Comparative genomics adopts the assumption that important biological processes are often conserved across related species. Based on that, scientists use animal models to infer human physiological and genetic properties.1–3
Sequence comparison is the most popular tool for comparative genomics. However, sequence similarity is not necessarily proportional to functional similarity.4
The biological functions of a gene not only rely on its molecular functions but also its spatiotemporal expression pattern. Changes in gene expression often mean changes in function.5
One example is that, for duplicate genes, which are usually associated with highly consistent coding sequences but diverse biological functions, there is only a weak correlation between rates of sequence divergences and rates of expression divergences.6
It is urgent to make the details of gene expression evolution clear for the aim of making proper functional inferences across species.
Microarrays, which can characterize the transcriptional profiles of tens of thousands of genes simultaneously, have been widely used in biomedical7–9
and comparative genomic10–12
studies. In the latter applications, studies of gene expression levels in different species often rely on cross-species hybridization.13–16
This method is limited to closely related species as it is based on the hybridization of target RNA and gene probes designed for other species,17
and when the probe and target RNA sequences are inconsistent to some extent, this method fails. Even in related species, several studies18,19
found that this approach may be problematic.
Using microarray data, some theories on gene expression evolution across genomes have been suggested. Yanai et al20
found that no expression conservation exists in human and mouse orthologous gene pairs because the evolution of expression profiles of orthologous gene pairs is comparable to that of randomly paired genes. Khaitovich et al14
suggested that the majority of expression divergences between species are selectively neutral and are of no functional significance. The above two studies deviated from the idea that genes should be expressed properly to conduct their functions and that basic biological processes are often conserved between related species. Jordan et al21
suggested that gene expression divergence among mammalian species is subject to the effects of purifying selective constraint, and it could also be substantially influenced by positive Darwinian selection. Liao and Zhang22
found that the expression profile divergence for the majority of orthologous genes between humans and mice is significantly lower than expected under neutrality and is correlated with the coding sequence divergence.
Another issue that should be addressed on the study of gene expression evolution is the relationship between promoter evolution and gene expression evolution. While the premise that the differences in upstream regulatory sequences represent gene expression divergence is widely accepted by researchers, several studies have shown that the changes in transcription factor binding sequences (TFBSs) have only little effect on gene expression evolution.23–26
The diverse conclusions on gene expression evolution may be due, in part, to the improper comparisons of gene expression patterns across genomes. Expression data should not be compared across probes directly.22
Some scientists seek indirect methods, which can make the expression data comparable across probes and even across platforms or species. The conservation of gene co-expression patterns across species has been widely surveyed.27–30
However, co-expression shows little information on the expression conservation or evolution of orthologous genes across species. To overcome these obstacles, Liao and Zhang22
introduced the relative mRNA abundance among tissues (RA) and extracted 26 common tissues between humans and mice to make cross-species expression comparisons possible; Dutilh et al31
and Tirosh and Barkai32
used either all or most one-to-one orthologs as referred sets for facilitating the gene expression comparisons across genomes.
In this study, we investigated several aspects of gene expression evolution between human and mouse genomes based on olignonucleotide microarray data of humans and mice generated by Su et al,33
which is widely used and is one of the largest data sets for humans and mice.21,22,34,35
Two methods presented by Liao and Zhang22
and Dutilh et al31
were adopted and compared for the aim of making reliable conclusions.