The brain is the most complex organ in mammals, characterised by diverse physiological functions, intricate cellular structures and extremely rich gene expression. Identifying the molecular composition of the brain across different regions, cell types and subcellular levels will significantly advance our understanding of brain function in both healthy and diseased states.
On 5 March (US time), a study titled “Maps of protein-coding genes in the brains of humans, pigs, and mice”, jointly conducted by the Karolinska Institute, the KTH Royal Institute of Technology, and BGI, was published in Science. The research integrates multiple transcriptomics approaches and antibody mapping technologies to provide a comprehensive molecular analysis of different brain regions and generate high-quality molecular maps of protein-coding genes.
The Human Protein Atlas (HPA) is a large-scale research initiative designed to integrate transcriptomics and proteomics to explore the spatial expression patterns of transcripts and proteins across different cells, tissues and organs. Currently, the Tissue Atlas includes only a limited number of human brain regions, such as the cerebral cortex, hippocampus, caudate nucleus and cerebellum. As a result, more in-depth research is required to better understand the complex molecular processes that drive cellular dynamics in the central nervous system.
This study successfully constructed a genetic map of the mammalian brain, providing an important complement to several existing brain mapping projects. To examine similarities and differences across evolutionary lineages, the researchers analysed gene expression across three representative species: primates (humans), artiodactyls (pigs), and rodents (mice).
A key component of the study was the integration of protein atlas data with transcriptomic datasets to analyse protein expression across multiple regions of the mouse brain. All datasets, including high-resolution images and metadata, were made publicly available to support further research.
The research team first normalised transcriptome data from the GTEx, CAGE, and HPA databases, together with pig and mouse transcriptome datasets, to remove batch effects and enable comparative analysis. The team then applied UMAP (Uniform Manifold Approximation and Projection), a dimensionality reduction algorithm, to analyse gene expression patterns across brain regions. Using the existing HPA brain atlas, pig and mouse transcriptome data were mapped to human homologous genes.
The brains of the three species were divided into ten regions: the olfactory bulb, cerebral cortex, hippocampus, amygdala, basal nucleus, hypothalamus, thalamus, midbrain, pons and medulla, and cerebellum. Clustering analysis showed that gene expression patterns in the brainstem regions (midbrain, thalamus, pons and medulla) were relatively similar across species, with smaller differences compared with the hypothalamus. In contrast, the cerebral cortex, hippocampus and amygdala displayed closer relationships with each other.
Based on gene expression specificity, the research team classified the molecular characteristics of different brain regions. Across all three species, the cerebellum contained the largest number of region-specific up-regulated genes. Further analysis grouped the genes according to major brain structures: brain, hypothalamus, brainstem and cerebellum. Hierarchical clustering of these genes supported existing theories of brain structure, grouping the hypothalamus and cerebellum together, while the brainstem clustered with the brain.
Figure 1. Gene expression characteristics in different brain regions of three species.
The research team identified 537 genes that are highly expressed in specific brain regions. These include well-known genes such as galanin, oxytocin and vasopressin in the hypothalamus, transcription factors TBR1, SATB2 and NEUROD6 in the brain, and Hox genes in the brainstem. Several other highly expressed genes were also identified, although their functions require further investigation.
Cross-species comparisons revealed additional insights. For example, the transcription factor TFAP2B, expressed in interneuronal cells, showed similar expression patterns across all three species. In contrast, secretagogin (SCGN), a calcium-binding protein expressed in the olfactory bulb, is also present in stellate cells in the human cerebellum but not in pigs or mice. These differences may be associated with species-specific brain functions.
Figure 2. Gene expression comparison between species.
Brain function relies on complex neural circuits composed of neurons with distinct chemical phenotypes. The researchers therefore analysed the distribution of cell-recognition genes across the three species, including 1,053 transcription factors, 63 genes related to neurotransmitter production, transport and clearance, and 118 known neurotransmitter and neuropeptide receptor genes.
Certain transcription factors were highly conserved across species, such as EMX1 and BHLHE22. However, other genes showed notable differences. For example, NEUROD1, an important gene involved in mouse cerebellar development, is expressed not only in the cerebellum but also in retinal cells of pigs and mice, while it is not expressed in human retinal cells.
Analysis of neurotransmitter-related genes revealed that the distribution of enzymes responsible for neurotransmitter synthesis is largely similar across the three species, reflecting the evolutionary conservation of mammalian brain structure. However, some differences were observed. For instance, genes related to G protein-coupled receptors (GPCRs) — which are targets for approximately 30% of prescription drugs — showed different expression patterns among species. These findings highlight the importance of considering species differences in animal models used for drug development.
Figure 3. Expression profiles of cell-recognition genes in the mammalian brain.
The researchers also compared brain gene expression with expression patterns across other tissues and organs in the body. Genes were classified into three groups: brain-enriched genes, genes enriched in other tissues and genes with low tissue specificity. Analysis of marker genes for neurons, astrocytes, oligodendrocytes and microglia suggested that some of these genes are also expressed in other tissues, emphasising the need for a broader systemic perspective in related research.
Figure 4. Expression patterns of characteristic genes across tissues in the brain.
Overall, the study integrates transcriptomics, protein mapping and other multidimensional datasets to provide a comprehensive analysis of protein-coding genes in the brains of humans, pigs and mice. The Human Brain Atlas (http://www.proteinatlas.org/brain) offers researchers worldwide an expanded resource for genomics studies and future exploration of mammalian brain biology.
MGI Sequencing Platform
This research was conducted using the MGISEQ-2000 and BGISEQ-500 sequencing platforms. The MGISEQ-2000 is a core model within the MGI sequencing portfolio, capable of generating 75–1440 Gb of data per run. The platform supports multiple read lengths, with a full-load PE150 run completed in as little as 38 hours, enabling applications in scientific research, clinical medicine, forensic science and agriculture.
The sequencer is based on MGI’s proprietary DNBSEQ technology, providing a flexible high-throughput sequencing platform that supports multiple sequencing modes while delivering high accuracy, low duplication rates and minimal index hopping. In some international markets, the MGISEQ-2000 is marketed as the DNBSEQ-G400.
