Trends in Genetics
ReviewTuning in to the signals: noncoding sequence conservation in vertebrate genomes
Section snippets
Plundering the junk in our genome
Our understanding of the functional elements within the human genome was until recently largely restricted to protein-coding sequences. However, this fraction represents just a few percent of the total amount of DNA, fuelling a fierce debate as to the extent of the function of the remainder (Box 1). Now that we have access not only to the sequence of the human genome, but to the genomes of many other vertebrates with which to compare it, we can begin to explore the remaining 98%. However, apart
Lessons from the ENCODE project
Recently, more sophisticated algorithms have been developed to profile multiple sequence alignments, primarily across mammals; these include PhastCons [33], BinCons [24] and GERP [34]. Although distinct from one another, they all use some form of phylogenetic modelling and compare local rates of evolution (substitutions), either at individual base positions or across small windows, with calculated neutral rates across the length of the alignment for each species. These methods were used to
‘Raising the bar’ – increasing the stringency, decreasing the numbers
The alternative to casting a wide net to try to capture all sequences is to target a much smaller set of sequences and aim to determine their function more thoroughly. In this way, one can design different, more intensive ways of understanding functional mechanisms and ultimately build paradigms that can be applied back to the genome as a whole.
Two different approaches have been used to identify a more focused subset of conserved sequences and some attempts have been made to characterize these
Conserved noncoding elements in invertebrate genomes
Although CNEs and noncoding UCEs are among the most highly conserved sequences in vertebrate genomes, they have no apparent orthologous sequences in invertebrate genomes 36, 38, 40. However, whole genome comparisons of invertebrate genomes have revealed that they also contain numerous highly conserved noncoding elements 33, 40, 41. The phenomenon of ultraconservation of noncoding DNA elements is therefore not a vertebrate novelty but a feature that might be common among all metazoan genomes.
In silico identification of motifs and transcription factor binding sites
The prediction of sequence signatures or language within different sets of conserved noncoding sequences is difficult because we have little knowledge of their subsequence composition (this assumes, given lengths of hundreds of bases, that each conserved sequence is not a single functional unit) and because for the majority, function is unknown. Nevertheless, recent studies have demonstrated that using limited functional information, such as the tissues in which a conserved element can act as a
The evolution of conserved noncoding sequences
When did functional noncoding sequences first appear and have they changed throughout evolution? Tracing the evolutionary ancestry of noncoding sequences has presented several challenges. Unlike the comparison of coding sequences, where the patterns of substitutions and insertions/deletions (indels) can be weighted according to their overall effect on the protein-coding product, the comparison of noncoding sequences generally lacks proven informative grammar. To overcome this, Kim and Pritchard
What is the relationship between sequence conservation and function?
Although homology-based searching has proved a powerful approach, it has led to an overemphasis on the importance of sequence conservation in defining functional elements in the genome. Doubtless, certain types of noncoding sequence, for example UCEs and CNEs, exhibit very high levels of identity. By contrast, other noncoding regions, such as promoters, are poorly conserved and yet clearly have important functional roles. A recent study deleted four different UCEs from mice without any apparent
Concluding remarks: learning the language(s)
Clearly we have a lot to learn about the nature of conserved noncoding elements in the genome. Their identification has provided a starting point for the identification of functional elements in vertebrate genomes. This has proved a robust and in some cases, highly successful, strategy. However, two common misconceptions arise from this somewhat adventitious approach. First, not all functional elements can be captured by this means, not least because sequence conservation has low resolving
Acknowledgements
T.V. is funded by a Marie Currie Intra-European Fellowship for Career Development.
References (63)
- et al.
Identification of regulatory regions which confer muscle-specific gene expression
J. Mol. Biol.
(1998) An atomic model of the interferon-beta enhanceosome
Cell
(2007)Embryonic epsilon and gamma globin genes of a prosimian primate (Galago crassicaudatus). Nucleotide and amino acid sequences, developmental regulation and phylogenetic footprints
J. Mol. Biol.
(1988)Defining a genomic radius for long-range enhancer action: duplicated conserved non-coding elements hold the key
Trends Genet.
(2006)Highly conserved regulatory elements around the SHH gene may contribute to the maintenance of conserved synteny across human chromosome 7q36.3
Genomics
(2005)Regulation of even-skipped stripe 2 in the Drosophila embryo
EMBO J.
(1992)The eve stripe 2 enhancer employs multiple modes of transcriptional synergy
Development
(1996)Genomic cis-regulatory logic: experimental and computational analysis of a sea urchin gene
Science
(1998)Deletion of ultraconserved elements yields viable mice
PLoS Biol.
(2007)Metrics of sequence constraint overlook regulatory sequences in an exhaustive analysis at phox2b
Genome Res.
(2008)