LAUSR

Background There are many types of cells in the brain. They form the structure and execute the function of the brain. Cell classification conventionally uses location, morphology, and electrophysiological characteristics, often combined with cell type-specific markers. The cell type-specific marker gene is important for identifying cells in the brain and understanding the cell-specific mechanism underlying psychiatric disorders. However, the definition of the cell type-specific marker is frequently inconsistent across sources and studies. The relationship between cell types and psychiatric disorders is not clear. Methods The transcriptome and proteomics data of brain cells from human and mouse, by RNA-Seq and microarray from acutely isolation cell and primary culture cell were collected. We followed the convention quality control and analytical processing on the raw data. We collected 543 “putative” marker genes for ten cell types that have been commonly used from literature, In Situ Hybridization (ISH) databases and antibody companies. We defined general marker gene as their expressions in claimed target cells is the highest across all cell types tested. The expression difference for the rigorous marker gene defined as dividing the second highest expression level across all the other cell types by the expression level in target cell type. We set the fold change threshold for rigorous marker gene as at least 2. Parietal cortex tissue specimens from the Stanley Medical Research Institute (SMRI) Neuropathology Consortium and Array collections included schizophrenia, bipolar disorder and control samples were used for Weighted Gene Co-Expression Network Analysis (WGCNA). Results With the datasets collected and the criteria provided, we found that 44 general marker genes showed stable specificity across all data collected. That included 29 neuron markers, eight astrocyte markers, seven oligodendrocyte markers. The averaged correlation values of specific marker genes between human and mouse, transcriptome and proteome, RNA-Seq and microarray, acutely isolation and primary culture were 0.50, 0.58, 0.51 and 0.43, respectively. According to the criterion for rigorous marker genes, 23 of the 44 marker genes showed more than two-fold changes in at least seven data sets evaluated. The most specific marker gene (fold change=710.31) is RELN, a neuron marker gene. It is not clear whether the non-specificity of rest of marker genes is related to data quality or technical artifacts in the data we can use for the assessment. In the WGCNA analysis, we found general marker genes of astrocyte were significantly enriched (p Discussion Based on transcriptome and proteomics of data of isolated cells, we confirmed a small set of 44 genes to be cell-type-specific in the brain while many other commonly-used marker genes will require additional studies to verify their specificity. Studies using these marker genes to tag cell types should exercise caution. Moreover, astrocyte marker genes enriched in disease-associated module co-expressed with known GWAS loci for schizophrenia and bipolar disorder. Further study and more data is need for evaluation of the rest non-specific marker genes and the co-expression of brain cell marker genes with psychiatric disorders.