Determining how transcriptional regulatory signals are encoded in vertebrate genomes is essential for understanding the origins of multi-cellular complexity; yet the genetic code of vertebrate gene regulation remains poorly understood. In an attempt to elucidate this code, we synergistically combined genome-wide gene expression profiling, vertebrate genome comparisons, and transcription factor binding site analysis to define sequence signatures characteristic of candidate tissue-specific enhancers in the human genome. We applied this strategy to microarray-based gene expression profiles from 79 human tissues and identified 7,187 candidate enhancers that defined their flanking gene expression, the majority of which were located outside of known promoters. We cross-validated this method for its ability to de novo predict tissue-specific gene expression and confirmed its reliability in 57 of the 79 available human tissues, with an average precision in enhancer recognition ranging from 32 percent to 63 percent, and a sensitivity of 47 percent. We used the sequence signatures identified by this approach to assign tissue-specific predictions to ~;328,000 human-mouse conserved noncoding elements in the human genome. By overlapping these genome-wide predictions with a large in vivo dataset of enhancers validated in transgenic mice, we confirmed our results with a 28 percent sensitivity and 50 percent precision. These results indicate the power of combining complementary genomic datasets as an initial computational foray into the global view of tissue-specific gene regulation in vertebrates.

Genomic sequence comparisons between human, mouse and pufferfish (Takifugu rubripes (Fugu))have revealed a set of extremely conserved noncoding sequences. While this high degree of sequence conservation suggests severe evolutionary constraint and predicts a lack of tolerance to change in order to retain in vivo functionality, such elements have been minimally explored experimentally. In this study, we describe the in-depth characterization of an ancient conserved enhancer, Dc2 located near the dachshund gene, which displays a human-Fugu identity of 84 percent over 424 basepairs (bp). In addition to this large overall conservation, we find that Dc2 is characterized by the presence of a large block of sequence (144 bp) that is completely identical between human, mouse, chicken, zebrafish and Fugu. Through the testing of reporter vector constructs in transgenic mice, we observed that the 424 bp Dc2 conserved element is necessary and sufficient for brain tissue enhancer activity. In vivo analyses also revealed that the 144 bp 100 percent conserved sequence is necessary, but not sufficient, to replicate Dc2 enhancer function. However, the introduction of two separate 16 bp insertions into the highly conserved enhancer core did not cause any detectable modification of its in vivo activity. Our observations indicate that the 144 bp 100 percent conserved element is tolerant of change at least at the resolution of this transgenic mouse assay and suggest that purifying selection on Dc2 sequence might not be as strong as we predicted or that some unknown property also constrains this highly conserved enhancer sequence.

The functional importance of the approximately 98 percent of mammalian genomes not corresponding to protein coding sequences remain largely un-scrutinized 1. To test experimentally whether some extensive regions of non-coding DNA, referred to as gene deserts 2-4, contain critical functions essential for the viability of the organism, we deleted two large non-coding intervals, 1,511 kb and 845 kb in length, from the mouse genome. Viable mice homozygous for the deletions were generated and were indistinguishable from wild-type littermates with regards to morphology, reproductive fitness, growth, longevity and a variety of parameters assaying general homeostasis. Further in-depth analysis of the expression of genes bracketing the deletions revealed similar expression characteristics in homozygous deletion and wild-type mice. Together, the two deleted segments harbour 1,243 non-coding sequences conserved between humans and rodents (>100bp, 70 percent identity). These studies demonstrate that some large-scale deletions of non-coding DNA can be well tolerated by an organism, bringing into question the role of many human-mouse conserved sequences 5,6, and further supports the existence of potentially "disposable DNAi" in the genomes of mammals.

Rationale

Objective

Methods and results

Conclusions

The identification of enhancers with predicted specificities in vertebrate genomes remains a significant challenge that is hampered by a lack of experimentally validated training sets. In this study, we leveraged extreme evolutionary sequence conservation as a filter to identify putative gene regulatory elements and characterized the in vivo enhancer activity of human-fish conserved and ultraconserved1 noncoding elements on human chromosome 16 as well as such elements from elsewhere in the genome. We initially tested 165 of these extremely conserved sequences in a transgenic mouse enhancer assay and observed that 48 percent (79/165) functioned reproducibly as tissue-specific enhancers of gene expression at embryonic day 11.5. While driving expression in a broad range of anatomical structures in the embryo, the majority of the 79 enhancers drove expression in various regions of the developing nervous system. Studying a set of DNA elements that specifically drove forebrain expression, we identified DNA signatures specifically enriched in these elements and used these parameters to rank all ~;3,400 human-fugu conserved noncoding elements in the human genome. The testing of the top predictions in transgenic mice resulted in a three-fold enrichment for sequences with forebrain enhancer activity. These data dramatically expand the catalogue of in vivo-characterized human gene enhancers and illustrate the future utility of such training sets for a variety of iological applications including decoding the regulatory vocabulary of the human genome.