Massively parallel decoding of mammalian regulatory sequences supports a flexible organizational model

Abstract

Despite continual progress in the cataloging of vertebrate regulatory elements, little is known about their organization and regulatory architecture. Here we describe a massively parallel experiment to systematically test the impact of copy number, spacing, combination and order of transcription factor binding sites on gene expression. A complex library of ~5,000 synthetic regulatory elements containing patterns from 1 2 liver-specific transcription factor binding sites was assayed in mice and in HepG2 cells. We find that certain transcription factors act as direct drivers of gene expression in homotypic clusters of binding sites, independent of spacing between sites, whereas others function only synergistically. Heterotypic enhancers are stronger than their homotypic analogs and favor specific transcription factor binding site combinations, mimicking putative native enhancers. Exhaustive testing of binding site permutations suggests that there is flexibility in binding site order. Our findings provide quantitative support for a flexible model of regulatory element activity and suggest a framework for the design of synthetic tissue-specific enhancers.

High-throughput functional comparison of promoter and enhancer activities

“…S13A,B; Supplemental Table S4). Several motifs produce no enhancer or promoter activity at all, consistent with a previous observation that only a subset of TFs can generate transcriptional activity on their own (Smith et al 2013). Motifs for EGR, CREB, and RFX families of TFs stand out as having the highest promoter activity, and they also generate enhancer activity.…”

“…The enhancer and promoter activities that we observe may be influenced by cryptic enhancers present in our specific MPRA plasmids, by the short tile lengths, or by the viral transduction used to introduce libraries into cells. We speculate that the ability of a specific transcriptional activator or coactivator to drive promoter activity may be context dependent, as is the ability to encode transcriptional activation generally (Smith et al 2013). …”

Linear models enable powerful differential activity analysis in massively parallel reporter assays

“…Characterization studies examine thousands of different putative regulatory elements that have a wide variety of sequence features and try to correlate these sequence features with measured activity levels (Grossman et al, 2017;Guo et al, 2017;Safra et al, 2017;Levo et al, 2017;Maricque, Dougherty, and Cohen, 2017;Groff et al, 2016;Ernst et al, 2016;White, Kwasnieski, et al, 2016;Ferreira et al, 2016;Fiore and Cohen, 2016;Farley et al, 2015;Kamps-Hughes et al, 2015;Dickel et al, 2014;Kwasnieski, Fiore, et al, 2014;Mogno, Kwasnieski, and Cohen, 2013;Gisselbrecht et al, 2013;White, Myers, et al, 2013;Smith et al, 2013). Typical statistical analyses use regression to study the impact of multiple features simultaneously.…”

Multiplex enhancer-reporter assays uncover unsophisticated TP53 enhancer logic

“…Whereas classically enhancer-reporter assays consist of cloning each enhancer one by one, first in vitro, later in vivo (Banerji et al 1981;O'Kane and Gehring 1987;Chiocchetti et al 1997;Dailey 2015), now hundreds to thousands of enhancers can be tested in parallel (Patwardhan et al 2009(Patwardhan et al , 2012Kwasnieski et al 2012;Melnikov et al 2012;Arnold et al 2013;Kheradpour et al 2013;Smith et al 2013;White et al 2013;Vanhille et al 2015). These methods can be broadly categorized in two groups, namely, massively parallel reporter assays (MPRA) utilizing barcodes as a measure of activity of synthesized enhancer fragments (Patwardhan et al 2009(Patwardhan et al , 2012Kwasnieski et al 2012;Melnikov et al 2012;Kheradpour et al 2013;Smith et al 2013;White et al 2013) and self-transcribing active regulatory region sequencing (STARR-seq) (Arnold et al 2013;Vanhille et al 2015).…”