ABSTRACT: Down syndrome (DS) is a neurodevelopmental disorder caused by an extra copy of human chromosome 21 (HSA21). People with Down syndrome have poor motor skills and speech, impaired adaptive behavior, and cognition deficits. (Carlesimo et al., 1997; Contestabile et al., 2010; Roizen and Patterson, 2003; Sherman et al., 2007; Vicari et al., 2000). At the brain anatomical level, DS is characterized by structural deficiencies in diverse regions. These deficits include hippocampal and cortical growth reduction, abnormal lamination and differentiation of neurons, decreased dendritic branching and spine density, and abnormal synaptic plasticity (Kazemi et al., 2016; Rachidi and Lopes, 2011). Even though the genetic cause of DS was discovered in 1959 (Megarbane et al., 2009), we still do not know the precise mechanisms by which triplication of HSA21 genes leads to neuroanatomical and neurobehavioral phenotypes in DS individuals. Moreover, a fundamental question in DS is to discover the impact of the triplication of chromosome 21 on the global expression regulation of non-triplicated genes (Letourneau et al., 2014) and how these profound changes at the genetic level translate to phenotypic changes in DS individuals. Finally, drug treatments targeting various putative molecular pathways are effective in rescuing cognition in DS animals. Nevertheless, while the preclinical studies performed in diverse mouse models of DS have reported improvement in cognitive deficits, the same drugs have failed to replicate all the good and promising effects when administered to people with DS during clinical trials (Rueda et al., 2020). Therefore, there is a notable need in strongly extending our understanding of DS in humans. Global gene expression analysis at the mRNA or protein level with different techniques have reported interesting findings on disparate DS-derived samples (fibroblasts, circulating immune cells, and amniocytes) (Guedj et al., 2016; Stamoulis et al., 2019; Waugh et al., 2019; Lanzillotta et al., 2020; Liu et al., 2017. Liu et al., 2018; Sommer et al., 2008) or iPSC-derived neurons obtained from DS individuals (Gonzales et al., 2018; Huo et al., 2018; Sobel et al., 2019). However, only a few studies have directly assessed global mRNA dysregulation in the adult DS brains by microarray hybridization (Lockstone et al., 2007; Olmos-Serrano et al., 2016). Although the latter studies have provided interesting information about the significant cellular pathways altered in trisomy, they lacked the depth, sensitivity, and isoform specificity that characterize modern gene expression analysis by next-generation sequencing (NGS). In this regard, another study used single-nucleus RNA sequencing (snRNA-seq) to profile aging DS brains (Palmer et al., 2021). Although this cutting-edge technique can deliver high-level information for the characterization of cellular diversity in brain tissues, this comes at the cost of profiling less number of genes (Bakken et al., 2018) and the possibility of missing the specific signature of important pathological processes (Thrupp et al., 2020). Moreover, none of these studies addressed the other side of the coin, assessing the corresponding changes in protein expression. In fact, the question of how well the transcriptome expression profiles match with those at the protein level remains largely unanswered. Large-scale analysis of transcriptome and proteome of the human brain would therefore provide a more comprehensive data-driven approach to identify dysregulated biological processes or genes/proteins co-expression networks as drivers of disease pathogenesis and prioritize the development of corresponding therapeutic interventions. Here, we took a multi-level data acquisition and interpretation approach to integrate in-depth transcriptomics and proteomics data in parallel from the same set of human brain samples to identify dysregulated expression networks in DS, a complex and multigenic neurodevelopmental disorder. We also provide unprecedented evidence of changes in mRNA transcript (isoform) expression, and to our knowledge, a first integrative analysis to identify alternative splicing, RNA binding proteins, and miRNAs as crucial intermediate regulatory steps in DS. Our data provide molecular information signatures that consistently repeat at gene, transcript, and protein levels and may function as critical players in DS pathophysiology. In particular, our multi-level data pinpoint to alteration in cell projection genes that translate into defects in axon and dendrites formation, which we find in DS iPSC derived neurons and primary hippocampal culture from a murine DS model.
Martina Bartolucci 
