sign in sign up
<<
Home , CAAT box, Lac repressor, Repressor, TATA box

Proc. Nati. Acad. Sci. USA Vol. 85, pp. 1394-1397, March 1988 Biochemistry TnlO-encoded tet can regulate an operator-containing (cauliflower mosaic 35S promoter/electroporation/transient chloramphenicol acetyltransferase assays) CHRISTIANE GATZ* AND PETER H. QUAILt Departments of Botany and Genetics, University of Wisconsin, Madison, WI 53706 Communicated by Folke Skoog, October 26, 1987 (receivedfor review July S, 1987)

ABSTRACT The TnlO-encoded tet repressor-operator The TnlO-encoded tet repressor regulates the expression system was used to regulate from the cauliflower of the Tc resistance by binding to nearly identical mosaic virus (CaMV) 35S promoter. Expression was moni- operator sequences that overlap with three divergent pro- tored in a transient assay system by using electric field- moters (14, 15). The of the tet operon are only mediated transfer ("electroporation") into tobacco pro- transcribed in the presence of the Tc, which pre- toplasts. The tet repressor, being expressed in the plant cells vents the repressor from binding to its operator sequences. under the control of signals, blocks The tet repressor was chosen for regulating a plant promoter transcription of a CaMV 35S promoter chloramphenicol ace- for two reasons. (i) With a native molecular mass of 48 kDa, tyltransferase (cat) fusion gene when the two tet operators diffusion into the nucleus seemed likely (16). (ii) The high flank the "TATA" box. In the presence of the inducer equilibrium association constant of the repressor-inducer tetracycline, expression is restored to full activity. Location of complex ensures efficient induction at sublethal Tc concen- the operators 21 base pairs downstream of the transcription trations (17), thus making the system useful as an on/off start site does not significantly affect transcription in the switch for the specific regulation of transferred genes. presence of the repressor. These experiments show that a prokaryotic regulatory can function in plant cells. The MATERIALS AND METHODS tet repressor-operator complex may be useful for specifically inducing transferred genes at different stages of plant devel- Electric Field-Mediated Gene Transfer ("Electroporation") and Transient Assay Conditions. Expression was measured in opment. electroporated Nicotiana tabacum (gift from T. Bradshaw, University of Washington, Seattle) protoplasts. For the In prokaryotes as well as in , regulation of tran- preparation of protoplasts, suspension cultured cells were scription initiation is mediated by that recognize taken on the third day after being subcultured. Culture specific DNA sites, thereby influencing RNA conditions were as follows. Cells were maintained in 4.3 g of activity (1). One of the well-characterized negative control Murashige-Skoog (MS) salts per liter, 30 g of sucrose per mechanisms in involves repressor proteins liter, 1 mg of thiamine per liter, 1 g of myo-inositol per liter, binding to operator DNA, thus preventing RNA polymerase 2 g of KH2PO4 per liter, and 0.2 mg of 2,4-dichlorophenox- from binding (2). In eukaryotes, however, the position of the yacetic acid (2,4-D) per liter at 280C with shaking at 150 rpm regulatory DNA binding sites is not limited to the region near and were subcultured every week with a 5% inoculum. Cells the start site of transcription. Negative (3-5) as well as were washed twice with the culture medium containing 0.4 positive (6) control sequences function somewhat indepen- M mannitol (MSM medium) and then incubated in this dently of their positions and orientations with respect to the medium with 500 mg of cellulase and 50 mg of macerozyme regulated promoter. In spite of these fundamental differ- per 50 ml of culture for 2 hr. Cells were pelleted, washed ences between eukaryotic and prokaryotic control mecha- twice with MSM medium, washed once in electroporation nisms, the E. coli lexA protein is able to repress gene buffer (10 mM Hepes, pH 7.2/150 mM NaCl/4 mM expression in yeast (7). Recently, Hu and Davidson (8) have CaCl2/0.4 M mannitol), and resuspended in the same buffer shown that a suitably engineered -operator at 4 x 106 protoplasts per ml. After mixing 0.5 ml of the system is functional in mouse cells. By using the TnlO- protoplast-containing solution with 0.5 ml of the same solu- encoded tet repressor-operator interaction, we have ap- tion containing 200 ,ug of supercoiled plasmid DNA, electro- proached the question of whether the plant eukaryotic poration was done as described (18). The electric pulse was transcription machinery can be inhibited by the same mech- delivered from a 490-,uF capacitor charged to 340 V with a anism as is E. coli RNA polymerase-namely, by steric 3-msec resistance-capacitance time constant. Chloramphen- interference with a repressor protein. Introducing the tet icol acetyltransferase (CAT) activity was measured after 24 regulatory elements that respond to the inducer tetracycline hr of incubation (18). Equal amounts of protein were added (Tc) (9, 10) into plant cells provides a unique tool to in each assay. The amount of acetylated chloramphenicol specifically regulate expression of transferred genes. By was determined by cutting the reaction product from a silica using this system rather than plant promoters regulated by plate and counting the radioactivity in a liquid scintillation light (11), stress (12), or (13), one has the advan- spectrometer. tage that only the expression of the transferred gene will be affected by the inducer. This contrasts with the regulatory Abbreviations: CAT, chloramphenicol acetyltransferase; CaMV, factors mentioned above, where pleiotropic effects on the cauliflower mosaic virus; 2,4-D, 2,4-dichlorophenoxyacetic acid; plant are expected. Tc, tetracycline; 01 and 02, operators 1 and 2. *Present address: Institut fuer Genbiologische Forschung, Ihnestr 33, 1 Berlin 63, F.R.G. The publication costs of this article were defrayed in part by page charge tTo whom reprint requests should be addressed at the present payment. This article must therefore be hereby marked "advertisement" address: Plant Center, 800 Buchanan Street, in accordance with 18 U.S.C. §1734 solely to indicate this fact. Albany, CA 94710.

1394 Downloaded by guest on September 30, 2021 Biochemistry: Gatz and Quail Proc. Natl. Acad. Sci. USA 85 (1988) 1395

Construction of Recombinant Plasmids. For constructing region of the tet repressor (tetR) gene was inserted as an pTET7 the cauliflower mosaic virus (CaMV) 35S promoter EcoRI fragment [from pWH305 (24)] between the CaMV 35S fragment [ 7017-7437 (19)] was subcloned from promoter and the nos poly(A) signal by replacing the nptII an intermediate construct containing these promoter se- gene of pCaMVNEO (20). pTETO contains the same EcoRI quences between the HindIII and the Pst I sites ofthe PiAN7 fragment in the reverse polarity and was used for control (Biolabs, Northbrook, IL) polylinker inserted into pUC18 experiments. (20). The promoter was fused to the cat gene by cloning the former as a Sma I/Bgl II fragment into pGA582 (21) that had been digested with Hpa I and Bgl II. This construct contains RESULTS nucleotides - 390 to + 1 relative to the start site oftranscrip- We inserted the tet operators (01 02) into selected sites of tion of the CaMV 35S promoter, 50 base pairs (bp) between the CaMV 35S promoter whose functional domains have + 1 and the start site ofcat , the cat coding region, been characterized by 5' deletions (25, 26). pTET8 (see Fig. and the nos poly(A) signal site. The promoter-cat-nos con- 1) contains the operators between the transcription start site struct was subsequently inserted as a HindIII/Sal I fragment and the initiation codon, with the first base of the left into pUC19 to yield pTET7. Plasmid pTET8 contains an operator (O1) being base + 22 of the cat-encoded mRNA. 80-bp-long EcoRI fragment with the two tet operator sites The inserted operator fragment contains one ATG initiation (14, 22) cloned into the Bgl II site ofpTET7 after filling in the codon that is in frame with the ATG of the cat gene, so that protruding ends. Plasmid pTET8 contains 3 bp of the TnJO- expression was not reduced by an upstream out-of-frame encoded sequence flanking operator 1 (01) and 20 bp flank- open reading frame (27). In pTET14 (Fig. 1A) 01 is posi- ing operator 2 (02). For constructing pTET14, two comple- tioned between the CAAT box and the TATA box and 02 is mentary oligonucleotides containing the "TATA" box ofthe between the TATA box and the start site of transcription. CaMV 35S promoter between the two operators were syn- We chose the position between the consensus boxes under thesized with Hga I and Sau III ends. The distance between the assumption that the CaMV 35S promoter has the same the "CAAT" box and the TATA box was maintained as in characteristics as the thymidine kinase promoter (28). the wild-type promoter. The spacing between the operators Linker scanning of this promoter has shown that the se- was 11 bp as in the E. coli operon. Because of multiple Hga quences between the CAAT and the TATA boxes can be I sites in pUC19, the plasmid was constructed by ligating an changed without eliminating . Taking into isolated promoter fragment from HindIII to Hga I ofpTET7, account that spacing between the TATA box and the up- with the hybridized oligonucleotides and the HindIII/Bgl II stream element (in this case, the CAAT box) is important for vector fragment of pTET7. The correct of the maximal promoter activity, at least in simian virus 40 early operators into pTET8 and pTET14 was confirmed by sub- promoter (29), we maintained the CaMV wild-type spacing. cloning the respective fragments into M13 and subsequent Similarly, we inserted the two operators in their normal sequence analysis (23). For constructing pTET3 the coding tandem arrangement-i.e., separated by 11 bp. Fig. 1B A TRANSCRIPTION DISTAL REGION CCAAT BOXES TATA BOX START SITE H/ / i B91QI eAT GENE Ed\\-LI El -\\--- pTET7 -89 -53 -24 +1

O °02 CAT GENE f - - \\-pTET8

°l CAT GENE LI LIL C _ I \\ p TET 14

o0 0 TRANSCRIPTION 2 START SITE FOR TEJ A - 35 -10 E-COLI: MEM_: =

B HgaI 01 0 TSS GATGACAA CCACTACTCTATCATTGATAGAGTCTCITATATAAGTCCCTATCAGTGATAGAGAGAGGATC

FIG. 1. Diagram of the operator insertions in the promoter regions. (A) pTET7 contains the wild-type CaMV 35S promoter fused to the cat gene in pUC19. In pTET8, the tet operators were inserted into the Bgl II site downstream of the transcription start site. pTET14 contains the operators flanking the TATA box. The different domains (TATA box, CAAT boxes, distal region) of the CaMV 35S promoter from the start site oftranscription are designated (26). The operators are schematically shown as black boxes. The lower panel ofA indicates the location ofthe two operators within the tetA promoter of E. coli (14). The - 35- and the - 10-bp consensus sequences are shown as open boxes. (B) Sequence of pTET14 from + 3 to - 69. The Hga I cleavage site_ is indicated by the arrows, the operator sequences are presented by bars with the asterisks marking the centers of the palindromes, and the CAAT and TATA sequences are indicated by boxes. TSS, transcriptional start site. Downloaded by guest on September 30, 2021 1396 Biochemistry: Gatz and Quail Proc. Natl. Acad. Sci. USA 85 (1988) shows the sequence of the altered CaMV 35S promoter in pTET14 pTET14. The promoter constructs were fused to the cat gene as a +pTET3 +pTETO reporter sequence for testing promoter activity. Expression REPRESSOR: was tested in a transient assay by using electroporation to introduce the DNA into tobacco protoplasts. Comparison of Tc (mg/I): o 0.2 2.0 0 0.2 2.0 lanes 1, 3, and 5 of Fig. 2 shows that the operator insertions alone did not dramatically affect promoter activity. 3-CM - *- - . .. To test the effect of the tet repressor on expression from 1-CM - _e & pTET7, pTET8, and pTET14, we constructed pTET3, a plasmid containing the repressor coding region between the CaMV 35S promoter and the nos poly(A) signal. This is the CM - 0 4 0 6 6 6 4 same arrangement ofeukaryotic transcription signals as used to express the prokaryotic cat gene in pTET7, which is functional in this assay. Prior to electroporation, 180 jug of pTET3 and 20 ,ug of the respective promoter-cat construct 1 2 3 4 5 6 7 were mixed. Activity of the promoter-cat constructs in the FIG. 3. Induction of CAT activity by Tc. Protoplasts were absence of the repressor was tested by electroporation of 20 electroporated and CAT assays were performed. pTET0 contains ,ug of pTET7, pTET8, and pTET14, respectively, with 180 the tetR gene in an inverse orientation with respect to the coding gg of pTET0, which differs from pTET3 by having the tetR region between the CaMV promoter and the nos poly(A) signal; coding region in the inverse orientation between the eukary- pTET3 is the same construct as pTET0 with the coding region in the otic transcriptional signals. By keeping the total amount of correct orientation. Lanes: 1, 200 ,ug of pTETO; 2-4, 20 ,ug of DNA constant in each experiment we avoided in pTET14 with 180 zg of pTET3; 5-7, 20 ,ug of pTET14 with 180 ,ug differences of pTET0. After the electroporation, protoplasts were incubated the levels of expression due to increases in the signal caused with different concentrations of Tc. Lanes: 1, 2, and 5, no Tc; 3 and by the addition of non-cat-containing DNA (30). Fig. 2 6, 0.2 mg of Tc per liter; and 4 and 7, 2 mg of Tc per liter. CM, shows the difference in the expression of the three tested chloramphenicol; 1-CM, 1-acetyl chloramphenicol; 3-CM, 3-acetyl promoter-cat constructs when coelectroporated with pTET3 chloramphenicol. or pTETO. Expression of the wild-type CaMV 35S promoter is independent of which plasmid was coelectroporated (Fig. activity driven by the pTET14 promoter in the presence or 2, lanes 1 and 2). Also, no significant decrease in expression absence of the repressor. Plasmid pTET14 was mixed with was observed when pTET8 was electroporated with pTET3 pTET3 (Fig. 3, lanes 2-4) or pTETO (Fig. 3, lanes 5-7) and instead of pTET0 (Fig. 2, lanes 3 and 4). However, CAT coelectroporated into protoplasts that were then incubated activity dropped by a factor of 10 when pTET14 was tested with two different concentrations of Tc: 0.2 mg/liter and 2.0 together with pTET3 (Fig. 2, lanes 5 and 6). The data mg/liter. At 0.2 mg ofTc per liter, the promoter was induced indicate that after transfer of pTET3 into the plant cells, 5-fold, whereas full was achieved when the active repressor is being made and is binding to the tet cells were incubated with 2.0 mg of Tc per liter (Fig. 3, lane operator sequences. The repressor apparently has no effect 4). The Tc concentration used does not affect transcription on the expression of the wild-type CaMV 35S promoter but in the absence of the repressor. Comparison of lanes 4 and 5 does inhibit expression of the CaMV 35S promoter with the (Fig. 3) shows that the inducer Tc restores expression to the tet operators flanking the TATA box in a fashion similar to level that occurs in the absence of repressor. In the re- that in the E. coli tetA promoter (see Fig. 1). We think that pressed state of the promoter (Fig. 3, lane 2), CAT activity the inhibition of transcription is due to steric interference was 3-fold higher than the background CAT activity detect- with proteins of the eukaryotic transcription initiation com- able in tobacco protoplasts electroporated with pTETO. We plex by the repressor protein bound to the operators. think that this additional activity may be due to immediate Fig. 3 depicts the effect of the inducer Tc on the CAT expression of the cat gene before enough tet repressor is pTET7 pTET8 pTET14 made to provide full repression.

REPRESSOR: - + - + - DISCUSSION In this paper we have approached the question of whether a 3-CM - prokaryotic repressor-operator interaction can regulate a plant promoter. We inserted the two operators of the TnlO- 1-CM - encoded tet operon into two selected sites of the CaMV 35S promoter. Placing the operators downstream of the tran- scription start site did not significantly reduce the promoter CM - 0 activity. This result might be considered surprising because the two palindromic operator sequences could lead to the formation of hairpin structures in the mRNA, thus affecting mRNA stability or translation initiation (31). These results 1 2 3 4 5 6 are in contrast to experiments in which hairpin structures generated upstream of the initiation codon by oligonucleo- FIG. 2. CAT activity in electroporated N. tabacum protoplasts tide insertions lead to a drastic reduction of translation. It as a function of tet repressor expression. In each electroporation was also unexpected that expression ofthis construct did not experiment, 20 ,g of pTET7, pTET8, and pTET14, respectively, change in the presence of the tet repressor. When the effect were coelectroporated with either 180 ,ug of pTET0 (lanes 1, 3, and 5) or 180 ,ug of pTET3 (lanes 2, 4, and 6). pTET3 contains the tetR of the lac repressor on transcription of a mammalian pro- gene between the CaMV 35S promoter and the nos poly(A) signal; moter was tested by Hu and Davidson (8), the location of the pTET0 is the same construct with the coding region in the inverse lac operators downstream of the transcription start site led orientation. CM, chloramphenicol; 1-CM, 1-acetyl chlorampheni- to a decrease in mRNA synthesis in the presence of the col; 3-CM, 3-acetyl chloramphenicol. repressor by a factor of 30, indicating that a transcribing Downloaded by guest on September 30, 2021 Biochemistry: Gatz and Quail Proc. Natl. Acad. Sci. USA 85 (1988) 1397 RNA polymerase II cannot pass the lac repressor protein We thank Dr. T. D. Sullivan for many fruitful discussions and for bound to the template. Also, in E. coli, location of one tet improving the manuscript, Dr. W. Hillen and K. H. Tovar for operator downstream of the strong bacteriophage promoter pWH305 and the 80-bp operator fragment, Drs. M. Fromm and V. Walbot for the CaMV 35S promoter and pCaMVNEO, Dr. G. An for PL inhibits gene expression in the presence of the tet pGA582, Dr. B. Sharrock for help with the electroporation experi- repressor (32). We do not know at present ifthe tet operators ments, Dr. R. Amasino for advice on plant tissue culture techniques, are not functional in this position because the tet repressor and W. Ealy for typing the manuscript. This work was supported by cannot interfere with elongation of RNA polymerase II or if National Institutes of Health Grant GM 36381. C.G. was supported the repressor cannot recognize the operator sequences in by the Alexander von Humboldt Stiftung. this position. Also, we cannot rule out that the additional 20 1. Ptashne, M. (1986) Nature (London) 322, 697-701. bp downstream of the operators in combination with the 42 2. Reznikoff, W. S. & Abelson, J. N. (1980) in The Operon, eds. bp of untranslated sequence (21) introduce an artificial Miller, J. H. & Reznikoff, W. S. (Cold Spring Harbor Lab., promoter. As the sequence does not indicate any possible Cold Spring Harbor, NY), pp. 221-243. TATA elements, we did not address this question by S1 3. Brand, A. H., Breeden, L., Abraham, J., Sternglanz, R. & nuclease mapping experiments. Nasmyth, K. (1985) Cell 41, 14-48. 4. Osborne, T. F., Goldstein, J. L. & Brown, M. S. (1985) Cell Changing the sequences between the CAAT box and the 42, 203-212. TATA box and the TATA box and the transcriptional start 5. Simpson, J., Schell, J., Van Montague, M. & Herrera-Estrella, site reduced promoter activity by a factor of only 2-3, which L. (1986) Nature (London) 323, 551-555. is analogous to the results obtained by linker scanning of the 6. Rogers, B. L. & Saunders, G. F. (1986) BioEssays 4, 62-64. thymidine kinase promoter (28). Apparently the tet repressor 7. Brent, R. & Ptashne, M. (1984) Nature (London) 312, 612-615. reduces transcription from this promoter by steric interfer- 8. Hu, M. C.-T. & Davidson, N. (1987) Cell 48, 555-566. 9. Jorgensen, R. A. & Reznikoff, W. S. (1979) J. Bacteriol. 138, ence with binding of RNA polymerase or some other factor 705-714. of the plant transcription machinery. These data imply that 10. Hillen, W., Klock, G., Kaffenberger, I., Wray, L. V. & the tet repressor is able to enter the nucleus and is also able Reznikoff, W. S. (1982) J. Biol. Chem. 257, 6605-6613. to recognize its operators in DNA, presumably packaged 11. Fluhr, R., Kuhlemeier, C., Nagy, F. & Chua, N. H. (1986) into chromatin (33). The repression (by a factor of 10) Science 232, 1106-1112. measured in the transient assay might not reflect the tight- 12. Schoffl, F. & Baumann, G. (1986) EMBO J. 4, 1119-1124. ness 13. Hagen, G. & Guilfoyle, T. J. (1985) Mol. Cell. Biol. 5, 1197- of this control mechanism, because the CAT protein 1203. might be synthesized before enough tet repressor is made to 14. Bertrand, K. P., Postle, K. P., Wray, L. V. & Reznikoff, provide full repression of the promoter. As in E. coli, the tet W. S. (1983) Gene 23, 149-156. repressor loses its specific DNA binding activity in plant 15. Hillen, W., Schollmeier, K. & Gatz, C. (1984) J. Mol. Biol. protoplasts in the presence of sublethal amounts of Tc. As 172, 185-201. compared to isopropyl f-D-thiogalactoside (IPTG), which 16. Bonner, W. M. (1978) in The Cell Nucleus 6, ed. Busch, H. induces the lac repressor-operator system in mammalian (Academic, New York), pp. 97-148. cells only to 60-80% (8), Tc is a more 17. Takahashi, M., Altschmied, L. & Hillen, W. (1986) J. Mol. efficient inducer. This Biol. 187, 341-348. is probably due to the fact that the equilibrium association 18. Fromm, M. E., Callis, J., Taylor, L. P. & Walbot, V. (1987) constant of the tet repressor-Tc complex is 1000-fold higher Methods Enzymol. 153, 351-366. than that determined for the -lac repressor com- 19. Hohn, T., Richards, K. & Lebeurier, G. (1982) Curr. Top. plex (17). In addition, the Tc might be taken up more easily Microbiol. Immunol. 96, 193-236. than IPTG in eukaryotic cells. In E. coli, the uptake of IPTG 20. Fromm, M. E., Taylor, L. P. & Walbot, V. (1986) Nature is facilitated by lac permease, which is not present in (London) 319, 791-793. eukaryotic cells. 21. An, G. (1987) Methods Enzymol. 153, 292-305. 22. Hillen, W. & Schollmeier, K. (1983) Nucleic Acids Res. 11, In conclusion, we have provided genetic evidence that the 525-539. tet repressor is able to regulate a plant promoter if the 23. Sanger, F., Nicklen, S. & Coulson, A. R. (1977) Proc. Natl. operators are suitably localized within the promoter. This Acad. Sci. USA 74, 5463-5467. system might be useful for characterizing a eukaryotic 24. Oehmichen, R., Klock, G., Altschmied, L. & Hillen, W. (1984) promoter in addition to and linker scanning analy- EMBO J. 3, 539-543. sis. Placing the operators in different positions between the 25. Odell, G. T., Nagy, F. & Chua, N. H. (1985) Nature (London) functional domains of a eukaryotic promoter might reveal 313, 810-812. positions that have to be left other 26. Ow, D. W., Jacobs, J. D. & Howell, S. H. (1987) Proc. Natl. unoccupied by proteins, Acad. Sci. USA 84, 4870-4874. to allow binding or interaction between transcription factors. 27. Kozak, M. (1978) Cell 15, 1105-1112. Moreover, the tet repressor-operator system is a very effi- 28. McKnight, S. L. & Kingsbury, R. (1982) Science 217, 316-324. cient on/off switch for the regulation of transferred genes. 29. Takahashi, K., Vigneron, M., Matthes, H., Wildeman, A., The high affinity of the repressor for the inducer allows Zenke, M. & Chambon, P. (1985) Nature (London) 313, efficient induction at sublethal concentrations of the antibi- 121-126. otic. Since Tc is taken up by and transported within the 30. Ecker, L. R. & Davis, R. W. (1986) Proc. Natl. Acad. Sci. plant (34), this system might be useful for turning genes on or USA 83, 5372-5376. off at defined developmental stages of the plant. This strat- 31. Kozak, M. (1986) Proc. Natl. Acad. Sci. USA 83, 2850-2854. egy could be used in mutant rescue to deter- 32. Unger, B., Becker, J. & Hillen, W. (1984) Gene 31, 103-108. experiments 33. Weintraub, H., Cheng, P. F. & Conrad, K. (1986) Cell 46, mine at which time during development a lesion has to be 115-122. complemented. Specific inhibition of gene expression by 34. Klein, M., Frederick, R. J. & Marahorosch, K. (1972) Phyto- inducing antisense RNA (35) might be helpful in studies on pathology 62, 111-115. the effect of a deficient in cases where perma- 35. Weintraub, H., Izant, J. G. & Harland, R. M. (1985) Trends nent mutants are lethal. Genet. 2, 22-25. Downloaded by guest on September 30, 2021