Injected embryos were screened during the first 5 days post-fertilization using a Nikon SMZ1500 fluorescence stereomicroscope. Specific expression patterns were documented at 48 hpf and levels of expression were quantified by counting the number of embryos showing enhancer-specific expression. In order to control for overall background activity from the construct (ie, promoter, backbone) an empty pDB896 vector containing gata2 zebrafish promoter linked to the reporter gene but lacking an enhancer sequence was used. Any tissue-specific enrichment shown by enhancer-containing vectors over the activity shown by the empty control vector was considered enhancer-specific. Additionally, three negative regions were also cloned to check the specificity of the enhancer selection process. These regions were chosen randomly from the human genome to have low conservation with zebrafish and no other enhancer-selective feature, that is, no DNase I hypersensitivity, no H3K4me1 or H3K27ac signals and CAGE signal only at noise levels. In parallel, 5 selected human enhancers were also analyzed. See Supplementary Table 9 for a summary of zebrafish validations, including expression patterns, signal strengths and primers.
