Thomas Bourgeron, Ph. D., Pasteur Institute (Pilot Project)
Autism is a developmental disorder characterized by impaired social interaction and communication, in addition to restricted range of interests and activities. The PARIS study, an international collaborative project coordinated by Pr. Marion Leboyer (Paris) and Pr. Christopher Gillberg (Goteborg), joined the efforts of clinicians in Europe and the United States with the purpose of gathering a large population of autistic sib pairs and trios to perform linkage and association analyses. During the systematic mutation screening and chromosomal analysis of autistic subjects, we identified mutations in cell adhesion molecules involved in synaptogenesis and chromosomal rearrangements associated with autism and Asperger syndrome. Synaptogenesis, the formation of functional synapses, is considered as the final step in the development of the central nervous system. Among the genes involved in this process, cell adhesion molecules are crucial factors for the identification of the appropriate partner cell. On the basis of these results, this project aims to analyze cell adhesion molecules involved in synaptogenesis and to identify the genes localized at or near the breakpoints of chromosomal translocations. For this, we will screen for mutations in candidate genes by direct sequencing and characterize translocated genes using Fluorescent In situ Hybridization (FISH). When disrupted genes will be identified, they will be systematically screened in all autistic subjects recruited by the PARIS study. In parallel, functional analysis of the mutations identified in the autistic subjects will be conducted to understand their biological roles and to evaluate the consequence of these mutations in the establishment of the neural connection pattern.
Elizabeth a. Hoffman, Ph.D., Georgetown University Medical School (Young Investigator)
In addition to the hallmark deficit of social communication, individuals with autism exhibit visual motion processing abnormalities. Despite the prevalence of this profile of dysfunction, there are no reports of brain changes associated with motion perception deficits and few studies have investigated the neurophysiology of impaired social perception in autistic individuals. Furthermore, the link between these two classes of impairment has not been established, although researchers have suggested that the ability to interpret motion cues may be necessary for later development of social communication. Neuroimaging studies have localized motion perception in normal adults to a region in the middle temporal lobe, area V5/MT. We will use functional magnetic resonance imaging (fMRI) to probe the motion system in typically developing and autistic children, ages 8-10 years. Subjects will view dynamic and static arrays of cartoon fish. In the dynamic condition, a proportion of the fish will "swim" coherently in one direction, with the remaining fish moving randomly. Subjects will determine whether or not two sequential arrays of fish are moving in the same direction. Subjects will view arrays of still fish in the static condition. This project will address the following questions:  what is the neural representation of motion perception in normal children?  How does this representation differ in age-matched individuals with autism?  Can we identify a neural network that compensates for a dysfunctional motion perception system?
First year funding partner: The Autism Coalition for Research and EducationCandidate Genes for Autism on Chromosome 7q
AHM Mahbubul Huq, M.D., Ph.D., Wayne State University (Pilot Project)
Several genome screens found evidence of linkage for autism on overlapping regions on chromosome 7q. The broad goal of this proposal is to identify autism susceptibility gene or genes on chromosome 7q. The hypothesis underlying this proposal is that common variants (>5% frequency) underlie susceptibility to autism. Our approach is direct analysis of the likely candidate genes and single nucleotide polymorphism (SNP)-based linkage disequilibrium mapping using a large sample of families with autism. We have selected 30 candidate genes from consensus regions of linkage based on known or presumed function and expression in brain. We will sequence exons and promoters of these genes in 20 individuals with autism to identify variants that rare in general population but are enriched in autism population. SNPs will also be selected from the databases focusing on coding sequence and promoter variants. We will perform multilocus linkage disequilibrium and haplotype analyses of the SNPs in the candidate genes. In order to improve our chance of identifying susceptibility alleles, we will also perform quantitative trait association studies of the candidate genes. We will genotype enough SNPs in the candidate genes to capture the common haplotypes. The genes that we have selected are less than 50 kb in size; so 10 SNPs in each gene will capture the most common haplotypes. There is an ongoing debate whether "few common variants" or "multiple rare variants" underlie susceptibility to complex traits. With our strategy and sample size, we will able to identify susceptibility alleles with frequency greater than 5%.
Tal Kenet, Ph.D., University of California, San Francisco (Young Investigator)
We believe one of the core weaknesses in autism is a general failure in gain control mechanisms. Such a failure would be manifested most prominently though abnormal processing of intensity. The aim of this study is to investigate this in both the auditory and visual domain in children with autism, using both behavioral and neuroimaging methods. Specifically, we will look at intensity dependent processing as a function of varying parameters in both the visual and auditory domains. We have two specific goals for this project. The first is to document in detail any intensity related sensory processing deficits revealed by behavioral and neurological measures in both these sensory domains. We shall study how these deficits vary as a function of elementary stimulus parameters, spectral characteristics in the auditory domain, and contrast and spatial frequency in the visual domain. The second goal is to compare observed impairments across modalities. Behavioral studies will span both magnitude estimation tasks, and forced choice tasks requiring the child to discriminate between stimuli based on intensity. Neuroimaging studies will consist of EEG and MEG imaging of the neuronal responses to the same stimuli. We will look at the growth of magnitude functions resulting from each of these methods, and will attempt to correlate (1) between behavioral and neurological findings (2) between deviation from control data and severity of impairment (verbal or functional). Finally, we hope to gain a deeper insight into the core deficits of autism by looking at primary deficits across modalities.
Bryan H. King, MD, Dartmouth Medical School (Psychometric Award)
As controlled clinical trials in children with autism are increasingly performed, an appreciation for the need for better outcomes measures is growing. In part, the difficulty for autism clinical trials derives from the fact that no two children with autism are alike. Further, behavioral problems that are particularly challenging will differ from family to family. Current assessment instruments may be inadequate to capture change that is important to patients and families, and thus recent reports document discordant findings. Investigators see change that parents do not; or parents see change that is lost on investigators. The aims of this project are to explore the utility of obtaining videotaped behavior samples for the assessment of treatment response in children with autism. The proposed studies examine factors that may influence raters, and which may separate parents from investigators with respect to their scoring of behavioral disturbance. This revised and downsized project will provide essential data to inform subsequent large scale studies in which videotaped behavioral samples may be utilized in single and double blind clinical trials. Funded in Partnership with Repligen Corporation
Michael Merzenich, Ph.D., Keck Center for Integrative Neuroscience, UCSF (Genius Award)
The primary objective of this research and development project is to design, produce, and test the application of new training tools designed to ameliorate the expressions of autism spectrum disorders (ASD) and to test new strategies designed to prevent the emergence of the full expression of ASD symptoms in at-risk infants. The research is multi-pronged, integrating animal studies of neural plasticity, human behavioral and imaging studies, and testing and development of a new targeted generation of re-training tools for ASD based upon experience with the Fast ForWord programs in ASD subjects.
First year funding partner: The Autism Coalition for Research and EducationQuantitative Psychophysiologic Evaluation of Sensory Processing in Children with Autism Spectrum Disorders
Lucy Jane Miller, Ph.D., OTR, University of Colorado Health Sciences Center (Psychometrics Award)
The proposed study will examine the reliability and validity of an established psychophysiologic assessment of sensory reactivity in 40 school-aged children with Autism Spectrum Disorders (ASD). The quantitative laboratory paradigm, called the Sensory Challenge Protocol has successfully documented sympathetic and parasympathetic nervous system functioning in children with fragile X syndrome, cognitive delays, sensory modulation dysfunction, and ADHD, and field-tested in a group of children with ASD. In this pilot feasibility study (n = 8), the Sensory Challenge Protocol was found to be practicable with ASD, however, a larger scale project evaluating the reliability and validity of the paradigm with children who meet stringent criteria for ASD is needed. The protocol is administered in a non-threatening, fun situation, and 'pretend spaceship during which 50 sensory stimuli are presented while electrodermal reactivity (sympathetic nervous system marker) and vagal tone (parasympathetic nervous system marker) data are collected continuously. The four specific aims are: to evaluate the reliability of the Sensory Challenge Protocol in school-aged children with ASD;to investigate the variability of sensory reactivity among children with ASD;to examine relations among sensory symptoms and core symptoms of autism (e.g., social behaviors, communication, and restricted behaviors); andto evaluate the convergent validity of the laboratory measure with parent-report measures of sensory and functional behaviors.Validation of a quantitative measure of sensory reactivity in children with ASD will further our understanding of individual differences among children with ASD and will be useful in future treatment outcomes research. Funded in Partnership with Repligen Corporation
Cynthia A. Molloy, M.D., M.S., Cincinnati Children's Hospital (Pilot Project)
Research on the etiology of autism spectrum disorders (ASD) has been hindered by the heterogeneity of the population with this behaviorally defined condition. Previous studies have suggested the possibility that distinct phenotypic subgroups of ASD may be defined, based on history of immune disorders and measurable immune responses. The adaptive immune response involves the proliferation of T helper cells along two divergent, antagonistic lines. The resulting subsets, TH1 and TH2, are distinct immune phenotypes characterized by the particular cytokines they produce and the immune response mediated by those cytokines. The TH1 immune phenotype is associated with organ specific autoimmunity. The TH2 phenotype predominates in an atopic response. Different investigators have reported either TH1 or TH2 predominance in ASD. This molecular epidemiologic pilot study seeks to address this apparent paradox. Our proposed study uses a case control design, comparing children with ASD (cases) to children with other developmental disorders (Control A) and children with neurotypical development (Control B). The specific aims of the study are to 1) compare the histories of atopy and familial autoimmunity between cases and controls, 2) measure peripheral blood cytokines and compare the distribution of TH1 and TH2 phenotypes between cases and controls and 3) examine the relationship between immune phenotype, history of atopy and familial autoimmunity, and the clinical characteristics of ASD. By addressing important methodologic issues such as unbiased selection of cases and controls, and measurement of both medical history and cytokine levels, this study will contribute significant new information about immune phenotypes in ASD.
Peter Mundy, Ph.D., University of Miami (Psychometrics Award)
Research indicates joint attention impairment is pathognomonic of autism, predicts cognitive and social outcomes in these children, and may be an index of neurodevelopmental components of the disorder. Consequently, joint attention has become an important dimension to consider in treatment research. However, current assessment and diagnostic tools lack the efficiency and precision necessary to measure joint attention development in intervention studies. Furthermore, because these skills can be low-frequency, context-specific behaviors in children with autism, the combination of both direct observations and parent or professional report data may be necessary to provide a reliable and valid assessment of this critical domain of social development. To this end we will develop and test the efficacy of the Early Social Communication Scales Live (ESCS-L), a live coding quantitative assessment of social communication skills, and the Pictorial Infant Communication Scale (PICS), a sixteen item measure employing both pictorial and written descriptions of critical behaviors to improve the reliability and validity of informant responses. The psychometric properties of this measurement system will be assessed in an 8-week longitudinal study of 40 children with autism and 20 comparison children. It is expected this research will lead to the development of a system of efficient quantitative measures of a critical domain of social development that will enable researchers to better monitor the growth development and treatment response of children with autism. Funded in Partnership with Repligen Corporation
Carlos A. Pardo-Villamizar, M.D., Johns Hopkins University School of Medicine (Pilot Project)
Current evidence suggests that neurobiological abnormalities in autism are associated with changes in cytoarchitectural and neuronal organization that can be determined by genetic, environmental, immunological and toxic factors. Since neuroglia have central roles during brain development for cortical organization, neuronal function and immune responses, we hypothesize that autism is, in part, an immune mediated disorder. In order to determine if neuroimmunological mechanisms contribute to the pathogenesis of autism, we propose to study the profile of neuroglial responses and the presence of cellular and humoral immunopathological reactions, in brain tissues from patients with autism as compared with those from age-matched normal and neurodevelopmental disorders such as Rett's syndrome and Down's syndrome. We also plan to study cytokine pathways involved in neuroinflammation to determine if they have a role in the pathogenesis of autism. These studies may be valuable for the understanding of autism and the design of new therapeutic approaches to treat it.
Paul H. Patterson, Ph.D., California Institute of Technology (Pilot Project)
We are developing a mouse model of autism based on epidemiological findings that maternal viral infection can increase the frequency of autism in the offspring. Using a respiratory infection of pregnant mice at E9.5, we find that the offspring display striking behavioral abnormalities in 4 tests relevant to autism. Moreover, these offspring have a deficit in Purkinje cells specifically in lobules VI and VII of the cerebellum, a finding that strikingly parallels the neuropathology in autism. We propose to (i) quantitate and further examine the cerebellar pathology, (ii) using microarray analysis, examine molecular changes in the brains of exposed offspring, at several time points in development and in adulthood, and (iii) investigate the cause of altered fetal brain development, specifically testing the hypothesis that virally-evoked cytokines in the maternal immune response are responsible. The latter experiments will involve the use of cytokine knockout mice, as well as injection of killed virus.
Antonio M. Persico, M.D., University Campus Bio-Medico (Pilot Project)
We have recently described an association, to our knowledge replicated in at least two independent studies, between autism and "long" alleles of a polymorphic GGC repeat located in the 5'UTR of the gene encoding Reelin, a pivotal protein for neuronal migration during neurodevelopment. We have also found that long GGC alleles yield reduced protein translation rates, both in vitro and in vivo, and that Reelin exerts a proteolytic activity potently inhibited by organophosphates, compounds routinely used as pesticides and insecticides. Within the framework of a gene-environment interactive model, genetically-vulnerable individuals producing lower amounts of Reelin, if prenatally exposed to organophosphates during critical periods in neurodevelopment, could undergo altered neuronal migration resulting in an autistic syndrome.
Paraoxonase is the enzyme responsible for detoxification of organophosphates, for hydrolysis of lipid peroxides involved in atherosclerosis and for modulation of cell-mediated immune responses. Impressive 40-fold interindividual differences in serum paraoxonase activity have been documented in humans, largely resulting from functional polymorphisms present in the paraoxonase-1 (PON1) gene, located on chromosome 7q31. The current project is aimed at (a) characterizing paraoxonase enzymatic activity in the serum of autistic patients from A.G.R.E. families; (b) characterizing functional polymorphisms and searching for mutations in the PON1 gene of the same patients, and (c) correlating enzymatic activity and genotypic information to verify whether PON1 polymorphisms/mutations may yield reduced levels of paraoxonase activity, thereby increasing the risk for developmental teratogenesis following prenatal exposure to organophosphates and explaining the enhanced cell-mediated immune responses previously described in many autistic patients.
Jonathan Pettegrew Memorial Young Investigator Award 2003:Language Related Working Memory and Executive Functions in Relatives of Individuals with Autism
Abraham Reichenberg, Ph.D., Mount Sinai School Medicine (Young Investigator)
Functions of planning, monitoring, inhibition and selective attention are collectively termed: Executive Functions (EF). The term Working Memory (WM) is used generally to refer to a limited-capacity system allowing the temporary storage and manipulation of information in complex cognitive tasks. WM and EF have been suggested as potential endophenotypes for autism. Speech abnormalities have also been characterized as a potential phenotype, and have generated promising genetic findings. Baddeley and Hitch (1974) proposed a four-component WM model that provides a conceptual framework for the central role of temporary storage of information in the development and performance of more complex cognitive tasks, including language acquisition, and Executive Functions. The proposed project will study the familiality of the language-associated components of this well characterized model of WM. In addition, the role of the supervisory component of the model in EF will also be investigated. Given the recent evidence for autism related genes specifically found among families in which the autistic members or their relatives have language deficits, this may be critical for understanding the core deficits in autism.Our hypothesis is that deficits in the language-associated components of the working memory model and the supervisory component will be more frequent among family members of autism probands than among family members of probands with ADHD or family members of normally developing children. Our secondary hypothesis is that EF measures will be directly associated with measures in the supervisory component, and indirectly (i.e., through the measures of the supervisory) with measures in the language-associated components.
Lawrence T. Reiter, Ph.D., University of California, San Diego (Young Investigator)
Loss-of-function of the UBE3A ubiquitin ligase in selected regions of the brain causes the severe mental retardation disorder Angelman syndrome, while elevated expression of this same gene has been strongly implicated autism. One phenotype of autism is a reduced ability to form long term memory. A critical step in understanding the mechanism by which altered levels of UBE3A lead to abnormal brain development or function is to identify the protein targets which are normally degraded by UBE3A dependent ubiquination. We propose to identify UBE3A candidate target proteins using the model system Drosophila melanogaster, which has a single highly conserved homologue of UBE3A. We will employ two complementary strategies to identify such UBE3A targets. First, we will search for genes in flies that alter the phenotypes of flies having reduced or elevated UBE3A activity. Second, we will identify Drosophila proteins that are down-regulated as a consequence of overexpressing human UBE3A in Drosophila. We will then assay for the function of these proteins in well established Drosophila assays for long term memory. The ultimate goal of these studies will be to determine whether human homologues of Drosophila UBE3A targets we identify are misregulated in autism and/or Angelman syndrome.
Donald C. Rojas, Ph.D., University of Colorado Health Sciences Center (Pilot Project)
The main goal of this proposal is to evaluate whether left hemisphere planum temporale (PT) volume reduction, assessed by morphometric analysis of magnetic resonance imaging (MRI) data, is present in the parents of children with autism. Such a finding would bolster support our hypothesis that PT volume reduction represents a familial risk factor for language delay and deviance often seen in individuals with autistic disorder. Preliminary evidence from our laboratory suggests that alteration in the volume of the planum temporale in both adults and children with autism. The planum temporale (PT) is an auditory processing brain region intimately associated with language processing in the left hemisphere in normal development. Measuring the PT in non-affected family members will allow us to assess whether the PT alteration in autism is familial, the next logical step in determining if PT volume change will be a productive endophenotype. We are proposing to measure PT volume and surface area in 20 parents of children with autism, 20 age-matched adults with autism and 20 adults with no personal or familial history of autism or other developmental disorders. We hypothesize that the volume of the left PT will be significantly smaller in the parent group compared to the controls, and that the PT volume in the adult autism group will be lower than that for the parent group (consistent with the view that not all of the parents in the autism group will carry the risk factor in question). We will also measure language function and assess the family history of language development in the same three groups of adults. Measures of receptive and expressive language, as well as phonology and pragmatics, will be obtained in all of the adults and correlated to the MRI findings.
Andres Ruiz-Linares, Ph.D., University College London (Pilot Project)
Autism is perhaps the most heritable of psychiatric disorders and has an estimated prevalence of 1/1000, making this condition a very important public health issue. To date there have been 7 genome-wide searches for autism susceptibility loci. These genome screens have suggested linkage to several regions, particularly on chromosomes 7 and 2. However, these candidate regions are very large and no gene conferring susceptibility to autism has yet been identified. In order to take these encouraging results forward it is essential to examine independent well-characterized study samples. A powerful complement to linkage analysis is the use of linkage disequilibrium mapping in population samples, particularly when drawn from genetically isolated areas. Here we propose to initiate a program of research aimed at identifying autism loci by studying unrelated patients from the founder population of Antioquia, in North West Colombia. We propose to collect a study sample of at least 100 carefully diagnosed autistic patients. The refined clinical assessments aim at enabling the analysis of genetic effects underlying specific autistic symptoms. We will subsequently employ this study sample in whole-genome screens so as to perform linkage disequilibrium mapping of autism loci. The diagnostic assessments that we will carry out in Antioquia are the same being employed in a parallel study of patients from the Central Valley of Costa Rica, a population genetically very similar to Antioquia. The standardization of clinical criteria across studies will facilitate the comparison of genetic mapping results and allow joint analyses of the data. The collaborative study of autism in two closely related population isolates should dramatically increase the power and resolution of the genetic analyse.
Sara Jane Webb, Ph.D., University of Washington (Young Investigator)
Autism is characterized by impairments in social functioning, communication, and a restricted range of activities. Specific deficits in complex information processing have been proposed as both core and broader phenotype autism deficits (e.g. Frith, 1989; Minshew et al., 1997). The weak "central coherence" hypothesis has focused on the manner in which individuals with autism process complex items. Brock et al. (2002) suggest that weak central coherence reflects a failure of feature binding, which can be indexed by EEG recordings of neural activity in the gamma band range (30 to 80 Hz). The primary goal of this proposal is to explore the relation between gamma activation, a neural index of feature binding, and neurocognitive functioning, in order to better define information processing deficits in individuals with autism. The following questions will be addressed: Are deficits in feature binding, as reflected in gamma band abnormalities, evident in individuals with autism? If so, are they evident across a wide range of stimuli, suggesting a global perceptual processing impairment? Or are these deficits more evident during tasks that involve social processing such as the perception of a face? If such deficits are present, are they correlated with performance on (1) central coherence tasks, and (2) tasks that are known to tap brain regions involved in feature binding (i.e. hippocampus)? This study will help better define the nature of information processing impairments in autism and their neural bases, and provide potentially useful information for genetic studies of broader phenotype.
Lonnie Zwaigenbaum, M.D., McMaster University (Pilot Project)
The objectives of this project are 1) to identify families co-segregating autism and language impairments, and to determine the specific domains of language function that are involved, and 2) To localize genes underlying shared liability to autism and language impairments in extended families. Evidence from family and genetic studies of autism suggests that language impairments in relatives result from the same susceptibility genes as autism itself, but likely involve fewer genes and less heterogeneity, which may result in greater power to localize these genes. From among families participating in an ongoing genetic study of autism, we have identified 30 families that include at least 2 autistic probands, and at least 2 relatives screening positive for language impairments. Relatives from these families will be assessed using a battery of measures covering domains of language functioning hypothesized to be genetically related to autism. Language measures which show familial aggregation will be used to define categorical and quantitative phenotypes, which will then be studied using linkage analysis to localize underlying genes. We will use non-parametric analyses, which are applicable to both categorical and continuous traits and do not require the specification of genetic transmission parameters. For categorical phenotypes (e.g., presence/absence of speech delay), we will test for excess allele sharing among sets of affected relatives, with non-parametric linkage (NPL) scores computed with the program ALLEGRO. Significant linkage will be declared when the observed NPL score exceeds 4.4 (p-value< 2.5x10-5) to ensure a genomewide type I error of 0.05. Quantitative traits will be localized using multipoint variance component linkage methods.
First year funding partner: The Autism Coalition for Research and Education