- Email: [email protected]
Cysteine 38 Holds the Key to NF-κB Activation
Molecular Cell
Previews Cysteine 38 Holds the Key to NF-kB Activation Neil D. Perkins1,* 1Institute for Cell and Molecular Biosciences, Newcastle University, Medical School, Catherine Cookson Building, Framlington Place, Newcastle Upon Tyne NE2 4HH, UK *Correspondence: [email protected] DOI 10.1016/j.molcel.2011.12.023
The importance of parallel signaling pathways controlling NF-kB subunit posttranslational modifications is demonstrated by Sen et al. (2012), who reveal that RelA (p65) sulfhydration, at its highly conserved cysteine 38 residue, regulates association with the coactivator RPS3, DNA binding, and antiapoptotic gene expression. The nuclear factor kB (NF-kB) family of transcription factors, comprising homoand heterodimers formed from RelA (p65), c-Rel, RelB, p50/p105 (NF-kB1), and p52/p100 (NF-kB2), are critical regulators of the cellular response to stress and infection (Hayden and Ghosh, 2008). The majority of NF-kB-inducing stimuli, such as inflammatory cytokines, bacterial and viral components, and cell stresses such as DNA damage, result in the activation of the inhibitor of NF-kB (IkB) kinase (IKK) complex, leading to the phosphorylation and ubiquitin-mediated degradation of a member of the IkB family of proteins (Hayden and Ghosh, 2008). This ‘‘classical’’ pathway of NF-kB activation typically results in the rapid nuclear localization of a p50/RelA heterodimer. The cellular consequences of NF-kB activation include resistance to apoptosis induced by tumor necrosis factor (TNF)-a and other stimuli, which is also an important component of the ability of aberrantly active NF-kB to drive the progression of inflammatory diseases and promote the survival of cancer cells (Fan et al., 2008). In addition to release from IkB, ‘‘activation’’ of NF-kB requires a wide variety of subunit posttranslational modifications (PTMs), including phosphorylation, acetylation, ubiquitylation, and methylation, which can control nuclear translocation, target gene specificity, transcriptional activity, and subunit degradation (Huang et al., 2010). In this edition of Molecular Cell, a manuscript from the Snyder laboratory reveals a new and critical RelA modification (referred to as p65 by Sen et al.) required for its ability to bind DNA and induce antiapoptotic target gene expression (Sen et al., 2012).
Almost all NF-kB subunits, with the exceptions being some isoforms from nonvertebrate species, contain a highly conserved cysteine residue in the N-terminal region of the Rel homology domain (RHD) (Cys-38 in human RelA; Figure 1). The importance of this residue has been appreciated for many years: it interacts with the phosphate backbone of NF-kB binding sites (Chen et al., 1998) (Figure 2) while its oxidation or nitrosylation are known to inhibit DNA binding (Kelleher et al., 2007). Moreover, it is the target of many naturally occurring NF-kB inhibitors such as the sesquiterpene lactones (Gilmore and Herscovitch, 2006). Here, Sen et al. demonstrate that RelA Cys-38 is also subject to hydrogen sulfide-linked sulfhydration (Sen et al., 2012). This modification, which creates a hydropersulfide ( SSH) by attaching an additional sulfur to the thiol ( SH) group, is performed by cystathionine
g-lyase (CSE), the expression of which is induced in an NF-kB- and IKK-independent manner following TNF stimulation (Sen et al., 2012). Sen et al. convincingly demonstrate that impairment of this modification inhibits RelA DNA binding and compromises its ability to induce antiapoptotic gene expression (Sen et al., 2012). These effects are at least partly explained by sulfhydration of Cys38 being required for RelA interaction with ribosomal protein S3 (RPS3), which can function as a NF-kB coregulator (Wan et al., 2007). The evolutionary conservation of Cys-38 suggests that this is a highly conserved mechanism, and it will be interesting to see if this pathway also functions across species. However, it is not currently clear if cysteine sulfhydration will regulate the other NF-kB subunits. Sen et al. did examine the p50 subunit and found that CSE did not
Figure 1. Alignment of RelA (p65) and NF-kB Subunit Sequences Showing the Conservation of the Cys-38 Residue Also shown is the conservation of p50 Cys-119 (human Cys-122), mutation of which allows sulfhydration of p50 Cys-59 (human Cys-62). The lack of ‘‘Cys-38’’ conservation in Relish is consistent in multiple insect species (data not shown). Alignments were performed using the ClustalW interface of DNADynamo analysis software.
Molecular Cell 45, January 13, 2012 ª2012 Elsevier Inc. 1
Molecular Cell
Previews
modify the Cys-38 equivalent, CysTNF activation of NF-kB can also 59, unless Cys-119 was mutated to be very rapid, occurring within alanine (Sen et al., 2012) (note: minutes and before induction of human numberings are Cys-62 the CSE gene will have had time to and -122). One possibility, noted affect CSE enzymatic activity. Sen by the authors, is that this might et al. do not look at time points prior result from disulfide bond formation, to 60 min, and so the status of RelA although analysis of the structure modification at these earlier time of p50/RelA bound to DNA suggests points is not known. It is possible this is unlikely to occur in that conthat ‘‘early’’ activated RelA is not formation of the protein (Figure 2). modified at this residue and that Moreover, Cys-119 is also conCSE induction has the effect of proserved in RelA, albeit with differing longing and establishing an NF-kB flanking residues (Figure 1). An alterresponse. Alternatively, pre-existing native possibility is that a CSE CSE may be sufficient to ensure docking motif on RelA or an associthis ‘‘early’’ RelA is appropriately ated protein is required. Some modified. It will also be interesting specificity determinant must exist, to see if Cys-38 sulfhydration is as Cys-38 is the only one of eight linked to other RelA PTMs. For cysteines in RelA to be modified example, Ser-276 phosphorylation by CSE. Analysis of the p50/RelA is a critical activator of RelA (Chen structure also reveals that while et al., 1998; Hayden and Ghosh, Cys-38 is closely associated with 2008) and may provide a conformational trigger to allow CSE the DNA, it is also partially exposed access to the Cys-38 residue. Sen and accessible for protein binding et al. also show that RelA Cys-38 to RPS3 (Figure 2). This suggests nitrosylation replaces sulfhydration that the docking site for RPS3 and that this is associated with will be in the ‘‘cleft’’ at the underside switching off the NF-kB response. of the NF-kB/DNA complex, with Sulfhydrated RelA can be nitrosythe potential to make contact with lated, indicating that active represthe exposed DNA surface, thus sion of RelA activity by this second stabilizing DNA binding (Figure 2). modification may occur (Sen et al., Although Sen et al. demonstrated 2012). However, it cannot be ruled that LPS stimulation also induces out that, in vivo, Cys-38 sulfhydrated RelA Cys-38 sulfhydration, it is not ‘‘active’’ RelA will be degraded, with yet clear how universally applicable de novo synthesized RelA being this mechanism of regulation will be subject to Cys-38 nitrosylation. to other inducers of NF-kB and other This study demonstrates that biological contexts. Indeed, deletion Figure 2. Structure of the RelA/p50 Dimer Bound to the there is still much to be learned or loss of CSE does not abolish IgH/HIV kB Site about NF-kB subunit function and NF-kB DNA binding but instead (A) The positions of RelA Cys-38 (red) and Cys-95 (purple), together with p50 Cys-59 (blue) and Cys-119 (green), are regulation. Moreover, this could reduces it (Sen et al., 2012), while shown. The DNA kB site is shown in gold. lead to new drugs and therapies. CSE knockout mice are viable (B) Space-filling structures of the RelA/p50/DNA complex from Many natural products covalently (Yang et al., 2008) in contrast to an ‘‘underside’’ (top image) and ‘‘side on’’ (bottom image) perspective, showing RelA Cys-38 (red) and p50 Cys-59 modify Cys-38 as a means of RelA null mice, which die in utero (blue). The likely binding site for RPS3 suggested by the work inhibiting or modulating NF-kB as a consequence of TNF-induced of Sen et al. is shown. Images were created from PDB file activity, implying that inhibiting liver apoptosis (Fan et al., 2008). 1VKW (Chen et al., 1998) using the Jmol interface of DNADyCSE or the pathway leading to its Therefore, in contexts where Cysnamo software. inducible expression may provide 38 sulfhydration does not occur, a way to indirectly target NF-kB as NF-kB promoter targeting may be directed toward variant kB sites where translocation (Wan et al., 2011), raising a treatment for inflammatory diseases RPS3 is not required or where the possibility that cytoplasmic RelA and cancer. That the CSE knockout other proteins, such as HMGA1, can fulfill Cys-38 sulfhydration results in the forma- mice are viable suggests that such drugs the role of a DNA binding coregulator. tion of the RelA/RPS3 complex, which will be tolerated (Yang et al., 2008). Given The order, location, and timing of these then moves to the nucleus. Alternatively, the likely problems of targeting IKKb, due regulatory events are not yet established. RPS3 might associate with RelA in the to its widespread effects and the toxicity For example, RPS3 undergoes IKKb- nucleus after induction or even bind and associated with its inhibition, this should dependent phosphorylation and nuclear stabilize a prebound RelA/DNA complex. warrant further investigation. 2 Molecular Cell 45, January 13, 2012 ª2012 Elsevier Inc.
Molecular Cell
Previews REFERENCES
Hayden, M.S., and Ghosh, S. (2008). Cell 132, 344–362.
Chen, F.E., Huang, D.B., Chen, Y.Q., and Ghosh, G. (1998). Nature 391, 410–413.
Huang, B., Yang, X.D., Lamb, A., and Chen, L.F. (2010). Cell. Signal. 22, 1282–1290.
Fan, Y., Dutta, J., Gupta, N., Fan, G., and Ge´linas, C. (2008). Adv. Exp. Med. Biol. 615, 223–250.
Kelleher, Z.T., Matsumoto, A., Stamler, J.S., and Marshall, H.E. (2007). J. Biol. Chem. 282, 30667– 30672.
Gilmore, T.D., and Herscovitch, M. (2006). Oncogene 25, 6887–6899.
Sen, N., Paul, B.D., Gadalla, M.M., Mustafa, A.K., Sen, T., Xu, R., Kim, S., and Snyder, S.H. (2012). Mol. Cell 45, this issue, 13–24.
Wan, F., Anderson, D.E., Barnitz, R.A., Snow, A., Bidere, N., Zheng, L., Hegde, V., Lam, L.T., Staudt, L.M., Levens, D., et al. (2007). Cell 131, 927–939. Wan, F., Weaver, A., Gao, X., Bern, M., Hardwidge, P.R., and Lenardo, M.J. (2011). Nat. Immunol. 12, 335–343. Yang, G., Wu, L., Jiang, B., Yang, W., Qi, J., Cao, K., Meng, Q., Mustafa, A.K., Mu, W., Zhang, S., et al. (2008). Science 322, 587–590.
Get Back TFIIF, Don’t Let Me Gdown1 Joaquı´n M. Espinosa1,2,* 1Howard
Hughes Medical Institute of Molecular, Cellular, and Developmental Biology The University of Colorado, Boulder, CO 80309, USA *Correspondence: [email protected] DOI 10.1016/j.molcel.2011.12.016 2Department
In this issue of Molecular Cell, papers by the Price and Roeder labs reveal how the Gdown1 protein antagonizes the general transcription factor TFIIF during RNAPII initiation and elongation and how the Mediator complex intervenes in this molecular tug-of-war to activate RNAPII. In 1969, The Beatles recorded ‘‘Get Back’’ and ‘‘Don’t Let Me Down,’’ two songs that were later released as a single with ‘‘Get Back’’ on the A side. Paul McCartney and John Lennon composed mostly individually at this late stage, but these songs still represent their intertwined genius. McCartney penned ‘‘Get Back,’’ but Lennon’s three guitar solos pillar the song; Lennon wrote ‘‘Don’t Let Me Down,’’ but McCartney’s backing vocals make it work. Fast forward 42 years to the January 2012 issue of Molecular Cell: the Price and Roeder labs release back-to-back papers providing major basic insights into the inner workings of the eukaryotic transcriptional machinery. Their reports contain important discoveries on the roles of the general transcription factor TFIIF, the protein Gdown1, and the Mediator coactivator complex in control of RNA polymerase II (RNAPII) activity. The two papers stand alone and may become classics on their own merit, but together they are better. They do not always agree, and the two labs reach some distinct conclusions. But, as is also true in art, disparate ap-
proaches and observations can reveal equally valid truths. TFIIF is one of the general transcription factors (GTFs) required for formation of the preinitiation complex (PIC), the molecular assembly that recruits RNAPII to promoters and facilitates transcription initiation. Additionally, TFIIF stimulates RNAPII elongation. Previous work by the Price team showed that the positive effects of TFIIF on RNAPII elongation can be blocked by a factor present in nuclear extracts (Cheng and Price, 2007). In their new report (Cheng et al., 2012), they characterize Gdown1 as the ‘‘TFIIF resistance factor.’’ Using elegant biochemical assays, they show that Gdown1 prevents binding of TFIIF to elongation complexes to cause RNAPII pausing, thus joining other ‘‘pausing factors’’ such as NELF and DSIF. They find that Gdown1 also inhibits the activity of the termination factor TTF2, thereby preventing the release of short transcripts and RNAPII dissociation. Importantly, the negative effects of Gdown1, NELF and DSIF are all relieved by P-TEFb, the positive elongation factor.
Interestingly, Gdown1 was first characterized by Roeder and colleagues as a polypeptide tightly bound to RNAPII that created a requirement for the Mediator coactivator complex during transcriptional activation (Hu et al., 2006). They found that transactivators stimulate RNAPII lacking Gdown1 in the absence of Mediator. However, if Gdown1 is present, RNAPII becomes refractory to the transactivators and Mediator becomes indispensable for activating transcription. In their current paper (Jishage et al., 2012), they expand their studies to show that Gdown1 also represses ‘‘basal’’ transcription (i.e., in the absence of transactivators), pointing to an effect on the general transcriptional machinery. Using a different set of biochemical assays than those employed by the Price team, they demonstrate that an excess of TFIIF relieves the negative effects of Gdown1 on basal transcription. Gdown1 assembles an alternative promoter complex containing TBP, TFIIB, and RNAPII, which is inert relative to the authentic PIC containing TFIIF. Thus, RNAPII exists in mutually exclusive complexes with either
Molecular Cell 45, January 13, 2012 ª2012 Elsevier Inc. 3
Previews Cysteine 38 Holds the Key to NF-kB Activation Neil D. Perkins1,* 1Institute for Cell and Molecular Biosciences, Newcastle University, Medical School, Catherine Cookson Building, Framlington Place, Newcastle Upon Tyne NE2 4HH, UK *Correspondence: [email protected] DOI 10.1016/j.molcel.2011.12.023
The importance of parallel signaling pathways controlling NF-kB subunit posttranslational modifications is demonstrated by Sen et al. (2012), who reveal that RelA (p65) sulfhydration, at its highly conserved cysteine 38 residue, regulates association with the coactivator RPS3, DNA binding, and antiapoptotic gene expression. The nuclear factor kB (NF-kB) family of transcription factors, comprising homoand heterodimers formed from RelA (p65), c-Rel, RelB, p50/p105 (NF-kB1), and p52/p100 (NF-kB2), are critical regulators of the cellular response to stress and infection (Hayden and Ghosh, 2008). The majority of NF-kB-inducing stimuli, such as inflammatory cytokines, bacterial and viral components, and cell stresses such as DNA damage, result in the activation of the inhibitor of NF-kB (IkB) kinase (IKK) complex, leading to the phosphorylation and ubiquitin-mediated degradation of a member of the IkB family of proteins (Hayden and Ghosh, 2008). This ‘‘classical’’ pathway of NF-kB activation typically results in the rapid nuclear localization of a p50/RelA heterodimer. The cellular consequences of NF-kB activation include resistance to apoptosis induced by tumor necrosis factor (TNF)-a and other stimuli, which is also an important component of the ability of aberrantly active NF-kB to drive the progression of inflammatory diseases and promote the survival of cancer cells (Fan et al., 2008). In addition to release from IkB, ‘‘activation’’ of NF-kB requires a wide variety of subunit posttranslational modifications (PTMs), including phosphorylation, acetylation, ubiquitylation, and methylation, which can control nuclear translocation, target gene specificity, transcriptional activity, and subunit degradation (Huang et al., 2010). In this edition of Molecular Cell, a manuscript from the Snyder laboratory reveals a new and critical RelA modification (referred to as p65 by Sen et al.) required for its ability to bind DNA and induce antiapoptotic target gene expression (Sen et al., 2012).
Almost all NF-kB subunits, with the exceptions being some isoforms from nonvertebrate species, contain a highly conserved cysteine residue in the N-terminal region of the Rel homology domain (RHD) (Cys-38 in human RelA; Figure 1). The importance of this residue has been appreciated for many years: it interacts with the phosphate backbone of NF-kB binding sites (Chen et al., 1998) (Figure 2) while its oxidation or nitrosylation are known to inhibit DNA binding (Kelleher et al., 2007). Moreover, it is the target of many naturally occurring NF-kB inhibitors such as the sesquiterpene lactones (Gilmore and Herscovitch, 2006). Here, Sen et al. demonstrate that RelA Cys-38 is also subject to hydrogen sulfide-linked sulfhydration (Sen et al., 2012). This modification, which creates a hydropersulfide ( SSH) by attaching an additional sulfur to the thiol ( SH) group, is performed by cystathionine
g-lyase (CSE), the expression of which is induced in an NF-kB- and IKK-independent manner following TNF stimulation (Sen et al., 2012). Sen et al. convincingly demonstrate that impairment of this modification inhibits RelA DNA binding and compromises its ability to induce antiapoptotic gene expression (Sen et al., 2012). These effects are at least partly explained by sulfhydration of Cys38 being required for RelA interaction with ribosomal protein S3 (RPS3), which can function as a NF-kB coregulator (Wan et al., 2007). The evolutionary conservation of Cys-38 suggests that this is a highly conserved mechanism, and it will be interesting to see if this pathway also functions across species. However, it is not currently clear if cysteine sulfhydration will regulate the other NF-kB subunits. Sen et al. did examine the p50 subunit and found that CSE did not
Figure 1. Alignment of RelA (p65) and NF-kB Subunit Sequences Showing the Conservation of the Cys-38 Residue Also shown is the conservation of p50 Cys-119 (human Cys-122), mutation of which allows sulfhydration of p50 Cys-59 (human Cys-62). The lack of ‘‘Cys-38’’ conservation in Relish is consistent in multiple insect species (data not shown). Alignments were performed using the ClustalW interface of DNADynamo analysis software.
Molecular Cell 45, January 13, 2012 ª2012 Elsevier Inc. 1
Molecular Cell
Previews
modify the Cys-38 equivalent, CysTNF activation of NF-kB can also 59, unless Cys-119 was mutated to be very rapid, occurring within alanine (Sen et al., 2012) (note: minutes and before induction of human numberings are Cys-62 the CSE gene will have had time to and -122). One possibility, noted affect CSE enzymatic activity. Sen by the authors, is that this might et al. do not look at time points prior result from disulfide bond formation, to 60 min, and so the status of RelA although analysis of the structure modification at these earlier time of p50/RelA bound to DNA suggests points is not known. It is possible this is unlikely to occur in that conthat ‘‘early’’ activated RelA is not formation of the protein (Figure 2). modified at this residue and that Moreover, Cys-119 is also conCSE induction has the effect of proserved in RelA, albeit with differing longing and establishing an NF-kB flanking residues (Figure 1). An alterresponse. Alternatively, pre-existing native possibility is that a CSE CSE may be sufficient to ensure docking motif on RelA or an associthis ‘‘early’’ RelA is appropriately ated protein is required. Some modified. It will also be interesting specificity determinant must exist, to see if Cys-38 sulfhydration is as Cys-38 is the only one of eight linked to other RelA PTMs. For cysteines in RelA to be modified example, Ser-276 phosphorylation by CSE. Analysis of the p50/RelA is a critical activator of RelA (Chen structure also reveals that while et al., 1998; Hayden and Ghosh, Cys-38 is closely associated with 2008) and may provide a conformational trigger to allow CSE the DNA, it is also partially exposed access to the Cys-38 residue. Sen and accessible for protein binding et al. also show that RelA Cys-38 to RPS3 (Figure 2). This suggests nitrosylation replaces sulfhydration that the docking site for RPS3 and that this is associated with will be in the ‘‘cleft’’ at the underside switching off the NF-kB response. of the NF-kB/DNA complex, with Sulfhydrated RelA can be nitrosythe potential to make contact with lated, indicating that active represthe exposed DNA surface, thus sion of RelA activity by this second stabilizing DNA binding (Figure 2). modification may occur (Sen et al., Although Sen et al. demonstrated 2012). However, it cannot be ruled that LPS stimulation also induces out that, in vivo, Cys-38 sulfhydrated RelA Cys-38 sulfhydration, it is not ‘‘active’’ RelA will be degraded, with yet clear how universally applicable de novo synthesized RelA being this mechanism of regulation will be subject to Cys-38 nitrosylation. to other inducers of NF-kB and other This study demonstrates that biological contexts. Indeed, deletion Figure 2. Structure of the RelA/p50 Dimer Bound to the there is still much to be learned or loss of CSE does not abolish IgH/HIV kB Site about NF-kB subunit function and NF-kB DNA binding but instead (A) The positions of RelA Cys-38 (red) and Cys-95 (purple), together with p50 Cys-59 (blue) and Cys-119 (green), are regulation. Moreover, this could reduces it (Sen et al., 2012), while shown. The DNA kB site is shown in gold. lead to new drugs and therapies. CSE knockout mice are viable (B) Space-filling structures of the RelA/p50/DNA complex from Many natural products covalently (Yang et al., 2008) in contrast to an ‘‘underside’’ (top image) and ‘‘side on’’ (bottom image) perspective, showing RelA Cys-38 (red) and p50 Cys-59 modify Cys-38 as a means of RelA null mice, which die in utero (blue). The likely binding site for RPS3 suggested by the work inhibiting or modulating NF-kB as a consequence of TNF-induced of Sen et al. is shown. Images were created from PDB file activity, implying that inhibiting liver apoptosis (Fan et al., 2008). 1VKW (Chen et al., 1998) using the Jmol interface of DNADyCSE or the pathway leading to its Therefore, in contexts where Cysnamo software. inducible expression may provide 38 sulfhydration does not occur, a way to indirectly target NF-kB as NF-kB promoter targeting may be directed toward variant kB sites where translocation (Wan et al., 2011), raising a treatment for inflammatory diseases RPS3 is not required or where the possibility that cytoplasmic RelA and cancer. That the CSE knockout other proteins, such as HMGA1, can fulfill Cys-38 sulfhydration results in the forma- mice are viable suggests that such drugs the role of a DNA binding coregulator. tion of the RelA/RPS3 complex, which will be tolerated (Yang et al., 2008). Given The order, location, and timing of these then moves to the nucleus. Alternatively, the likely problems of targeting IKKb, due regulatory events are not yet established. RPS3 might associate with RelA in the to its widespread effects and the toxicity For example, RPS3 undergoes IKKb- nucleus after induction or even bind and associated with its inhibition, this should dependent phosphorylation and nuclear stabilize a prebound RelA/DNA complex. warrant further investigation. 2 Molecular Cell 45, January 13, 2012 ª2012 Elsevier Inc.
Molecular Cell
Previews REFERENCES
Hayden, M.S., and Ghosh, S. (2008). Cell 132, 344–362.
Chen, F.E., Huang, D.B., Chen, Y.Q., and Ghosh, G. (1998). Nature 391, 410–413.
Huang, B., Yang, X.D., Lamb, A., and Chen, L.F. (2010). Cell. Signal. 22, 1282–1290.
Fan, Y., Dutta, J., Gupta, N., Fan, G., and Ge´linas, C. (2008). Adv. Exp. Med. Biol. 615, 223–250.
Kelleher, Z.T., Matsumoto, A., Stamler, J.S., and Marshall, H.E. (2007). J. Biol. Chem. 282, 30667– 30672.
Gilmore, T.D., and Herscovitch, M. (2006). Oncogene 25, 6887–6899.
Sen, N., Paul, B.D., Gadalla, M.M., Mustafa, A.K., Sen, T., Xu, R., Kim, S., and Snyder, S.H. (2012). Mol. Cell 45, this issue, 13–24.
Wan, F., Anderson, D.E., Barnitz, R.A., Snow, A., Bidere, N., Zheng, L., Hegde, V., Lam, L.T., Staudt, L.M., Levens, D., et al. (2007). Cell 131, 927–939. Wan, F., Weaver, A., Gao, X., Bern, M., Hardwidge, P.R., and Lenardo, M.J. (2011). Nat. Immunol. 12, 335–343. Yang, G., Wu, L., Jiang, B., Yang, W., Qi, J., Cao, K., Meng, Q., Mustafa, A.K., Mu, W., Zhang, S., et al. (2008). Science 322, 587–590.
Get Back TFIIF, Don’t Let Me Gdown1 Joaquı´n M. Espinosa1,2,* 1Howard
Hughes Medical Institute of Molecular, Cellular, and Developmental Biology The University of Colorado, Boulder, CO 80309, USA *Correspondence: [email protected] DOI 10.1016/j.molcel.2011.12.016 2Department
In this issue of Molecular Cell, papers by the Price and Roeder labs reveal how the Gdown1 protein antagonizes the general transcription factor TFIIF during RNAPII initiation and elongation and how the Mediator complex intervenes in this molecular tug-of-war to activate RNAPII. In 1969, The Beatles recorded ‘‘Get Back’’ and ‘‘Don’t Let Me Down,’’ two songs that were later released as a single with ‘‘Get Back’’ on the A side. Paul McCartney and John Lennon composed mostly individually at this late stage, but these songs still represent their intertwined genius. McCartney penned ‘‘Get Back,’’ but Lennon’s three guitar solos pillar the song; Lennon wrote ‘‘Don’t Let Me Down,’’ but McCartney’s backing vocals make it work. Fast forward 42 years to the January 2012 issue of Molecular Cell: the Price and Roeder labs release back-to-back papers providing major basic insights into the inner workings of the eukaryotic transcriptional machinery. Their reports contain important discoveries on the roles of the general transcription factor TFIIF, the protein Gdown1, and the Mediator coactivator complex in control of RNA polymerase II (RNAPII) activity. The two papers stand alone and may become classics on their own merit, but together they are better. They do not always agree, and the two labs reach some distinct conclusions. But, as is also true in art, disparate ap-
proaches and observations can reveal equally valid truths. TFIIF is one of the general transcription factors (GTFs) required for formation of the preinitiation complex (PIC), the molecular assembly that recruits RNAPII to promoters and facilitates transcription initiation. Additionally, TFIIF stimulates RNAPII elongation. Previous work by the Price team showed that the positive effects of TFIIF on RNAPII elongation can be blocked by a factor present in nuclear extracts (Cheng and Price, 2007). In their new report (Cheng et al., 2012), they characterize Gdown1 as the ‘‘TFIIF resistance factor.’’ Using elegant biochemical assays, they show that Gdown1 prevents binding of TFIIF to elongation complexes to cause RNAPII pausing, thus joining other ‘‘pausing factors’’ such as NELF and DSIF. They find that Gdown1 also inhibits the activity of the termination factor TTF2, thereby preventing the release of short transcripts and RNAPII dissociation. Importantly, the negative effects of Gdown1, NELF and DSIF are all relieved by P-TEFb, the positive elongation factor.
Interestingly, Gdown1 was first characterized by Roeder and colleagues as a polypeptide tightly bound to RNAPII that created a requirement for the Mediator coactivator complex during transcriptional activation (Hu et al., 2006). They found that transactivators stimulate RNAPII lacking Gdown1 in the absence of Mediator. However, if Gdown1 is present, RNAPII becomes refractory to the transactivators and Mediator becomes indispensable for activating transcription. In their current paper (Jishage et al., 2012), they expand their studies to show that Gdown1 also represses ‘‘basal’’ transcription (i.e., in the absence of transactivators), pointing to an effect on the general transcriptional machinery. Using a different set of biochemical assays than those employed by the Price team, they demonstrate that an excess of TFIIF relieves the negative effects of Gdown1 on basal transcription. Gdown1 assembles an alternative promoter complex containing TBP, TFIIB, and RNAPII, which is inert relative to the authentic PIC containing TFIIF. Thus, RNAPII exists in mutually exclusive complexes with either
Molecular Cell 45, January 13, 2012 ª2012 Elsevier Inc. 3