sign in sign up
<<
Home , DNA glycosylase, Histone methylation

(2010) 29, 3691–3702 & 2010 Macmillan Publishers Limited All rights reserved 0950-9232/10 www.nature.com/onc ORIGINAL ARTICLE LSD1-mediated demethylation of H3 4 triggers Myc-induced

S Amente1,2, A Bertoni3, A Morano3, L Lania1,2, EV Avvedimento2,3 and B Majello1

1Department of Structural and Functional Biology, University of Naples ‘Federico II’, Naples, Italy; 2Naples Oncogenomic Center (CEINGE), Naples, Italy and 3Department of Biology, Cellular and Molecular Pathology, ‘L. Califano’ University of Naples ‘Federico II’, Naples, Italy

Myc is a transcription factor that significantly contributes (Kouzarides, 2007). Furthermore, lysine to progression by modulating the expression of residues can be mono-, di- or trimethylated, and it has important genes through binding to a DNA sequence, been suggested that each methylation state represents a CACGTG, called E-box. We find that on Myc binding to specific epigenetic marker with a precise biological , the lysine-demethylating , LSD1, meaning in transcription (Shi, 2007; Cloos et al., 2008; triggers a transient demethylation of lysine 4 in the Kirmizis et al., 2009). . In addition, we demonstrate that Myc binds The c-myc proto-oncogene is an essential activator of and recruits LSD1 to the E-box chromatin and the cell proliferation and one of the genes most commonly formation of this complex is stimulated by cAMP-PKA. deregulated in cancer (Cowling and Cole, 2006; Eilers Demethylation by LSD1 produces H2O2, which locally and Eisenman, 2008; Meyer and Penn, 2008). The Myc oxidizes and induces the recruitment of 8-oxo- binds to a multitude of sites in the guanine–DNA (OGG1) and of the genome; it has been estimated that 15% of all Ape1 on the E-box chromatin. Inhibition of oxidation regions in the human genome are potentially Myc- or silencing of LSD1, OGG1 or Ape1 significantly binding targets (Fernandez et al., 2003; Guccione et al., reduce transcription and inhibit mRNA accumulation of 2006; Zeller et al., 2006). Once bound to its targets, Myc Myc-target genes. Collectively, these data highlight the introduces changes on chromatin. It is known that Myc role of transient LSD1-mediated demethylation of H3K4 can interact with several histone modifiers resulting in leading to local DNA oxidation as driving force in the hyperacetylation of and transcriptional activa- assembly of the Myc-induced transcription initiation tion (McMahon et al., 2000; Frank et al., 2001; Frank complex. et al., 2003; Vervoorts et al., 2003; Secombe et al., 2007). Oncogene (2010) 29, 3691–3702; doi:10.1038/onc.2010.120; Despite the large bulk of information on Myc-induced published online 26 April 2010 histone acetylation, little is known about the effects of Myc on histone methylation (Secombe et al., 2007). Keywords: Myc; LSD1; transcription; histone; Large-scale or quantitative chromatin immunoprecipi- methylation; oxidative burst tation (qChIP) with probes specific for various histone post-transcriptional modifications shows that Myc- binding targets can be classified in specific groups, characterized by different histones marks (Guccione Introduction et al., 2006; Martinato et al., 2008). To find the link between histone modifications The basic building blocks of chromatin are the nucleo- induced by Myc and transcription, we have examined somes composed of DNA wrapped around a histone the consequences of specific changes of methylation octamer. The N-terminal tails of histones are subjected marks in the histone H3 tail, by using a conditional Myc to different post-translational modifications: phospho- expression system. We focused our attention on two rylation, ubiquitination, acetylation and methylation, Myc-target genes, Ncl (nucleolin) and Cad (carbamoyl- which affect chromatin structure and transcription -dihydroorotase), selected from high-affinity (Berger, 2007; Kouzarides, 2007). It has been shown E-box human Myc-regulated promoters with high levels that transcriptional activation is linked to H3K4, of H3K4 methylation (Guccione et al., 2006; Martinato H3K36 and H3K79 methylation, whereas transcrip- et al., 2008). tional repression is associated with H3K9 and H3K27 Our data indicate that Myc binds to the E-boxes sequences of Ncl and Cad and triggers a transient Correspondence: Dr B Majello, Department of Structural and demethylation of H3K4me2 before significant changes Functional Biology, University of Naples Federico II, Via Cinthia, in local histone acetylation become apparent. We find Naples 80126, Italy. E-mail: [email protected] that Myc binds and facilitates the recruitment of the Received 5 November 2009; revised 8 March 2010; accepted 16 March lysine-specific , LSD1 (Shi et al., 2004; Shi, 2010; published online 26 April 2010 2007), to the E-box chromatin. Both the binding and LSD1-induced H3 demethylation drives Myc transactivation S Amente et al 3692 recruitment of LSD1 are stimulated by cAMP-PKA or present in the absence of Myc on the TSS (0). As soon as serum. LSD1 is a flavin adenine dinucleotide (FAD)- Myc binds to the E-box, RNAPII moves from the 0 containing enzyme, which generates H2O2 during the promoter site and progressively accumulates at 30 on demethylation process (Forneris et al., 2005; Anand and the CRs (compare the signal at 0 and 30 on the TSS and Marmorstein, 2007; Lan et al., 2008). Oxidation of CR). At 600 after Myc activation, RNAPII is again guanine by H2O2 recruits components of the base recruited on the TSS. Concomitantly, after 1 h of Myc excision repair system, the 8-oxoguanine–DNA glyco- induction, stable Ncl and Cad mRNA begins to sylase, OGG1, and the Ape1 (Sidorenko et al, accumulate (Supplementary Figure 2c). Comparing the 2007; Tell et al., 2008) on the E-box and promoter relative site occupancy of RNAPII on TSS and CR, we chromatin regions of Myc-target genes. According to can speculatively suggest a transcription cycle induced these results, inhibition of oxidation or silencing of by Myc with a period of 45–600. Acetylated LSD1, OGG1 or Ape1 significantly affects Myc-induced steadily accumulates on the TSS and CR of both genes, transcription. peaking 2400 after Myc activation. Collectively, these data suggest a mechanism for the We have analyzed the methylation marks on these initiation of transcription induced by Myc and highlight regions following Myc binding, focusing our attention the role of LSD1-mediated demethylation of histone on 4 and 9 of histone H3. Figure 2 (Ncl gene) H3K4 as a driving force in the assembly of the shows that and levels remain transcription initiation complex of Myc targets. stable except for a slight demethylation of H3K4me3 on the E-box chromatin after Myc binding (0–30). H3K4me2, on the other end, undergoes demethylation on the TTS and E-box on Myc binding. H3K4me2 levels Results return to the basal level 4 h after Myc activation. On the US chromatin, H3K4me2 levels decrease at 300 and Myc induces H3K4 demethylation cycles on the E-box rapidly return to the basal level at 600. Similar changes in chromatin H3K4me2 and me3 are detectable on the chromatin of Myc binds the E-box regions in the genome containing Cad gene (Supplementary Figure 1). the hexanucleotide 50-CANNTG-30 (Cowling and Cole, The description of the changes in H3K4 methylation, 2006). The chromatin context, the so-called ‘histone following Myc binding on the E-box chromatin, code’, determines the binding and possibly the trans- suggests that Myc-induced transcription is accompanied criptional outcome. However, notwithstanding extensive by cycles of methylation and demethylation of lysine 4 analysis of chromatin regions activated or repressed by in histone H3, before the robust burst of H4 acetylation. Myc, up to now there is no clear definition of the These data confirm and extends the notion that H3K4 chromatin domains recognized and modified by the methylated chromatin characterizes high-affinity Myc protein. promoters (Guccione et al., 2006). qChIP with various probes targeting histone post- transcriptional modifications indicates that Myc-respon- sive regions are characterized by specific histone marks. Myc recruits LSD1 on the chromatin and triggers H3K4 High-affinity Myc-binding sites are marked by methyla- demethylation tion of lysine 4 in the histone H3 tail, a signature of The results shown in Figure 1 indicate that, on Myc transcriptional activation (Guccione et al., 2006; Zeller binding, H3K4me2 is subjected to demethylation. As et al., 2006; Martinato et al., 2008). We have focused H3K4me2 is a substrate of the lysine-specific demethy- our attention on two prototypic Myc-target genes, Ncl lase, LSD1 (Forneris et al., 2008), we have investigated and Cad, showing high levels of H3K4 methylation the recruitment of this enzyme on Ncl and Cad and we have analyzed the methylation changes on Myc chromatin. Figure 2 and Supplementary Figure 1 show expression. that LSD1 is transiently recruited on Ncl and Cad We have used the Rat-MycER cell line (Littlewood sequences shortly after Myc activation. Although et al., 1995; Gargano et al., 2007) in which the MycER Myc binding increases over time essentially on the protein is activated on treatment of the cells with E-box, LSD1 binding is observed predominantly on tamoxifen (OHT). The cells are serum starved for 2 days the E-box and TSS regions but it is also seen on the and then are treated with OHT and serum. US sequence and CRs. The recruitment of LSD1 to To assess Myc and RNAPII occupancy on Ncl and the chromatin of these Myc-induced promoters is an Cad genomic sites in a precise temporal frame after Myc early and transient event, which overlaps to the wave activation, we carried out qChIP analysis on upstream of H3K4me2 demethylation. This result suggests a (US), transcription start site (TSS), E-box and coding functional relation between LDS1 recruitment and regions (CRs) of Ncl and Cad genes at 0, 30, 60 and 2400 H3K4me2 demethylation. after Myc activation. Figure 1 and Supplementary To gain further support to this hypothesis, we Figure 1 show that Myc is recruited (30–60) on the knocked down LSD1 expression with specific small E-box of both genes and it accumulates progressively with interfering RNAs (siRNAs) and analyzed Myc binding the time. Stabilization of the MycER protein following and H3K4me2 levels on the E-box chromatin. We found OHT treatment (Supplementary Figures 2a and b) that the specific siRNA did not affect Myc recruitment might account for this behavior. RNAPII is already to the E-box chromatin (Figure 3a); conversely,

Oncogene LSD1-induced H3 demethylation drives Myc transactivation S Amente et al 3693 Ncl Exon 1 Exon 2 Exon 3

US TSS E-box CR 1.6 1.6 1.6 1.6 Myc 1.2 1.2 1.2 1.2

0.8 0.8 0.8 0.8 % Input % Input % Input % Input 0.4 0.4 0.4 0.4

0 0 0 0 03060 240 03060 240 03060 240 030 60 240

1.6 1.6 1.6 1.6 RNAPII 1.2 1.2 1.2 1.2

0.8 0.8 0.8 0.8 % Input % Input % Input % Input 0.4 0.4 0.4 0.4

0 0 0 0 0 30 60 240 0 30 60 240 030 60 240 030 60 240

50 50 50 50 H4Ac 40 40 40 40 30 30 30 30 20 20 20 20 % Input % Input % Input % Input 10 10 10 10 0 0 0 0 030 60 240 03060 240 03060 240 030 60 240 Figure 1 RNAPII recruitment and H4 pan-acetylation on Ncl chromatin following Myc induction. RNAPII and pan-acetylated H4 histones levels are analyzed by qChIP in Rat-MycER serum-starved cells (0) and after Myc activation (OHT/serum) at the indicated times (30, 60 and 240 min). Amplicons used for the analysis of the Ncl locus occupancy are shown below the simplified structure of the Ncl gene. Amplicons correspond to sequences upstream of the transcription start site, US (À770), transcription start site (TSS), Myc- E-box ( þ 592) and coding regions (CRs) ( þ 1526). The ACHR promoter amplicon was used as a negative control in all experiments (see Supplementary Figure 6). The values reported were calculated as fold percentage of the amount of immunoprecipitated Ncl DNA relative to that present in total input chromatin. qChIP data are presented as mean of three independent biological experiments each analyzed by triplicate±s.d. Po0.01 values at single time points versus the basal 0.

demethylation of H3K4me2 was significantly inhibited whether Myc and LSD1 coexist within the same protein in samples in which LSD1 was silenced. In LSD1- complex resident in the Ncl E-box sequence, we carried silenced cells (Figure 3b, lanes 5 and 6), Myc failed to out sequential ChIP. In such assay, an initial ChIP was induce the transcription of Ncl and Cad genes. performed with the Myc antibody. The precipitated Furthermore, LSD1-silenced cells transfected with the chromatin–DNA complex was washed and eluted and a LSD1 expression vector recovered Myc-induced tran- second immunoprecipitation (IP) was carried out with scription (Figure 3b, lanes 7 and 8). Note that transient the same Myc antibody, the LSD1 antibody and a expression of LSD1 in control cells stimulated the levels control IgG. As illustrated in Figure 3d, sequential ChIP of Cad and Ncl mRNA only in the presence of Myc shows the presence of LSD1 in the same Myc protein– (Figure 3b, lanes 3 and 4). DNA complex. These data indicate that Myc-induced transcription As LSD1 is present also in other regions different requires LSD1 and suggest that Myc recruits LSD1 on from the E-box, it is likely that additional factors, E-box chromatin. To test this hypothesis, we immuno- perhaps the RNAPII, are able to recruit LSD1 to the precipitated Myc with specific antibodies and probed the chromatin. Nevertheless, the re-ChIP experiment, show- immunoprecipitated for the presence of LSD1. ing that LSD1 and MycER co-occupy the E-box, and Figure 3c shows that Myc specifically interacts with the coimmunoprecipitation of Myc and LSD1 indicate LSD1, suggesting that Myc binds to LSD1 and recruits that Myc and LSD1 form a tight complex on the E-box the enzyme on the E-box chromatin. To determine chromatin.

Oncogene LSD1-induced H3 demethylation drives Myc transactivation S Amente et al 3694 US TSS E-box CR H3K4me3 40 40 40 40

30 30 30 30

20 20 20 20 % Input % Input % Input 10 % Input 10 10 10

0 0 0 0 030 60 240 030 60 240 0 30 60 240 0 30 60 240

H3K9me2 4 4 4 4

3 3 3 3

2 2 2 2 % Input % Input % Input % Input 1 1 1 1

0 0 0 0 030 60 240 030 60 240 030 60 240 030 60 240

H3K4me2 16 16 16 16

12 12 12 12

8 8 8 8 % Input % Input % Input 4 4 % Input 4 4

0 0 0 0 0 30 60 240 0 30 60 240 030 60 240 0 30 60 240

LSD1 0.4 0.4 0.4 0.4

0.3 0.3 0.3 0.3

0.2 0.2 0.2 0.2 % Input % Input % Input % Input 0.1 0.1 0.1 0.1

0 0 0 0 0 30 60 240 0 30 60 240 0 30 60 240 0 30 60 240 Figure 2 Histone H3K4/K9 methylation and LSD1 occupancy on the Ncl chromatin following Myc activation. Rat-MycER serum- starved cells (0) are compared with cells induced with OHT and serum for Myc activation at the indicated times (30, 60 and 240 min). qChIP was carried out using specific antibodies recognizing H3K4me3, H3K4me2, H3K9me2 and LSD1. The qChIP data are presented as described in the text and the amplicons used are represented on the top of Figure 1. qChIP data are presented as described in the legend of Figure 1, ±s.d (n ¼ 3).

Mitogenic stimuli cooperate with Myc to recruit LSD1 to regions is efficient, but slower in the absence of serum Myc targets (2 versus 1 h) (Figure 4a). On the other hand, serum does One important question left unanswered by the data not substitute for Myc in the induction of transcription presented here is the following: is Myc by itself able to of target genes, because in rat cells with inactive Myc assemble a productive transcription initiation complex (MycÀ/À), serum is unable to stimulate transcription of by recruiting LSD1 on the E-box chromatin or other Myc-induced genes (Supplementary Figures 3a and b; complementary factor(s) are needed to complete and Frank et al., 2001). Taken together, these data suggest trigger the transcription? We find that Myc in the that other factors present in the serum cooperate with absence of serum poorly stimulates the transcription of Myc to efficiently assemble a productive transcription Ncl or Cad genes (Supplementary Figures 3a and b and initiation complex. Frank et al., 2001; Schlosser et al., 2005). Similarly, in LSD1 is phosphorylated and phosphorylation can be Rat-MycER cells, OHT induces a slower Myc accumu- modulated by growth factors (Olsen et al., 2006 and AB lation on the E-box, when compared with cells et al., manuscript in preparation). In the LSD1 protein, stimulated with OHT and serum (Supplementary there is a sequence at the NH-terminus, very similar to Figure 3c and Frank et al., 2001). Comparing the ChIP the PKA consensus motif, which is phosphorylated data in cells treated with OHT or OHT þ serum, we find in vivo (Olsen et al., 2006; and A B et al., manuscript in that recruitment of LSD1 to the E-box and promoter preparation). This suggests that cAMP-PKA may

Oncogene LSD1-induced H3 demethylation drives Myc transactivation S Amente et al 3695 Myc H3K4me2 0.8 12

8 0.4 siRNA LSD1

% Input 4 % Input siRNA N-T

0 0 03060 03060

8 7 Cad 6 Ncl 5 4 3

Fold Change 2 1 0 -++ - - + - + Myc --+ + -- + + LSD1wt --+- - +++ siRNA LSD1 + + ------siRNA N-T

WB: LSD1

WB: Actin

12345678

IP MycLSD1 IgG 2nd ChIP Ab st IMyc IgG Input MycMyc Myc IgG 1 ChIP Ab

WB:LSD1 Nuc

WB:Myc ACHR

Figure 3 LSD1 knockdown prevents H3K4me2 demethylation and inhibits Myc-induced transcription. (a) qChIP experiments on Ncl E-box chromatin from Rat-MycER-induced cells, pretreated with siRNA-targeting LSD1 or scrambled RNAs. Chromatin was prepared at the indicated times (minutes) after Myc activation and analyzed with the Myc antibody and anti-H3K4me2 antibody as indicated. qChIP data are presented as mean of three independent biological experiments±s.d. (b) Serum-starved Rat-MycER cells were treated for 4 h with OHT and serum ( þ Myc). LSD1 expression was inhibited with specific siRNAs. siRNA nontargeting (siRNA N-T) was used as scrambled RNAs. LSD1wt expression vector was transiently transfected in samples 3, 4, 7 and 8, whereas an empty vector was used in lanes 1, 2, 5 and 6. mRNA expression was quantified by qPCR compared with quiescent cells. mRNA levels were normalized to b-glucuronidase (GUS) mRNA levels. All values represent the average of at least three independent experiments. Error bars indicate s.d. LSD1 interacts with Myc. (c) Rat-MycER cells were transfected with LSD1wt vector 48 h before Myc activation (OHT þ serum). Whole-cell extracts were prepared after 1 h of induction and subjected to IP with anti-Myc or control IgG; the IP reactions were carried out with the indicated antibodies. (d) SMyc and LSD1 are found in the same complex at the Ncl E-box region. Sequential ChIP assay were carried out after an initial IP with anti-Myc. The Myc/LSD1 ChIP signal is eightfold lower of the Myc/Myc ChIP enrichment.

cooperate with Myc to induce transcription. To test this chromatin of Ncl and Cad (Figure 4c and Supplemen- hypothesis, we stimulated the cells with cAMP or serum tary Figure 4b). As in cells stimulated with cAMP and in the presence of Myc and we have measured the OHT the recruitment of LSD1 on chromatin is more expression of Cad and Ncl genes. Figure 4b and rapid and efficient compared with OHT cells (Figure 4c Supplementary Figure 4b show that cAMP-PKA is and Supplementary Figure 4b), we have investigated required for efficient Myc-induced transcription of Ncl whether cAMP stimulates the binding of LSD1 to Myc. and Cad, because H89, a PKA inhibitor, suppresses the Figure 4d (lanes 1–3) shows that cAMP stimulates the cooperative effect of serum. In addition, cAMP sub- interaction between LSD1 and Myc and that the stitutes for serum in the recruitment of LSD1 on E-box catalytic subunit of PKA is also present in the complex.

Oncogene LSD1-induced H3 demethylation drives Myc transactivation S Amente et al 3696 * LSD1 OHT + serum * 5 * 0.4 * OHT * 4 *

0.3 3

2 0.2 Fold Change

% input 1

01 0 cAMP + + + 0 H89 + + + Time hr 0 1 2 3 4 0 OHT OHT+serum

LSD1 LSD1wt LSD1PKAtmut 0.6 * cAMP + + serum + + 0.5 * 1 2 3 4 5 6 αLSD1 0.4

input αPKA 0.3

% input αMyc 0.2

αLSD1 0.1

αPKA 0 IP 0 - cAMP serum αMyc OHT

10

8

6

4 Fold Change 2

0 LSD1wt +++ LSD1PKAmut +++ 0 OHT+cAMP OHT+serum Figure 4 cAMP cooperates with Myc to induce transcription. (a) Quiescent Rat-MycER cells were stimulated with either OHT alone or with OHT þ serum at the indicated times. qChIP on Ncl E-box was carried out with anti-LSD1. qChIP data are presented as means of three independent biological experiments (n ¼ 3)±s.d. *Po0.001. (b) mRNA expression of Ncl, quantified by qPCR, from Rat- MycER serum-starved cells (0) or treated with OHT or OHT þ serum for 4 h for Myc activation. Samples in which cAMP or H89 were used are indicated ( þ ). Po0.01 values at single time points versus the basal 0. *Po0.001. (c) qChIP on Ncl E-box was carried out with anti-LSD1 on quiescent Rat-MycER cells stimulated for 1 h with OHT alone or in combination with cAMP or serum. (d) cAMP-PKA is required for efficient Myc-LSD1 interaction. Rat-MycER cells were transfected with LSD1wt or LSD1PKAmut expression vectors. Total cell extracts were prepared and immunoprecipitated with Myc. The IP were blotted with the indicated antibodies. (e) Rat- MycER cells were transfected with LSD1wt or LSD1PKAmut expression vectors, and expression of Ncl was determined in quiescent cells (0) and after Myc activation for 4 h. mRNA expression was quantified as in Figure 3.

To demonstrate a direct effect of the putative PKA site by substituting the threonine with alanine and we have at the NH-terminus of the protein, we have introduced a tested the mutant protein for binding to Myc and PKA. point in the LSD1 PKA consensus sequence Figure 4d (lanes 4–5) shows that the LSD1PKAmut

Oncogene LSD1-induced H3 demethylation drives Myc transactivation S Amente et al 3697 shows a reduced affinity for Myc and PKA. Finally, we H2O2 (Forneris et al, 2005; Anand and Marmorstein, tested the ability of cAMP to cooperate with Myc to 2007; Lan et al., 2008) during lysine demethylation. induce transcription. Figure 4e and Supplementary We looked for signs of DNA oxidation induced by Myc Figure 4c shows that expression of wild-type LSD1 by measuring the presence of 8-oxidized dG in the nuclei amplifies cAMP and Myc effects, whereas the expression of cells induced by OHT. To this end, Rat-MycER of the LSD1PKAmut inhibits cAMP and Myc-induced cells were treated with OHT and after 450 the cells were transcription of Ncl and Cad genes (Figure 4e and labeled with the fluorescein tagged 8-oxodeoxyguanine Supplementary Figure 4c). (8-oxodG)-binding protein. Figure 5a shows a signifi- Collectively, these data indicate that cAMP, induced cant labeling of the nuclei (8-oxodG foci) after the by different metabolic signals, cooperates with Myc to addition of OHT. OHT treatment of isogenic Rat1 induce transcription of target genes. cells did not induce nuclear labeling, suggesting that the effect observed was dependent on Myc. Accordingly, DNA oxidation marks Myc-target sites silencing of Myc expression with specific siRNA signifi- Myc recruits LSD1 on E-box and promoter chromatin. cantly reduced nuclear 8-oxodG labeling. Furthermore, LSD1 is an FAD-containing enzyme, which generates Myc-induced production of 8-oxodG was inhibited

Starved OHT-Induced

H2O2 siMyc Parg siLSD1 Rat-MycER

% of 8-oxo-Gs 5±2 84±8 fluorescent cells 74±12 12±4 >1 10±4

1.2 Myc 1 Rat1 LSD1 0.8 0.6 % of 8-oxo-Gs 0.4 6±387±8 7±3 fluorescent cells 0.2 Fold Change 0 +-siRNA N-T - + siRNA

4 Cad 7 Cad Ncl 6 Ncl 3 5 4 2 3 Fold Change 1 Fold Change 2 1 0 0 Myc - +-+ - + - + - + OHT NAC - - + + - - + +++ Parg --- - + + H2O2 Figure 5 Myc induces a burst of nuclear 8-oxodG. (a) Rat-MycER and the isogenic Rat1 cells were serum starved and induced with OHT alone. After 450, 8-oxodG was measured by labeling the cells with the fluorescein-tagged 8-oxodG binding protein. Rat-MycER 0 cells were treated with H2O2 for 15 as positive control. The monoamine oxidase inhibitor pargyline (Parg) was included in the media before Myc induction. For siRNA experiments, Rat-MycER were transfected with siRNA-targeting LSD1, Myc or scrambled RNAs, then serum deprived for 48 h before activation of Myc. An example of fluorescence microscopy of the cells stained with the fluorescent oxodG binding protein is shown, (b) the percentage of positive labeled cells derived from three independent experiments is shown. The efficiency of siRNA treatments was measured by qPCR using primers for Myc and LSD1 transcripts. (c) Reactive oxygen species (ROS) scavenging inhibits Myc-induced transcription. Rat-MycER cells were serum starved (À) or treated with OHT ( þ ) in the presence or absence of the ROS scavenger N-acetyl cysteine ( þ /À NAC, 1 mM for 4 h). Ncl and Cad mRNAs expression levels were normalized to b-glucuronidase (GUS) mRNA levels. The values are presented relative to RNA levels in quiescent Rat-MycER levels. All values represent the average of at least three independent experiments. Error bars indicate s.d. (n ¼ 3–4). (d)H2O2 rescues pargyline-inhibited transcription only in the presence of Myc. Rat-MycER cells were serum starved (À) or treated for Myc induction ( þ ) in the presence or absence of the momoamine oxidase inhibitor pargyline ( þ /À Parg) or H2O2 as indicated. Ncl and Cad mRNAs expression levels were quantified as in (c).

Oncogene LSD1-induced H3 demethylation drives Myc transactivation S Amente et al 3698 by pargyline, a monoamine oxidase LSD1 inhibitor to remove 8-oxodG lesions (Sidorenko et al, 2007). in vivo, as well as by siRNA against LSD1 (Figures 5a To find signs of localized DNA oxidation induced by and b). Myc, we immunoprecipitated the chromatin from Myc- To find a cause–effect relationship between induc- induced cells with antibodies against OGG1 or tion of transcription by Myc and oxidation, we treated 8-oxodG. We show that both 8-oxodG and OGG1 were the cells with short pulses of the reducing agent engaged on Ncl (Figure 6) and Cad (Supplementary N-acetylcysteine (NAC) and analyzed Ncl and Cad Figure 5) TSS and E-boxes concomitantly with Myc. As expression following Myc activation. Figure 5c shows the apurinic site generated by OGG1 is further that Myc induction of Ncl and Cad was abolished by processed by the nuclease Ape1, which recognizes and NAC. Note that NAC does not influence demethylation cleaves the phosphodiester bond at the 50 end of the of H3K9 or K4 (Perillo et al, 2008, and data not shown). apurinic site (Tell et al., 2008), we looked for its presence In addition, we carried out the complementary experi- on the E-box and promoter sites following Myc ment: we inhibited demethylation with pargyline and activation. Figure 6 and Supplementary Figure 5 show treated the cells with low concentrations of H2O2. that Ape1 is recruited at the same RNAPII and Myc- H2O2 did not stimulate transcription per se, but reversed occupied chromatin. The occupancy of Ape1 on the very efficiently pargyline inhibition of Ncl and Cad chromatin follows OGG1 recruitment: it slightly de- expression only in the presence of Myc (Figure 5d). creases 300 after Myc activation (when OGG1 and These data suggest that oxidation substitutes for 8-oxodG are already on the site) and robustly accumu- H3K4me2 demethylation, providing that Myc is present lates 600 after Myc activation. Similar to OGG1 and on the target site. 8-oxodG, APE1 disappears from these chromatin sites 2400 after Myc induction. To demonstrate a relevant role on Myc-induced are recruited to the transcription of OGG1 and Ape1, we silenced OGG1 transcription initiation complex induced by Myc or Ape1 expression in Rat-MycER cells. Figures 7a and DNA base oxidation produces 8-oxodG, which is b show that silencing of OGG1 or Ape1 inhibits Myc- recognized by the 8-oxoguanine–DNA glycosylase-1 induced Ncl and Cad gene expression. Moreover, to (OGG1). This enzyme initiates the base excision repair demonstrate that LSD1 initiates the chain of events

US TSS E-box CR 0.10 0.10 0.10 0.10

8-oxo-dG 0.05 0.05 0.05 0.05 % Input % Input % Input % Input

0 0 0 0 0 30 60 240 0 30 60 240 0 30 60 240 0 30 60 240

0.06 0.06 0.06 0.06

0.04 0.04 0.04 0.04 OGG1 % Input % Input % Input

% Input 0.02 0.02 0.02 0.02

0 0 0 0 0 30 60 240 0 30 60 240 0 30 60 240 0 30 60 240

0.16 0.16 0.16 0.16

0.12 0.12 0.12 0.12

Ape1 0.08 0.08 0.08 0.08 % Input % Input % Input % Input 0.04 0.04 0.04 0.04

0 0 0 0 0 30 60 240 0 30 60 240 0 30 60 240 0 30 60 240 Figure 6 Recruitment of oxidation repair enzymes on Ncl chromatin. Rat-MycER cells, starved (0) or treated with OHT for 30, 60 and 2400, were analyzed by qChIP using specific antibodies recognizing 8-oxodG, the 8-oxoguanine–DNA glycosylase-1, OGG1, and the AP endonuclease Ape1. The qChIP data are presented as described in the text and the amplicons used are represented on the top of Figure 1. Error bars indicate s.d. (n ¼ 3).

Oncogene LSD1-induced H3 demethylation drives Myc transactivation S Amente et al 3699 Cad 4 4 Ncl

3 3

2 2 Fold Change Fold Change 1 1

0 0 Myc Myc - + - + - + - + - - + + siRNA OGG1 - - + + siRNA Ape1 + + - - siRNA N-T + + - - siRNA N-T

OGG1 Ape1 1.2 1.2 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 Fold Change Fold Change 0.2 0.2 0 0 WB: OGG1 WB: Ape1

- + - + Myc - + - + Myc - - + + siRNA OGG1 - - + + siRNA Ape1 + + - - siRNA N-T + + - - siRNA N-T

siRNA LSD1 siRNA N-T % input

Time min Figure 7 OGG1 or Ape1 knockdown impairs the expression of Myc-target genes. (a) Serum-deprived Rat-MycER cells (ÀMyc) were treated for 4 h with OHT þ serum ( þ Myc). OGG1 or Ape1 expression was silenced with specific siRNAs; Ncl and Cad mRNAs expression levels were quantified by qPCR. Error bars indicate s.d. (n ¼ 3). (b) The efficiency of siRNA treatments was measured by qPCR using primers for OGG1 and Ape1 transcripts. Error bars indicate s.d. (n ¼ 3). Protein expression from a single experiment is shown in immunoblots carried out with the OGG1 and Ape1 antibodies. (c) LSD1 knockdown prevents OGG1 recruitment on E-box chromatin. qChIP experiments on Ncl E-box chromatin prepared from Rat-MycER-induced cells, pretreated with siRNA-targeting LSD1 or scrambled RNAs (siRNA N-T). Chromatin was prepared at the indicated times (minutes) after Myc activation and analyzed with the OGG1 antibody.

leading to recruitment of OGG1 and Ape1 on E-box Discussion chromatin, we silenced LSD1 and analyzed OGG1 on E-box chromatin in the presence of OHT. Figure 7c Histone H3 K4 methylation cycles drive Myc induction shows that in the absence of LSD1, OGG1 is absent of transcription on the E-box chromatin although Myc is still bound (see Methylated H3K4 marks the promoter and the E-box Figure 3). chromatin in Myc-inducible genes. This configuration These data indicate that LSD1 initiates the recruit- was found in a large-scale qChIP analysis of histone ment of oxidative repair enzymes on Myc targets and marks of Myc-target genes (Guccione et al., 2006; Zeller that this event is essential for the induction of et al., 2006; Martinato et al., 2008). On activation of transcription. Myc, H3K4 is rapidly and transiently demethylated by

Oncogene LSD1-induced H3 demethylation drives Myc transactivation S Amente et al 3700 LSD1 (30–600) (Figure 2). In cells, exposed to Myc for producing 8-oxodG, which is recognized by OGG1. The 4 h, the H3K4 methylation profile is identical to that oxidized site is further processed by Ape1, the nuclease found in uninduced cells (Figures 1 and 2, and that cleaves the apurinic site (30-phospho-ab-unsatu- Supplementary Figure 1), whereas high levels of H4 rated aldehyde) and generates the 30-OH end primer acetylation are detectable on these regions. Thus, the suitable for repair (Tell et al., 2008; Mantha et al., 2008). wave of H3K4 demethylation occurs before changes in Our data provide evidence that both OGG1 and Ape1 H4 acetylation. The enzyme responsible for demethyla- are essential for Myc-mediated transcription, as their tion of H3K4 is LSD1. LSD1 is a demethylase, whose silencing inhibits the expression of Myc targets. More- chromatin association is extensively modulated by its over, demethylation by LSD1 of H3K9me2 driving interacting partners and local histone marks (Metzger recruitment on target promoters of the DNA et al., 2005; Perillo et al., 2008; Shi, 2007; Wissmann glycosylase (Chen et al., 2003; Garcia-Bassets et al., et al, 2007). By interacting with diverse cofactors and 2007) and OGG1 (Perillo et al., 2008) has been reported catalyzing demethylation of mono- and di-methylated to be required for estrogen-driven transcription. H3K4 or H3K9, LSD1 is capable of either repressing or Furthermore, our data indicate that OGG1 recruitment activating target genes (Metzger and Schule, 2007; Shi of E-box chromatin requires functional LSD1 2007). The demethylation of H3K4 induced by Myc is (Figure 7c) and Ape1 occupancy on TSS and E-box rapidly reversed after 600; 4 h after Myc induction H3K4 sequences, 300 after Myc induction, do not overlap. methylation profile is similar to control uninduced cells Thus, OGG1 precedes at least by 300 Ape1 in the (Figure 2). Not only the H3 methylation marks after 4 h binding to the chromatin sites (Figure 6). Ape1 may be are similar to that found in uninduced state, but also important during the initiation process, because the LSD1-OGG1-Ape1 disappear from chromatin. H3K4 introduction of nicks in the E-box chromatin and TSS demethylation cycles, as well as the recruitment of the might dynamically dissipate transcription-induced associated factors (LSD1/OGG1/Ape1), can be detected supercoiling (Bhakat et al., 2008; Chattopadhyay only under stringent synchronized conditions of tran- et al., 2008). Together, Ape1 may represent the physical scription initiation. Thus, the stochastically desynchro- link between transcription and repair as it recognizes the nized transcription might pose severe limitations in oxodG-OGG1 complex and targets positive RNAPII detecting early transcriptional events. co-activators, such as p300 to the site (Bhakat et al., 2003a, b, 2008). Intriguingly, Ape1 has been reported as a Myc target (Guccione et al., 2006) and it will be of cAMP signals cooperate with Myc to induce transcription interest to investigate whether Myc is involved in its Serum is known to be an essential factor that amplifies transcriptional activity. Finally, our data are consistent Myc effects on transcription and proliferation (Frank with a recent report showing that a base excision process et al., 2001; Schlosser et al., 2005). We found that involving the glycosylase, MBD4, and APE1 was cAMP-PKA stimulates Myc-dependent transcription in essential for the hormonal control of transcription the absence of serum and concomitantly increases Myc- (Kim et al., 2009). LSD1 binding (Figure 4). Interestingly, the catalytic In conclusion, our data describe a mechanism under- subunit of PKA is also present in the Myc-LSD1 lying Myc-induced transcription that links histone H3 complex, and mutation of the putative PKA site at the demethylation to oxidation and DNA repair. Myc N-terminus of LSD1 impairs Myc and LSD1 ability to recruits LSD1 to the target E-box chromatin and activate transcription. We shall note that, in the absence initiates with other factors, including serum or cAMP- of serum, Myc alone, with a slower kinetics, is still able PKA, the events that trigger transcription initiation. to recruit LSD1 on the chromatin (Figure 4). We Although it remains to be seen whether this represents a speculate that the mechanism is still PKA dependent as specific case of a set of Myc-activated promoters or can it has been shown that Myc induces PKA catalytic be extended to all transcribed genes, we propose that subunit expression (Wu et al., 2002). cAMP and PKA Myc-induced genes use methylated H3K4 as basic signal efficiently substitute for serum in stimulating LSD1 for transient demethylation to trigger oxidation and to recruitment to the target chromatin. On the other hand, assemble the initiation complex. serum or cAMP does not substitute for Myc in the induction of transcription, as rat MycÀ/À cells do not activate transcription of Ncl and Cad (Supplementary Materials and methods Figure 5). These data suggest a role of cAMP-PKA to achieve maximal Myc biological effects, and such and drugs cooperative function can be at least in part due to Rat1 cells expressing a 4-hydroxytamoxifen (OHT)-inducible cAMP-mediated increase of LSD1-Myc functional and MycER chimera, Rat1 and rat MycÀ/À cells were cultured in physical interaction. Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum. Cells were made quiescent by contact inhibition followed by serum removal for 2 days. To induce Recruitment of DNA oxidation repair enzymes to the entry into the cycle, synchronized G0 arrested cells were transcription initiation complex treated with OHT (1 mM) alone or OHT plus serum or serum LSD1 catalyzes histone demethylation through an alone as indicated in the text and harvested at the indicated fFAD-dependent oxidative reaction. Local oxidation times. Pargyline treatment (3 mM): cells were pretreated for generated by LSD1 promotes modification of the Gs, 24 h before OHT þ serum induction of Myc, collected after 4 h

Oncogene LSD1-induced H3 demethylation drives Myc transactivation S Amente et al 3701 and analyzed by quantitative PCR (qPCR). NAC treatment: indicated in the text, cell extracts were prepared and cells were treated for 4 h in the presence or absence of OHT immunoprecipitated with anti-Myc or control IgG. The IP and serum. H2O2 was added to the cells at 300 mM for 4 h in material was analyzed by western blots with anti-Myc and presence or absence of Myc activation. 8-CPT-cAMP and H89 anti-LSD1 antibodies. treatment: cells were treated with 20 mM for 900 and with 10 mm for 4 h, respectively. mRNA quantification by qPCR cDNA was prepared from total RNA with Quantitect Reverse Detection of 8-oxo-dG by fluorescence Transcription Kit (Qiagen, Milan, Italy) according to the Rat-MycER and the isogenic cell line Rat1 were serum starved manufacturer’s instructions. Each sample was assayed in and induced with OHT alone. After 450 of Myc activation, cells triplicate. Oligoprimers are described in Supplementary were fixed for 150 with 4% paraformaldehyde in phosphate- Methods. The qPCR data were normalized to the expression buffered saline. The slides were then washed with TBS/ of the housekeeping b-glucuronidase (GUS) gene, and after Tween-20 and permeabilized by serial washes in methanol normalization, the data were presented as fold change relative solutions, before washing with TBS/Tween-20, blocked for 1 h to the 0 point. at 37 1C and incubated with fluorescein isothiocyanate-labeled protein, which binds to 8-oxo-guanine, for 15 h at 4 1C qChIP (Biotrin OxyDNA Test, Biotrin, Dublin, Ireland). Cover slips qChIP experiments were carried out essentially as described were mounted in MoWiol (Sigma Aldrich, Milan, Italy) and (Frank et al., 2001; Gargano et al., 2007) and presented in the viewed by fluorescence on a Zeiss axiomatic photomicroscope Supplementary Methods. Normal serum and input DNA with a  63 objective (Oberkochen, Germany). As control, values were used to subtract or normalize the values from starved cells were treated with H O (100 mM,150). For 2 2 qChIP samples by using: % Input ¼ 2DCt  3; DCt ¼ Ct(input)– pargyline treatment (3mM), cells were pretreated for 24 h Ct(cIP) (Frank et al., 2001). All qChIP data were derived from before OHT activation (450). For siRNA interference ON- at least three independent experiments and for each experiment TARGETplus SMARTpool targeting LSD1 and Myc were qPCR was carried out in triplicate. used with the protocol described below. Cells were analyzed for 8-oxo-dG by fluorescence after OHT induction (450). Statistical analysis Transfections and siRNA All qChIP data are presented as means of three independent To carry out transient transfection experiments in Rat-MycER biological experiments each analyzed by triplicate ± cells, we used MicroPorator Digital Bio Technology, a pipette- (qPCR s.d.). Statistical significance was determined using type electroporation (Seoul, Korea). Indicated plasmids, the matched pairs test. Student’s t test was carried out for each or siRNA were introduced into each 3  106 dissociated pair of data separately. Po0.001 nonspecific versus specific cells in 100 ml volume according to the manufacturer’s antibody. instructions. Pulse width was determined according to applied voltages: 1400V, 30 ms, 1 pulse. Electroporated cells were then seeded into 100-mm culture dishes containing 5 ml of culture Conflict of interest media. After 5 h of recovery, the cells were serum deprived for 48 h. For siRNA treatments, ON-TARGETplus SMARTpool The authors declare no conflict of interest. (L-105863-01) LSD1, (L-003282-00-0005) Myc, (L-093481-01) OGG1, (L-100326-01) Ape1 and ON-Targetplus ON-TAR- GETplus nontargeting pool (D-001810-10-5) were obtained from Dharmacon (Lafayette, CO, USA). A final concentration Acknowledgements of 100 nM of the pools was used for each transfection. The expression of proteins was determined by western blot. In We thank R Dalla Favera for critical reading of the paper and transfection experiments, 3 mg of LSD1wt (Perillo et al., 2008) A Fusco for stimulating discussions and support. This work or LSD1PKAmut, in which the LSD1 threonine 112 was was supported by grants from the Italian Association substituted with alanine (AB et al., manuscript in preparation), for Cancer Research (AIRC) and by a core grant to NOGEC expression vectors were transfected into Rat-MycER by (AIRC). SA is a recipient of a postdoctoral fellowship electroporation. Cells were serum starved and at the times from AIRC.

References

Anand R, Marmorstein R. (2007). Structure and mechanism of Chattopadhyay R, Das S, Maiti AK, Boldogh I, Xie J, Hazra TK et al. lysine-specific demethylase enzymes. J Biol Chem 282: 35425–35429. (2008). Regulatory role of human AP-endonuclease (APE1/Ref-1) Berger SL. (2007). The complex language of chromatin regulation in YB-1-mediated activation of the multidrug resistance gene during transcription. Nature 447: 407–412. MDR1. Mol Cell Biol 28: 7066–7080. Bhakat KK, Izumi T, Yang SH, Hazra TK, Mitra S. (2003a). Role of Chen D, Lucey MJ, Phoenix F, Lopez-Garcia J, Hart SM, Losson R et al. acetylated human AP-endonuclease (APE1/Ref-1) in regulation of (2003). T:G mismatch-specific thymine-DNA glycosylase potentiates the parathyroid hormone gene. EMBO J 22: 6299–6309. transcription of estrogen-regulated genes through direct interaction Bhakat KK, Mantha AK, Mitra S. (2008). Transcriptional regu- with estrogen receptor alpha. JBiolChem278: 38586–38592. latory functions of mammalian AP-endonuclease (APE1/Ref-1), an Cloos PA, Christensen J, Agger K, Helin K. (2008). Erasing the methyl essential multifunctional protein. Antioxid Redox Signal 11: 621–638. mark: histone at the center of cellular differentiation Bhakat KK, Yang SH, Mitra S. (2003b). Acetylation of human AP- and disease. Genes Dev 22: 1115–1140. endonuclease 1, a critical enzyme in DNA repair and transcription Cowling VH, Cole MD. (2006). Mechanism of transcriptional regulation. Methods Enzymol 371: 292–300. activation by the Myc oncoproteins. Semin Cancer Biol 16: 242–252.

Oncogene LSD1-induced H3 demethylation drives Myc transactivation S Amente et al 3702 Eilers M, Eisenman RN. (2008). Myc’s broad reach. Genes Dev 22: McMahon SB, Wood MA, Cole MD. (2000). The essential 2755–2766. TRRAP recruits the histone acetyltransferase hGCN5 to c-Myc. Fernandez PC, Frank SR, Wang L, Schroeder M, Liu S, Greene J et al. Mol Cell Biol 20: 556–562. (2003). Genomic targets of the human c-Myc protein. Genes Dev 17: Metzger E, Schule R. (2007). The expanding world of histone lysine 1115–1129. demethylases. Nat Struct Mol Biol 14: 252–254. Forneris F, Binda C, Battaglioli E, Mattevi A. (2008). LSD1: oxidative Metzger E, Wissmann M, Yin N, Muller JM, Schneider R, Peters AH chemistry for multifaceted functions in chromatin regulation. et al. (2005). LSD1 demethylates repressive histone marks to promote Trends Biochem Sci 33: 181–189. androgen-receptor-dependent transcription. Nature 437: 436–439. Forneris F, Binda C, Vanoni MA, Mattevi A, Battaglioli E. (2005). Meyer N, Penn LZ. (2008). Reflecting on 25 years with MYC. Nat Rev Histone demethylation catalysed by LSD1 is a flavin-dependent Cancer 8: 976–990. oxidative process. FEBS Lett 579: 2203–2207. Olsen JV, Blagoev B, Gnad F, Macek B, Kumar C, Mortensen P et al. Frank SR, Parisi T, Taubert S, Fernandez P, Fuchs M, Chan HM (2006). Global, in vivo, and site-specific phosphorylation dynamics et al. (2003). MYC recruits the TIP60 histone acetyltransferase in signaling networks. Cell 127: 635–648. complex to chromatin. EMBO Rep 4: 575–580. Perillo B, Ombra MN, Bertoni A, Cuozzo C, Sacchetti S, Sasso A et al. Frank SR, Schroeder M, Fernandez P, Taubert S, Amati B. (2001). (2008). DNA oxidation as triggered by H3K9me2 demethylation Binding of c-Myc to chromatin mediates mitogen-induced drives estrogen-induced gene expression. Science 319: 202–206. acetylation of histone H4 and gene activation. Genes Dev 15: Schlosser I, Holzel M, Hoffmann R, Burtscher H, Kohlhuber F, 2069–2082. Schuhmacher M et al. (2005). Dissection of transcriptional Garcia-Bassets I, Kwon YS, Telese F, Prefontaine GG, Hutt KR, programmes in response to serum and c-Myc in a human B-cell Cheng CS et al. (2007). Histone methylation-dependent mechanisms line. Oncogene 24: 520–524. impose ligand dependency for gene activation by nuclear receptors. Secombe J, Li L, Carlos L, Eisenman RN. (2007). The Trithorax group Cell 128: 505–518. protein Lid is a trimethyl histone H3K4 demethylase required for Gargano B, Amente S, Majello B, Lania L. (2007). P-TEFb is a crucial dMyc-induced cell growth. Genes Dev 21: 537–551. co-factor for Myc transactivation. Cell Cycle 6: 2031–2037. Shi Y. (2007). Histone lysine demethylases: emerging roles in Guccione E, Martinato F, Finocchiaro G, Luzi L, Tizzoni L, Dall’ development, physiology and disease. Nat Rev Genet 8: 829–833. Olio V et al. (2006). Myc-binding-site recognition in the human Shi Y, Lan F, Matson C, Mulligan P, Whetstine JR, Cole PA et al. genome is determined by chromatin context. Nat Cell Biol 8: (2004). Histone demethylation mediated by the nuclear amine 764–770. oxidase homolog LSD1. Cell 119: 941–953. Kim MS, Kondo T, Takada I, Youn MY, Yamamoto Y, Takahashi S Sidorenko VS, Nevinsky GA, Zharkov DO. (2007). Mechanism of et al. (2009). DNA demethylation in hormone-induced transcrip- interaction between human 8-oxoguanine-DNA glycosylase and AP tional derepression. Nature 461: 1007–1012. endonuclease. DNA Repair (Amst) 6: 317–328. Kirmizis A, Santos-Rosa H, Penkett CJ, Singer MA, Green RD, Tell G, Quadrifoglio F, Tiribelli C, Kelley MR. (2008). The many Kouzarides T. (2009). Distinct transcriptional outputs associated functions of APE1/Ref-1: not only a DNA repair enzyme. Antioxid with mono- and dimethylated histone H3 2. Nat Struct Mol Redox Signal 11: 601–620. Biol 16: 449–451. Vervoorts J, Luscher-Firzlaff JM, Rottmann S, Lilischkis R, Kouzarides T. (2007). Chromatin modifications and their function. Walsemann G, Dohmann K et al. (2003). Stimulation of c-MYC Cell 128: 693–705. transcriptional activity and acetylation by recruitment of the Lan F, Nottke AC, Shi Y. (2008). Mechanisms involved in the regulation cofactor CBP. EMBO Rep 4: 484–490. of histone lysine demethylases. Curr Opin Cell Biol 20: 316–325. Wissmann M, Yin N, Muller JM, Greschik H, Fodor BD, Jenuwein T Littlewood TD, Hancock DC, Danielian PS, Parker MG, Evan GI. et al. (2007). Cooperative demethylation by JMJD2C and LSD1 (1995). A modified oestrogen receptor ligand-binding domain as an promotes androgen receptor-dependent gene expression. Nat Cell improved switch for the regulation of heterologous proteins. Nucleic Biol 9: 347–353. Acids Res 23: 1686–1690. Wu K-J, Mattioli M, Morse HC, Dalla Favera R. (2002). c-Myc Mantha AK, Oezguen N, Bhakat KK, Izumi T, Braun W, Mitra S. activates protein kinase A (PKA) by direct transcriptional activa- (2008). Unusual role of a cysteine residue in substrate binding and tion of the PKA catalytic subunit beta (PKA-Cb) gene. Oncogene activity of human AP-endonuclease 1. J Mol Biol 379: 28–37. 21: 7872–7882. Martinato F, Cesaroni M, Amati B, Guccione E. (2008). Analysis of Zeller KI, Zhao X, Lee CW, Chiu KP, Yao F, Yustein JT et al. (2006). Myc-induced histone modifications on target chromatin. PLoS Global mapping of c-Myc binding sites and target gene networks in ONE 3: e3650. human B cells. Proc Natl Acad Sci USA 103: 17834–17839.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)

Oncogene