The objective of this study, was to analyse and to compare the transcriptome of brain samples from individuals with DS and euploid controls. For that purpose, we used data from a DNA microarray experiment GSE59630 the contained log2 expression values of 17537 human genes from postmortem brain samples of individuals with DS and samples from euploid controls. Here we found differences in gene expression along the whole transcriptome obtained from brain samples, not only in the genes from the chromosome 21; also, the analysis per brain areas showed that the hippocampus and the cerebellar cortex had the most different gene expression pattern when compared to the brain as a whole.
Our findings support the hypothesis of a systemic imbalance of brain protein homeostasis, or proteostasis network of cognitive and neuroplasticity process as an important effect of trisomy not only in loci of chromosome 21 but also in genes located in other chromosomes [29–30]. It is possible that an accumulation of toxic protein aggregates caused by a failed degradative system in DS neurons, negatively affects neuroplasticity processes in brain structures [31–37]. In this sense, our results extended the current knowledge frontier of the neurophysiological mechanisms involved in the disturbance of extensive gene expression, that are remodeling the functional gene networks interaction architecture in DS brains.
One of the most important findings for this study is the global over-expression of over 5000 genes across the transcriptome in every chromosome, not only in chromosome 21 nor even in the called “Down Syndrome Critical Region” as could be expected given that only that DS samples used had a full trisomy confirmed. There are some studies addressing this issue but mainly in murine models. Kahlem et al. [38] found in their study in mice that a significant fraction of genes was differentially regulated in a few tissues, suggesting additional mechanisms affecting gene expression in specific cell types. One of the possible explanations that we propose is the “cascade effect” in which over-expressed transcription factors or epigenome regulators such as HMGN1, located in chromosome 21, affect the expression of other genes located in different chromosomes and thus there is a loss of protein homeostasis in the brain. This could explain how the triplication of one of the smallest chromosomes with approximately 346 genes can cause the over-expression of 482 genes in the brain of Down sndrome. In fact, Kahlem et al. [38] found that most triplicated genes coding for DNA binding proteins, including transcription factors, chromatin proteins, and RNA binding proteins, were overexpressed by a factor of about 1.5-fold. It is worth noticing that not all genes from chromosome 21 are affected, this is given the dose-compensation presented and documented in trisomy 21, where we find genes that are dose-sensitive and others not sensitive.
Another interesting finding in the present study, was that GO-categories biological processes associated with the overexpressed genes were mainly focused on epigenetic processes such as DNA methylation and histone deacetylation. Nowadays it would be a mistake disregard the effect that epigenetics have, in combination with genetics, in the development of syndromes, and Down syndrome is not the exception [30]. Genome-wide methylation studies have identified epigenetic marks in different sample tissues from individuals with DS including skin fibroblasts, liver, placenta and brain among others [39].
According to our results, chromosome 21 had the highest percentage over-expressed genes in comparison to the total of protein coding genes found in the chromosome, followed by chromosome 18, 8 and X. These results suggest that full trisomy of chromosome 21 affects not only the expression of the genes within chromosome 21, but also the expression of other genes of another different chromosomes. The dysregulation found across the transcriptome could be a “cascade effect”, initially, due to the anomalous expression of genes in chromosome 21 that regulate the expression of other genes i.e. transcription factors. Specifically, the over-expressed genes in chromosome 21 and 18 are involved mainly in mitochondrial processes. Izzo et al. [40] report how mitochondrial dysfunction might affect the phenotype found in individuals with DS in aspects such as muscle hypotonia, intellectual disability and neurodegeneration, heart defects, type 2 diabetes and obesity, and immune disorders [41–42]. The study by Piccoli et al. [43] showed how in human primary lines of DS fetal fibroblasts, trisomy 21 perturbed the expression of genes involved in mitochondrial pathways, decreasing oxygen consumption and ATP content and increasing mtCa2+ load and ROS production. Likewise, Izzo et al. [40] in their study shows how ooverexpressed of human genes on chromosome 21 are directly or indirectly responsible for the pathogenesis of DS phenotypic features, given that, as we stated above, many genes located in chromosome 21 can affect the expression of other genes from different chromosomes. They focused specifically on the involvement of over-expressed genes such as DYRK1A, RCAN1, NRIP1 and ATP in mitochondrial function and energy conversion, leading to mitochondrial dysfunction and chronic oxidative stress which is consistently observed in individuals with DS [44].
According to our results, the expression pattern in the brain of individuals with DS during the pre-gestational period is completely different from the pattern during the late 30 to 40 years old as it would be expected. The brain during the embryogenesis is still in formation, rearranges in synaptic connection are made throughout the brain by changes in gene expression. In contrast, when a person reaches their 30–40 years, the brain is completely formed and even though they can learn new things and new synapsis connections can be made, the expression pattern does not change drastically. This difference was visible with the negative correlation found when these two age-ranks where compared. The epigenetic here plays a crucial role, the macro and microenvironment that surrounds both age-groups are completely different as shown in our results.