Northern blots

Northern blot analysis was performed as described elsewhere (16). Probes: Myc, cDNA of exon 2 and 3; human Bop1, accession number D50914, PCR fragment from nucleotides 160–899; human Nop56, V12065, nucleotides 199–1226; human fibrillarin, X56597, nucleotides 245–840; Cbf5p, U59151, nucleotides 724–1387; human spermidine synthase, M64231, nucleotides 163–861; human cdk4, M1450, nucleotides 286–487; human GAPDH cDNA; 7SK, X05490, nucleotides 231–330; 5′-ETS, U13369, nucleotides 1–50; rat Bop1, U77415, nucleotides 1951–2246; rat Nop56, BI295000, nucleotides 93–579; rat fibrillarin, AI171263, nucleotides 97–508; rat transferrin receptor, M58040, nucleotides 1504–2491; rat ODC, M19157, nucleotides 453–1618; rat Ssecks, RNU 23146, nucleotides 4–889; rat Gadd45, L32591, nucleotides 19–691; rat GAPDH, NM_017008, nucleotides 21–746.

High-density oligonucleotide arrays

Target cRNA was prepared and hybridised (45°C, 16 h) to the human HG95 chip set and the rat U34 chip set according to the manufacturer’s directions (Affymetrix). Arrays were washed and stained using the GeneChip fluidics station 400 and scanned using the Agilent GeneArray scanner. Signals were analysed using Microarray Suite 5.0 software (Affymetrix). Conditional c-Myc induction was performed in three independent experiments. For the expression analysis, P493-6 cells were arrested (see above) and re-stimulated with 10% FCS for 8 h in the presence or absence of c-Myc (control). Total RNA was isolated and prepared for the chip analysis. The Ho15.19MycER cells were grown to confluence to arrest the cells. After 2 days at confluence, MycER was activated by adding tamoxifen for 8 h. Non-induced cells served as a control. The replicates were used to calculate means and standard deviations for the signal values of all probe sets from each cell line and condition. The signal was calculated using the One-Step Tukey’s Biweight Estimate. For the comparison analysis the Wilcoxon’s signed-rank test was used (for details see Statistical Algorithms Reference Guide, Affymetrix). The P-value for the change call ‘increase’ was set at least by 0.002, for the change call ‘decreased’ by at least 0.998. The criteria for a differentially expressed gene after c-Myc induction were as follows. Induced target genes must have an ‘induced’ change call and must be ‘present’ on the induced array, whereas they can be either present or absent on the un-induced array. Repressed target genes must have a ‘decreased’ change call and must be ‘present’ on the un-induced array, whereas they can be either present or absent on the induced array. These criteria were applied to all possible comparisons between the triplicate experiments to determine c-Myc target genes. Because of the highly stringent P-value, genes that have a differential expression less than the factor 2 were also included (Table ​(Table11).

Nuclear run-ons

Isolation of nuclei, nuclear run-on reaction and hybridisation were carried out as previously described (17). Labelled RNA was hybridised to pan oligonucleotides of the rDNA gene U13369 (18S: nucleotides 4014–4063, 28S: nucleotides 9561–9610, 5.8S: nucleotides 6690–6739), 7SK (X05490: nucleotides 231–330), GAPDH (X01677: nucleotides 351–450) and actin (X00351: nucleotides 151–260) that were bound on a nylon membrane (Hybond N+, Amersham). Signal intensities were analysed with a PhosphoImager (Fuji-BAS 1000).

c-Myc-dependent production of the 18S and 28S rRNA

A large number of the nucleolar genes listed in Table ​Table11 are directly involved in rRNA processing. To test whether processing of the 47S rRNA precursor in P493-6 cells is regulated by c-Myc, we performed pulse–chase labelling experiments. Cells were starved for 72 h at 0.25% serum in the presence of tet (arrested cells) and were subsequently stimulated for 12 h with 10% serum in the presence or absence of tet. Cells were then pulse-labelled for 30 min with ³H-uridine followed by a chase for indicated time points (Fig. ​(Fig.2).2). In the absence of tet, high levels of mature 18S and 28S rRNA accumulated during the chase kinetic (Fig. ​(Fig.2A).2A). In contrast, only a little increase of mature forms of rRNA was seen if cells were labelled in the presence of tet. After 6 h chase, the amount of processed 28S rRNA was ~5-fold higher in cells expressing c-myc (Fig. ​(Fig.2B).2B). When arrested cells were stimulated only by c-Myc, rRNA processing was almost as efficient as in the presence of c-Myc plus FCS (Fig. ​(Fig.2C).2C). Arrested cells showed no labelling of rRNA (data not shown). Thus, P493-6 cells appear to have a severe defect in production of mature 18S and 28S rRNA when c-Myc is down-regulated by tet. This defect could be due to a decreased processing of the 47S precursor or a reduced transcription rate of rDNA genes.

Open in a separate window

Open in a separate window

Open in a separate window

Figure 2

c-Myc-driven rRNA maturation in P493-6 cells. (A) Cells were arrested for 72 h and subsequently serum stimulated in the absence or presence of c-Myc. After 12 h, the cells were pulse-labelled with 2.5 µCi/ml ³H-uridine for 30 min and chased for the indicated times. Labelled RNAs were isolated, separated on a 1% agarose formaldehyde gel, transferred to a nylon membrane and visualized by fluorography. The same blot was hybridised with a probe specific for c-myc. Results of a representative experiment are shown. (B) Quantification of signals of the incorporated ³H-uridine in 28S rRNA. (C) Cells were arrested for 72 h, stimulated for 12 h in the presence of c-Myc, FCS or both, pulsed for 30 min with ³H-uridine and chased for 4 h. Signals for c-myc were detected by northern analysis.

c-Myc and serum can stimulate production of the 47S rRNA precursor

c-Myc had a strong positive effect on the production of the mature 18S and 28S rRNAs. To investigate the level at which this regulation occurs, we measured the amount of total 47S precursor rRNA by northern analysis and compared it with the amount of radioactively labelled precursor in pulse–chase experiments (Fig. ​(Fig.3A).3A). Arrested cells were stimulated with 10% FCS in the absence or presence of tet for the indicated time points. The 47S precursor was detected with a 5′-ETS leader probe in northern analysis. Since removal of this leader sequence from the 47S precursor is the first step during rRNA processing, only the newly transcribed full-length precursor hybridises to this probe. Therefore, the abundance of the 47S precursor is often used as a measure for transcription of rDNA genes. Surprisingly, the amount of precursor detected with this probe did not significantly change in arrested and stimulated P493-6 cells (Fig. ​(Fig.3A,3A, 5′-ETS).

Open in a separate window

Figure 3

Kinetics of 47S rRNA production. (A) P493-6 cells were arrested for 72 h and serum stimulated in the absence or presence of c-Myc for the indicated times. Subsequently, cells were pulse-labelled with ³H-uridine for 30 min and chased for 1 h. The 47S RNAs was visualised by fluorography as described. The same blot was hybridised with a probe specific for the 5′-ETS of the 47 S precursor and c-myc. (B) Northern blot with the 5′-ETS probe. EREB2-5 cells were depleted for 72 h of FCS and oestrogen (A) and then serum stimulated in the absence or presence of estrogen for the indicated times.

However, a pulse–chase labelling experiment in P493-6 cells indicated that the turn-over rate of the 47S rRNA precursor is affected by FCS and c-Myc. Arrested cells did not incorporate ³H-uridine into the 47S precursor (Fig. ​(Fig.3A,3A, lanes 1 and 9), whereas significant incorporation of ³H-uridine into the 47S precursor was already detectable 2 h after serum stimulation of P493-6 cells (lanes 2 and 10). Incorporation of ³H-uridine into the 47S precursor was also induced by c-Myc in the absence of FCS (Fig. ​(Fig.2C,2C, lane 3). In conclusion, the data show that FCS and c-Myc can independently stimulate the synthesis of the 47S rRNA precursor.

As mentioned above, the high abundance of the 47S precursor detected by the 5′-ETS probe in northern analysis was unexpected in un-stimulated P493-6 cells. To investigate whether this observation was specific for P493-6 cells, we tested the parental cell line EREB2-5 (14). EREB2-5 cells are primary B-lymphocytes infected with an EBV mutant expressing a conditional EBNA2-ER. In the absence of oestrogen and serum, EREB2-5 cells enter a quiescent, non-proliferative state. The 47S precursor was not detectable in arrested cells in northern analysis, but was up-regulated in stimulated cells (Fig. ​(Fig.3B,3B, 5′-ETS). Thus, the high level of the 47S precursor was specific for arrested P493-6 cells, but was not observed in quiescent EREB2-5 cells. The high abundance of the 47S rRNA precursor may be due to a leakiness of the tet-regulated c-myc construct in P493-6 cells (Discussion).

The density of RNA polymerase I complexes on ribosomal genes is independent of mitogenic stimulation and c-Myc

The pulse–chase labelling experiment above indicated that radiolabel is incorporated into the 47S precursor only after stimulation of cells. We performed nuclear run-on experiments to measure the density of pol I molecules on rDNA genes before and after stimulation of P493-6 cells. Strong transcription signals were obtained for oligonucleotide probes corresponding to the 5.8S, 18S and 28S rDNA sequence in arrested cells (Fig. ​(Fig.4A,4A, lane 1). These signals increased only slightly if cells were stimulated for 12 h with c-Myc/FCS (lane 2) or FCS alone (lane 3). Signals of the pol II transcribed genes GAPDH and actin could only be detected after six times longer exposure (lower panel). All signals were suppressed >95% when cells were pre-treated for 1 h with 2 µg/ml Actinomycin-D, an inhibitor of transcription (data not shown). The high abundance of pol I on ribosomal genes in arrested cells was unexpected and could be the reason for the high abundance of the 47S precursor observed in these cells. Therefore we also analysed arrested EREB2-5 cells, Ho15.19MycER cells and human primary fibroblast (Fig. ​(Fig.4B–D).4B–D). In all experiments, we found high levels of pol I complexes on ribosomal genes in arrested cells, which did not significantly increase after stimulation. The results strongly indicate that in arrested cells pol I pauses at multiple positions on ribosomal genes. In isolated nuclei, the paused pol I complexes become activated if the run-on reaction is carried out. The results suggest that transcription of rDNA genes is mainly regulated at the level of rRNA elongation.

Open in a separate window

Open in a separate window

Figure 4

Density of pol I complexes on ribsomal genes. (A) Nuclei were prepared from arrested P493-6 cells and from cells 12 h after serum stimulation in the absence or presence of c-Myc. Nuclear run-on reactions were performed. Labelled RNAs were purified and hybridised to probes specific for the indicated genes. A six-times longer exposure is shown in the lower panel. (B) Run-on assay in EREB2-5 cells. Arrested cells were stimulated for 20 h with oestrogen and run-on reactions were carried out as described. (C) Run-on assay in Ho15.19 MycER cells. Arrested cells were stimulated for 6 h with Myc/FCS, FCS or Myc. (D) Human primary fibroblasts were arrested for 72 h in 0.25% FCS. Run-on reactions were carried out in arrested, confluent and serum-stimulated cells.

Notably, c-Myc has recently been reported to regulate the expression of several pol III-specific genes (21). However, no difference for the transcription signal of the pol III-specific 7SK gene was detected after stimulation with c-Myc in P493-6 and Ho15.19MycER cells (Fig. ​(Fig.4A4A and C). We could also not detect any changes in the steady-state levels of 7SK RNA after up-regulation of c-Myc (Fig. ​(Fig.1A).1A). Thus, 7SK seems to be an example of a pol III transcribed gene that is not regulated by c-Myc.