Transcriptional "silencer" element in rat repetitive sequences

The enhancer elements from either simian virus 40 or murine sarcoma virus activate the expression of a transfected rat insulin 1 (r11) gene when placed within 2.0 kilobases or less of the rl1 gene cap site. Inclusion of 4.0 kilobases of upstream r11 sequence, however, results in a substantial reduction in the enhancer-dependent insulin gene expression. These observations suggested that a negative transcriptional regulatory element was present between 2.0 and 4.0 kilobases of the r11 sequence. To test this notion, we employed a heterologous enhancer-dependent transcription assay in which the simian virus 40 72-base-pair repeat is linked to a human f8-globin gene. Addition of the upstream rI1 element to this system decreased the level of enhancer-dependent fl-globin transcription by a factor of 5 to 15. This rIb "silencer" element functions in a manner relatively independent of position and orientation and requires a cis-dependent relationship to the transcription unit on which it acts. Thus, the silencer sequence seems to have a number of the characteristics of enhancer elements, and we suggest that it may function by the converse of the enhancer mechanism. The rIb silencer sequence was identified as a member of a long interspersed rat repetitive family. Thus, a potential role for certain repetitive sequences interspersed throughout the eukaryotic genome may be to regulate gene expression by retaining transcriptional activity within defined domains. Enhancers are cis-dependent DNA sequences that activate the transcription of viral and cellular genes in a manner relatively independent of position and orientation (1, 2). This transcriptional activation can occur over distances as great as 6 kilobases (kb) (3, 4) from either the 5' or 3' end of the gene. What appears to be more important than the distance between an enhancer and the promoter being assayed for transcription are the specific sequences that intervene. Thus, for example, if an enhancer is flanked on both sides by active promoters, its effect on more distal promoter elements is likely to be significantly reduced (4-6). In studying the sequences upstream of the rat insulin 1 (rIb) gene, we have identified an element that appears to suppress enhancer-dependent transcriptional activity by a distinct mechanism. In fact, this negative regulatory element, which we will refer to as a silencer (see ref. 7), appears to have many of the characteristics of enhancers including positionand orientation-independent function on a heterologous gene. The identification of this rI, silencer element as a member of a family of long interspersed rat repetitive sequences (LINES; refs. 8 and 9) suggests a potential role for some of these repetitive elements in gene regulation. MATERIALS AND METHODS Plasmid Constructions. The recombinant plasmids used in this study were constructed according to methods that use standard recombinant DNA technology (10). The parental vector irSVHPA128 (kindly provided by M. Green and T. Maniatis) contains the human P3-globin gene with its promoter including 128 base pairs (bp) of 5' sequence upstream of the cap site and 500 bp of 3' flanking sequence. This plasmid also contains irVx sequences with the simian virus 40 (SV40) origin fragment including the enhancer element (Pvu II to HindIII) at the 3' end of the P-globin gene (see Fig. 1). Between the SV40 ori and the 5' j3-globin flanking sequences are approximately 900 bp of prokaryotic DNA sequence necessary for ampicillin and tetracycline selection in the bacterial host MC1061/P3. Two derivative plasmids were constructed that contained Bgl II restriction sites on either the 5' side of P-globin at bp -128 relative to the cap site (rrSVx5'Bgl II) or the 3' side of bp 500 from the 18-globin polyadenylylation signal (rrSVx3'Bgl II). These two plasmids were used to construct the derivative plasmids shown in Fig. 1. Briefly, irSVxA contains the 3.2-kb upstream flanking sequences from the rat insulin 1 gene (Pvu II-Pvu II) at the 5' end of the 3-globin coding sequence (Bgl II site of 7rSVxS'Bgl II) in the sense (wSVxAs) or antisense (7rSVxAa) orientations; wrSVxB contains the same fragment at the 3' end of the P-globin coding sequences (Bgl II site of nrSVx3'Bgl II) in the sense (irSVxBs) or antisense (irSVxBa) orientation. Two derivative plasmids were constructed from rrSVxAs with internal deletions of either the Sac I sites (irSVxASac) or the Bgl II sites (rSVxABgl II) (Fig. 1). The plasmid rSVxD contains the BamHI-Bgl II fragment from the upstream sequences of the rat insulin gene, and irSVxE contains the 700-bp Bgl II fragment from the adjacent upstream region inserted into rrSVxS'Bgl II. The plasmid irSVxC contains the 1-kb HindII fragment from the region immediately adjoining the insulin enhancer (ref. 11 and L.L., unpublished data) inserted into the parental plasmid irSVxS'Bgl II. As a control plasmid (irSVxMP3G), the 3.2-kb Bgl II fragment from the 3' end of the mouse P-globin gene was isolated and inserted into the Bgl II site of nrSVx5'Bgl II. The plasmids described in Table 1, pSV40.rI-B, pSV40.rI-R, pSV40.rI,-H contain the segments of the rI, gene extending from an Xba I site at the 3' end to BamHI, EcoRI, and HindlI, respectively, at the 5' end, inserted into plasmid pA10. The SV40 72-bp monomer enhancer was located at the 5' end of these constructs. The plasmids pMSV.rI,-B and pMSV.rI,-H contain the fragments deAbbreviations: SV40, simian virus 40; MSV, murine sarcoma virus; kb, kilobase(s); bp, base pair(s); LINES, long interspersed rat repetitive family. *Present address: Howard Hughes Medical Institute and Department ofMolecular Genetics and Cell Biology, The University of Chicago, Chicago, IL 60637. 3151 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. Proc. Natl. Acad. Sci. USA 83 (1986)