Fig.9.2 Transcriptional repression by DNA-binding repressors (R) and corepressors (Co-R). (A) Distal DNA-binding activators (A) can turn on transcription by contacting the preinitiation complex formed on the core promoter. (B) Quenching: a complex containing a short-range repressor (R) and a corepressors (Co-R) inhibits an activator-A (A) bound to an enhancer within 100 bp away from the repressor, while an activator-B (B), located more than 100 bp away from the repressor, can initiate transcription. (C) Direct repression: the repressor/corepressor complex binds within 100 bp from the transcription start site and directly blocks the promoter activity. (D) Competition: activators and repressors can compete for common binding sites. This repression does not require corepressors. (E) Long-range repression. The repressor/corepressor complex blocks transcription over the distances of more than 1 kb from its binding site. (F) Inhibition of activators by their heterodimeric partners.

Many liver specific genes are activated by a nuclear receptor HNF-4 (Hayashi et al., 1999). The homodimeric form of HNF-4 binds a direct repeat of hexamer sequence, typically AGGTCA or related sites (Jiang and Sladek, 1997). Other nuclear receptors, COUP-TFI and COUP-TFII (also known as ARP-1), wich are ubiquitously expressed, can recognize the HNF-4 binding sites (Kimura et al., 1993; Ladias et al., 1992; Mietus-Snyder et al, 1992). COUP-TFs inhibit HNF-4 dependent activation by competing for binding to the common sites in cotransfection assays. COUP-TFs can also occupy binding sites for other nuclear receptors, PPAR, VDR, TR, RAR, and SF1, leading to repression of their target genes (Park et al., 2003; Pereira et al., 2000).

Spl family proteins, Spl, Sp3, and Sp4, regulate gene expression through GC and GT boxes in cellular and viral promoters (Lania et al., 1997; Suske, 1999). These factors contain three highly conserved Cys2His2 zinc fingers that bind DNA with similar affinities. Spl and Sp3 are ubiquitously expressed, while Sp4 is a brain-specific factor. Spl and Sp4 were originally characterized as activators, while Sp3 acts as both a repressor and a weak activator. Cotransfection assays in Drosophila Schneider SL2 cells lacking endogenous Sp activity demonstrated that Sp3 can repress expression of a reporter gene activated by Spl via competition (Majello et al., 1997; Yu et al, 2003).

Brinker has a helix-turn-helix (HTH) DNA-binding domain and negatively regulates TGFbeta (Dpp) signaling pathway in Drosophila embryos and wing imaginal discs (Campbell and Tomlinson, 1999; .lazwinska et al., 1999; Minami et al, 1999). This signaling pathway triggers phosphorylation of the DNA-binding factor Mad (a member of the Smad family) and the activated Mad then turns on downstream target genes. Several lines of evidence indicate that Mad binding sites in the enhancers controlling zerknüllt (zen) and Ultrabitorax (Ubx) genes are also recognized by Brinker which therefore competes with Mad (Kirkpatrick et al., 2001; Rushlow et al, 2001; Sailer and Bienz, 2001).

Bicoid, a homeodomain protein, governs development of the anterior region of the early Drosophila embryo by activating gap genes, pair-rule genes, and other target genes (Ochoa-Espinosa et al., 2005). Some of the Bicoid binding sites located within embryonic enhancers overlap with binding sites for the gap gene products, Knirps, Giant, and Krüppel. Cotransfection assays show that increasing amounts of Knirps gradually eliminates Bicoid-mediated activation in a 16-bp element found within a Krüppel enhancer which contains overlapping sites for Bicoid and Knirps

(Hoch et al, 1992). In transgenic embryos, the 16-bp element fused to the lacZ reporter is repressed by overexpression of Knirps and is not expressed in the absence of Bicoid activity, consistent with a mechanism of repression by competition. Furthermore, in vitro gel shift assays, cotransfection assays, and reporter assays in transgenic embryos show that transcriptional activation by some Bicoid sites within the eve stripe 2 enhancer is repressed by Krüppel and Giant through competition (Arnosti et al., 1996a; Small et al, 1992; Small et al., 1991; Stanojevic et al, 1991). For the competition activity of Krüppel, the dCtBP corepressor is not required in transgenic embryos (Nibu et al., 2003).

It should be noted that most of the repressors mentioned in this section have active repression domains that are known to interact with corepressors. For example, N-CoR/SMRT is a corepressor for COUP-TFs, dCtBP is a corepressor for Knirps, Krüppel, and Brinker, and Groucho is a corepressor for Brinker.

C: Long-range Repression

Long-range repressors, such as Hairy (bHLH) and Dorsal, can function over distances of more than 1 kb to silence the transcription complex in the Drosophila embryo, thereby resulting in simple on/off patterns of gene expression (Fig. 9.2E) (Barolo and Levine, 1997; Cai et al., 1996). Dorsal is the major activator for genes expressed along the dorsal-ventral axis, but it also mediates long-range repression on the 600 bp zen ventral repression element together with additional DNA-binding factors, Cut, Dead Ringer, and Capicua (Dubnicoff et al., 1997; Jimenez et al, 2000; Valentine et al, 1998). A corepressor for long-range repressors is Groucho, a WD-repeat containing protein (Paroush et al., 1994; Dubnicoff et al., 1997), which interacts genetically and physically with dRpd3 (HDAC1) (Chen et al., 1999), although a genetic interaction between dRpd3 and hairy has not been observed (Rosenberg and Parkhurst, 2002). The Courey group has proposed a spreading model to explain how Groucho mediates repression (Courey and Jia, 2001; Song et al, 2004). A ternary complex containing repressors, Groucho, and dRpd3 removes acetyl residues from nearby histones. Subsequently, Groucho is polymerized and covers the chromatin template through the deacetylated histones, as it has been shown that Groucho predominantly binds to the deacetylated histones in vitro. The corepressor polymer ultimately leads chromatin to a repressed state.

dCtBP-dependent short-range and Groucho-dependent long-range repression are qualitatively different. Dorsal/Groucho mediated long-range repression is clearly dCtBP-independent (Nibu et al, 1998a). Repression of zen expression is normal in the dCtBP mutant embryo. Hairy/Groucho mediated long-range repression and Kriippel/dCtBP mediated short-range repression are also different (Nibu et al., 2001). For example, when Krüppel is misexpressed in the ventral region it represses the hairy 6 stripe enhancer in a dCtBP-dependent manner. In fact, disruption of the PEDLSMH motif, which is necessary for interaction with dCtBP, abolishes this repression. Addition of the Hairy repression domain including the Groucho interaction motif (WRPW) to the Krüppel protein leads to the repression of both stripes 5 and 6, indicating long-range repression. Thus, it is clear that Groucho and dCtBP mediate two separate modes of repression, long-range and short-range, respectively.

Recently, it has been clearly demonstrated that Drosophila Heterochromatin Protein 1 (HP1) mediates long-range repression (silencing) within euchromatin by two mechanisms (Danzer and Wallrath, 2004). HP1 spreads along DNA bidirectionally from its entry site and alters chromatin structure. The chromo domain of HP1 binds a methylated lysine 9 of histone H3 generated by the histone methyltransferase SU(VAR) 3-9. This study also demonstrates that repression by HP1 up to 1.9 kb is likely due to self-propagation, but repression over longer distances is SU(VAR)3-9-dependent.

0 0

Post a comment