X-Chromosome Inactivation and Autosomal Random Monoallelic Expression as "Faux Amis"

Abstract

X-chromosome inactivation (XCI) and random monoallelic expression of autosomal genes (RMAE) are two paradigms of gene expression regulation where, at the single cell level, genes can be expressed from either the maternal or paternal alleles. X-chromosome inactivation takes place in female marsupial and placental mammals, while RMAE has been described in mammals and also other species. Although the outcome of both processes results in random monoallelic expression and mosaicism at the cellular level, there are many important differences. We provide here a brief sketch of the history behind the discovery of XCI and RMAE. Moreover, we review some of the distinctive features of these two phenomena, with respect to when in development they are established, their roles in dosage compensation and cellular phenotypic diversity, and the molecular mechanisms underlying their initiation and stability.

Keywords: LINE-1 elements; X-chromosome inactivation; cellular diversity; dosage compensation; epigenetic silencing; random monoallelic expression; stochasticity.

FIGURE 1

Schematic representation illustrating the features of different types of random monoallelic expression: X-chromosome inactivation (XCI) and random monoallelic autosomal expression (RMAE). Xp, X chromosome of paternal origin, Xm, X chromosome of maternal origin; p, paternal autosome, m, maternal autosome.

FIGURE 2

A schematic representation of mechanisms responsible for X-chromosome inactivation (XCI) and mechanisms possibly responsible for random monoallelic autosomal expression (RMAE). The gray arrow represents a potential parallel between XCI and RMAE associated with LINE-1 (L1) elements.

FIGURE 3

Intersections of autosomal gene collections identified as random monoallelically expressed in the genome-wide studies described in Table 2 (except Jeffries et al., 2012, which is not publicly available). (A) Half-matrix showing all pairwise intersections. (B) Upset plot (Conway et al., 2017) showing gene collection intersections of the same studies. The lower part of the panel has a horizontal bar plot showing the number of elements on each study collection, and a right section with a dot matrix. Each dot represents unique gene intersections, i.e., each gene is represented only once in the dot matrix. The upper vertical bar plot is related to the dot matrix, showing the number of unique genes in each intersection (for instance, there are 500 MAE genes in the Gimelbrant et al. (2007) dataset, but only 388 of those are uniquely present in that dataset; similarly, the Li et al. (2012) dataset shares more than 40 MAE genes with Eckerley-Maslin (2014) NPC dataset, but those 40 are uniquely shared between those two sets). Intersections of size smaller than 4 are not represented. For a complete description of the intersections and gene listing, see the Supplementary File provided with this review. ASL, Astrocyte-like cells; NSC, Neural stem cells; NPC, Neural progenitor cells; ESC, Embryonic stem cells; SPC01, Clonal Neural stem cells (before epigenetic reprogramming); iPSC, induced Pluripotent stem cells after epigenetic reprogramming of SPC01. Note that “NPC” on Jeffries et al. (2016) are derived from iPSC. Colors represent instances where a different cell/tissue type was studied more than once. To obtain intersections, gene ids were manually curated for immediate inconsistencies (e.g., gene name-to-date conversions when data was originally provided in microsoft excel format). All gene sets were then parsed with the gprofiler2 R package (Raudvere et al., 2019) for gene id consistency, using transcript ids as query whenever possible, and ENSEMBL gene ids as target (performed July 12th, 2021). Orthology conversion (from human to mouse) was performed with the same package for datasets involving human data. For Gimelbrant et al. (2007) and Zwemer et al. (2012) gene collections, MAE classes I, II and III were used to retrieve RMAE genes, and for Gendrel et al. (2014), the “NPC_random_catalog” classification was retrieved as RMAE.