Understanding how gene regulation evolves and the effect of DNA sequence changes in controlling when genes are switched on.

Promoter sequences are rapidly evolving

Figure 1. Distributions of core promoter corrected substitution rate ratios for primary TSSs that are orthologous between human and mouse (n = 1,190). The vertical black/red bar shows the distribution median (vertical black/red bar). (Schroder et al 2012).

Our research group has a well-established interest in the evolution of promoter sequences (Taylor et al 2006; Taylor et al 2008; Semple and Taylor 2009). We are members of the international FANTOM consortium, which we have been contributing to for over a decade (Okazaki et al 2002; Carninci et al 2005; Carninci et al, 2006; Suzuki et al, 2009; Forrest et al 2014). We also reported that divergent expression between humans and mice was not associated with concomitant sequence changes at the core promoter (Figure 1; Schroder et al 2012). Similarly, those promoters which show conserved expressed in the testis show an increased rate of nucleotide substitution upstream of the core promoter, which is consistent with an increased mutation rate at these regions (Young et al 2015). We are currently investigating the molecular mechanisms which drive this mutational pressure within the mammalian germline.



Figure 2. Proportion of promoters displaying various evolutionary outcomes in human and mouse. Samples are ordered by rank of human:mouse average promoter count per sample. The white line denotes the number of promoters with that tisuse bias or expression profile (right axis). (Young et al 2015).

We have extended our work on promoter evolution to investigate the complete gain and loss of functional promoters in both the human and mouse lineages (Forrest et al 2014), where we found that is remarkably common and can be detected at over 50% of genes shared between human and mouse (Figure 2; Young et al 2015). These turnover events are concentrated within the testis and immune systems and are enriched at genes which show evidence for positive selection acting on their coding sequences, which suggests that some of these may be evolving adaptively (Young, 2016). Our current work is using promoter catalogues from ever more distant species to determine the frequency and importance of promoter birth and death within conserved, ancestral DNA sequence.

The gain and loss of long-range regulatory elements

The gain and loss of long-range regulatory elements is also a common occurrence in the genome as demonstrated in macrophages by our recent work in collaboration with Wendy Bickmore’s lab at the Human Genetics Unit and David Hume’s lab at the Roslin Institute. Unlike at promoters, the turnover of binding sites for the glucocorticoid receptor is directly associated with divergent transcriptional responses to glucocorticoid stimulation (Jubb, Young 2016) in both human and mouse macrophages.

The role of enhancer RNAs (eRNAs) in gene regulation


Figure 3. Bidirectional transcription initiates around DHSs but is not a specific mark of active enhancers. Solid lines consider transcription initiation from the positive strand around DHS midpoints (x = 0) and dashed lines show transcription initiation from the negative strand. (bioRxiv).

Bidirectional transcription has been widely reported to emanate from enhancers, a specific class of noncoding regulatory element which positively drive gene expression from a distance. We have observed this transcriptional signal to be neither specific nor sufficient for identifying active regulatory regions with enhancer activity (Figure 3; bioRxiv preprint).  An ongoing project in the group is further investigating the precise temporal relationships between eRNA transcription, enhancer activity and gene regulation.

Nuclear export

We work in collaboration with researchers in Grzegorz Kudla’s group  to study post-transcriptional regulation of gene expression. By combining transcript sequence features, e.g. GC content, with genome-wide measures of gene expression and nuclear export of transcripts we hope to understand the rules which regulate differential gene expression across the genome, and how these can be disrupted in disease