Regulation of RNA polymerase III transcription during hypertrophic growth

Abstract

The cell division-independent growth of terminally differentiated cardiomyocytes is commonly associated with cardiovascular disease. We demonstrate that it is accompanied by a substantial rise in transcription by RNA polymerase (pol) III, which produces essential components of the biosynthetic apparatus, including 5S rRNA and tRNAs. This increase in transcription is achieved by changes in both the activity and level of the essential pol III-specific transcription factor TFIIIB. Erk and c-Myc, which directly activate TFIIIB in proliferating fibroblasts, also induce pol III transcription in growing cardiomyocytes. Furthermore, hypertrophic stimulation increases expression of the essential TFIIIB subunit Brf1, an effect not seen when fibroblasts proliferate. Erk mediates this induction of Brf1 expression and therefore contributes in at least two ways to pol III transcriptional activation during hypertrophy. Increased production of tRNA and 5S rRNA will contribute to the enhanced translational capacity required to sustain hypertrophic growth.

Figure 1

Hypertrophic stimuli activate pol III transcription. (A–F) Cultured cardiomyocytes were serum-starved for 24 h, then either maintained in serum-free media (control) or exposed to 10% FCS, 100 nM ET-1 or 100 μM PE for 16 h, as indicated. (A) Protein and DNA synthesis rates were determined by measuring incorporation of [³⁵S]methionine/cysteine or [³H]thymidine, respectively. The data represent the mean of five separate experiments, each with three replicates per condition. (B) Myocyte volume and number were assayed using a Z2 Coulter counter. The fold increases represent the mean of three separate experiments, each with two replicates per condition. (C) Specific primers for the transcripts indicated were used to PCR amplify cDNAs generated from total RNA of cultured cardiomyocytes treated as specified. (D) RT–PCR transcript intensities were quantified and the level of class III transcripts normalised to ARPP P0. The average fold increases are represented. (E) Northern analysis was performed using total RNA from cardiomyocytes treated as indicated. Northern blots were probed for B2, then stripped and re-probed for ARPP P0. (F) B2 and ARPP P0 RNA levels were quantified by densitometry, and B2 normalised to ARPP P0. The average fold increases are represented. (G) Whole-cell extracts (20 μg), prepared from serum-starved or FCS-stimulated cardiomyocytes, were tested for their ability to transcribe 5S rRNA, tRNA^Leu, B2 and VAI templates in vitro. (H) Average fold increases in transcriptional activity are displayed graphically. (I) Activation of pol III transcription accompanies hypertrophic growth in situ. RNA was extracted from hearts derived from mice subjected to TAC or sham operation. RT–PCR was performed using pre-tRNA^Tyr-, ANF- and ARPP P0-specific primers. RT–PCRs using RNA derived from two separate animals per condition are shown. (J) Six hearts were analysed as in (I) for each condition, and densitometry used to quantify transcript levels. The average level of pre-tRNA^Tyr, normalised to ARPP P0, is shown. For all bar charts, error bars indicate the standard deviation from the mean (^*significantly higher than control, P<0.05; ^**no significant difference from control).

Figure 2

Hypertrophic stimulation enhances interaction of TFIIIB and pol III, but not TFIIIC, with class III genes in vivo. Cultured cardiomyocytes were treated with 10% FCS for 16 h to induce hypertrophic growth. Control cells were maintained in serum-free medium. ChIPs were performed with antibodies against TFIIIB (Brf1 and Bdp1), TFIIIC (110 and 220) and pol III (RPC155), as indicated. Negative control ChIPs used a TFIIB antibody and beads alone. (A) Association of each factor with 5S rRNA and tRNA^Leu genes, in unstimulated and hypertrophic cells, was analysed by PCR with gene-specific primers. Input genomic DNA (10 and 2% of that used in ChIPs) was analysed in parallel. (B) PCR products were quantified by densitometry and normalised to the appropriate input. The average fold increases in factor binding are represented (n=3; ^*significantly higher than control, P<0.05; ^**no significant difference from control).

Figure 3

Brf1 levels limit pol III transcription in unstimulated cardiomyocytes, but increase specifically after hypertrophic stimulation. (A) Whole-cell lysates of rat1A fibroblasts that had been serum-starved for 24 h then either maintained in serum-free media or exposed to 10% FCS for 16 h were resolved by SDS–PAGE, and immunoblotted using antibodies against the proteins indicated. (B) The same procedure as outlined in (A) was used to analyse expression of the indicated proteins in cardiomyocytes. (C) Cultured cardiomyocytes were treated as in (A), then total RNA was extracted and used to generate cDNAs, which were amplified by PCR using primers specific for Brf1 or ARPP P0. (D) Cultured cardiomyocytes were infected with adenoviruses expressing HA-Brf1 (B) or GFP (G). Cells were serum-starved or stimulated with 10% FCS for 16 h before extracting whole-cell RNA. RNA was analysed by Northern blotting using B2- and ARPP P0-specific probes. (E) Northern blots were quantified by densitometry and B2 levels normalised to ARPP P0. The fold increases in normalised B2 expression relative to serum-starved, GFP control infected cells are shown. (n=8; ^*significantly different from SF GFP, P<0.05; ^**not significantly different from FCS GFP.)

Figure 4

Pol III transcripts increase within 2 h of hypertrophic stimulation, followed by Brf1 induction within 6 h. Serum-starved cardiomyocytes were stimulated with 10% FCS for the times indicated (0–48 h). Whole-cell protein or RNA was then extracted. (A) B2 and ARPP P0 RNA levels were analysed by Northern blotting, and (B) quantified by densitometry (n=4; ^*significantly different from control, P<0.05). (C) Protein lysates were resolved by SDS–PAGE and Western blotting performed using antibodies against the indicated proteins. (D) The level of Brf1 at each time point was quantified and normalised to actin (n=4; ^*significantly different from control, P<0.05).

Figure 5

Known TFIIIB regulators respond to hypertrophic stimulation. (A) Whole-cell lysates were prepared from cultured cardiomyocytes treated as in Figure 3B, and analysed by Western blotting using antibodies against the proteins, or phosphoproteins (‘phos-'), specified. The phos-RB antibodies used specifically detect RB phosphorylated on the indicated sites. (B) Cardiomyocytes were cultured as described in Figure 4, and protein lysates analysed by Western blotting to detect the indicated proteins.

Figure 6

c-Myc regulates pol III transcription in cardiomyocytes. (A, B) Cultured cardiomyocytes were infected with adenovirus expressing Lac Z (control) or c-Myc, as indicated. Cells were grown in the absence of hypertrophic stimuli, then total RNA extracted and used to generate cDNAs by reverse transcription. cDNAs were amplified by PCR using primers specific for the indicated transcripts. (A) Representative PCRs are shown. (B) The average fold increases in transcript levels (normalised to ARPP P0), induced by c-Myc, are represented graphically (n=3; ^*significantly higher than LacZ-expressing control, P<0.05). (C) c-Myc induces pol III transcription in the heart. RNA was extracted from hearts derived from wild-type adult mice (lanes 1 and 2) or transgenic littermates expressing MycER specifically in cardiomyocytes (MycER Tg, lanes 3 and 4). Mice had either been exposed to 4-hydroxytamoxifen (OHT) (lanes 2 and 4) or vehicle (lanes 1 and 3) for 1 week before removal of hearts. RNA was analysed by RT–PCR using pre-tRNA- and ARPP P0-specific primers, as specified. (D–G) Cells were cultured in the absence of serum (SF) or stimulated with 10% FCS for 24 h. c-Myc inhibitor (‘+') or vehicle (‘−') were also included for 24 h where indicated. Whole-cell RNA was analysed by (D) Northern blotting or (E) RT–PCR. (F) tRNA and B2 levels were quantified by densitometry and normalised to ARPP P0, as represented graphically (n=3; ^*significantly higher than SF+vehicle, P<0.05; ^**significantly lower than FCS+vehicle, P<0.05). (G) Cell volume was assayed using a Z2 coulter counter and the average fold changes in cell size are shown. (n=3; ^*significantly higher than SF+vehicle, P<0.05; ^**significantly lower than FCS+vehicle, P<0.05).

Figure 7

Erk induces pol III transcription and Brf1 expression in cardiomyocytes. (A) Cultured cardiomyocytes were infected (at a multiplicity of infection (m.o.i.) of 20, 50 or 80) with adenovirus expressing CAMEK or β-gal (negative control), as indicated. Cells were serum-starved for 48 h before harvesting. As a positive control for Erk activation, uninfected cells were incubated without (SF) or with 10% FCS for 16 h. Whole-cell protein was analysed by Western blotting using the antibodies indicated. (B) Total RNA from cells treated as in (A) was analysed by Northern blotting using the indicated probes. (C) Infections were carried out as in (A), using an m.o.i of 50 or 80, then total RNA analysed by RT–PCR using pre-tRNA^Leu- and ARPP P0-specific primers. (D–G) Cells were cultured in the absence of serum or stimulated with 10% FCS for 16 h. PD98059 or vehicle were also included for 16 h where indicated. Whole-cell RNA was analysed by (D) RT–PCR or (E) Northern blotting. (F) Protein and DNA synthesis rates were determined by measuring the incorporation of [³⁵S]methionine/cysteine or [³H]thymidine, respectively. The fold increases represent the mean of four separate experiments, each with three replicates per condition (^*significantly higher than SF+vehicle, P<0.05; ^•significantly lower than FCS+vehicle, P<0.05; ^**not significantly different from SF+vehicle). (G) Protein lysates were analysed by Western blotting using antibodies against the proteins indicated. (H) Total RNA was extracted from cardiomyocytes infected (at an m.o.i of 50) with adenovirus expressing CAMEK or β-gal (negative control), as indicated. Cells were serum-starved for 48 h before harvesting. RT–PCR was performed using the primers specified. (I) Whole-cell lysates, treated as in (A), were analysed by Western blotting using the indicated antibodies (^*nonspecific band).

Figure 8

Model for activation of pol III transcription following induction of hypertrophic growth. Hypertrophic stimulation leads to the rapid activation of Erk and c-Myc, which are likely to target TFIIIB directly (as documented in other cell types), leading to the initial induction of pol III transcription. However, limiting levels of Brf1, and sequestration of TFIIIB by hypophosphorylated RB, may restrict the extent to which these molecules activate transcription at early time points. The increase in Brf1 abundance by 6 h, and the subsequent hyperphosphorylation of RB, would relieve these restrictions, thus allowing maximum pol III output. Decreased c-Myc levels and RB dephosphorylation might account for the downregulation of pol III transcription occurring 24–48 h after hypertrophic stimulation. Abbreviations used: TFIIIB, IIIB; TFIIIC, IIIC; c-Myc, Myc. Bold arrows below transcription complexes indicate active class III gene transcription; dashed arrows indicate transcription at a reduced rate.