1

P2 TATA-box

E2F binding site

Nucleosomes 12 and 13

Fig. 3 (A) Nucleosomal structure of active and inactive human c-myc genes. Adapted from Pullner ei a/. (1996, Fig. 8, p. 31456). Shown are the nucleosomes 1-16, the promoters P0, P1, and P2, the DNAse I-hypersensitive sites II2, III1, and III2 and the nuclease resistant nucleosomal spacer region (black box) between nucleosomes 4 and 5. Dashed nucleosomes were identified only in submolar amounts (Pullner ei a/., 1996). Human promyelocytic HL-60 leukemia cells were differentiated toward granulocytes with DMSO (Pullner ei a/., 1996). The activity of the c-myc gene is indicated (+ = expressed; - = not expressed). The DNAse I-hypersensitive sites II2, III1, and III2 were only present on active c-myc genes of undifferentiated HL-60 cells (Pullner ei a/., 1996). (B) Hierarchal control of transcription from the c-myc P1 and P2 promoters. See text (Section III.D) for a description. SoB = sodium butyrate.

promoter (Bentley and Groudine, 1986a; Dyson et al., 1985; Eick and Bornkamm, 1986; Pullner et al., 1996; Siebenlist et al., 1988). The spacer region between nucleosomes 4 and 5 was found to be nuclease resistant only on active c-myc genes in undifferentiated proliferating HL-60 cells (Fig. 3A; Pullner et al., 1996). Similarly, it was protected from nuclease cleavage in serum-stimulated c-myc expressing human Hs68 primary fibroblasts but cleaved by micrococcal nuclease in c-myc non-expressing IMR32 neuroblastoma cells (Liu et al., 2006a). Furthermore, in two B-cell lines with stably transfected episomal c-myc genes it was nuclease resistant in the RF266C3 cells capable of expressing c-myc but nuclease sensitive in the MA76 cells, in which c-myc is silent (Michelotti et al., 1996b). This nucleosomal spacer region colocalizes with the FUSE (Pullner et al., 1996) that is single stranded, if c-myc is expressed, but duplex DNA, if c-myc is not expressed (Kouzine et al., 2004; Michelotti et al., 1996b). The single-stranded FUSE, but not its duplex form, is bound by the single-strand-specific transcriptional activator FBP (FUSE binding protein), whose expression profile parallels the one of c-Myc so that expression of both FBP and c-Myc decreases with similar kinetics during induced HL-60 cell differentiation (Bazar et al., 1995a,b; Braddock et al., 2002; Davis-Smyth et al., 1996; Duncan et al., 1994; Kouzine et al., 2004; Michelotti et al., 1996b). Since FBP binds to the FUSE, only if c-myc is expressed (Avigan et al., 1990; Bazar et al., 1995a; Duncan et al., 1994; Kouzine et al., 2004; Michelotti et al., 1996b), binding of FBP to the FUSE could explain the nuclease resistance of the spacer region between nucleosomes 4 and 5 only on active c-myc genes (Pullner et al., 1996). Alternatively, this nuclease resistance could reflect movement and heterogeneous positioning of nucleosome 5 (Liu et al., 2006a).

In c-myc non-expressing IMR23 neuroblastoma cells, the double-stranded FUSE (Michelotti et al., 1996b) is wrapped over nucleosome 5 and not occupied by FBP or FIR (FBP interacting repressor) (Liu et al., 2006a). Vice versa, the FUSE is nucleosome free and occupied by FBP and FIR in continuously c-myc expressing U2OS osteosarcoma and SW13 adrenal cortical carcinoma cells (Liu et al., 2006a). This chromatin remodeling at the FUSE is BRG-1-independent because SW13 cells are BRG-1-deficient.

A hierarchical control regulates P1 transcription, binding of paused Pol II complexes to the P2 promoter and processive transcriptional elongation by these P2-initiated Pol II complexes (Fig. 3B; Albert et al., 1997, 2001; Pullner et al., 1996): At episomal c-myc promoters, P1 transcription and transcriptional elongation by P2-initated Pol II complexes are activated by the HDAC inhibitor SoB (sodium butyrate) and thus probably repressed by HDAC (Albert et al., 1997, 2001; Pullner et al., 1996; Strobl et al., 1993; Wolf et al., 1995). This effect on transcription elongation at P2 depends at least in part on the E2F-binding site strongly suggesting the involvement of E2F-pocket protein-HDAC complexes (Albert et al., 2001; Harbour and Dean, 2000). In contrast, the effect on P1 transcription depends at least in part on the P2 TATA-box whereas the E2F-binding site has no influence on transcription from P1 (Albert et al., 2001). It has to be noted that the E2F-binding site of the c-myc promoter colocalizes with the binding sites for ETS-1/2, STAT3 (signal transducer and activator of transcription 3), NFATc1 (nuclear factor of activated T cells c1), KLF11 (Kruppel-like factor 11), and METS (mitogenic Ets transcriptional suppressor) (Fig. 4) so that the effects of mutation of the E2F site could be caused by changes in E2F, ETS-1/2, STAT3, NFATc1, KLF11, or METS binding. The nucleosomes 12 and 13 are present at inactive, uninducible c-myc promoters whereas both active and inactive, but inducible c-myc promoters lack them (Albert et al., 1997, 2001; Pullner et al., 1996). These two nucleosomes repress P1 transcription as well as transcription elongation by P2-initiated Pol II complexes and prevent induction of transcription from P1 and P2 by SoB (Albert et al., 1997,2001; Pullner et al., 1996). However, the absence of these two nucleosomes does not automatically result in activation of c-myc expression because both P1 transcription and transcriptional elongation by P2-initiated Pol II complexes still require induction by SoB at episomal c-myc promoters (Albert et al., 1997, 2001; Pullner et al., 1996). Nevertheless, binding of paused Pol II complexes to the P2 promoter requires neither the absence of nucleosomes 12 and 13 nor induction by SoB so that Pol II complexes pausing at the P2 promoter are also found at inactive c-myc promoters (Albert et al., 1997, 2001). The nucleosomes 12 and 13 seem to be under the control of the element ME1a1. Deletion of ME1a1 results in their appearance so that activation of c-myc transcription should depend on ME1a1 to eliminate them (Albert et al., 2001). Moreover, ME1a1 seems to control an additional nucleosome that appears at the P2 promoter if ME1a1 is deleted. The presence of this additional nucleosome results in complete inhibition of the c-myc promoter, that is, inhibition of Pol II binding to the P2 promoter and thus inhibition of transcriptional elongation by P2-initated Pol II complexes, inhibition of P1 transcription and prevention of SoB-induced transcription from P1 and P2 (Albert et al., 2001). Consequently, ME1a1 appears to function as a major regulator for the global chromatin structure at the c-myc promoter and thus for c-myc transcription (Albert et al., 2001). In summary, activation of c-myc expression should include ME1a1-mediated elimination of the additional nucleo-some at P2 allowing binding of promoter proximal paused Pol II complexes to P2 as well as ME1a1-mediated elimination of the nucleosomes 12 and 13 allowing SoB induction of both P1 transcription (at the level of initiation) and processive transcriptional elongation by P2-initiated Pol II complexes (Fig. 3B; Albert et al., 1997, 2001; Pullner et al., 1996).

Multiple c-Myb binding sites

Was this article helpful?

0 0

Post a comment