Transcriptional activation of NFI/CTF1 depends on a sequence motif ...

Nucleic Acids Research, 1994, Vol. 22, No. 13 2601-2603

Transcriptional activation of NFI/OTF1 depends on a sequence motif strongly related to the carboxyterminal domain of RNA polymerase lI Wolfgang Wendler, Herbert Altmann and Ernst-Ludwig-Winnacker* Institut fur Biochemie der Ludwig-Maximilians-Universitat Munchen, Am Klopferspitz 18a, 82152 Martinsried, Germany Received March 17, 1994; Revised and Accepted April 22, 1994

ABSTRACT Initiation of RNA polymerase Il-directed transcription is mediated by DNA sequence specific activator proteins interacting with components of the basal transcription machinery. NFI/CTF is a family of such binding proteins which have been shown to stimulate transcription via proline-rich activation domains. In order to identify residues crucial for its activator function, a pool of CTF1 mutants was cloned and fused to the bacterial repressor LexA. Transcriptional activation of these constructs was monitored in a Saccharomyces cerevisiae reporter assay. Our studies reveal the existence of a core domain in CTF1 between residues 463 and 508 essential for transcriptional activation functions. It contains the sequence motif SPTSPSYSP, which is strongly related to the heptapeptide repeat YSPTSPS present in the carboxyterminal domain (CTD) of RNA polymerase 11. Removal of the entire CTD related motif, as well as substitution of key amino acids therein, abolish CTF1 mediated transcriptional activation. The different members of the NFI/CTF family are derived from a single gene by alternative splicing (1). The N-terminal 220 amino acids mediate specificDNA-binding, dimerization and stimulation of adenovirus DNA replication, whereas the Cterminus is harbouring a proline-rich activation domain capable of enhancing transcription in Drosophila and HeLa cells (2, 3). Previously we have shown that different splice variants of NFI/CTF can be functionally expressed in S. cerevisiae and activate transcription of reporter genes with varying effectivity (4). This is in accordance with other reports demonstrating that the CTF1 C-terminus stimulates transcription in yeast at least from TATA-proximal positions (5, 6). Our recent studies also suggest a bipartite structure of the CTF activation domain which lacks the most proline-rich portion of the C-terminal region. Therefore we wanted to know, whether there are key amino acids playing a crucial role in the transactivation domain of CTF 1. To address this question we cloned a set of different internal deletions *

To whom correspondence should be addressed

and substitutions within the CTF1 carboxyterminal domain. The constructs were expressed in yeast as fusion proteins together with the DNA binding and dimerization domain of the bacterial repressor LexA(I-202)* The transactivating activities of the different CTF1-mutants were analysed in yeast reporter strain EGY48 which harbours a LEU2 gene carrying LexA binding sites in its promoter region (7). Yeast cells of this strain are capable of growing on leucine-deficient media, only if transcriptionally active LexA-fusion proteins are expressed. The results of these experiments are shown in figure 1. First we analysed the activity of CTF1 mutants in which stretches of 10 wild-type amino acids have been substituted for the motif PARHGLANSG (written in one letter-code). Two mutants with substitutions between residue 461 and 481, failed to activate transcription completely. However, the mutant with a substitution between aa 453 to 462 mediated growth on leucinedeficient media. This demonstrates that the simultaneous exchange of proline residues at positions 455, 456 and 460 in the most proline-rich CTF 1-region has no detrimental effects on its activating properties. This confirms our recent observation on the strong activating splice variant CTF7 which lacks the most proline-rich stretches at the CTF carboxy-terminus (4). We also assayed some internal deletions between residues 469 to 481, harbouring the sequence-motif SPTSPSYSP, which is related to the heptapeptide repeat YSPTSPS present in the carboxyterminal domain (CTD) of RNA polymerase II (8). Deletions removing this element do not activate transcription (A469 to 481 and A469 to 479). Also smaller deletions of only 4 or 5 amino acids abolished transcriptional activation (A469 to 472 and A469 to 473). To identify essential amino acids we mutated smaller portions of the protein down to the level of single amino acids. Exchanging sequences SPT at position 470 to 472 as well as PDT between residues 479 to 481 for the tripeptide AGA abolished transactivation activity altogether. Interestingly, single substitutions of T, S, Y and T at positions 469, 470, 476 and 481 into alanine do not interfere with transcriptional activation. However, transcriptional activity of full length CTF1 is lost when P at position 471 is exchanged into A. To

.~ .~~ ~~ ~-..

2602 Nucleic Acids Research, 1994, Vol. 22, No. 13

CTD-related motif

4150 ,

WT-CTF1

470 1

460 ,

480 ~~~~ ~~~9110 ~

P R L A L P P A T K P A T T S E G G A T S P T S P S Y S P

A R H G L A IP

F * 10 fold substitutions L

.

N S

GI

.

.

.

.

.

.

Xi :.,,.. -V i,:: 41~~~~~~~~~~~~..:0

3 fold r substitutions

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

*

p

_ , .

.,... ., ......,,. ', .....

0... .

.

.

.

.

.

.

........ .

...

+

...

A..R.H.G.L..A N .......

_.

..

...

....

.,

G

..

: ..:.::..:.. ,- .:..

A

[ 1 fold substitutions[

+

...

. ..,. .

.

P D T S P A N

*.x0Zte0+ 00Wx:t;. . . .. -.: .:.- .:. -.... -..

.

.

...,,.

:-::: -:: .0

IP A R H G LG|A N S

.*

.

Transcriptional activated activity

:.:.

'..

+

....

* * * @ @ @ s @ @ * @ - @ - a * a @ @ l @~~~~~~~~~~......... llXss * e @ Z @ @ , :*- ...:t:

....

deletions

F

C:::-:::::*::+::w:*::::::*:::*:t::-:*::: : X.................................... ::::: :::

Figure 1. Mutation analysis of the transcription activation domain of NFI/CTF1. Line 'WT-CTF1' displays a portion of the amino acid sequence of wild-type CTFl written in one letter-code. The amino acid positions are indicated as well as the splice site between exon 9 to 10 and the CTD-related motif SPTSPSYSP. Mutations are marked as boxes and wild-type amino acids are symbolized with dots. In the case of substitution mutants the newly generated sequences are written within the boxes, whereas deleted amino acids are marked as lines. All mutated CTFl-proteins were fused to the DNA-binding and dimerization domain of LexA(O -202) and expressed in yeast reporter strain EGY48 (4, 7). Transcriptional activation activities were monitored as the ability (+) or the inability (-) of transformed yeast cells to grow on leucine-deficient media.

demonstrate that nonactivating mutants are also functionally expressed in transformed yeast cells we used a LexA driven repressor assay described elsewhere (9). In summary, our results indicate (a) that the transcriptional activation domain of CTFl is localized within the carboxyterminal region between aa 463 and 508, (b) that the CTD-like motif has an essential function in CTF1 mediated transcription, (c) that proline residues in the most proline-rich CTF1-stretches are dispensible, whereas (d) prolines within (residue 471) or close to the CTD-like sequence motif (position 479) are essential for transcriptional activation. The importance of the CTD-related motif for the transcriptional stimulation of CTF1 is in agreement with data showing that multiple copies of the CTD heptamer repeat are able to stimulate RNA polymerase II transcription (10). In addition, our data are supported by a recent deletion analysis showing that residues 460 to 479 are important for transcriptional activation of CTF 1 (1 1). It appears attractive to look at relationships between structure, function and possible modifications of this short element. First it is notable that the sequence 'SPXX' is overrepresented in generegulatory proteins (12). Interestingly, most of the SP motifs in CTF1 are localized in the C-terminal half of the protein and 4 of them are found within or close to the CTD-related motif.

However, not all SP elements appear to be essential for transcriptional activation because, for once, S at position 470 can be substituted for A without loss of stimulatory activity. Moreover, the inactivity of the P to A mutant at position 471 indicates, that it is a single proline residue which is crucial for function and not the SP duplet. Nevertheless, more data are necessary to determine the role of the other SP elements. Another interesting aspect relates to posttranslational modification of the CTD-related element. It has often been shown that the RNA polymerase II CTD region is phosphorylated at serine and threonine residues by a set of different kinases (see (8) and references therein). Furthermore, it was observed that copies of the YSPTSPS element are phosphorylated when fused to a specific DNA-binding domain (13). Phosphorylation of transcription factors provides an excellent mechanism for regulation of their activity and it will be interesting to show if transcriptional activation properties of CTF 1 are influenced by such means.

ACKNOWLEDGEMENTS This work was supported by the Deutsche Forschungsgemeinschaft grant FA 138/6-1. We

are

grateful to Roger Brent for

Nucleic Acids Research, 1994, Vol. 22, No. 13 2603 providing us with the appropriate reporter strain EGY48 and the LeXA(I-202) fusion plasmid.

REFERENCES 1. Meisteremst, M., Rogge, L., Foeckler, R., Karaghiosoff, M. and Winnacker, E.-L. (1989) Biochemistry 28, 8191-8200. 2. Mermod, N., O'Neill, E.A., Kelly, T.J. and Tjian, R. (1989) Cell 58, 741-753. 3. Knox, J.J., Rebstein, P.J., Manoukian, A. and Gronostajski, R.M. (1991) Mol. Cell. Biol. 11, 2946-2951. 4. Altmann, H., Wendler, W. and Winnacker, E.-L. (1994) Proc. Natl. Acad. Sci. U.S.A. 91, 3901-3905. 5. Kunzler, M., Braus, G.H. Georgiev, O., Seipel, K. and Schaffner, W. (1994) EMBO 13, 641-645. 6. Kim, T.K. and Roeder, R.G. (1993) J. Biol. Chem. 268, 20866-20869. 7. Gyuris, E., Golemis, E., Chestkov, H. and Brent, R. (1993) Cell 75, 791-804. 8. Drapkin, R., Merino, A. and Reinberg, D. (1993) Curr. Opin. Cell Bio. 5, 469-476. 9. Lech, K., Anderson, K. and Brent, R. (1988) Cell 52, 179-184. 10. Seipel, K., Georgiev, 0. Gerber, H.P. and Schaffner, W. (1993) Nucl Acids Res. 21, 5509-5615. 11. Kim, T.K. and Roeder, R.G. (1994) Nucl. Acids Res. 22, 251. 12. Suzuki, M. (1989) J. Mol. Biol. 207, 61-84. 13. Peterson, S.R., Dvir, A., Anderson, C.W. and Dynan, W.S. (1992) Genes and Dev. 6, 426-438.