By the early 1990s, a few examples of insulator sequences had been discovered and characterized in eukaryotic cells. These insulators were mostly found in Drosophila. To extend the work to vertebrates, Chung, Whitely, and Felsenfeld focused on features of the β-globin gene cluster that were conserved in chickens, mice, and humans. They noted that locus control regions (LCRs) near the 5′ end of the β-globin gene cluster served to attract enzymes that opened up the chromatin from the 5′ to the 3′ end of the cluster, a distance encompassing a few hundred thousand base pairs of the human genome. In principle, the LCR can serve that function—attracting enzymes to open up the chromatin to prepare it for transcription—in either direction. However, at the 5′ end of the cluster and beyond, the chromatin remained condensed. Something was blocking the chromatin remodeling in that direction. The investigators focused on a prominent and constitutive nuclease-hypersensitive site (HS) at the 5′ end of the cluster. The location of that site for the human and chicken systems is indicated by the black arrows in Figure 1, which shows the genes of the β-globin gene cluster in each organism.

756

The investigators constructed a series of vectors in which the constitutive hypersensitive region (denoted c in Figure 2) was placed at sites on either side of a gene conferring resistance to the antibiotic neomycin (γ-NEO in Figure 2). For eukaryotes, researchers generally use the related antibiotic geneticin, or G418, to kill cells. The γ-NEO gene confers resistance. The constructs are labeled X and Y in Figure 2. As a control, the investigators replaced the constitutive hypersensitive site (c) with a fairly random DNA segment of comparable length derived from λ phage (construct Z in Figure 2). They transfected human erythroleukemia cells with these constructs, isolated stably transfected cell lines (in which the construct had integrated at some random site in the genome), and counted the number of colonies produced when the cells were suspended in semisolid agar medium containing G418. As shown in the graph in Figure 2a, the number of G418-resistant colonies decreased when one or two of the chicken hypersensitive sites were inserted between the neomycin-resistance gene and the LCR.

The investigators made a second series of constructs, shown in Figure 2b. Here, they added a gene for resistance to the antibiotic hygromycin (TK-HYG). This was set up in each construct so that the gene would be expressed constitutively. The constructs were again transfected into the same human cell line, and the numbers of colonies growing in media containing either hygromycin or G418 were tallied. The ratio again revealed that the number of neomycin-resistant colonies was greatly reduced when the neomycin-resistance gene was flanked by the chicken hypersensitive site.