Comparative transcriptomics in human and mouse

Abstract

Cross-species comparisons of genomes, transcriptomes and gene regulation are now feasible at unprecedented resolution and throughput, enabling the comparison of human and mouse biology at the molecular level. Insights have been gained into the degree of conservation between human and mouse at the level of not only gene expression but also epigenetics and inter-individual variation. However, a number of limitations exist, including incomplete transcriptome characterization and difficulties in identifying orthologous phenotypes and cell types, which are beginning to be addressed by emerging technologies. Ultimately, these comparisons will help to identify the conditions under which the mouse is a suitable model of human physiology and disease, and optimize the use of animal models.

Figure 1.. Homology of human and mouse genes and genomic elements. Orthologous genes between human and mouse can be identified based on sequence homology of coding exons.

Orthologous genes tend to have conserved exonic structure and exon lengths, but introns are generally shorter in mouse. There is some degree of conservation of alternative splicing patterns (Box 2), but species specific splicing events exist (green gene). Orthologous genes may have conserved expression profiles between the two species (green) or diverged expression (orange). The bar chart represents expression levels of the genes in different organs. Genes with homologous sequence within the same species are called paralogous. Paralogous genes may originate from gene duplication events and their exonic structure, sequence and expression may diverge with evolutionary time. Promoter sequences (upstream from genes) are less conserved than gene body sequences. Regulatory motifs may differ although regulatory networks may be conserved. Orthologous genomic regions (and elements) can be identified through whole genome alignments (pink). However, some elements cannot be aligned to the other species (different shades of grey), or can map in multiple locations (brown). Finally, some genomic regions can be aligned, but their function may not be conserved (blue).

Figure 2.. Simplified clustering of human and mouse tissue samples and variance decomposition of gene expression.

Samples can be clustered based on their transcriptional profiles. If a human organ (for example, liver or heart) has a more similar gene expression profile to the homologous mouse organ than to another human organ, the clustering is organ-dominated (a). Vice versa, if human organs have more similar gene expression profiles between each other than compared to their homologous mouse organs, the clustering is species-dominated (b). The variation of expression for each gene can be decomposed into the most contributing factors, in this case species and organs (c). Genes are distributed in a continuous way along these proportions of variation. Nonetheless, genes at the extremes of this distribution can be identified as genes with proportionally higher variation across species and lower across organs (orange) and genes with proportionally higher variation across species and lower across organs (green). If only the expression of one or the other set of genes is used for clustering, genes with proportionally higher variation across species or organs lead to a more species-dominated clustering, or organ-dominated clustering, respectively. d∣ Hierarchical clustering based on real gene expression data from different organs across mammals and chicken, performed with the entire set of orthologous genes across species, reveal organ-dominated clustering. Distances between samples can be visually represented also on a 2-D space through several dimensionality reduction techniques, such tSNE (e, same input as d, perplexity=4, iterations=1000), MDS (f, same input as d, euclidean distance) and PCA (g).

Figure 3.. Cellular composition of human (a) and mouse (b) pancreatic islets.

Humans and mice have a different composition of pancreatic islets of Langerhans. Insulin-producing beta cells make up to 80% of mouse islets, whereas they constitute only up to 50% of the human islets. By contrast, glucagon-producing alpha cells compose up to 40% of the human islets. Fluorescent-stained images are taken from. The expression of a given gene may appear different when the whole anatomical structure is profiled, whereas what actually changes is the relative abundance of cells of different types expressing that gene, and not the expression of a gene in a particular cell type.

Figure 4.. Multidimensional complexity of omics-layers integration across species.

Four-dimensional matrix illustrating possible experimental combinations of genomics features profiled in different sample types, across species and in dynamic conditions. Colored triangles represent combinations of factors for which experiments are already available. This information is just figurative and might not reflect the current status of published experiments across all public or private databases. This figure is adapted from.