BrewerIX enables allelic expression analysis of imprinted and X-linked genes from bulk and single-cell transcriptomes

Abstract

Genomic imprinting and X chromosome inactivation (XCI) are two prototypical epigenetic mechanisms whereby a set of genes is expressed mono-allelically in order to fine-tune their expression levels. Defects in genomic imprinting have been observed in several neurodevelopmental disorders, in a wide range of tumours and in induced pluripotent stem cells (iPSCs). Single Nucleotide Variants (SNVs) are readily detectable by RNA-sequencing allowing the determination of whether imprinted or X-linked genes are aberrantly expressed from both alleles, although standardised analysis methods are still missing. We have developed a tool, named BrewerIX, that provides comprehensive information about the allelic expression of a large, manually-curated set of imprinted and X-linked genes. BrewerIX does not require programming skills, runs on a standard personal computer, and can analyze both bulk and single-cell transcriptomes of human and mouse cells directly from raw sequencing data. BrewerIX confirmed previous observations regarding the bi-allelic expression of some imprinted genes in naive pluripotent cells and extended them to preimplantation embryos. BrewerIX also identified misregulated imprinted genes in breast cancer cells and in human organoids and identified genes escaping XCI in human somatic cells. We believe BrewerIX will be useful for the study of genomic imprinting and XCI during development and reprogramming, and for detecting aberrations in cancer, iPSCs and organoids. Due to its ease of use to non-computational biologists, its implementation could become standard practice during sample assessment, thus raising the robustness and reproducibility of future studies.

Chemical conversion of human conventional PSCs to TSCs following transient naive gene activation

“…Among them, we found placental transcripts, such as FAM20A, PDE4D and ZC3H12C (Okae et al , 2014 ; Fig 1G ). Placental imprinted genes are highly expressed in the trophoblast cells of the placenta (Hanna, 2020 ; Martini et al , 2022 ), so we asked whether other markers of the TE lineage were induced after chemical resetting. We observed that 35 out of 107 expressed genes ( P ‐value = 4.27 × 10 −6 , Fisher's exact test) specifically expressed by TE cells in the human embryo (Xiang et al , 2020 ) were highly induced after chemical resetting (Fig 1H ), including KRT19, DAB2, LRP2 and TFAP2A (Fig 1H ).…”

Variable allelic expression of imprinted genes at the Peg13, Trappc9, Ago2 cluster in single neural cells

“…Analysis of imprinted gene expression on a bulk tissue level raises the question of whether the observed allelic bias is reflected in every cell of the lysate in the same way, or whether individual cells differ in their mono-/bi-allelic transcriptional status of the gene and, thus, deviate from the tissue average. Large-scale single-cell imprinted gene expression analysis is still in its infancy, but novel approaches indicate that not all cells of a tissue might show the same allelic expression status ( Martini et al, 2022 ). To address this question specifically for the Peg13 imprinting cluster, we isolated single NSCs from C57BL/6J × Cast/EiJ neurospheres as well as single neurons differentiated from these in vitro .…”

“…Due to advances in technology, especially single-cell transcriptomics and highly sensitive in situ hybridization methods, it has now become possible to investigate imprinted gene expression on the cellular level ( Varrault et al, 2020 ; Martini et al, 2022 ). Instead of a single-cell RNA-seq approach, we used the sc-GEM method ( Lorthongpanich et al, 2013 ; Cheow et al, 2015 ; Cheow et al, 2016 ) for a more limited analysis of the genes of this imprinting cluster in single cultured NSCs and differentiated neurons.…”

The role of imprinting genes’ loss of imprints in cancers and their clinical implications

“…Analysis of existing RNA-seq datasets can be used to identify ASE of imprinted genes beyond evaluation of gene expression, thereby detecting the LOI of imprinted genes ( 150 ). However, when heterogeneous populations of cells, such as cancer samples, are analyzed, only single-cell measurements allowed the detection of widespread LOI events ( 151 ). Therefore, the use of effective and appropriate data analysis methods to analyze single-cell transcriptomic data will provide a major advantage in the analysis of tumor epigenetic aberrations.…”

“…Therefore, the use of effective and appropriate data analysis methods to analyze single-cell transcriptomic data will provide a major advantage in the analysis of tumor epigenetic aberrations. For example, BrewerIX, a standardized approach for the analysis of known imprinted genes, can be used to analyze RNA-seq data from single breast cancer cells to identify LOI of imprinted genes ( 151 ). Differential allelic expression using single-cell data (DAESC), a powerful method for differential ASE analysis using single-cell RNA sequencing (scRNA-seq) from multiple individuals, is capable of analyzing genes with differential ASE in pancreatic endocrine cells from patients with type 2 diabetes and controls, taking into account the effect of allelic switching, although it is not suitable for estimating cancer cells ( 152 ).…”