The transactivators tTA and rtTA are fusion proteins containing an N-terminal domain consisting of the TetR protein from transposon Tn10 and a C-terminal herpes simplex virus VP16-derived activation domain. The activation domains are identical in both proteins, whereas the TetR domains differ from each other by four amino acids (Fig. 1A). We have used viral evolution to optimize the function of the Tet-On system, and identified several amino acid substitutions in the rtTA protein (rtTA2^S-S2 6) that greatly enhance its transcriptional activity and dox-sensitivity 2426 (Fig. 1A). To test whether these amino acid substitutions could similarly improve tTA, we introduced the V9I, F67S, F86Y and G138D mutations and compared the activity of these variants with that of the wild-type tTA (tTA2^S 6). The tTA expression plasmids were transfected into HeLa X1/6 cells containing chromosomally integrated copies of the CMV-7tetO luciferase-reporter construct 28. Transfected cells were cultured at different dox concentrations for two days. We subsequently determined the intracellular luciferase level, which reflects tTA activity (Fig. 2). Wild-type tTA shows high transcriptional activity without dox, and its activity is gradually reduced with increasing dox concentrations. For example, the activity is reduced to 30% at 0.1 ng/ml dox, and 0.2% residual activity is observed at 2 ng/ml dox. All variants show a transcriptional activity similar to wild-type tTA in the absence of dox, but they respond differentially to the presence of dox. While the V9I and F67S variants are similarly inhibited by dox as wild-type tTA, the F86Y and G138D variants are less efficiently inhibited. The F86Y and G138D variants show more than 80% residual activity at 0.1 ng/ml dox, and 50% and 7% residual activity, respectively, at 2 ng/ml dox. These results demonstrate that mutations beneficial for rtTA do not necessarily improve tTA activity. In fact, the F86Y and G138D mutations reduce the dox-sensitivity of tTA.

In sc-rtTA, two TetR moieties are connected head to tail by a peptide linker, and a single activation domain is fused to the C-terminus (Fig. 1). The mutations that did improve rtTA activity are all positioned within the TetR part of the transactivator. To test whether these beneficial mutations can also improve the activity and dox-sensitivity of sc-rtTA, we introduced them into a single (N- or C-terminal) or both TetR moieties. The activity of these three sets of variants was analyzed in HeLa X1/6 cells and compared with the activity of the wild-type sc-rtTA (sc-rtTA2-S2) and the regular rtTA (rtTA2^S-S2) (Fig. 3). Both the wild-type sc-rtTA and the regular rtTA show no background activity without dox, and their activity gradually increases with increasing dox levels. However, the induced activity of sc-rtTA is lower than that of rtTA at all dox concentrations tested. For example, sc-rtTA is about 40-fold less active than rtTA at 1000 ng/ml dox. Introduction of the F86Y mutation in the N-terminal TetR domain increased sc-rtTA activity ~10-fold at all dox levels, but did not affect background activity (Fig. 3A). The additional introduction of the V9I mutation into the F86Y variant only marginally improved sc-rtTA activity, whereas the addition of the F67S, G138D, or V9I plus G138D mutations further improved sc-rtTA activity ~2-fold at all dox levels. The background activity of these variants was not increased.

Similar results were obtained upon introduction of the mutations in the C-terminal TetR part (Fig. 3B). However, none of these variants are as active as their counterparts with mutations in the N-terminal TetR moiety. The F86Y mutation increased sc-rtTA activity ~3-fold, and the addition of the F67S, G138D, or V9I plus G138D mutations further increased activity ~2-fold. These results demonstrate that the activity of sc-rtTA can be improved by mutations in either TetR moiety, but mutations in the N-terminal TetR domain have a larger impact on activity than the same mutations in the C-terminal domain.

Introduction of the mutations in both TetR parts resulted in the most active sc-rtTA variants (Fig. 3C). All these variants demonstrate a higher transcriptional activity than the corresponding variants with mutations in only one of the two TetR moieties. For instance, the sc-rtTA with the F86Y mutation in both TetR moieties is ~13-fold more active than wild-type sc-rtTA at 1000 ng/ml dox, whereas the same mutation in the N-terminal or in the C-terminal TetR domain increased sc-rtTA activity ~10-fold and ~3-fold, respectively. The variants containing the F67S, G138D, or V9I plus G138D mutations in addition to the F86Y mutation in both TetR moieties demonstrate an up to 30-fold increased activity at 1000 ng/ml dox. Moreover, these variants are particularly more sensitive to low dox levels. To compare the dox-sensitivity of the sc-rtTA variants, we calculated the dox concentration that each variant needs to reach an activity similar to that of the wild-type sc-rtTA at 1000 ng/ml dox (set at 100% in Fig. 4). The lower the required dox concentration, the more sensitive the variant is toward dox. For example, the variant carrying the F67S and F86Y mutations in both TetR moieties requires only 16 ng/ml dox to reach this 100% activity level, which reflects a 62-fold higher dox-sensitivity than the wild-type sc-rtTA. This variant thereby becomes the most dox-sensitive and most active sc-rtTA (30-fold more active than the wild-type, Fig. 4). In fact, both the transcriptional activity and dox-sensitivity of this sc-rtTA variant are similar to that of the original dimeric rtTA.

In the HeLa X1/6 cell line, the CMV-7tetO luciferase-reporter construct is stably integrated into the chromosome and shows no background activity in the absence of dox. However, when this reporter construct is transiently transfected into cells, multiple copies will be present, which may increase background activity and thus reduce inducibility (i.e. reduce the ratio between dox-induced and uninduced promoter activity). We therefore tested the new sc-rtTA variants with mutations in both TetR domains upon transient cotransfection of C33A cells with the CMV-7tetO luciferase-reporter construct (Fig. 5). As observed with HeLa X1/6 cells, the dox-induced activity of the new sc-rtTA variants in C33A cells is higher than that of the wild-type sc-rtTA, and the new variants are almost as active as the regular rtTA, while the background activity is not affected by the introduced mutations. However, this background promoter activity is relatively high (~7%) and as a consequence the inducibility is low with all rtTA and sc-rtTA variants (16–69 fold).

Since the background expression from tetO-controlled minimal promoters can be actively suppressed by tTS 89, we cotransfected C33A cells with a tTS expression plasmid (tTS^S 18). The presence of tTS indeed reduced the background activity more than 20-fold (~0.3% residual activity in the absence of dox; Fig. 5). Administration of dox inactivates tTS and activates the (sc-)rtTAs. The dox-induced activity of rtTA and the wild-type sc-rtTA was however ~5-fold reduced by the presence of tTS. This may be due to heterodimerization between the tTS and rtTA or sc-rtTA molecules 29. The dox-induced activity of the new sc-rtTA variants was not affected by the presence of tTS. The low background and high dox-induced activity observed when combining the new sc-rtTA variants with tTS results in a high inducibility (809–1193 fold), whereas combining the wild-type sc-rtTA or regular rtTA with tTS yields a much lower inducibility (99 and 273 fold, respectively). The highest inducibilty is obtained with the F67S/F86Y-mutated sc-rtTA variant, which was also the most active and dox-sensitive variant when tested without tTS.

The plasmid pCMV-tTA contains the tTA2^S gene cloned in the expression vector pUHD141-1/X 6. Mutations were introduced in pCMV-tTA by mutagenesis PCR 32 with the mutagenic primers (primer M) tTA-V9I (5'-GGACAAGAGCAAA

A

C

A

A

The plasmids pCMV-rtTA and pCMV-sc-rtTA contain the rtTA2^S-S2 and sc-rtTA2-S2 genes, respectively, cloned in the expression vector pUHD141-1/X 618. The sc-rtTA gene contains two TetR moieties and a single activation domain. To introduce mutations into the N-terminal tetR gene (with synthetic humanized codon usage), the EcoRI-BfuAI fragment of pCMV-sc-rtTA was replaced with the corresponding fragment of the appropriate pCMV-rtTA plasmid 26. Mutations were introduced into the C-terminal tetR gene (with original bacterial codon usage) by mutagenesis PCR 32 on pCMV-sc-rtTA with the mutagenic primers (primer M) sc-rtTA-V9I (5'-GGCTCTAGATCTCGTTTAGATAAAAGTAAA

A

C

C

A

AC

The activity of tTA, tTS, rtTA and sc-rtTA was assayed in C33A cells (ATCC HTB31) 33 and HeLa X1/6 cells 28, which are HeLa-derived cells containing chromosomally integrated copies of the CMV-7tetO luciferase reporter construct pUHC13-3 4. Cells were cultured at 37°C with 5% CO₂ in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, minimal essential medium nonessential amino acids, penicillin (100 U/ml), and streptomycin (100 μg/ml). Cells were grown in 2-cm² wells to 60% confluency and transfected with the pCMV-tTA, pCMV-tTS (pWHE122sB, 18), pCMV-rtTA or pCMV-sc-rtTA expression plasmids and the plasmid pRL-CMV (Promega) by the calcium phosphate precipitation method. pRL-CMV expresses Renilla luciferase from the CMV promoter and was used as an internal control to allow correction for differences in transfection efficiency. 1 μg of the DNA mixture in 15 μl water was mixed with 25 μl of 50 mM HEPES (pH 7.1)-250 mM NaCl-1.5 mM Na₂HPO₄ and 10 μl of 0.6 M CaCl₂, incubated at room temperature for 20 min, and added to the culture medium. The DNA mixture contained 8 ng pCMV-tTA and 2.5 ng pRL-CMV for the tTA activity assay, 20 ng pCMV-sc-rtTA or pCMV-rtTA and 2 ng pRL-CMV for the sc-rtTA or rtTA activity assay in HeLa X1/6 cells, and 5 ng pCMV-sc-rtTA or pCMV-rtTA, 20 ng pCMV-7tetO-luciferase (pUHC13-3) 4, 0.5 ng pRL-CMV and 200 ng pCMV-tTS (when indicated) for the sc-rtTA or rtTA activity assay in C33A cells. The DNA mixture was completed to 1 μg with pBluescript as carrier DNA. Cells were cultured after transfection for 48 hours at different dox (D-9891, Sigma) concentrations and then lysed in Passive Lysis Buffer (Promega). Firefly and Renilla luciferase activities were determined with the Dual-Luciferase Reporter Assay (Promega). The expression of firefly and Renilla luciferase was within the linear range and no squelching effects were observed. The activity of the transactivators was calculated as the ratio of the firefly and Renilla luciferase activities, and corrected for between-session variation as described 34.