A total of 38 brain cytoarchitectonic subdivisions, representing subcortical and cortical structures, cerebellum, and brainstem, were examined in 4- to 60-year-old subjects diagnosed with autism and control subjects (a) to detect a global pattern of developmental abnormalities and (b) to establish whether the function of developmentally modified structures matches the behavioral alterations that are diagnostic for autism. The volume of cytoarchitectonic subdivisions, neuronal numerical density, and total number of neurons per region of interest were determined in 14 subjects with autism and 14 age-matched controls by using unbiased stereological methods. The study revealed that significant differences between the group of subjects with autism and control groups are limited to a few brain regions, including the cerebellum and some striatum and amygdala subdivisions. In the group of individuals with autism, the total number and numerical density of Purkinje cells in the cerebellum were reduced by 25% and 24%, respectively. In the amygdala, significant reduction of neuronal density was limited to the lateral nucleus (by 12%). Another sign of the topographic selectivity of developmental alterations in the brain of individuals with autism was an increase in the volumes of the caudate nucleus and nucleus accumbens by 22% and 34%, respectively, and the reduced numerical density of neurons in the nucleus accumbens and putamen by 15% and 13%, respectively. The observed pattern of developmental alterations in the cerebellum, amygdala and striatum is consistent with the results of magnetic resonance imaging studies and their clinical correlations, and of some morphometric studies that indicate that detected abnormalities may contribute to the social and communication deficits, and repetitive and stereotypical behaviors observed in individuals with autism.

Several morphometric studies have revealed smaller than normal neurons in the neocortex of autistic subjects. To test the hypothesis that abnormal neuronal growth is a marker of an autism-associated global encephalopathy, neuronal volumes were estimated in 16 brain regions, including various subcortical structures, Ammon’s horn, archicortex, cerebellum, and brainstem in 14 brains from individuals with autism 4 to 60 years of age and 14 age-matched control brains. This stereological study showed a significantly smaller volume of neuronal soma in 14 of 16 regions in the 4- to 8-year-old autistic brains than in the controls. Arbitrary classification revealed a very severe neuronal volume deficit in 14.3% of significantly altered structures, severe in 50%, moderate in 21.4%, and mild in 14.3% structures. This pattern suggests desynchronized neuronal growth in the interacting neuronal networks involved in the autistic phenotype. The comparative study of the autistic and control subject brains revealed that the number of structures with a significant volume deficit decreased from 14 in the 4- to 8-year-old autistic subjects to 4 in the 36- to 60-year-old. Neuronal volumes in 75% of the structures examined in the older adults with autism are comparable to neuronal volume in age-matched controls. This pattern suggests defects of neuronal growth in early childhood and delayed up-regulation of neuronal growth during adolescence and adulthood reducing neuron soma volume deficit in majority of examined regions. However, significant correction of neuron size but limited clinical improvements suggests that delayed correction does not restore functional deficits.

Background: Epigenetic modifications, such as cytosine methylation in CpG-rich regions, regulate multiple functions in mammalian development. Maternal nutrients affecting one-carbon metabolism during gestation can exert long-term effects on the health of the progeny. Using C57BL/6 J mice, we investigated whether the amount of ingested maternal folic acid (FA) during gestation impacted DNA methylation in the offspring’s cerebral hemispheres. Reduced representation bisulfite sequencing at single-base resolution was performed to analyze genome-wide DNA methylation profiles. Results: We identified widespread differences in the methylation patterns of CpG and non-CpG sites of key developmental genes, including imprinted and candidate autism susceptibility genes (P <0.05). Such differential methylation of the CpG and non-CpG sites may use different mechanisms to alter gene expressions. Quantitative real time reverse transcription-polymerase chain reaction confirmed altered expression of several genes. Conclusions: These finding demonstrate that high maternal FA during gestation induces substantial alteration in methylation pattern and gene expression of several genes in the cerebral hemispheres of the offspring, and such changes may influence the overall development. Our findings provide a foundation for future studies to explore the influence of gestational FA on genetic/epigenetic susceptibility to altered development and disease in offspring.

Characterization of the type and topography of structural changes and their alterations throughout the lifespan of individuals with autism is essential for understanding the mechanisms contributing to the autistic phenotype. The aim of this stereological study of neurons in 16 brain structures of 14 autistic and 14 control subjects from 4 to 64 years of age was to establish the course of neuronal nuclear and cytoplasmic volume changes throughout the lifespan of individuals with autism. Our data indicate that a deficit of neuronal soma volume in children with autism is associated with deficits in the volume of the neuronal nucleus and cytoplasm. The significant deficits of neuronal nuclear and cytoplasmic volumes in 13 of 16 examined subcortical structures, archicortex, cerebellum, and brainstem in 4- to 8-year-old autistic children suggest a global nature of brain developmental abnormalities, but with region-specific differences in the severity of neuronal pathology. The observed increase in nuclear volumes in 8 of 16 structures in the autistic teenagers/young adults and decrease in nuclear volumes in 14 of 16 regions in the age-matched control subjects reveal opposite trajectories throughout the lifespan. The deficit in neuronal nuclear volumes, ranging from 7% to 42% in the 16 examined regions in children with autism, and in neuronal cytoplasmic volumes from 1% to 31%, as well as the broader range of interindividual differences for the nuclear than the cytoplasmic volume deficits, suggest a partial distinction between nuclear and cytoplasmic pathology. The most severe deficit of both neuronal nucleus and cytoplasm volume in 4-to 8-year-old autistic children appears to be a reflection of early developmental alterations that may have a major contribution to the autistic phenotype. The broad range of functions of the affected structures implies that their developmental and age-associated abnormalities contribute not only to the diagnostic features of autism but also to the broad spectrum of clinical alterations associated with autism. Lack of clinical improvement in autistic teenagers and adults indicates that the observed increase in neuron nucleus and cytoplasm volume close to control level does not normalize brain function.

The increased life expectancy of individuals with Down syndrome (DS) is associated with increased prevalence of trisomy 21–linked early-onset Alzheimer’s disease (EOAD) and dementia. The aims of this study of 14 brain regions including the entorhinal cortex, hippocampus, basal ganglia, and cerebellum in 33 adults with DS 26–72 years of age were to identify the magnitude of brain region–specific developmental neuronal deficits contributing to intellectual deficits, to apply this baseline to identification of the topography and magnitude of neurodegeneration and neuronal and volume losses caused by EOAD, and to establish age-based staging of the pattern of genetically driven neuropathology in DS. Both DS subject age and stage of dementia, themselves very strongly correlated, were strong predictors of an AD-associated decrease of the number of neurons, considered a major contributor to dementia. The DS cohort was subclassified by age as pre-AD stage, with 26–41-year-old subjects with a full spectrum of developmental deficit but with very limited incipient AD pathology, and 43–49, 51–59, and 61–72-year-old groups with predominant prevalence of mild, moderately severe, and severe dementia respectively. This multiregional study revealed a 28.1% developmental neuronal deficit in DS subjects 26–41 years of age and 11.9% AD-associated neuronal loss in DS subjects 43–49 years of age; a 28.0% maximum neuronal loss at 51–59 years of age; and a 11.0% minimum neuronal loss at 61–72 years of age. A total developmental neuronal deficit of 40.8 million neurons and AD-associated neuronal loss of 41.6 million neurons reflect a comparable magnitude of developmental neuronal deficit contributing to intellectual deficits, and AD-associated neuronal loss contributing to dementia. This highly predictable pattern of pathology indicates that successful treatment of DS subjects in the fourth decade of life may prevent AD pathology and functional decline.