The JNKs differentially regulate RNA polymerase III by coordinately modulating the expression of all TFIIIB subunits

Shuping Zhong and Deborah L. Johnson1

Department of Biochemistry and Molecular Biology, Keck School of Medicine, University of Southern California, and the Norris Comprehensive Cancer Center, 2011 Zonal Avenue, Los Angeles, CA 90033

Edited by Robert G. Roeder, The Rockefeller University, New York, NY, and approved June 12, 2009 (received for review May 4, 2009) RNA polymerase (pol) III-dependent transcription is subject to strin- III transcription has been extensively observed in transformed and gent regulation by tumor suppressors and oncogenic and tumor cells supporting the idea that it plays a crucial role in enhanced RNA pol III transcription is essential for cellular transfor- tumorigenesis (11). Recent studies have demonstrated that en- mation and tumorigenesis. Since the c-Jun N-terminal kinases (JNKs) hanced RNA pol III transcription is required for promoting onco- display both oncogenic and tumor suppressor properties, the roles of genic transformation (12, 13). Thus, understanding the molecular these proteins in regulating RNA pol III transcription were examined. pathways by which this transcription process is controlled may In both mouse and human cells, loss or reduction in JNK1 expression provide useful therapeutic approaches for cancer. represses RNA pol III transcription. In contrast, loss or reduction in The TATA binding (TBP) associates with specific JNK2 expression induces transcription. The JNKs coordinately regu- proteins, TBP-associated factors (TAFs), to form at least 3 late expression of all 3 TFIIIB subunits. While JNK1 positively regulates distinct complexes, SL1, TFIID, and TFIIIB, which specify the TBP expression, the RNA pol III-specific factors, Brf1 and Bdp1, JNK2 role of TBP in the expression of directed by RNA pol I, negatively regulates their expression. Brf1 is coregulated with TBP II, and III, respectively. TFIIIB and the multiprotein complex, through the JNK target, Elk-1. Reducing Elk-1 expression decreases TFIIIC, are required to form specific transcription initiation Brf1 expression. Decreasing JNK1 expression reduces Elk-1 occupancy complexes on RNA pol III-dependent promoters. TFIIIB, com- at the Brf1 promoter, while decreasing JNK2 expression enhances posed of TBP and the RNA pol III-specific factors, Brf1 and recruitment of Elk-1 to the Brf1 promoter. In contrast, regulation of Bdp1, is often the target of regulatory responses that give rise to Bdp1 occurs through JNK-mediated alterations in TBP expression. altered RNA pol III transcription activity. Phosphorylation of Altered TBP expression mimics the effect of reduced JNK1 or JNK2 Brf1 and Bdp1 has been shown to regulate the assembly or levels on Bdp1 expression. Decreasing JNK1 expression reduces the function of the TFIIIB complex (7, 14–16). TBP expression is occupancy of TBP at the Bdp1 promoter, while decreasing JNK2 up-regulated by oncogenic Ras signaling and required to medi- expression enhances recruitment of TBP to the Bdp1 promoter. ate Ras transforming function (17). Together, these results provide a molecular mechanism for regulating TBP expression is induced through the activation of the epider- RNA pol III transcription through the coordinate control of TFIIIB mal growth factor receptor (EGFR1), requiring the activation of subunit expression and elucidate opposing functions for the JNKs in Ras and all 3 classes of MAPKs (18). JNK1 induces TBP expression, regulating a large class of genes that dictate the biosynthetic capacity while JNK2 represses TBP expression (19). This is accomplished of cells. through the opposing functions of the JNKs to modulate the phosphorylation state of Elk-1 and its recruitment to the TBP c-jun N-terminal kinases ͉ Elk-1 ͉ TATA-binding protein promoter. The ability of the JNKs to regulate TBP expression determines the proliferation rates of mouse embryo fibroblasts he c-jun N-terminal kinases (JNKs) are members of the mito- (MEFs). TBP-mediated changes in MEF proliferation are due, at Tgen-activated protein kinase (MAPK) family. JNKs are acti- least in part, to its ability to regulate c-jun transcription. TBP vated in response to stress, proinflammatory stimuli, and mitogenic expression can also be induced through AP-1 via the recruitment factors (1, 2). Three distinct genes the JNKs. JNK1 and of c-jun/c-fos to the TBP promoter (20). How alterations in cellular JNK2 are ubiquitously expressed, while JNK3 is more selectively TBP concentrations might affect the expression of other genes expressed in brain, heart, and testis (3). In addition, alternative involved in growth and proliferation remains to be determined. splicing gives rise to at least 10 JNK isoforms. JNK1 and JNK2 have As TBP is a component required for the transcription of RNA been shown to phosphorylate transcription factors such as c-Jun, pol III-dependent genes, we explored the possibility that the JNKs ATF-2, c-Fos, p53, Elk-1, and c-Myc. While initial studies indicated might modulate the activity of this class of genes. Our studies reveal that JNK1 and JNK2 possess redundant functions, subsequent that the JNKs differentially regulate RNA pol III transcription. studies support the idea that these proteins also have distinct JNK1 induces, while JNK2 represses, RNA pol III transcription. functions. The JNKs can induce both apoptotic and proliferative Notably, the expression of all 3 TFIIIB components, TBP, Bdp1, responses, depending on the physiological context and the time and Brf1, are coordinately regulated by these JNKs. Regulation of course of activation (4). The roles of these JNKs in tumorigenesis Bdp1 expression is modulated through JNK-mediated changes in are controversial (3). Depending on the cell type and stimulus, the TBP expression and by altered recruitment of TBP to the Bdp1 JNKs have been shown to stimulate oncogenic transformation or promoter. In contrast, Brf1 expression is regulated via JNK- act as tumor suppressors. Identifying the downstream molecular events selectively modulated by JNK1 and JNK2 will importantly clarify their roles in determining the oncogenic state of cells. Author contributions: S.Z. and D.L.J. designed research; S.Z. performed research; S.Z. and RNA polymerase (pol) III-dependent transcription and the D.L.J. analyzed data; and S.Z. and D.L.J. wrote the paper. production of its major products, 5S rRNAs and tRNAs, is tightly The authors declare no conflict of interest. regulated, controlling the translational and growth capacity of cells. This article is a PNAS Direct Submission. RNA pol III transcription is induced by oncogenic proteins, such as 1To whom correspondence should be addressed. E-mail: [email protected]. Ras (5), c-myc (6), and PI3 kinase (7), and repressed by tumor This article contains supporting information online at www.pnas.org/cgi/content/full/ suppressors Rb (8, 9), p53 (10), and PTEN (7). Enhanced RNA pol 0904843106/DCSupplemental.

12682–12687 ͉ PNAS ͉ August 4, 2009 ͉ vol. 106 ͉ no. 31 www.pnas.org͞cgi͞doi͞10.1073͞pnas.0904843106 Downloaded by guest on September 25, 2021 e mediated recruitment of Elk-1 to the Brf1 promoter. Together, p - - y / / Jnk2 -/- A t 1- 2- B Jnk1 -/- d k k il n n these results support the idea that JNK1 acts to promote RNA pol J J W -- III transcription, while JNK2 negatively regulate this process, JNK1 ++ JNK2 -- ++ thereby differentially controlling the biosynthetic capacity of cells. 5.0 6 1.5

Results 4 1.0 JNK1 and JNK2 Have Opposing Roles in Regulating RNA pol III 2.5 2 0.5 transcription gene transcription Transcription. The potential for JNK1 and JNK2 to regulate RNA gene transcription Fold change in tRNA in change Fold Fold changein tRNA polymerase (pol) III-dependent transcription was first assessed tRNA in change Fold 0.0 0 0.0 Jnk1-/- Jnk2 -/- WT JNK1 -+ JNK2 - + using MEFs that were deficient for either Jnk1 or Jnk2. A tRNA Jnk1-/- Jnk2-/- Ϫ/Ϫ gene reporter was transiently expressed in wild-type, Jnk1 , and JNK1 +- JNK2 +- Ϫ Ϫ Jnk2 / cells, and the amount of transcription was measured by a HA-JNK1 HA-JNK2 ribonuclease protection assay (Fig. 1A). The RNA produced con- β-actin β-actin tains a 12-bp insert that does not permit processing of the precursor C wild type Jnk1 -/- Jnk2 -/- transcript and allows it to be distinguished from endogenous Anisomycin - + - + - + transcripts (10). Compared with tRNA gene transcription activity in the wild-type cells, cells lacking Jnk1 displayed a decrease in RNA 3 3 3 pol III transcription, whereas transcription was enhanced in cells lacking Jnk2. To determine whether this difference in transcription 2 2 2 was a result of JNK status, expression plasmids for JNK1 or JNK2 1 1 1

were used to re-express these proteins. Expression of JNK1 in the transcriptiongene Ϫ/Ϫ Fold change in tRNA Jnk1 cells enhanced RNA pol III transcription, whereas expres- 0 0 0 Ϫ/Ϫ sion of JNK2 in the Jnk2 cells decreased transcription (Fig. 1B). Anisomycin - + -+ -+ MEFs were treated with anisomycin, a potent inducer of the JNKs, e p /- /- D ty 1- 2- to determine the role of activated JNKs on tRNA gene transcription ld K K i N N W J J (Fig. 1C). Anisomycin induced transcription in both the wild-type Nuclear exract and Jnk2Ϫ/Ϫ cells, but no effect on transcription was observed in the Ϫ/Ϫ 3 wild type Jnk1 cells. This suggests that while anisomycin induces tran- Jnk1-/- scription through JNK1, JNK2 acts to repress transcription in the Jnk2-/- absence of anisomycin-induced stimulation. To examine whether 2 JNK-mediated effects on transcription could be reproduced in 1

vitro, nuclear extracts were prepared, and transcription reactions gene transcription were performed. Extracts prepared from Jnk1Ϫ/Ϫ cells displayed a tRNA in change Fold decreased capacity for transcription, whereas extracts from the 0 Ϫ/Ϫ Jnk2 cells exhibited an increased capacity for transcription, 5 5.0 wild type E Jnk1-/- compared to extracts prepared from the wild-type cells (Fig. 1D). 4 Jnk2-/- Leu Furthermore, transcription was induced in extracts derived from 3 wild-type and Jnk2Ϫ/Ϫ cells that were treated with anisomycin, but 2.5 Ϫ Ϫ 2 in 7SL RNA Fold change / Fold change

no induction was observed in extracts derived from Jnk1 cells ( pre-tRNA in Fig. S1). RT-qPCR was used to assess the amounts of endogenous 1 precursor tRNALeu and 7SL RNA transcripts in the MEFs (Fig. 0 0.0 1E). Consistent with the tRNA gene reporter activity, the RNA pol Fig. 1. RNA pol III transcription is differentially regulated by JNK1 and Ϫ/Ϫ III transcription was decreased in the Jnk1 cells, and increased JNK2 in MEFs. (A) RNA pol III transcription is altered in MEFs deficient for Ϫ Ϫ BIOCHEMISTRY in the Jnk2 / cells, relative to the wild-type MEFs. Jnk1 or Jnk2. Wild-type, Jnk1Ϫ/Ϫ, and Jnk2Ϫ/Ϫ MEFs were transiently trans- The ability of the JNKs to modulate RNA pol III transcription fected with a tRNA gene reporter plasmid. RNA was extracted, and RNase in a human hepatoma cell line was examined. Using small protection assays were carried out. Quantification of 3 independent ex- interfering (siRNAs) specific for either JNK1 or JNK2, periments is shown, and an example of an autoradiogram is shown above expression of these isoforms was selectively decreased (Fig. 2A). the graph. (B) Expression of JNK1 increases, while JNK2 represses, tran- The amounts of endogenous precursor tRNALeu and 7SL RNA scription. MEFs were cotransfected with expression plasmids for HA-JNK1 or HA-JNK2 with the tRNA gene reporter. RNA analysis was performed as transcripts were measured when JNK1 or JNK2 expression was in A Upper, and immunoblot analysis was used to detect the HA-JNKs using reduced (Fig. 2B). Transcription of both RNA pol III products HA antibodies (Lower). Membranes were stripped and reprobed with was decreased upon reduced JNK1 expression, while transcrip- antibodies against ␤-actin. (C) Anisomycin induction of tRNA transcription tion was increased when JNK2 expression was reduced. To- requires JNK1. MEFs were transfected with a tRNA gene reporter and gether, these results support the idea that JNK1 positively treated with or without anisomycin before isolation. tRNA gene transcrip- regulates RNA pol III transcription, whereas JNK2 negatively tion was measured as in A.(D) JNK-mediated regulation of tRNA gene regulates transcription of these genes in both human and mouse transcription can be reproduced in vitro. Nuclear proteins were prepared cell lines. from MEFs, and in vitro transcription reactions were carried out using the tRNA reporter template and 10 or 20 ␮g nuclear extract. (E) Endogenous RNA pol III transcription is altered in MEFs deficient for Jnk1 or Jnk2. JNK1 and JNK2 Coordinately Regulate the Expression of All TFIIIB RT-qPCR was performed using RNA isolated from wild-type, Jnk1Ϫ/Ϫ, and Subunits. To determine the mechanism by which the JNKs regulate Jnk2Ϫ/Ϫ MEFs and with specific primers for precursor (pre) tRNALeu or 7SL RNA pol III transcription, we examined potential changes in RNA. Quantification of 3 independent experiments from 3 independent TFIIIB as it has been shown to be a target of various tumor RNA isolations is shown. suppressors and oncogenes (21). The amounts of each of the 3 subunits, Brf1, Bdp1, and TBP were analyzed by immunoblot analysis using lysates derived from Huh-7 cells and MEFs when expression is reduced or eliminated in both cell lines (Fig. 2C). expression of the JNKs was altered. Consistent with our earlier Notably, the expression of both RNA pol III-specific components, results (19), TBP expression was decreased when JNK1 expression Brf1 and Bdp1, paralleled that of TBP. In contrast, TFIIIC63 is reduced or absent, and its expression was increased when JNK2 expression was not affected by repressing either JNK1 or JNK2

Zhong and Johnson PNAS ͉ August 4, 2009 ͉ vol. 106 ͉ no. 31 ͉ 12683 Downloaded by guest on September 25, 2021 A Huh 7 cells modulate RNA pol III transcription. Expression of TBP in JNK2 siRNA -- + Huh-7 cells and MEFs was altered, and the resultant effect on JNK1 siRNA --+ Leu mm siRNA + -- the amounts of precursor tRNA and 7SL RNA transcripts JNK1 were measured. Cells transfected with a TBP expression plasmid

JNK2 to increase the cellular amounts of TBP displayed an increase in β-actin RNA pol III transcription (Fig. 3A and B, Left), indicating that TBP is limiting for transcription in these cells. In contrast, B 3 3 reduction of TBP expression by transfection of a specific TBP siRNA into Huh-7 cells or by expression of an antisense con-

Leu Ϫ/Ϫ 2 2 struct corresponding to the C-terminal region of TBP in Jnk2 cells, each resulted in decreased amounts of RNA pol III in 7SL RNA 7SL in

Fold change transcripts (Fig. 3 A and B, Right). Thus, modulating the levels Fold change 1 1 in pre-tRNAin of TBP in these cells regulates RNA pol III transcription.

0 0 We next ascertained whether alterations in cellular TBP amounts siRNA mm JNK1 JNK2 siRNA mm JNK1 JNK2 could be responsible for the JNK-mediated changes in Brf1 and Bdp1 expression observed. The amounts of these proteins in lysates Huh 7 cells MEFs

C e derived from Huh-7 and MEFs that had either increased or JNK2 siRNA -- + p - - y / / t 1- 2- JNK1 siRNA --+ d k k decreased amounts of TBP were determined (Fig. 3C). While il n n + -- J J mm siRNA W manipulating TBP expression in either MEFs or Huh-7 cells TBP changed neither Brf1 nor TFIIIC expression, Bdp1 expression Bdp1 63 was increased in cells where TBP expression was enhanced, while Brf1 Bdp1 expression was decreased in cells where TBP amounts were TFIIIC63 reduced. Similarly, analysis of the mRNAs for these proteins β-actin revealed that altering TBP expression regulates Bdp1 mRNA levels; 3 mm siRNA wild type JNK1 siRNA JNK1-/- however, neither Brf1 or TFIIIC63 mRNA levels were modulated JNK2 siRNA 2 JNK2 -/- by changes in TBP expression (Fig. 3 D and E). These results 2 support the idea that Bdp1 expression, but not that of Brf1, is 1 regulated by cellular TBP concentrations. 1 Fold changeFold protein in Fold changeFold protein in JNK-Mediated Regulation of Bdp1 Expression Occurs Through Altered 0 0 Bdp1 Brf1 Bdp1 Brf1 Recruitment of TBP to the Bdp1 Promoter. The above results support the idea that JNK1 and JNK2 differentially regulate Bdp1 expres- sion through their ability to affect cellular TBP amounts. As D mm siRNA Wild type 3 JNK1 siRNA 3 Jnk1-/- alterations in cellular TBP affect Bdp1 mRNA levels, this suggested JNK2 siRNA Jnk2-/- that TBP might directly regulate Bdp1 transcription. To investigate 2 2 this possibility, the transcription start site was first determined by RACE analysis (Fig. 4A). The start site was mapped to a position 1 1 that is 202 bp upstream of the translation start site. To determine Fold change in mRNA

Fold change in mRNA whether JNK-mediated regulation of Bdp1 occurs through its 0 0 ability to regulate cellular TBP amounts and ultimately affect the Bdp1 Brf1 Bdp1 Brf1 recruitment of TBP to the Bdp1 promoter, immunopre- Fig. 2. Expression of all TFIIIB subunits are coordinately regulated by the cipitation (ChIP) assays were performed. Huh-7 cells were trans- JNKs. (A) siRNAs specifically repress JNK1 or JNK2 expression. Protein lysates fected with either mismatch RNA or siRNAs directed against either derived from Huh-7 cells transfected with siRNAs against JNK1 or JNK2 were JNK1 or JNK2. Using primers that span the transcription start site subjected to immunoblot analysis using antibodies against JNK1, JNK2, or of the Bdp1 promoter (Fig. 4A), the occupancy of TBP was ␤-actin as indicated. (B) RNA pol III transcription is differentially regulated by significantly decreased when JNK1 expression was reduced (Fig. JNK1 or JNK2 in Huh-7 cells. Huh-7 cells were transfected with siRNAs for JNK1, 4B). In contrast, recruitment of TBP was increased when JNK2 JNK2, or mismatch RNA. RNA was isolated from the transfected cells, and real time RT-qPCR was performed using specific primers for precursor (pre) tRNALeu expression was reduced. This can be compared with the occupancy (Left) or 7SL RNA (Right). (C) JNK1 and JNK2 differentially regulate expression of histone H3 that remained unchanged when either JNK1 or JNK2 of TFIIIB subunits, TBP, Bdp1, and Brf1. Immunoblot analysis was preformed expression was repressed. Interrogating sequences upstream of the using lysates derived from either siRNA transfected Huh-7 cells (Upper Left)or Bdp1 start site, there was no detectable occupancy of TBP over this from MEFs (Upper Right). Antibodies against TBP, Bdp1, Brf1, TFIIIC63,or region, whereas the amount of bound histone H3 in this region was ␤-actin were used as indicated. Quantification of protein amounts from 3 unchanged when JNK expression was altered. These results support independent experiments relative to ␤-actin is shown Bdp1 and Brf1 for results the idea that JNK1 enhances the recruitment of TBP, while JNK2 obtained using Huh-7 (Lower Left) and MEFs (Lower Right). (D) JNK1 and JNK2 negatively regulates TBP occupancy, on the Bdp1 promoter. differentially regulate Bdp1 and Brf1 mRNAs. RT-qPCR was used to measure Bdp1 and Brf1 mRNAs using RNA derived from siRNA transfected Huh-7 cells Brf1 Expression Is Regulated Through the JNKs by Altered Recruitment or from MEFs as described in B. of Elk-1 to the Brf1 Promoter. To determine how the JNKs regulate Brf1 expression, we considered our previous studies demonstrating expression. Similarly, analysis of Brf1 and Bdp1 mRNAs revealed that the JNKs regulate TBP expression through their opposing that JNK1 positively regulated their expression, while JNK2 neg- effects on Elk-1 phosphorylation and its subsequent recruitment to atively regulated their expression in both Huh-7 cells and MEFs the TBP promoter (19). We therefore examined whether Brf1 (Fig. 2D). These results support the idea that the JNKs coordinately might also be regulated through Elk-1. When Elk-1 expression was regulate expression of all TFIIIB subunits. reduced in Huh-7 cells using siRNAs specific for Elk-1, this resulted in a reduction of Brf-1 expression (Fig. 5A). To determine whether The JNKs Differentially Regulate Bdp1 Expression Through Their Elk-1 might be directly regulating the Brf1 promoter, ChIP assays Ability to Modulate TBP Expression. We next assessed whether were performed. Primers that span the previously characterized JNK-mediated changes in cellular TBP amounts, alone, could Brf1 promoter were used (Fig. 5B) (22). Elk-1 was found to occupy

12684 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0904843106 Zhong and Johnson Downloaded by guest on September 25, 2021 A Huh7 cells A Bdp1 promoter -2725 / -2581 -141 / +31 3 3 1.5 1.5

-3000bp +200bp 2 2 1.0 1.0 TSS

ChIP primers: -141 / +31 1 1 0.5 0.5 B 3 1.5 mm siRNA Fold change pre-tRNA Fold change 7SL RNA 7SL change Fold Fold change 7SLRNA Fold changepre-tRNA JNK1 siRNA 0 0 0.0 0.0 JNK2 siRNA TBP +- TBP +- TBP siRNA - + TBP siRNA - + 2 1.0 B wild type MEFs Jnk2-/- MEFs occupancy

3 3 1.5 1.5 TBP Relative 1 0.5 Relative H3 occupancy 2 2 1.0 1.0 0 0.0

ChIP primers: -2725 / -2581 1 1 0.5 0.5 3 1.5 mm siRNA Fold chnag pre-tRNA chnag Fold Fold change 7SLRNA Fold change 7SLRNA Fold change pre- tRNA pre- change Fold JNK1 siRNA 0 0 0.0 0.0 JNK2 siRNA -+ TBP TBP -+ AS TBP - + AS TBP - + 2 1.0

wt MEFs Jnk2-/- MEFs Huh7 cells

C occupancy -+TBP -+AS TBP -+TBP -+TBP siRNA TBP Relative 1 0.5 TBP TBP TBP TBP Relative H3 occupancy H3 Relative

Bdp1 Bdp1 Bdp1 Bdp1 0 0.0 Brf1 Brf1 Brf1 Brf1 Fig. 4. JNK1 and JNK2 differentially regulate TBP occupancy on the Bdp1

β-actin β-actin HA-TBP TFIIIC63 promoter. (A) Schematic of the Bdp1 promoter. The diagram depicts the start site for transcription on the human Bdp1 promoter (TSS) identified by using a HA-TBP β-actin β-actin RACE kit (Invitrogen). The 2 primer pairs were used to amplify the PCR products, 1 at (Ϫ142 bp/ϩ31 bp) encompassing the TSS and an upstream TBP Bdp1 TFIIIC D 1.5 1.5 1.5 Brf1 1.5 63 fragment (Ϫ2725 bp/-2581 bp). (B) TBP occupancy on the Bdp1 promoter is regulated by the JNKs. Huh-7 cells were transfected with siRNAs for JNK1, 1.0 1.0 1.0 1.0 JNK2, or mismatch siRNAs. ChIP assays were performed using antibodies against TBP or histone H3 and the primer pair -141/31 encompassing the

0.5 0.5 0.5 0.5 transcription start site (Upper) or the primer pair -2725/-2581 upstream of the transcription start site. The bars represent ϮSE of at least 3 independent Fold change in mRNA in change Fold determinations from 3 separate chromatin preparations. The occupancy of 0.0 0.0 0.0 0.0 TBP siRNA -+ -+ - + -+ TBP or histone H3 was calculated relative to the percent occupancy observed in cells transfected with mismatch RNA. E 4 TBP 3 Bdp1 3 Brf1 3 TFIIIC63 3 2 2 2 the Brf1 promoter in Huh-7 cells, and decreasing Elk-1 expression 2 reduced its occupancy on the promoter. Reducing JNK1 expression 1 1 1 1 resulted in diminished Elk-1 binding, whereas decreased JNK2

Fold change in mRNA expression enhanced Elk-1 recruitment to the Brf1 promoter (Fig. 0 0 0 0 5C). In contrast, sequences upstream of the Bdp1 promoter did not TBP - + - + - + - +

bind to Elk-1, and histone H3 binding was unaffected with changes BIOCHEMISTRY Fig. 3. Changes in the cellular level of TBP modulates Bdp1 expression and in JNK expression. These results support the idea that Brf1 RNA pol III transcription. (A) Changes in TBP expression modulate transcrip- expression is differentially regulated by the JNKs through their tion of tRNALeu and 7SL RNA in Huh-7 cells. Cells were transfected with a ability to affect the recruitment of Elk-1 to the Brf1 promoter. human TBP expression plasmid (Left) or a TBP siRNA (Right). Real-time RT-PCR was performed using isolated RNA and primers specific for pretRNALeu or 7SL Discussion RNA, and the fold change was calculated by normalizing to the amount of GAPDH mRNA. (B) Changes in TBP expression modulate transcription of These studies define opposing roles for the JNKs in regulating RNA tRNALeu and 7SL RNA in MEFs. Wild-type MEFs were transfected with TBP pol III-dependent gene transcription. As the major products of expression plasmid and amounts of pretRNALeu and 7SL RNA were measured these genes, tRNAs and 5S rRNAs, are essential components of the (Left), and Jnk2Ϫ/Ϫ MEFs were transfected with antisense TBP construct and protein synthesis machinery, these results further identify a link the amount of pretRNALeu and 7SL RNA was measured as in A.(C) Alter- between the stress-induced JNKs and translational capacity. We ations in TBP expression changes Bdp1, but not Brf1 expression. Wild-type demonstrate that JNK1 serves to induce this class of genes, whereas Ϫ/Ϫ MEFs were transfected with a TBP expression plasmid and Jnk2 MEFs JNK2 acts to negatively regulate transcription. While both JNK1 with an antisense TBP construct (Left). Huh-7 cells were transfected with and JNK2 serve to regulate RNA pol III-dependent transcription, either an siRNA against TBP or an expression plasmid for HA-tagged TBP (Right). Resulting lysates were subjected to immunoblot analysis using JNK1 acts to regulate these genes following stress induction by antibodies against TBP, Bdp1, Brf1, TFIIIC63, ␤-actin, or HA-TBP as desig- anisomycin whereas JNK2 negatively regulates these genes in nated. (D) Reduction in cellular TBP expression decreases Bdp1, but not nonstimulated cells. Given that previous studies have shown that Brf1, mRNA levels in Huh-7 cells. Cells were transfected with an siRNA JNK2 negatively impairs RNA pol I-dependent transcription (23), against TBP, RNA was isolated, and RT-qPCR was conducted to measure the it is clear that the JNKs produce pleiotropic effects on gene amounts of TBP, Bdp1, Brf1, and TFIIIC63 mRNA. Three independent exper- expression that affect the biosynthetic capacity of cells. iments were performed, and the fold change was normalized to GAPDH Translational control has emerged as an important step in mRNA. (E) Enhanced expression of TBP increases Bdp1, but not Brf1, mRNA levels in Huh-7 cells. Cells were transfected with a TBP expression plasmid, oncogenic transformation. Enhanced production of RNA pol III and RT-qPCR was performed to quantify the amounts of mRNAs for TBP, transcripts is strongly associated with a transformed phenotype. Bdp1, Brf1, and TFIIIC63. Three independent experiments were performed, This is consistent with the transcription of these products being and the fold change was normalized to GAPDH mRNA. tightly controlled by tumor suppressors and oncogenic proteins

Zhong and Johnson PNAS ͉ August 4, 2009 ͉ vol. 106 ͉ no. 31 ͉ 12685 Downloaded by guest on September 25, 2021 A Elk-1 siRNA - + (26). Bdp1 expression is induced in cells transformed with Elk-1 papovaviruses (27), while an increase in Brf1 mRNA is observed in response to the integration of high-risk human papillomavirus Brf1 (28). Our current studies define a mechanism whereby the β-actin expression of all 3 TFIIIB components is coordinately regulated. The ability of the JNKs to differentially control RNA pol III B Brf1 Promoter transcription is mediated through altered expression of the -2983 / -2864 -366 / -195 TFIIIB complex. JNK-mediated regulation of all 3 TFIIIB -3200bp +200bp subunits is orchestrated through Elk-1. Although a role for Elk-1 TSS in human cancer has not been clearly established, initial results C ChIP primers: -366 / -195 indicate that reduced expression of Elk-1 suppresses tumorige- 2.0 2.0 mm siRNA nicity of human hepatocellular carcinoma cells (29). While Bdp1 JNK1 siRNA expression is not directly regulated by Elk-1, the resultant 1.5 1.5 JNK2 siRNA Elk-1-mediated change in TBP expression serves to regulate 1.0 1.0 Bdp1 promoter activity and cellular Bdp1 concentrations. This occupancy

Relative Elk-1 Relative demonstrates that cellular TBP amounts regulate the expression 0.5 0.5 of 1 of its associated factors. It will be interesting to determine Relative H3 occupancy 0.0 0.0 whether changes in TBP expression also serve to regulate the expression of its other associated factors present in the RNA pol ChIP primers -2983 / -2864 I-specific, SL1, or RNA pol II-specific, TFIID, complexes. 2.0 2.0 mm siRNA JNK1 siRNA Our previous studies demonstrated that the JNKs have opposing JNK2 siRNA 1.5 1.5 affects on Elk-1 phosphorylation, which is required for its DNA binding activity (19). The ability of JNK1 to induce, and JNK2 to 1.0 1.0 repress, Elk-1 phosphorylation subsequently controls Elk-1 recruit- occupancy Relative Elk-1 Relative 0.5 0.5 ment to the TBP promoter, dictating the amount of cellular TBP

Relative H3 occupancy H3 Relative expressed. Our current studies demonstrate that Elk-1 also directly 0.0 0.0 regulates the Brf1 promoter allowing synchronized expression of Fig. 5. JNK1 and JNK2 differentially regulate Elk-1 occupancy on the Brf1 these proteins. Accordingly, the Brf1 promoter contains a consen- promoter. (A) Decreased Elk-1 expression reduces Brf1 expression. Huh-7 cells sus Ets binding site at -671 relative to the transcription start site. As were transfected with siRNAs specific for Elk-1 (ϩ) or mismatch RNA (-), and Brf1 expression is increased in cardiomyocytes when MEK1, an immunoblot analysis was used to determine the amount of Elk-1, Brf1, and activator of ERK, is overexpressed (30), it is likely that the ␤ -actin. (B) Schematic of the Brf1 promoter. The diagram depicts the start site for activation of multiple signaling pathways converge on Elk-1 to transcription on the human Brf1 promoter identified previously (22). The 2 primer pairs used to amplify the ChIP PCR products at -366/-195 and an upstream regulate Brf1 expression. However, we cannot rule out the possi- fragment at -2983/-2864 are depicted. (C) The JNKs have opposing affects on Elk-1 bility that other JNK-targeted transcription factors regulate the occupancy on the Brf1 promoter. Huh-7 cells were transfected with siRNAs for Brf1 promoter and RNA pol III transcription. JNK1, JNK2, or mismatch siRNAs. ChIP assays were performed using antibodies Although initial in vitro studies defined common JNK substrates against Elk-1 or histone H3 and the primer pair encompassing -366/-195 (top)or such as c-jun, p53, and Elk-1, subsequent studies demonstrated that the -2983/-2864 (bottom) relative to the transcription start site. The bars represent JNK1 and JNK2 actually have opposing functions on the stability or ϮSE of 4 independent determinations from 3 separate chromatin preparations. phosphorylation of these substrates in vivo. Both JNKs phosphor- The occupancy of Elk-1 or histone H3 was calculated relative to the percent ylate c-jun, yet JNK1 positively regulates, whereas JNK2 negatively occupancy observed in cells transfected with mismatch RNA. regulates, c-jun stability (31). Another substrate of these kinases, p53, is also differentially regulated by the JNKs (32). Inhibition of (21, 24). A causative role for RNA pol III-dependent transcrip- JNK1 induces p53 expression, whereas inhibition of JNK2 represses tion in determining the oncogenic state of cells was recently p53 expression in primary human fibroblasts. While Elk-1 was established (12, 13). Preventing TBP- or c-myc-mediated in- originally shown to be a substrate for both JNK1 and JNK2 in vitro creases in RNA pol III transcription abrogated their transform- (33), only JNK1 was shown to induce phosphorylation of Elk-1 in ing activity (12). Increased RNA pol III transcription resulted in vivo, while JNK2 negatively regulates Elk-1 phosphorylation (19). the preferential translation of mRNAs encoding growth- Whether JNK2 antagonizes the function of JNK1 or indirectly promoting proteins, suggesting that there is not a uniform mediates the dephosphorylation of Elk-1 or other transcription quantitative increase in transcription or protein synthesis (13). factors remains to be determined. The ability of the JNKs to The ability of JNK1 to induce RNA pol III transcription supports differentially regulate the function or expression of a variety of its function in promoting an oncogenic state, whereas the ability transcription factor substrates has the potential for producing of JNK2 to repress expression of this class of genes is consistent pleiotropic consequences on cellular . Consistent with its role in suppressing a transformed phenotype. with this idea, our studies demonstrate that the opposing affect of Modulation of TFIIIB activity has been shown to play a central the JNKs on Elk-1 phosphorylation serves to simultaneously con- role in regulating RNA pol III transcription. The tumor sup- trol the expression of TBP and its associated factors, Bdp1 and Brf1, pressors RB and p53 bind directly to TFIIIB and prevent its which ultimately dictates the transcription of a large class of genes. recruitment to promoters (8–10). PTEN negatively regulates Our studies demonstrate that TBP is a limiting component for Brf1 phosphorylation and its ability to associate with TBP to RNA pol III-dependent transcription in Huh-7 and MEFs. JNK- form functional TFIIIB complexes (7). Bdp1 is phosphorylated mediated changes in cellular TBP concentrations, alone, can mod- during , causing its selective dissociation from chromatin ulate RNA pol III transcription in these cells. Our previous studies (14, 16). Maf1 acts to repress RNA pol III transcription by both demonstrated that JNK-mediated alterations in cellular TBP ex- repressing TBP expression and by inhibiting TFIIIB recruitment pression dictate the proliferation rates of MEFs (19). Furthermore, to promoters (25). In contrast, oncogenic proteins and viruses these changes in TBP concentrations regulate c-Jun levels and its can induce RNA pol III transcription via enhanced expression of activity, which have been shown to drive the proliferation rates of individual transcription components. The hepatitis B virus X MEFs (29). Our finding that the JNKs regulate RNA pol III- protein increases TBP expression, requiring the activation of Ras dependent transcription through its affects on TBP and its associ-

12686 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0904843106 Zhong and Johnson Downloaded by guest on September 25, 2021 ated RNA pol III factors identifies a mechanism by which it controls as described previously: TBP, Bdp1, Brf1, TFIIIC63, and ␤-actin (25); and JNK1, the proliferative capacity of these cells. JNK2, and Elk-1 (19). While JNK targets are being defined, analysis of the roles of the JNKs in cellular growth responses has proven to be more complex. Transcription Assays. MEFs were transfected with the pArg-maxi gene for 48 h Animal studies have begun to examine the functions of these JNKs and then serum-starved in 0.1% FBS-MEM for 3 h. Cells were treated with or without anisomycin (50 ng/mL) at 37 °C for 30 min, and total RNA was isolated. in cancer. Paradoxically, disruption of Jnk1 in mice inhibited RNase protection assays were carried out as described previously (10). Nuclear transformation of pre-B cells by Bcr-Abl (34), while these mice were extracts were prepared as previously described (18). The transcription reaction more susceptible to phorbol ester-induced skin tumors than wild- contained 10 or 20 ␮g nuclear extract protein, 0.8 ␮g DNA template, and 0.1 mM Ϫ Ϫ type mice (35). However, Jnk1 / mice are markedly less suscep- [␣-32P]GTP (6 Ci/mmol) in a reaction volume of 60 ␮L. Transcription products were tible to diethylnitrosamine-induced hepatocarcinogenesis (36). In analyzed by electrophoresis on 8 M urea-8% polyacrylamide gels and visualized contrast, Jnk2-deficient mice displayed a decreased propensity for by autoradiography, and the resultant bands were quantified by densitometry phorbol ester-induced papillomas compared to wild-type mice (37). using a phosphoimager. The 5Ј-RACE system (8374–058; Invitrogen) was used to As differences in functions observed for the JNKs are likely identify the transcription start site of human Bdp1 following the protocol. mediated by their substrates and cellular context, defining JNK1- ␤ and JNK2-specific targets is crucial for understanding how they RT-PCR and PCR Analysis. Quantification of TBP, Bdp1, Brf1, TFIIIC63, -actin mRNAs, as well as pretRNALeu and 7SL RNAs was performed by RT-qPCR according elicit distinct cellular responses. Further studies will be needed to to the protocol of SuperScript One-Step RT-PCR with Platinum tag (Invitrogen). determine the context-dependent functions of the different JNK The primer sequences used are shown in Table S1. JNK1 and JNK2 primer se- isoforms in suppressing or inducing an oncogenic state. quences were described previously (19). Quantitative PCR was performed using SYBR green supermix (Bio-Rad Laboratories) on an MX3000P system (Strategene). Methods Plasmids and Reagents. The human TBP promoter and HA human TBP (HA-hTBP) ChIP Assay. ChIP assays were performed as described previously (20). Cells were expression plasmids were described previously (17). The antisense TBP expression transiently transfected with mismatch, JNK1, JNK2, or Elk-1 siRNA for 48 h before construct contains a 991-bp fragment of the 3Ј-end of the mouse TBP cDNA as cross-linking with formaldehyde. Real-time PCR was performed using the isolated previously described (19). HA-JNK1 and HA-JNK2 expression plasmids were from DNA and SYBR green supermix (Bio-Rad Laboratories) on an MX3000P system Anning Lin (University of Chicago). The synthesized human siRNA oligonucleo- (Strategene). The primer sequences that were used are shown in Table S2. The tide sequences were described previously for JNK1, JNK2, and Elk-1 (19). The sense fold change in promoter occupancy was calculated by setting the level of pro- sequence of human TBP siRNA is 5Ј-UUGAAUAGUGAGACGAGUUTT-3Ј and an- moter occupancy in the cells transfected with mismatch (mm) siRNA at 1. tisense sequence is 5Ј-AACUCGUCUCACUAUUCAATT-3Ј. All experiments using these siRNAs were also performed with other target siRNA sequences to verify the ACKNOWLEDGMENTS. This work was supported by National Institutes of results. The antibodies used for immunoblot analysis and ChIP assays were used Health Grants CA108614 and CA74138 (to D.L.J.).

1. Chang L, Karin M (2001) Mammalian MAP kinase signalling cascades. Nature 410:37–40. 21. Johnson DL, Johnson SAS (2008) Cell biology. RNA metabolism and oncogenesis. 2. Davis RJ (2000) Signal transduction by the JNK group of MAP kinases. Cell 103:239–252. Science 320:461–462. 3. Bode AM, Dong Z (2007) The functional contrariety of JNK. Mol Carcinog 8:591–598. 22. Cabarcas S, Jacob J, Veras I, Schramm L (2008) Differential expression of the TFIIIB 4. Ventura JJ, Hu¨bner A, Zhang C, Flavell RA, Shokat KM, Davis RJ (2006) Chemical genetic subunits Brf1 and Brf2 in cancer cells. BMC Mol Biol 9:74–78. analysis of the time course of signal transduction by JNK. Mol Cell 5:701–710. 23. Mayer C, Bierhoff H, Grummt I (2005) The nucleolus as a stress sensor: JNK2 inactivates the 5. Wang H-D, Trivedi A, Johnson DL (1997) Hepatitis B virus X protein induces RNA transcription factor TIF-IA and down-regulates rRNA synthesis. Genes Dev 19:933–941. polymerase III-dependent gene transcription and increases cellular TATA-binding 24. Marshall L, White RJ (2008) Non-coding RNA production by RNA polymerase III is protein by activating the Ras signaling pathway. Mol Cell Biol 17:6838–6846. implicated in cancer. Nat Rev Cancer 8:911–914. 6. Gomez-Roman N, Grandori C, Eisenman RN, White RJ (2003) Direct activation of RNA 25. Johnson SAS, Zheng C, Fromm J, Willis I, Johnson DL (2007) Mammalian Maf1 is a polymerase III transcription by c-Myc. Nature 421:290–294. negative regulator of transcription by all three nuclear RNA polymerases. Mol Cell 7. Woiwode A, et al. (2008) PTEN represses RNA polymerase III-dependent transcription 26:1–13. by targeting the TFIIIB complex. Mol Cell Biol 12:4204–4214. 26. Wang HD, Trivedi A, Johnson DL (1997) Hepatitis B virus X protein induces RNA 8. Hirsch HA, Jawdekar GW, Lee KA, Gu L, Henry RW (2004) Distinct mechanisms for polymerase III-dependent gene transcription and increases cellular TATA-binding repression of RNA polymerase III transcription by the retinoblastoma tumor suppressor protein. Mol Cell Biol 24:5989–5999. protein by activating the Ras signaling pathway. Mol Cell Biol 17:6838–6846. 27. Felton-Edkins ZA, White RJ (2002) Multiple mechanisms contribute to the activation of 9. Sutcliffe JE, Brown TR, Allison SJ, Scott PH, White RJ (2000) BIOCHEMISTRY disrupts interactions required for RNA polymerase III transcription. Mol Cell Biol RNA polymerase III transcription in cells transformed by papovaviruses. J Biol Chem 20:9192–9202. 277:48182–48191. 10. Crighton D, et al. (2003) p53 represses RNA polymerase III transcription by targeting 28. Daly NL, et al. (2005) Deregulation of RNA polymerase III transcription in cervical TBP and inhibiting promoter occupancy by TFIIIB. EMBO J 22:2810–2820. epithelium in response to high-risk human papillomavirus. Oncogene 24:880–888. 11. White RJ (2005) RNA polymerases I and III, growth control and cancer. Nat Rev Mol Cell 29. Ying TH, Hsieh YH, Hsieh YS, Liu JY (2008) Antisense oligonucleotide Elk-1 suppresses Biol 1:69–78. the tumorigenicity of human hepatocellular carcinoma cells. Cell Biol Int 32:210–216. 12. Johnson SA, Dubeau L, Johnson DL (2008) Enhanced RNA polymerase III-dependent 30. Goodfellow SJ, Innes F, Derblay LE, MacLellan WR, Scott PH, White RJ (2006) Regulation of transcription is required for oncogenic transformation. J Biol Chem 283:19184–19191. RNA polymerase III transcription during hypertrophic growth. EMBO J 25:1522–1533. 13. Marshall L, Kenneth NS, White RJ (2008) Elevated tRNA (iMet) synthesis can drive cell 31. Sabapathy K, Hochedlinger K, Nam SY, Bauer A, Karin M, Wagner EF (2004) Distinct proliferation and oncogenic transformation. Cell 1:78–89. roles for JNK1 and JNK2 in regulating JNK activity and c-Jun-dependent cell prolifer- 14. Fairley JA, Scott PH, White RJ (2003) TFIIIB is phosphorylated, disrupted and selectively ation. Mol Cell 15:713–725. released from tRNA promoters during mitosis in vivo. EMBO J 21:5841–5850. 32. Tafolla E, Wang S, Wong B, Leong J, Kapila YL (2005) JNK2 oppositely regulate p53 in 15. Felton-Edkins ZA, Fairley JA, Graham EL, Johnston IM, White RJ, Scott PH (2003) The signaling linked to apoptosis triggered by an altered fibronectin matrix: JNK links FAK mitogen-activated protein (MAP) kinase ERK induces tRNA synthesis by phosphorylat- and p53. J Biol Chem 280:19992–19999. ing TFIIIB. EMBO J 10:2422–32. 33. Yang SH, Whitmarsh AJ, Davis RJ, Sharrocks AD (1998) Differential targeting of MAP 16. Hu P, Samudre K, Wu S, Sun Y, Hernandez N (2004) CK2 phosphorylation of Bdp1 executes kinases to the ETS-domain transcription factor Elk-1. EMBO J 17:1740–1749. cell cycle-specific RNA polymerase III transcription repression. Mol Cell 16:81–92. 34. Hess P, Pihan G, Sawyers CL, Flavell RA, Davis RJ (2002) Survival signaling mediated by 17. Johnson SAS, Mandavia N, Wang HD, Johnson DL (2000) Transcriptional regulation of c-Jun NH(2)-terminal kinase in transformed B lymphoblasts. Nat Genet 32:201–205. the human TATA-binding protein by Ras cellular signaling. Mol Cell Biol 20:5000–5009. 35. She QB, Chen N, Bode AM, Flavell RA, Dong Z (2002) Deficiency of c-Jun-NH(2)-terminal 18. Zhong S, Zhang C, Johnson DL (2004) Epidermal growth factor enhances cellular TATA binding protein levels and induces RNA polymerase I- and III-dependent gene activity. kinase-1 in mice enhances skin tumor development by 12-O-tetradecanoylphorbol-13- Mol Cell Biol 12:5119–5129. acetate. Cancer Res 62:1343–1348. 19. Zhong S, Fromm J, Johnson DL (2007) TBP is differentially regulated by c-Jun N-terminal 36. Sakurai T, Maeda S, Chang L, Karin M (2006) Loss of hepatic NF-kappa B activity kinase 1 (JNK1) and JNK2 through Elk-1, controlling c-Jun expression and cell prolif- enhances chemical hepatocarcinogenesis through sustained c-Jun N-terminal kinase 1 eration. Mol Cell Biol 1:54–64. activation. Proc Natl Acad Sci USA 103:10544–10551. 20. Fromm J, Johnson SAS, Johnson DL (2008) EGFR1 and its variant, EGFRvIII, regulate 37. Chen N, et al. (2001) Suppression of skin tumorigenesis in c-Jun NH(2)-terminal kinase- TATA-binding protein expression through distinct pathways. Mol Cell Biol 28:4204–4214. 2-deficient mice. Cancer Res 61:3908–3912.

Zhong and Johnson PNAS ͉ August 4, 2009 ͉ vol. 106 ͉ no. 31 ͉ 12687 Downloaded by guest on September 25, 2021