Combined study of postmortem brain tissue and a brain organoid model reveals shared molecular changes in dup15q syndrome and autism

By Azra Jaferi, Ph.D.


A recent study (not yet peer-reviewed) found overlap in gene-expression changes in specific cell types in dup15q syndrome and autism, combining insights from two approaches, postmortem brain samples from the Autism BrainNet bank and an organoid model of early brain development.


Autism encompasses a wide range of conditions with differences observed in social skills, repetitive behaviors and/or verbal and non-verbal communication. The genetic bases for autism are highly heterogeneous. In most cases of autism, termed idiopathic autism, the exact molecular mechanisms by which differences at the level of the genome are manifested in the brain and behavioral differences in autistic individuals are unknown. However, autism is also associated with several genetic conditions (syndromes) which are characterized by a change in a single genetic locus or chromosome. In such cases, the change in the genome is thought to be the cause of autism in that individual. For example, in a rare genetic syndrome known as dup15q syndrome, which accounts for about 3% of all autism cases1, an abnormal duplication or rearrangement of genetic material on chromosome 15 (specifically, the 15q11.2-q13.1 region) leads to changes in brain development. Dup15q syndrome thus offers a unique opportunity for researchers to investigate the neurobiological mechanisms of autism in the specific context of a syndrome that is genetically defined.

Much of the research autism examining differences in brain tissue between individuals with autism and neurotypical individuals has not considered the many different types of cells, neurons and non-neuronal cells like glia, that exist in the human brain. Identifying cell-type specific changes would help to better understand the early molecular changes that may be driving ASD-related differences in brain development. In a recent not yet peer-reviewed study2, Arnold Kriegstein and colleagues at the University of California, San Francisco investigated how molecular changes in specific cell types compare across different subtypes of autism (dup15q syndrome and idiopathic autism) using postmortem brain tissue obtained from Autism BrainNet. They studied how these changes may unfold over the course of development using brain organoids, an in vitro cellular model for the molecular and cellular changes occurring during the early stages of brain development.


Kriegstein and colleagues measured gene-expression changes at the level of single cells using RNAseq, a technique that detects the complete set of RNA sequences in cells (known as the transcriptome). Differences in the transcriptomic profile of cells is used to identify different cell types in the brain. Comparing RNAseq data from autistic and non-autistic individuals can thus help to identify molecular pathways and cell types that could be altered in autism.

The researchers performed single-nucleus RNA sequencing of postmortem brain tissue samples from 11 dup15q patients (ages 8–39 years old) with an autism diagnosis, obtained from the Autism BrainNet brain bank, and 17 samples from a non-autistic group of individuals without neurodevelopmental conditions. Using stem cells from three dup15q patients (and from three non-autistic individuals without dup15q, as a control group) they generated cortical organoids and conducted single-cell RNAseq at three different timepoints that roughly correspond to different stages of early brain development.

Findings and implications for autism

In both postmortem tissue and organoids, the researchers found numerous gene expression changes in the dup15q groups compared to control groups. When further exploring the degree to which gene-expression changes in dup15q affect different cell types, they found that all excitatory brain cells (neurons) were highly affected, especially those that project over long distances to other brain areas. The types of excitatory neurons found to be most affected in dup15q postmortem samples were upper-layer and deep-layer projection neurons, while in cortical organoids, it was the early radial glia (neuronal progenitors) and deep-layer projection neurons, and these changes were maintained at postnatal stages of organoid development. The study also found that in organoids, glycolysis, a major energy-producing pathway in cells that is required for proper forebrain development is abnormally increased in excitatory deep-layer neurons in dup15q. This metabolic dysfunction, in turn, led to abnormal changes in the identities of deep-layer neurons during organoid cortical development. Collectively, the study’s findings showed that excitatory cortical neurons, especially projection neurons, in dup15q syndrome are particularly vulnerable to gene-expression changes during development.

Out of all the genes differentially expressed in dup15q, many overlapped with genes that have been strongly implicated in autism3, especially postnatally in excitatory upper-layer neurons, and in genes implicated in synaptic signaling. The researchers found that upper-layer projection neurons are similarly preferentially affected in idiopathic autism4. In addition to synaptic signaling, many of them were strongly involved in regulating gene expression by remodeling the structure of genetic material in the cell. When taking a closer look at gene-expression changes in dup15q depending on the specific brain region, the researchers found 472 genes across all cell types that were differentially expressed between dup15q and controls in a region-specific manner, with the majority found in the PFC, followed by the TC and ACC, and most occurring in excitatory neurons. Overall, this suggests that the mechanisms underlying both dup15q syndrome and autism are preferentially associated with gene-expression changes in specific cortical cell types, especially excitatory neurons involved in synaptic signaling, across different stages of development.

In sum, Kriegstein and colleagues combined the use of postmortem cortical samples and an organoid model of the early developing human cortex and uncovered molecular changes occurring not only in dup15q syndrome but also those that are shared between dup15q syndrome and idiopathic autism. Importantly, they pinpointed the specific cell types in which these changes are occurring during development, helping to set the stage for finding specific targets in the brain for potential autism treatments.


  1. Scoles H.A., Urraca N., Chadwick S.W., et al. Mol. Autism. 2, 19 (2011) PubMed
  2. Perez Y., Velmeshev D., Wang L., et al. bioRxiv [Preprint] Published online September 22, 2023:2023.09.22.559056. PubMed
  3. SFARI Gene. Accessed December 1, 2023.
  4. Velmeshev D., Schirmer L., Jung D., et al. Science 364 (6441):685-689 (2019) PubMed