Transposable elements (TEs) are repetitive genetic sequences that can move from one location in the genome to another. TE-derived sequences are abundant in eukaryotes and make up substantial fractions of their genomic DNA. TEs have long been dismissed as selfish DNA elements that make little or no contribution to the function of their host genomes 
. This idea was supported by theoretical demonstrations that TEs can persist and proliferate in a genome without providing any function or benefit to the host 
. In the last couple of decades however, a number of anecdotal cases of TEs contributing regulatory or coding sequences to the host genome were reported. This has led to the development of a more nuanced view of TE sequences, whereby the relationship between TEs and the host genome can be characterized as a continuum ranging from extreme parasitism to obligate mutualism with their host 
. Indeed, TEs have been implicated in numerous functions that benefit the human genome. One way in which TEs can provide functional utility to the host genome is by donating enhancer sequences that can regulate the expression of host genes.
Enhancers are distal regulatory sequences, found outside of proximal promoter regions, which can increase the expression of genes by interacting with transcription factors. There are a handful of studies that provide experimental evidence for the exaptation of TE sequences as functional enhancers in the human genome. The first example comes from a study in 1993 by Hambor et al.
which shows that an Alu element serves as part of an enhancer that up-regulates the CD8 alpha gene in accordance with its role in differentiation along the hematopoietic lymphoid lineage 
. A few years later another study reported that an L1 element sequence donates an enhancer to up-regulate the expression of the APOC (Apolipoprotein) gene by more than 10-fold in cultured hepatocyte cells 
. Similarly, ancient SINE elements have been shown to serve as enhancers in mammalian specific brain formation. Santangelo et al. demonstrated the selection of a MAR1 element as an enhancer for the POMC (Proopiomelanocortin) gene expressed in the pituitary gland of jawed vertebrates 
. Another gene FGF8 (fibroblast growth factor 8) has also been shown to be regulated by the AmnSINE1 element in mammalian neuronal tissues 
. A final study by Bejerano et al. showed that an ancient SINE element drives the expression of ISL1 (insulin gene enhancer protein) in an in-vivo
mouse enhancer assay 
In addition to the experimental evidence showing that individual TE sequences provide functional enhancers to host genomes, we previously found evidence to suggest that human TEs may provide numerous enhancer sequences genome-wide. Our prior analysis showed that TE sequences reside in a substantial fraction of DNaseI hypersensitive (DHS) sites 
. The location of DHS sites signal ‘open chromatin’ regions which are involved in the regulation of transcription such as promoters and enhancers 
. The genome-wide analysis of DHS revealed that 23% of these sites contain TE sequences and are associated with higher expression levels of nearby genes in CD4+
. These data suggested that TEs may provide a large number of regulatory sequences that can increase the expression of genes in various tissues. Given the evidence from the experimental cases of TE-derived enhancers and the presence of TE sequences in DHS sites genome-wide, our goal in this study was to further explore the contribution of TEs in donating enhancers to various human cell types.
Experimentally characterized active enhancers display a distinct pattern of chromatin modifications that is significantly different from other regulatory regions as well as the genomic background 
. Specifically, functionally active enhancers are enriched for a suite of individual histone modifications – H3K4me1, H3K4me2, H3K4me3, H3K9ac, H3K27ac – and their enrichment patterns can be used to predict novel enhancers 
. We used the chromatin signature of active enhancers to guide the search for putative TE-derived enhancers in two human hematopoietic cell lines, GM12878 and K562, characterized as part of the ENCODE project 
. We employed a computational approach to identify novel enhancers by building a training set based on ChIP-Seq tag counts of the five enhancer-characteristic histone modifications found over a set of previously defined enhancer regions. Using genome-wide histone modification maps for the GM12878 and K562 cell lines, we identified hundreds of enhancers donated by TEs in each cell line. We also investigated the functional effect of these enhancers on gene expression and observed that TE-derived enhancers play a role in regulating gene expression in a cell type specific manner.