- Email: [email protected]
Activator-Specific Requirement for the General Transcription Factor IIE in Yeast
Biochemical and Biophysical Research Communications 261, 734 –739 (1999) Article ID bbrc.1999.1113, available online at http://www.idealibrary.com on
Activator-Specific Requirement for the General Transcription Factor IIE in Yeast Hiroshi Sakurai* ,1 and Toshio Fukasawa† *School of Health Sciences, Faculty of Medicine, Kanazawa University, 5-11-80 Kodatsuno, Kanazawa, Ishikawa 920-0942, Japan; and †Tamura Laboratory, Graduate School of Science, Chiba University, 1-33 Yayoi-cho, Inage-ku, Chiba 263-8522, Japan
Received July 5, 1999
The general transcription factor (TF) IIE is required for mRNA synthesis of many, but not all, genes in yeast. In the transcription process, TFIIE regulates TFIIH kinase activity that phosphorylates the carboxy-terminal domain (CTD) of the largest subunit of RNA polymerase II. The CTD and the CTD kinase Kin28, a subunit of TFIIH, have been shown to be dispensable for activation of several heat shock genes and the copper metallothionein gene CUP1. Here we analyzed requirement of TFIIE for transcription of these genes and found that TFIIE is necessary for activation of the heat shock genes by heat shock transcription factor Hsf1. By contrast, transcription of CUP1 mediated by both Hsf1 and copper-activated transcription factor Ace1 was inducible after inactivating TFIIE. These results show that both TFIIE and the CTD/the CTD kinase exhibit “gene specificities” which are overlapping, but not identical to each other, and thereby suggest that TFIIE functions with or without involvement of the CTD/the CTD kinase depending on the gene to be transcribed. © 1999 Academic Press
The largest subunit of RNA polymerase II (RNAPII) has a unique carboxy-terminal domain (CTD) composed of tandem repeats of a heptapeptide sequence. Phosphorylation and dephosphorylation of the CTD associate with the transcription cycle: RNAPII with unphosphorylated CTD is recruited to the core promoter and assembles to form the preinitiation complex together with a set of the general transcription factors (TFs). The CTD is then phosphorylated by TFIIH and/or other kinases, which in turn triggers transition from initiation to elongation in the process of mRNA synthesis. Phosphorylated form of RNAPII is dephosphorylated by the CTD phosphatase upon completion 1
To whom correspondence should be addressed. Fax: 81-76-2344360. E-mail: [email protected]. 0006-291X/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
of the elongation and reused for initiation of the next cycle (1). In vivo, an important fraction of RNAPII is present in a large complex called “holoenzyme” of RNAPII together with several general transcription factors and “Mediator,” a protein complex of more than 20 components which transmits regulatory signals of sequence-specific transcription factors to RNAPII (2, 3). The phosphorylation cycle of the CTD regulates interaction of core RNAPII with the Mediator in the holoenzyme. Thus, unphosphorylated CTD binds the Mediator, and phosphorylation of the CTD causes dissociation of the RNAPII-Mediator complex (4). The phosphorylated CTD binds capping enzymes resulting in addition of 7-methyl-guanosine to the 59 end of RNAPII-transcribed RNA (5, 6). The general transcription factor TFIIE of the yeast Saccharomyces cerevisiae is a heterodimer of 66- and 43-kDa subunits which are encoded by the TFA1 and TFA2 genes, respectively (7). We have previously shown that TFIIE binds Gal11 protein, a component of the Mediator, and that TFIIE-Gal11 interaction enhances TFIIH-catalyzed phosphorylation of the CTD in vitro when holo-RNAPII but not core RNAPII was used as substrate (8). On the other hand, physical as well as functional interactions between TFIIE and TFIIH have been disclosed in yeast and mammalian cells (see for review Refs. 9 and 10). The above observations altogether support the idea that TFIIE and TFIIH function at a common step in transcription, that is, transition from initiation to elongation. Analysis of temperature-sensitive (ts) mutant of TFA1 (tfa1-21) has shown that TFIIE is required for transcription of many genes in yeast (11). However, constitutive transcription of several genes was not significantly affected in the mutant cells after shifting to the restrictive temperature. Those TFIIE-independent genes lack the canonical TATA sequence, suggesting that the gene-specific requirement for TFIIE is related to the core promoter structure (11). Recent reports (12, 13) have uncovered that the CTD, the Kin28 CTD
734
Vol. 261, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Transcription of heat-inducible genes in tfa1-21 and kin28-ts3 cells. Yeast cells were grown at 28°C then the temperature was shifted to 39°C. At the indicated times, aliquots of cells were removed and total RNA was prepared. Cells were grown in YPD medium for analyses of SSA4 and HSP104 mRNAs or in ESD medium lacing uracil for transcripts of SSA4-GAL7-lacZ’ and HSE-GAL7-lacZ’. The mRNA of SSA4 was analyzed by S1 nuclease mapping, whereas transcripts of HSP104, SSA4-GAL7-lacZ’, and HSE-GAL7-lacZ’ were analyzed by primer extension. Yeast strains used are HS33 (wild-type), HS46 (tfa1-21), JGV-4 (kin28-ts3), and Y262 (rpb1-1).
kinase, and the Mediator are dispensable for activation of a subset of genes, such as the SSA4 or HSP82 heat shock genes and the CUP1 metallothionein gene. To know functional relationship between TFIIE and the CTD/the CTD kinase, we have analyzed requirement of TFIIE for the transcription that does not require the CTD/the CTD kinase. We found that transcription of CUP1 but not the heat shock genes was activated in the absence of TFIIE function, and that the TFIIEindependent activation of CUP1 was conferred by its upstream activation sequence (UAS). These results suggest that the dependence of transcription of a gene on TFIIE is determined not only by the structure of its core promoter (11), but also by UAS-activator interaction, and further that TFIIE functions in either CTDrelated or CTD-unrelated fashion depending on the gene to be transcribed. MATERIALS AND METHODS Yeast strains and media. TFA1 wild-type (HS33) and tfa1-21 ts mutant (HS46) yeasts were described previously (11). KIN28 wildtype (GF262-2) and kin28-ts3 (JGV-4) strains were gifts from G. Faye (14). Cells bearing rpb1-1 ts mutation (Y262) were gifts from R. Young (15). Strains HS73 and HS74 were ace1 : : LEU2 derivatives of HS33 and HS46, respectively. Rich medium containing 2% glucose (YPD) and enriched synthetic medium containing 2% glucose (ESD) or 2% sodium lactate plus 3% glycerol (ESGlyLac) were prepared as described (11, 16). Reporter genes. Plasmid pI145Z was an integrating-type plasmid bearing GAL7-lacZ’ and URA3 (16). To construct reporter genes, the GAL7 promoter region of pI145Z was replaced by the promoter fragments as follows. The SSA4-GAL7-lacZ’ hybrid gene contained the SSA4 UAS (from 2251 to 2106), which was amplified by polymerase chain reaction (PCR) from genomic DNA of W303-1a, and the GAL7 core promoter (from 274 to 143). To construct HSE-GAL7lacZ’, a Hsf1-binding sequence TCGACTTCTAGAAGCTTCAAGAGC (17) was inserted upstream of the GAL7 core promoter. Reporter gene CUP1-lacZ’ contained the region from 2250 to 175 of CUP1, which was amplified by PCR from the genomic DNA. The CUP1GAL7-lacZ’ reporter was a hybrid gene of the CUP1 UAS (from 2250 to 298) and the GAL7 core promoter. The GAL7-CUP1-lacZ’ reporter contained the GAL7 UAS (from 2271 to 269) and the CUP1 core promoter (from 293 to 175). Reporter gene CUP1(HSE-M)lacZ’ was
a derivative of CUP1-lacZ’ bearing two mutations, C at 2162 to T and G at 2159 to A, which abolish binding of Hsf1 to the CUP1 UAS (18). The reporter genes were digested with SmaI and integrated into the ura3 locus of yeast. RNA analysis. Cells were grown under the conditions described in the figure legends. Preparation of total RNA was described previously (11, 16). The mRNA of SSA4 was analyzed by S1 nuclease mapping with the probe fragment containing the region from 2251 to 1157 of SSA4 with 59 end-labeling at 1157. The transcripts of other genes were analyzed by primer extension with primers HSP104, 59-GTTGATGATCCGAAGCCAAT-39; lacI, 59-CAGTGAGACGGGCAACAGCC-39; CUP1, 59-TACCACATTGGCATTGGCAC-39. The relative amount of mRNA was measured by a BAS-1000 imaging analyzer (Fuji Photo Film, Co.).
RESULTS AND DISCUSSION Effect of tfa1-21 and kin28-ts3 mutations on heatinducible transcription by Hsf1. By using a ts mutant named tfa1-21 which carries two missense mutations in the large subunit of TFIIE (11), we studied requirement of TFIIE for transcription of a set of genes, whose activation is known to be independent of the CTD and Kin28 (12, 13). The temperature shift of tfa1-21 cells to the restrictive temperature (37°C) results in degradation of both subunits of TFIIE and in the stop of mRNA synthesis from most of the genes, suggesting that this mutation significantly, if not completely, inactivates TFIIE function (11). Previously, Tijerina and Sayre (19) argued that TFIIE was necessary for maximal rate transcription of several heat shock genes, such as HSP26, HSP104, and SSA4, by using a ts mutant named tfa1-1 which carries a missense mutation in the conserved zinc-finger motif of the large subunit. However, the effect of that mutation was relatively weak; transcription of all the genes tested by themselves did not completely stop after shifting to the restrictive temperature, and activation of the heat shock genes was observed in tfa1-1 cells to 15–20% of the wild-type level at 15 min after the temperature shift (19). To know whether TFIIE is indeed dispensable for heat shock response, we reexamined this issue by using our ts mutant that exhibits much
735
Vol. 261, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
stronger effect than theirs. Figure 1 shows effect of tfa1-21 mutation on activation of SSA4, HSP104, or their hybrid genes with GAL7-lacZ’ in response to heat. Induction of SSA4 encoding heat shock protein 70 was brought about by up-shifting the temperature of yeast cultures to 39°C in the wild-type but not in tfa1-21 cells: the amount of SSA4 mRNA in tfa1-21 cells was less than 5% of the wild-type control (after 15 min at 39°C), which was comparable to those in rpb1-1 cells bearing a mutation in the largest subunit of RNAPII. The HSP104 gene encoding heat shock protein 104 was also induced at 39°C in the wild-type, but not in tfa1-21 nor in rpb1-1 cells. Consistent with the previous reports (12, 13), heat-induction of SSA4 and HSP104 occurred in kin28-ts3 cells, but the mRNA levels were lower (25–30%) than the wild-type control. Heat-inducible transcription of both SSA4 and HSP104 is mediated by heat shock factor Hsf1, which binds the heat shock element (HSE) of the respective gene at the UAS (20 –22). To know whether Kin28independent transcription of SSA4 could be conferred by the UAS, we constructed a hybrid reporter gene containing the SSA4 UAS and the GAL7 core promoter (SSA4-GAL7-lacZ’) and integrated it into the yeast chromosome at the ura3 locus. Heat-induction of the hybrid gene was observed in kin28-ts3 (30% of the wild-type cells), but not in tfa1-21 nor in rpb1-1 yeast. We further found that Kin28 but not TFIIE was dispensable for heat shock response of a reporter gene containing a synthetic HSE upstream of the GAL7 core promoter (HSE-GAL7-lacZ’). Note that the heat shock response of HSE-GAL7-lacZ’ was transient in the wildtype cells, but was sustained in kin28-ts3 cells. Since the functions of TFIIE or Kin28 protein are severely impaired in tfa1-21 or kin28-ts3 cells, respectively, at the high temperature (11, 14), the above results strongly suggest that TFIIE is necessary for activation of heat shock genes by Hsf1, for which Kin28 is dispensable, however. Copper-inducible transcription of CUP1 is independent of TFIIE and Kin28. Next we tested requirement of TFIIE for activation of the CUP1 gene in response to copper ion. When cells were grown at 28°C, addition of copper sulfate (0.2 mM) caused increase of CUP1 mRNA more than 15-fold over the uninduced level in the wild-type as well as in ts mutant yeasts, tfa1-21, kin28-ts3 or rpb1-1 (Fig. 2A). When, however, cells were grown at 37°C for 1 h before the addition of copper sulfate, induction of CUP1 was repressed to variable extents with yeast strains. This repression was significantly restored by increasing the concentration of copper sulfate to 1.0 mM in the wild-type, tfa121, and kin28-ts3. Thus the amounts of CUP1 mRNA in the wild-type, tfa1-21, and kin28-ts3 reached the respective levels about 90%, 55%, and 65% of the control (the value at 28°C at 0.2 mM of copper sulfate). In
FIG. 2. Copper-inducible transcription of CUP1 in tfa1-21 and kin28-ts3 cells. (A) Cells were grown in YPD medium at 28°C or 37°C for 1 h. Then, copper sulfate was added and cells were cultured for a further 30 min. Total RNA was prepared and CUP1 mRNA was analyzed by primer extension. Relative levels of CUP1 mRNA were determined by Image-analyzer and normalized by the levels in cells grown in the presence of 0.2 mM copper sulfate at 28°C. (B) Cells were grown in YPD medium at 28°C, then transcription of CUP1 was induced by the addition of 1.0 mM copper sulfate for 30 min. After shifting to 37°C, total RNA was prepared at the indicated times and analyzed as above. Relative levels of CUP1 mRNA were determined by Image-analyzer and normalized to the 0 time value as 100%.
rpb1-1 yeast, as expected, essentially no copperinduction of CUP1 was observed. The increase of CUP1 mRNA level at the restrictive temperature in tfa1-21 or in kin28-ts3 therefore indicates that de novo synthesis of CUP1 mRNA requires neither TFIIE nor Kin28. That inactivation of Kin28 does not significantly affect the induced transcription of CUP1 has been reported previously (12). A kinetic profile of induced level of CUP1 mRNA in the wild-type, tfa1-21, or rpb1-1 cells is shown in Fig. 2B. The respective yeast was grown at 28°C in the presence of 1.0 mM copper sulfate for 30 min to fully induce CUP1, and then shifted to 37°C. Total RNA was isolated at the indicated times after the temperature shift, and the amount of CUP1 mRNA was determined in each sample. In rpb1-1 cells, the amount of CUP1 mRNA decreased continuously with a half life of approximately 20 min. The amount of CUP1 mRNA decreased in tfa1-21 cells to a similar extent with rpb1-1
736
Vol. 261, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 3. Transcription of CUP1/GAL7 hybrid genes in tfa1-21 and kin28-ts3 cells. Transformants harboring the reporter gene CUP1-lacZ’ or CUP1-GAL7-lacZ’ were grown in ESD medium lacking uracil at 28°C or 37°C for 1 h then grown in the presence of 1.0 mM copper sulfate for a further 30 min. For analysis of transcripts of GAL7-lacZ’ or GAL7-CUP1-lacZ’, cells were grown in ESGlyLac medium lacking uracil and 2% galactose was added for induction as above. Total RNA prepared was subjected to primer extension with lacI primer.
for the initial 30 min, slightly increased at 1 h, and then decreased very slowly with a slope similar to that seen in the wild-type cells. It appears that abrupt inactivation of TFIIE caused immediate cessation of CUP1 transcription followed by its re-initiation. We imagine that the induced transcription of CUP1 normally involves TFIIE function but can take place in its absence as well. Upstream activation sequence is a determinant of TFIIE- and Kin28-independent transcription of CUP1. To identify sequence elements that confer TFIIE- and Kin28-independent transcription of CUP1, we constructed CUP1-lacZ’, GAL7-lacZ’, and their hybrids as shown in Fig. 3. Transcription of CUP1-lacZ’ was successfully induced by copper sulfate after inactivation of either TFIIE or Kin28, indicating that the region from 2250 to 175 of CUP1 is sufficient for both TFIIE- and Kin28-independent activation. However, galactoseinduced transcription of GAL7-lacZ’ required both factors. The hybrid gene containing the CUP1 UAS and the GAL7 core promoter (CUP1-GAL7-lacZ’) was induced by copper sulfate in either ts mutant at 37°C. By contrast, GAL7-CUP1-lacZ’ containing the GAL7 UAS and the CUP1 core promoter was induced by galactose in neither mutant. These results suggest that activated transcription mediated by the CUP1 UAS, at least in part, is independent of both TFIIE and Kin28. Requirements of Ace1 and Hsf1 for TFIIE-independent transcription. The CUP1 UAS contains two activating sequences; the binding site of Ace1 which mediates copper-inducible transcription (23) and the binding site of Hsf1, HSE which is involved in constitutive as well as heat-inducible transcription (17, 18, 24, 25). It has been reported that both the CTD and Kin28 are dispensable not only for copper-induction
but also for heat-induction of CUP1 (13). We then examined requirement of TFIIE for heat shock response of CUP1 (Fig. 4A). In the wild-type cells, more than 10-fold increase of the CUP1 mRNA level was observed by shifting the growth temperature to 39°C. By contrast, the temperature shift to 39°C (or 37°C; data not shown) caused rather decrease of CUP1 mRNA level in tfa1-21 cells, indicating that TFIIE is required for heatinduction by Hsf1 in the context of the upstream sequence of CUP1 as well. Although Ace1 is the major activator for copperinducible transcription, a stress caused by a high concentration of copper ion also activates CUP1 transcription by the mediation of Hsf1 in ace1 mutant when Hsf1 is over-expressed (26). It was suspected therefore that the copper-induction of CUP1 observed after inactivating TFIIE could be mediated by Hsf1 instead of Ace1. To test this possibility, we carried out series of experiments by using cells lacking Ace1 protein. As shown in Fig. 4B, addition of 1.0 mM copper sulfate to TFA1 ace1 null (wild type in the figure) or tfa1-21 ace1 null yeast (tfa1-21 in the figure) grown at 28°C showed no effect on the CUP1 mRNA level. When TFA1 ace1 cells were grown at 37°C, CUP1 mRNA synthesis was induced presumably by the mediation of Hsf1/HSE. By contrast, the constitutive synthesis of CUP1 mRNA was arrested in tfa1-21 ace1 cells at that temperature. Under these conditions, copper sulfate unaffected the CUP1 transcription in both types of yeast. From these results we conclude that copper-activated Ace1 is responsible for the TFIIE-independent activation of CUP1 in response to copper sulfate, as is seen in Fig. 2A. To know whether constitutive activity of Hsf1 is involved in the activation pathway of CUP1, we em-
737
Vol. 261, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
tion by Hsf1. It would be presumable that Hsf1 recruits the preinitiation complex to the core promoter without involvement of the CTD and TAFs, and yet, that the initiation step still requires TFIIE as an integral component of the basal initiation machinery. On the other hand, we have also shown that activation of CUP1 mediated by the combination of Ace1 and Hsf1 is not significantly affected by inactivation of TFIIE. We suggest that the TFIIE-independent activation of CUP1 may be partially conferred by CTD function, since TFIIE is involved in regulation of CTD phosphorylation (8). The CTD-independent pathway of CUP1 activation mediated by Hsf1 may be further modified by Ace1 so that the requirement for TFIIE is bypassed. Elucidation of the molecular mechanism underlying the activation-specific requirement for TFIIE should furnish a clue to get further insight to the central problem. How is the activation signal conveyed to the basal transcription machinery? ACKNOWLEDGMENTS FIG. 4. Activation of CUP1 in tfa1-21 cells. (A) Heat-inducible transcription of CUP1 in tfa1-21 cells was analyzed as in Fig. 1 by using CUP1 primer. (B) Copper-inducible transcription of CUP1 was analyzed as in Fig. 2A, except that ace1 null strains HS73 (ace1 : : LEU2) and HS74 (ace1 : : LEU2 tfa1-21) were used. (C) Copperinducible transcription of the reporter gene CUP1(HSE-M)lacZ’ in HS33 (wild-type) and HS46 (tfa1-21) was analyzed as in Fig. 3.
ployed a reporter gene CUP1 (HSE-M)lacZ’ that contains a mutation in the HSE of the CUP1 UAS (18). Transcription of this reporter was induced by copper sulfate (Fig. 4C) but not by heat (see Ref. 18, data not shown). The observed activation mediated by Ace1 was significantly inhibited by tfa1-21 mutation at 37°C (Fig. 4C). We also analyzed transcription of a reporter gene having a copy of Ace1-binding sequence upstream of the GAL7 core promoter and found that copperinduction of this reporter was abolished by inactivation of TFIIE (data not shown). These results indicate that activation of CUP1 mediated by either Ace1 or Hsf1 is dependent on TFIIE, but that the presence of both activators on the CUP1 UAS circumvents the requirement for TFIIE. Our previous study suggested an importance of the core promoter structure for the dependence on TFIIE (11). Here we demonstrated that the DNA sequencespecific activator is also a determinant of the requirement for TFIIE. It has been reported that activation of heat shock genes and CUP1 is independent of various components of the RNAPII transcription machinery including the CTD, Kin28, CTD-associated Mediator components such as Srb4 and Srb6, or TATA-binding protein associated factors (TAFs) (12, 13, and for review, see Ref. 27). In the present study, we have shown that TFIIE is necessary for heat-inducible transcrip-
We thank Drs. G. Faye and R. A. Young for providing yeast strains. This work was supported in part by Grants-in-Aid for Scientific Research from the Ministry of Education, Sciences, Sports, and Culture.
REFERENCES 1. Dahmus, M. E. (1996) J. Biol. Chem. 271, 19009 –19012. 2. Kim, Y.-J., Bjorklund, S., Li, Y., Sayre, M. H., and Kornberg, R. D. (1994) Cell 77, 599 – 608. 3. Bjorklund, S., and Kim, Y.-J. (1996) Trends Biochem. Sci. 21, 327–335. 4. Svejstrup, J. Q., Li, Y., Fellows, J., Gnatt, A., Bjorklund, S., and Kornberg, R. D. (1997) Proc. Natl. Acad. Sci. USA 94, 6075– 6078. 5. McCracken, S., Fong, N., Rosonina, E., Yankulov, K., Brothers, G., Siderovski, D., Hessel, A., Foster, S., Amgen EST Program, Shuman, S., and Bentley, D. (1997) Genes Dev. 11, 3306 –3318. 6. Cho, E.-J., Takagi, T., Moore, C., and Buratowski, S. (1997) Genes Dev. 11, 3319 –3326. 7. Feaver, W. J., Henry, N. L., Bushnell, D. A., Sayre, M. H., Brickner, J. H., Gileadi, O., and Kornberg, R. D. (1994) J. Biol. Chem. 269, 27549 –27553. 8. Sakurai, H., and Fukasawa, T. (1998) J. Biol. Chem. 273, 32663– 32669. 9. Svejstrup, J. Q., Vichi, P., and Egly, J.-M. (1996) Trends Biochem. Sci. 21, 346 –350. 10. Zawel, L., and Reinberg, D. (1995) Annu. Rev. Biochem. 64, 533–561. 11. Sakurai, H., Ohishi, T., and Fukasawa, T. (1997) J. Biol. Chem. 272, 15936 –15942. 12. Lee, D., and Lis, J. T. (1998) Nature 393, 389 –392. 13. McNeil, J. P., Agah, H., and Bentley, D. (1998) Genes Dev. 12, 2510 –2521. 14. Valay, J. G., Simon, M., Dubois, M. F., Bensaude, O., Facca, C., and Faye, G. (1995) J. Mol. Biol. 249, 535–544. 15. Nonet, M., Scafe, C., Sexton, J., and Young, R. A. (1987) Mol. Cell. Biol. 7, 1602–1611. 16. Sakurai, H., Ohishi, T., and Fukasawa, T. (1996) FEBS Lett. 398, 113–119.
738
Vol. 261, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
17. Santoro, N., Johansson, N., and Thiele, D. J. (1998) Mol. Cell. Biol. 18, 6340 – 6352. 18. Tamai, K. T., Liu, X., Silar, P., Sosinowski, T., and Thiele, D. J. (1994) Mol. Cell. Biol. 14, 8155– 8165. 19. Tijerina, P., and Sayre, M. H. (1998) J. Biol. Chem. 273, 1107– 1113. 20. Boorstein, W. R., and Craig, E. A. (1990) J. Biol. Chem. 265, 18912–18921. 21. Halladay, J. T., and Craig, E. A. (1995) Mol. Cell. Biol. 15, 4809 – 4897.
22. Sorger, P. K. (1991) Cell 65, 363–366. 23. Furst, P., Hu, S., Hackett, R., and Hamer, D. (1988) Cell 55, 705–717. 24. Silar, P., Butler, G., and Thiele, D. J. (1991) Mol. Cell. Biol. 11, 1232–1238. 25. Yang, W., Gahl, W., and Hamer, D. (1991) Mol. Cell. Biol. 11, 3676 –3681. 26. Sewell, A. K., Yokoya, F., Yu, W., Miyagawa, T., Murayama, T., and Winge, D. R. (1995) J. Biol. Chem. 270, 25079 –25086. 27. Hahn, S. (1998) Cell 95, 579 –582.
739
Activator-Specific Requirement for the General Transcription Factor IIE in Yeast Hiroshi Sakurai* ,1 and Toshio Fukasawa† *School of Health Sciences, Faculty of Medicine, Kanazawa University, 5-11-80 Kodatsuno, Kanazawa, Ishikawa 920-0942, Japan; and †Tamura Laboratory, Graduate School of Science, Chiba University, 1-33 Yayoi-cho, Inage-ku, Chiba 263-8522, Japan
Received July 5, 1999
The general transcription factor (TF) IIE is required for mRNA synthesis of many, but not all, genes in yeast. In the transcription process, TFIIE regulates TFIIH kinase activity that phosphorylates the carboxy-terminal domain (CTD) of the largest subunit of RNA polymerase II. The CTD and the CTD kinase Kin28, a subunit of TFIIH, have been shown to be dispensable for activation of several heat shock genes and the copper metallothionein gene CUP1. Here we analyzed requirement of TFIIE for transcription of these genes and found that TFIIE is necessary for activation of the heat shock genes by heat shock transcription factor Hsf1. By contrast, transcription of CUP1 mediated by both Hsf1 and copper-activated transcription factor Ace1 was inducible after inactivating TFIIE. These results show that both TFIIE and the CTD/the CTD kinase exhibit “gene specificities” which are overlapping, but not identical to each other, and thereby suggest that TFIIE functions with or without involvement of the CTD/the CTD kinase depending on the gene to be transcribed. © 1999 Academic Press
The largest subunit of RNA polymerase II (RNAPII) has a unique carboxy-terminal domain (CTD) composed of tandem repeats of a heptapeptide sequence. Phosphorylation and dephosphorylation of the CTD associate with the transcription cycle: RNAPII with unphosphorylated CTD is recruited to the core promoter and assembles to form the preinitiation complex together with a set of the general transcription factors (TFs). The CTD is then phosphorylated by TFIIH and/or other kinases, which in turn triggers transition from initiation to elongation in the process of mRNA synthesis. Phosphorylated form of RNAPII is dephosphorylated by the CTD phosphatase upon completion 1
To whom correspondence should be addressed. Fax: 81-76-2344360. E-mail: [email protected]. 0006-291X/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
of the elongation and reused for initiation of the next cycle (1). In vivo, an important fraction of RNAPII is present in a large complex called “holoenzyme” of RNAPII together with several general transcription factors and “Mediator,” a protein complex of more than 20 components which transmits regulatory signals of sequence-specific transcription factors to RNAPII (2, 3). The phosphorylation cycle of the CTD regulates interaction of core RNAPII with the Mediator in the holoenzyme. Thus, unphosphorylated CTD binds the Mediator, and phosphorylation of the CTD causes dissociation of the RNAPII-Mediator complex (4). The phosphorylated CTD binds capping enzymes resulting in addition of 7-methyl-guanosine to the 59 end of RNAPII-transcribed RNA (5, 6). The general transcription factor TFIIE of the yeast Saccharomyces cerevisiae is a heterodimer of 66- and 43-kDa subunits which are encoded by the TFA1 and TFA2 genes, respectively (7). We have previously shown that TFIIE binds Gal11 protein, a component of the Mediator, and that TFIIE-Gal11 interaction enhances TFIIH-catalyzed phosphorylation of the CTD in vitro when holo-RNAPII but not core RNAPII was used as substrate (8). On the other hand, physical as well as functional interactions between TFIIE and TFIIH have been disclosed in yeast and mammalian cells (see for review Refs. 9 and 10). The above observations altogether support the idea that TFIIE and TFIIH function at a common step in transcription, that is, transition from initiation to elongation. Analysis of temperature-sensitive (ts) mutant of TFA1 (tfa1-21) has shown that TFIIE is required for transcription of many genes in yeast (11). However, constitutive transcription of several genes was not significantly affected in the mutant cells after shifting to the restrictive temperature. Those TFIIE-independent genes lack the canonical TATA sequence, suggesting that the gene-specific requirement for TFIIE is related to the core promoter structure (11). Recent reports (12, 13) have uncovered that the CTD, the Kin28 CTD
734
Vol. 261, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Transcription of heat-inducible genes in tfa1-21 and kin28-ts3 cells. Yeast cells were grown at 28°C then the temperature was shifted to 39°C. At the indicated times, aliquots of cells were removed and total RNA was prepared. Cells were grown in YPD medium for analyses of SSA4 and HSP104 mRNAs or in ESD medium lacing uracil for transcripts of SSA4-GAL7-lacZ’ and HSE-GAL7-lacZ’. The mRNA of SSA4 was analyzed by S1 nuclease mapping, whereas transcripts of HSP104, SSA4-GAL7-lacZ’, and HSE-GAL7-lacZ’ were analyzed by primer extension. Yeast strains used are HS33 (wild-type), HS46 (tfa1-21), JGV-4 (kin28-ts3), and Y262 (rpb1-1).
kinase, and the Mediator are dispensable for activation of a subset of genes, such as the SSA4 or HSP82 heat shock genes and the CUP1 metallothionein gene. To know functional relationship between TFIIE and the CTD/the CTD kinase, we have analyzed requirement of TFIIE for the transcription that does not require the CTD/the CTD kinase. We found that transcription of CUP1 but not the heat shock genes was activated in the absence of TFIIE function, and that the TFIIEindependent activation of CUP1 was conferred by its upstream activation sequence (UAS). These results suggest that the dependence of transcription of a gene on TFIIE is determined not only by the structure of its core promoter (11), but also by UAS-activator interaction, and further that TFIIE functions in either CTDrelated or CTD-unrelated fashion depending on the gene to be transcribed. MATERIALS AND METHODS Yeast strains and media. TFA1 wild-type (HS33) and tfa1-21 ts mutant (HS46) yeasts were described previously (11). KIN28 wildtype (GF262-2) and kin28-ts3 (JGV-4) strains were gifts from G. Faye (14). Cells bearing rpb1-1 ts mutation (Y262) were gifts from R. Young (15). Strains HS73 and HS74 were ace1 : : LEU2 derivatives of HS33 and HS46, respectively. Rich medium containing 2% glucose (YPD) and enriched synthetic medium containing 2% glucose (ESD) or 2% sodium lactate plus 3% glycerol (ESGlyLac) were prepared as described (11, 16). Reporter genes. Plasmid pI145Z was an integrating-type plasmid bearing GAL7-lacZ’ and URA3 (16). To construct reporter genes, the GAL7 promoter region of pI145Z was replaced by the promoter fragments as follows. The SSA4-GAL7-lacZ’ hybrid gene contained the SSA4 UAS (from 2251 to 2106), which was amplified by polymerase chain reaction (PCR) from genomic DNA of W303-1a, and the GAL7 core promoter (from 274 to 143). To construct HSE-GAL7lacZ’, a Hsf1-binding sequence TCGACTTCTAGAAGCTTCAAGAGC (17) was inserted upstream of the GAL7 core promoter. Reporter gene CUP1-lacZ’ contained the region from 2250 to 175 of CUP1, which was amplified by PCR from the genomic DNA. The CUP1GAL7-lacZ’ reporter was a hybrid gene of the CUP1 UAS (from 2250 to 298) and the GAL7 core promoter. The GAL7-CUP1-lacZ’ reporter contained the GAL7 UAS (from 2271 to 269) and the CUP1 core promoter (from 293 to 175). Reporter gene CUP1(HSE-M)lacZ’ was
a derivative of CUP1-lacZ’ bearing two mutations, C at 2162 to T and G at 2159 to A, which abolish binding of Hsf1 to the CUP1 UAS (18). The reporter genes were digested with SmaI and integrated into the ura3 locus of yeast. RNA analysis. Cells were grown under the conditions described in the figure legends. Preparation of total RNA was described previously (11, 16). The mRNA of SSA4 was analyzed by S1 nuclease mapping with the probe fragment containing the region from 2251 to 1157 of SSA4 with 59 end-labeling at 1157. The transcripts of other genes were analyzed by primer extension with primers HSP104, 59-GTTGATGATCCGAAGCCAAT-39; lacI, 59-CAGTGAGACGGGCAACAGCC-39; CUP1, 59-TACCACATTGGCATTGGCAC-39. The relative amount of mRNA was measured by a BAS-1000 imaging analyzer (Fuji Photo Film, Co.).
RESULTS AND DISCUSSION Effect of tfa1-21 and kin28-ts3 mutations on heatinducible transcription by Hsf1. By using a ts mutant named tfa1-21 which carries two missense mutations in the large subunit of TFIIE (11), we studied requirement of TFIIE for transcription of a set of genes, whose activation is known to be independent of the CTD and Kin28 (12, 13). The temperature shift of tfa1-21 cells to the restrictive temperature (37°C) results in degradation of both subunits of TFIIE and in the stop of mRNA synthesis from most of the genes, suggesting that this mutation significantly, if not completely, inactivates TFIIE function (11). Previously, Tijerina and Sayre (19) argued that TFIIE was necessary for maximal rate transcription of several heat shock genes, such as HSP26, HSP104, and SSA4, by using a ts mutant named tfa1-1 which carries a missense mutation in the conserved zinc-finger motif of the large subunit. However, the effect of that mutation was relatively weak; transcription of all the genes tested by themselves did not completely stop after shifting to the restrictive temperature, and activation of the heat shock genes was observed in tfa1-1 cells to 15–20% of the wild-type level at 15 min after the temperature shift (19). To know whether TFIIE is indeed dispensable for heat shock response, we reexamined this issue by using our ts mutant that exhibits much
735
Vol. 261, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
stronger effect than theirs. Figure 1 shows effect of tfa1-21 mutation on activation of SSA4, HSP104, or their hybrid genes with GAL7-lacZ’ in response to heat. Induction of SSA4 encoding heat shock protein 70 was brought about by up-shifting the temperature of yeast cultures to 39°C in the wild-type but not in tfa1-21 cells: the amount of SSA4 mRNA in tfa1-21 cells was less than 5% of the wild-type control (after 15 min at 39°C), which was comparable to those in rpb1-1 cells bearing a mutation in the largest subunit of RNAPII. The HSP104 gene encoding heat shock protein 104 was also induced at 39°C in the wild-type, but not in tfa1-21 nor in rpb1-1 cells. Consistent with the previous reports (12, 13), heat-induction of SSA4 and HSP104 occurred in kin28-ts3 cells, but the mRNA levels were lower (25–30%) than the wild-type control. Heat-inducible transcription of both SSA4 and HSP104 is mediated by heat shock factor Hsf1, which binds the heat shock element (HSE) of the respective gene at the UAS (20 –22). To know whether Kin28independent transcription of SSA4 could be conferred by the UAS, we constructed a hybrid reporter gene containing the SSA4 UAS and the GAL7 core promoter (SSA4-GAL7-lacZ’) and integrated it into the yeast chromosome at the ura3 locus. Heat-induction of the hybrid gene was observed in kin28-ts3 (30% of the wild-type cells), but not in tfa1-21 nor in rpb1-1 yeast. We further found that Kin28 but not TFIIE was dispensable for heat shock response of a reporter gene containing a synthetic HSE upstream of the GAL7 core promoter (HSE-GAL7-lacZ’). Note that the heat shock response of HSE-GAL7-lacZ’ was transient in the wildtype cells, but was sustained in kin28-ts3 cells. Since the functions of TFIIE or Kin28 protein are severely impaired in tfa1-21 or kin28-ts3 cells, respectively, at the high temperature (11, 14), the above results strongly suggest that TFIIE is necessary for activation of heat shock genes by Hsf1, for which Kin28 is dispensable, however. Copper-inducible transcription of CUP1 is independent of TFIIE and Kin28. Next we tested requirement of TFIIE for activation of the CUP1 gene in response to copper ion. When cells were grown at 28°C, addition of copper sulfate (0.2 mM) caused increase of CUP1 mRNA more than 15-fold over the uninduced level in the wild-type as well as in ts mutant yeasts, tfa1-21, kin28-ts3 or rpb1-1 (Fig. 2A). When, however, cells were grown at 37°C for 1 h before the addition of copper sulfate, induction of CUP1 was repressed to variable extents with yeast strains. This repression was significantly restored by increasing the concentration of copper sulfate to 1.0 mM in the wild-type, tfa121, and kin28-ts3. Thus the amounts of CUP1 mRNA in the wild-type, tfa1-21, and kin28-ts3 reached the respective levels about 90%, 55%, and 65% of the control (the value at 28°C at 0.2 mM of copper sulfate). In
FIG. 2. Copper-inducible transcription of CUP1 in tfa1-21 and kin28-ts3 cells. (A) Cells were grown in YPD medium at 28°C or 37°C for 1 h. Then, copper sulfate was added and cells were cultured for a further 30 min. Total RNA was prepared and CUP1 mRNA was analyzed by primer extension. Relative levels of CUP1 mRNA were determined by Image-analyzer and normalized by the levels in cells grown in the presence of 0.2 mM copper sulfate at 28°C. (B) Cells were grown in YPD medium at 28°C, then transcription of CUP1 was induced by the addition of 1.0 mM copper sulfate for 30 min. After shifting to 37°C, total RNA was prepared at the indicated times and analyzed as above. Relative levels of CUP1 mRNA were determined by Image-analyzer and normalized to the 0 time value as 100%.
rpb1-1 yeast, as expected, essentially no copperinduction of CUP1 was observed. The increase of CUP1 mRNA level at the restrictive temperature in tfa1-21 or in kin28-ts3 therefore indicates that de novo synthesis of CUP1 mRNA requires neither TFIIE nor Kin28. That inactivation of Kin28 does not significantly affect the induced transcription of CUP1 has been reported previously (12). A kinetic profile of induced level of CUP1 mRNA in the wild-type, tfa1-21, or rpb1-1 cells is shown in Fig. 2B. The respective yeast was grown at 28°C in the presence of 1.0 mM copper sulfate for 30 min to fully induce CUP1, and then shifted to 37°C. Total RNA was isolated at the indicated times after the temperature shift, and the amount of CUP1 mRNA was determined in each sample. In rpb1-1 cells, the amount of CUP1 mRNA decreased continuously with a half life of approximately 20 min. The amount of CUP1 mRNA decreased in tfa1-21 cells to a similar extent with rpb1-1
736
Vol. 261, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 3. Transcription of CUP1/GAL7 hybrid genes in tfa1-21 and kin28-ts3 cells. Transformants harboring the reporter gene CUP1-lacZ’ or CUP1-GAL7-lacZ’ were grown in ESD medium lacking uracil at 28°C or 37°C for 1 h then grown in the presence of 1.0 mM copper sulfate for a further 30 min. For analysis of transcripts of GAL7-lacZ’ or GAL7-CUP1-lacZ’, cells were grown in ESGlyLac medium lacking uracil and 2% galactose was added for induction as above. Total RNA prepared was subjected to primer extension with lacI primer.
for the initial 30 min, slightly increased at 1 h, and then decreased very slowly with a slope similar to that seen in the wild-type cells. It appears that abrupt inactivation of TFIIE caused immediate cessation of CUP1 transcription followed by its re-initiation. We imagine that the induced transcription of CUP1 normally involves TFIIE function but can take place in its absence as well. Upstream activation sequence is a determinant of TFIIE- and Kin28-independent transcription of CUP1. To identify sequence elements that confer TFIIE- and Kin28-independent transcription of CUP1, we constructed CUP1-lacZ’, GAL7-lacZ’, and their hybrids as shown in Fig. 3. Transcription of CUP1-lacZ’ was successfully induced by copper sulfate after inactivation of either TFIIE or Kin28, indicating that the region from 2250 to 175 of CUP1 is sufficient for both TFIIE- and Kin28-independent activation. However, galactoseinduced transcription of GAL7-lacZ’ required both factors. The hybrid gene containing the CUP1 UAS and the GAL7 core promoter (CUP1-GAL7-lacZ’) was induced by copper sulfate in either ts mutant at 37°C. By contrast, GAL7-CUP1-lacZ’ containing the GAL7 UAS and the CUP1 core promoter was induced by galactose in neither mutant. These results suggest that activated transcription mediated by the CUP1 UAS, at least in part, is independent of both TFIIE and Kin28. Requirements of Ace1 and Hsf1 for TFIIE-independent transcription. The CUP1 UAS contains two activating sequences; the binding site of Ace1 which mediates copper-inducible transcription (23) and the binding site of Hsf1, HSE which is involved in constitutive as well as heat-inducible transcription (17, 18, 24, 25). It has been reported that both the CTD and Kin28 are dispensable not only for copper-induction
but also for heat-induction of CUP1 (13). We then examined requirement of TFIIE for heat shock response of CUP1 (Fig. 4A). In the wild-type cells, more than 10-fold increase of the CUP1 mRNA level was observed by shifting the growth temperature to 39°C. By contrast, the temperature shift to 39°C (or 37°C; data not shown) caused rather decrease of CUP1 mRNA level in tfa1-21 cells, indicating that TFIIE is required for heatinduction by Hsf1 in the context of the upstream sequence of CUP1 as well. Although Ace1 is the major activator for copperinducible transcription, a stress caused by a high concentration of copper ion also activates CUP1 transcription by the mediation of Hsf1 in ace1 mutant when Hsf1 is over-expressed (26). It was suspected therefore that the copper-induction of CUP1 observed after inactivating TFIIE could be mediated by Hsf1 instead of Ace1. To test this possibility, we carried out series of experiments by using cells lacking Ace1 protein. As shown in Fig. 4B, addition of 1.0 mM copper sulfate to TFA1 ace1 null (wild type in the figure) or tfa1-21 ace1 null yeast (tfa1-21 in the figure) grown at 28°C showed no effect on the CUP1 mRNA level. When TFA1 ace1 cells were grown at 37°C, CUP1 mRNA synthesis was induced presumably by the mediation of Hsf1/HSE. By contrast, the constitutive synthesis of CUP1 mRNA was arrested in tfa1-21 ace1 cells at that temperature. Under these conditions, copper sulfate unaffected the CUP1 transcription in both types of yeast. From these results we conclude that copper-activated Ace1 is responsible for the TFIIE-independent activation of CUP1 in response to copper sulfate, as is seen in Fig. 2A. To know whether constitutive activity of Hsf1 is involved in the activation pathway of CUP1, we em-
737
Vol. 261, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
tion by Hsf1. It would be presumable that Hsf1 recruits the preinitiation complex to the core promoter without involvement of the CTD and TAFs, and yet, that the initiation step still requires TFIIE as an integral component of the basal initiation machinery. On the other hand, we have also shown that activation of CUP1 mediated by the combination of Ace1 and Hsf1 is not significantly affected by inactivation of TFIIE. We suggest that the TFIIE-independent activation of CUP1 may be partially conferred by CTD function, since TFIIE is involved in regulation of CTD phosphorylation (8). The CTD-independent pathway of CUP1 activation mediated by Hsf1 may be further modified by Ace1 so that the requirement for TFIIE is bypassed. Elucidation of the molecular mechanism underlying the activation-specific requirement for TFIIE should furnish a clue to get further insight to the central problem. How is the activation signal conveyed to the basal transcription machinery? ACKNOWLEDGMENTS FIG. 4. Activation of CUP1 in tfa1-21 cells. (A) Heat-inducible transcription of CUP1 in tfa1-21 cells was analyzed as in Fig. 1 by using CUP1 primer. (B) Copper-inducible transcription of CUP1 was analyzed as in Fig. 2A, except that ace1 null strains HS73 (ace1 : : LEU2) and HS74 (ace1 : : LEU2 tfa1-21) were used. (C) Copperinducible transcription of the reporter gene CUP1(HSE-M)lacZ’ in HS33 (wild-type) and HS46 (tfa1-21) was analyzed as in Fig. 3.
ployed a reporter gene CUP1 (HSE-M)lacZ’ that contains a mutation in the HSE of the CUP1 UAS (18). Transcription of this reporter was induced by copper sulfate (Fig. 4C) but not by heat (see Ref. 18, data not shown). The observed activation mediated by Ace1 was significantly inhibited by tfa1-21 mutation at 37°C (Fig. 4C). We also analyzed transcription of a reporter gene having a copy of Ace1-binding sequence upstream of the GAL7 core promoter and found that copperinduction of this reporter was abolished by inactivation of TFIIE (data not shown). These results indicate that activation of CUP1 mediated by either Ace1 or Hsf1 is dependent on TFIIE, but that the presence of both activators on the CUP1 UAS circumvents the requirement for TFIIE. Our previous study suggested an importance of the core promoter structure for the dependence on TFIIE (11). Here we demonstrated that the DNA sequencespecific activator is also a determinant of the requirement for TFIIE. It has been reported that activation of heat shock genes and CUP1 is independent of various components of the RNAPII transcription machinery including the CTD, Kin28, CTD-associated Mediator components such as Srb4 and Srb6, or TATA-binding protein associated factors (TAFs) (12, 13, and for review, see Ref. 27). In the present study, we have shown that TFIIE is necessary for heat-inducible transcrip-
We thank Drs. G. Faye and R. A. Young for providing yeast strains. This work was supported in part by Grants-in-Aid for Scientific Research from the Ministry of Education, Sciences, Sports, and Culture.
REFERENCES 1. Dahmus, M. E. (1996) J. Biol. Chem. 271, 19009 –19012. 2. Kim, Y.-J., Bjorklund, S., Li, Y., Sayre, M. H., and Kornberg, R. D. (1994) Cell 77, 599 – 608. 3. Bjorklund, S., and Kim, Y.-J. (1996) Trends Biochem. Sci. 21, 327–335. 4. Svejstrup, J. Q., Li, Y., Fellows, J., Gnatt, A., Bjorklund, S., and Kornberg, R. D. (1997) Proc. Natl. Acad. Sci. USA 94, 6075– 6078. 5. McCracken, S., Fong, N., Rosonina, E., Yankulov, K., Brothers, G., Siderovski, D., Hessel, A., Foster, S., Amgen EST Program, Shuman, S., and Bentley, D. (1997) Genes Dev. 11, 3306 –3318. 6. Cho, E.-J., Takagi, T., Moore, C., and Buratowski, S. (1997) Genes Dev. 11, 3319 –3326. 7. Feaver, W. J., Henry, N. L., Bushnell, D. A., Sayre, M. H., Brickner, J. H., Gileadi, O., and Kornberg, R. D. (1994) J. Biol. Chem. 269, 27549 –27553. 8. Sakurai, H., and Fukasawa, T. (1998) J. Biol. Chem. 273, 32663– 32669. 9. Svejstrup, J. Q., Vichi, P., and Egly, J.-M. (1996) Trends Biochem. Sci. 21, 346 –350. 10. Zawel, L., and Reinberg, D. (1995) Annu. Rev. Biochem. 64, 533–561. 11. Sakurai, H., Ohishi, T., and Fukasawa, T. (1997) J. Biol. Chem. 272, 15936 –15942. 12. Lee, D., and Lis, J. T. (1998) Nature 393, 389 –392. 13. McNeil, J. P., Agah, H., and Bentley, D. (1998) Genes Dev. 12, 2510 –2521. 14. Valay, J. G., Simon, M., Dubois, M. F., Bensaude, O., Facca, C., and Faye, G. (1995) J. Mol. Biol. 249, 535–544. 15. Nonet, M., Scafe, C., Sexton, J., and Young, R. A. (1987) Mol. Cell. Biol. 7, 1602–1611. 16. Sakurai, H., Ohishi, T., and Fukasawa, T. (1996) FEBS Lett. 398, 113–119.
738
Vol. 261, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
17. Santoro, N., Johansson, N., and Thiele, D. J. (1998) Mol. Cell. Biol. 18, 6340 – 6352. 18. Tamai, K. T., Liu, X., Silar, P., Sosinowski, T., and Thiele, D. J. (1994) Mol. Cell. Biol. 14, 8155– 8165. 19. Tijerina, P., and Sayre, M. H. (1998) J. Biol. Chem. 273, 1107– 1113. 20. Boorstein, W. R., and Craig, E. A. (1990) J. Biol. Chem. 265, 18912–18921. 21. Halladay, J. T., and Craig, E. A. (1995) Mol. Cell. Biol. 15, 4809 – 4897.
22. Sorger, P. K. (1991) Cell 65, 363–366. 23. Furst, P., Hu, S., Hackett, R., and Hamer, D. (1988) Cell 55, 705–717. 24. Silar, P., Butler, G., and Thiele, D. J. (1991) Mol. Cell. Biol. 11, 1232–1238. 25. Yang, W., Gahl, W., and Hamer, D. (1991) Mol. Cell. Biol. 11, 3676 –3681. 26. Sewell, A. K., Yokoya, F., Yu, W., Miyagawa, T., Murayama, T., and Winge, D. R. (1995) J. Biol. Chem. 270, 25079 –25086. 27. Hahn, S. (1998) Cell 95, 579 –582.
739