year-old boy with epilepsy (dup[15]; Case 2). Pathologic alterations extended from the head to the tail of the hippo-campal formation, with loss of neurons in the range of 80% in the head and 50% in the body and tail. These findings might be the result of severe and frequent seizures.

            An area of marked subpial gliosis was found within a sulcus between the inferior frontal and the orbital gyri in the brain of a dup(15) female with epilepsy and seizure-related asphyxia at the age of 26 years (Case 8). Almost complete focal loss of the granular layer was associated with gliosis, thickening of the affected molecular layer, degeneration of astrocytes, and deposition of many corpora amylacea. These findings most likely represent Chaslin gliosis, indicative of epilepsy-related brain damage. This pathologic condition coexisted with hyperconvolution of the DG, focal thinning, and duplication of the granular cell layer, considered developmental abnormalities contributing to abnormal electrical activity and seizures.

            Discussion

            Knowledge of the clinical phenotype and genetic factors in autism is based on examination of thousands of individuals with idiopathic autism; however, between 1980 and 2003, only 58 brains of individuals with idiopathic autism were examined postmortem (50). Knowledge of the clinical and genetic characteristics of the dup(15) syndrome is based on examination between 1994 and 2006 of approximately 160 cases (47, 5154), but the neuropathology of dup(15) with and without autism has not been studied. Results of the application of an extended neuropathologic protocol were previously reported for 13 brains of subjects with idiopathic autism (49). The current study characterizes qualitative neu-ropathologic changes in the brains of 9 individuals with dup(15), including 7 diagnosed as having an ASD (78%). This autism prevalence is in the highest range reported in clinical studies. The association with autism in some of the earlier individual reports (i.e. 4 [33%] of 12 [55] or 6 [36%] of 17) was not based on use of standardized screening (56). A standardized assessment of autistic manifestations in 29 children and adults with a supernumerary idic(15) detected in 20 individuals (69%) with a high probability of ASD (29). All studies reported a significant variability in the autistic phenotype, severity of autistic features, delayed development and/or ID, and seizures among subjects with dup(15) (15).

            Major Neuropathologic Differences Between dup(15) Autism and Idiopathic Autism

            Numerous studies indicate that autism is associated with a short period of increased brain size (5759) and more neurons (60). Macrocephaly was detected in 37% of children with autism younger than 4 years (61) and in 42% of the 19 twins diagnosed with idiopathic autism younger than 16 years (6). Postmortem studies (62, 63) and imaging studies (64) also provided converging evidence of increased brain volume in autism. Microcephaly has been observed in only 15.1% of 126 children with autism aged 2 to 16 years (65) and is usually associated with severe pathology (66, 67), ID, and other medical disorders (65).

            This study revealed a high prevalence of microcephaly in the dup(15) autism cohort examined postmortem, that is, the mean weight of the brains of subjects with dup(15) autism was 300 g less than that of subjects with idiopathic autism and 189 g less than that in the controls (p < 0.001 for both). The characteristics of head circumference and brain volume in dup(15) cohorts have been studied less comprehensively than in idiopathic autism, but published reports also show a strong prevalence of microcephaly. A summary of records from 107 supernumerary inv dup(15) cases revealed that only 3 subjects had macrocephaly (2.8%), but 6 times more cases (n = 18, 16.8%) had microcephaly (15). Battaglia (68) detected microcephaly in radiologic evaluations of 1 in 4 subjects with dup(15) whose age ranged from 4 to 8 years. These data suggest the failure of mechanisms controlling brain growth in autism, resulting in the prevalence of macrocephaly in idiopathic autism and of microcephaly in dup(15) autism.

            Our extended neuropathologic protocol revealed several striking differences between the pattern of developmental alterations in dup(15) autism and idiopathic autism. Neuronal migration defects in the hippocampus resulting in heteroto-pias in the alveus, CA4, and DG were 8 times more common in dup(15) autism (89% of subjects) than in idiopathic autism (10%; p < 0.001). The second developmental abnormality distinguishing dup(15) autism from idiopathic autism was the dysplasia that occurred 8 times more often in the DG of subjects with dup(15) and the different types of developmental abnormalities that occurred 22 times more often in the DG of subjects with dup(15) (p < 0.001). The third factor differentiating these 2 cohorts was the absence of cerebral cortex dysplasia in dup(15) autism and the presence of this pathology in 50% of subjects with idiopathic autism (p < 0.03). The increased number of developmental alterations and the topographic differences suggest significant differences between mechanisms contributing to abnormal neuronal migration and altered cytoarchitecture in these 2 cohorts.

            Linkage and gene mapping analysis, molecular reports, and clinical studies revealed the link between de novo, maternally derived proximal 15q chromosome alterations, and autism (13, 15, 27, 33, 56, 6972). This postmortem study suggests that neuropathologic profile with microcephaly and multiple focal developmental defects is another marker of maternally derived proximal 15q chromosome alterations contributing to autistic phenotype.

            Although ASD and epilepsy are heterogeneous disorders, they often occur together. This may indicate that these disorders share some underlying mechanisms and that epi-leptogenesis affects brain structure and function, which modify the clinical manifestations of autism. Approximately 30% of children with autism are diagnosed with epilepsy and 30% of children with epilepsy are diagnosed with autism (73). Significant cognitive impairment is present in approximately 50% of all individuals who have autism (74). Early onset of seizures contributes to clinical regression, enhanced severity of autistic phenotype, and enhanced mortality. The rate of death among subjects with autism is 5.6 times higher than expected (75), and epilepsy- and cognitive impairment-related accidents account for most of the deaths (7578). The diagnosis of epilepsy in 78% (7/9) of postmortem-examined individuals with dup(15) indicates that epilepsy is an important component of the clinical phenotype in most individuals diagnosed with dup(15).

            Microcephaly might be one of several indications of brain immaturity that increases the risk of epilepsy. The immature brain exhibits increased excitation, diminished inhibition, and increased propensity for seizures in infancy and early childhood (79). The reduced volume of neurons in most subcortical structures and some cortical regions in the brains of 4- to 8-year-old children with autism seems to reflect brain immaturity in early childhood (80). In the normally developing brain, maturation of the frontal and temporal cortex is associated with differential expression of 174 genes; however, none of these genes are differentially expressed in ASD (81). Altered development of neurons resulting in brain immaturity may contribute to an increased tendency for the seizures and epileptogenesis observed in the examined cohort. Very early onset of intractable epilepsy (at 10, 8, and 10 months, respectively) was reported in all 3 of the youngest individuals diagnosed with dup(15), who died as a result of SUDEP at 9, 10, and 11 years. These findings suggest that very early onset of seizures and very severe seizures may increase the risk of SUDEP in this cohort. Severe ID in all 3 of these individuals indicates that brain immaturity and a profound ID are another SUDEP risk factor in the examined cohort of subjects with dup(15). The combination of all these factors may contribute to death at a very early age (∼10 years).

            Sudden unexpected death in childhood (SUDC) is the sudden death of a child older than 1 year that, despite a review of the clinical history, circumstances of death, and complete autopsy with appropriate ancillary testing, remains unexplained.

            It occurs throughout childhood (<18 years) but occurs most commonly between the ages of 1 and 4 years (82, 83). Sudden unexpected death in childhood is an unexpected and unexplained death that occurs in patients with known epilepsy, including children, and is typically associated with sleep (84, 85). Kinney et al (86) reported several types of hippo-campal anomalies in SUDC cases, including hyperconvolution of the DG, focal duplication of the DG, granular nodular heterotopia, abnormal folding of the subiculum, and focal clustering of pyramidal neurons in the cornu Ammonis. These developmental anomalies are considered a cause of seizure-related autonomic and/or respiratory dysfunction and sudden death (8789). Kinney et al (86) proposed that these anomalies represent an epileptogenetic focus that, when triggered by fever, trivial infection or minor head trauma at a susceptible age, may result in unwitnessed seizure, cardiopulmonary arrest, and sudden death. The presence of these developmental anomalies in the examined brains of individuals with dup(15) may explain SUDEP in 5 cases and SUDC in 2 other cases. The presence of these changes in the hippocampus of subjects with dup(15) who died of causes other than SUDEP or SUDC suggests that they also were at higher risk for sudden unexpected death. Collectively, these data indicate that the risk of SUDC is much higher in the dup(15) cohort than in the general population with SUDC in which there is an overall rate of 57 of 100,000 deaths per year (90).

            Microdysgenesis is not specific for dup(15), epilepsy, or autism; it has been reported in ID (91), psychosis (92), and dyslexia (93), as well as in some control subjects. None of these developmental alterations can be considered pathognomonic of an “epileptic” brain (83), but changes in the cohort affected by dup(15) or idiopathic autism reveal significant differences. This indicates that not a single lesion but, instead, a complex pattern of developmental defects distinguishes these subjects. A 2.3-fold higher prevalence of these developmental abnormalities, 2.3 times higher prevalence of epilepsy, and 6 times higher prevalence of epilepsy-related death in the dup(15) cohort compared with the idiopathic autism group suggest that the mechanisms leading to developmental structural alteration in the hippocampal formation are a major contributor to epilepsy and SUDEP/SUDC in dup(15). These seem to be a much weaker contributor to epilepsy and SUDEP in idiopathic autism. The results presented in this report reinforce the hypothesis that additional copies of the critical 15q11-q13 region are causally related to the autism phenotype and developmental abnormalities contributing to epilepsy and to an increased risk of SUDEP and SUDC. Future studies of the expression and distribution of proteins encoded by GABAA receptor subunit genes (α5, β3, and γ3) and the gene for juvenile epilepsy located near D14S165 on chromosome 15 may explain the role of duplication or triplication of these genes in autism and the enhanced susceptibility to seizures in dup(15) syndrome.

            In conclusion, despite the common clinical diagnosis of autism in the dup(15) and idiopathic cohorts, significant differences in brain growth and focal developmental defects of neuronal migration and cytoarchitecture indicate that the dup(15) autistic phenotype is a product of unique genetic, molecular, and neuropathologic alterations.

            Acknowledgments

            The authors thank Maureen Marlow for editorial corrections and Jadwiga Wegiel, Cathy Wang, and En Wu Zhang for histology. Tissue and clinical records acquisition was coordinated by the Autism Tissue Program (Princeton, NJ; Directors: Jane Pickett, PhD and Daniel Lightfoot, PhD). Carolyn Komich Hare, MSc, provided results of postmortem application of ADI-R. We are deeply indebted to the Dup15q Alliance, and the families of the tissue donors who have made this study possible.

            Footnotes

            • The tissue was obtained from the Harvard Brain Tissue Resource Center, Belmont, MA, supported in part by PHS Grant R24-MH 068855; the National Institute of Child Health and Human Development Brain and Tissue Bank for Developmental Disorders at the University of Maryland, Baltimore, MD; and the Brain and Tissue Bank for Developmental Disabilities and Aging of the New York State Institute for Basic Research in Developmental Disabilities, Staten Island, NY.

            • This study was supported in part by funds from the New York State Office for People With Developmental Disabilities, a grant from the US Department of Defense Autism Spectrum Disorders Research Program (AS073234 Program Project; J.W., T.W., A.C.), a grant from Autism Speaks (Princeton, NJ; J.W.), and an Autism Center of Excellence (National Institutes of Health P50 HD055751; E.H.C.). Clinical and molecular investigations of the subjects with chromosome 15 duplications were supported by the Collaborative Programs for Excellence in Autism Research (National Institutes of Health U19 HD35470; N.C.S.) and Nemours Biomedical Research, duPont Hospital for Children.

            • Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Web site (www.jneuropath.com).

            References

            1. 1.
            2. 2.
            3. 3.
            4. 4.
            5. 5.
            6. 6.
            7. 7.
            8. 8.
            9. 9.
            10. 10.
            11. 11.
            12. 12.
            13. 13.
            14. 14.
            15. 15.
            16. 16.
            17. 17.
            18. 18.
            19. 19.
            20. 20.
            21. 21.
            22. 22.
            23. 23.
            24. 24.
            25. 25.
            26. 26.
            27. 27.
            28. 28.
            29. 29.
            30. 30.
            31. 31.
            32. 32.
            33. 33.
            34. 34.
            35. 35.
            36. 36.
            37. 37.
            38. 38.
            39. 39.
            40. 40.
            41. 41.
            42. 42.
            43. 43.
            44. 44.
            45. 45.
            46. 46.
            47. 47.
            48. 48.
            49. 49.
            50. 50.
            51. 51.
            52. 52.
            53. 53.
            54. 54.
            55. 55.
            56. 56.
            57. 57.
            58. 58.
            59. 59.
            60. 60.
            61. 61.
            62. 62.
            63. 63.
            64. 64.
            65. 65.
            66. 66.
            67. 67.
            68. 68.
            69. 69.
            70. 70.
            71. 71.
            72. 72.
            73. 73.
            74. 74.
            75. 75.
            76. 76.
            77. 77.
            78. 78.
            79. 79.
            80. 80.
            81. 81.
            82. 82.
            83. 83.
            84. 84.
            85. 85.
            86. 86.
            87. 87.
            88. 88.
            89. 89.
            90. 90.
            91. 91.
            92. 92.
            93. 93.
            View Abstract