An episomal vector for stable tetracycline-regulated gene expression
An episomal vector for stable tetracycline-regulated gene expression Monika Jost, Csaba Kari and Ulrich Rodeck*
The Wistar Institute of Anatomy and Biology, 3601 Spruce Street, Philadelphia, PA 19104, USA
Received March 12, 1997;Revised and Accepted May 5, 1997
ABSTRACT
The recently introduced tetracycline (Tc)-regulatable eukaryotic gene expression system based on the Escherichia coli Tn10 tetracycline operon has proven to be a powerful tool for controlled expression of a variety of genes in vitro as well as in vivo. Control elements of this expression system are contained in two separate plasmid vectors. The tTA vector encodes a transactivator protein and the tetP vector contains a responsive operator-promoter element (tetP) that controls gene expression depending on tTA binding. Establishment of cell lines expressing a gene of interest under tetP control requires two subsequent rounds of transfection and clonal selection after each transfection. Here we describe a modification of this system in which the tetP element is placed in an episomal EBNA-based plasmid that can be stably maintained in primate but not in rodent cells. Using HeLa and human melanoma cells, we show that upon transient or stable transfection a reporter gene is expressed in a Tc-regulated manner similar to the original system. Thus, this expression system combines the advantages of episomal vectors, such as high efficiency of transfection and time-efficient selection of mass cultures, with tight control of gene expression provided by the Tc-regulatable system.
INTRODUCTION
The tetracycline (Tc)-regulatable eukaryotic gene expression system, as originally developed by Gossen and Bujard (1 ), has been shown to be a versatile tool for controlled gene expression in vitro as well as in vivo (4 -9 ). It is composed of two plasmids (1 ). One plasmid (ptTA) encodes a fusion protein of the sequence-specific DNA binding tetracycline repressor (TetR) and the C-terminal domain of the herpes simplex virus VP16 transactivator. The second plasmid (pTetP) contains seven copies of the Escherichia coli Tn10 tetracycline operator (tetO) contiguous with a CMV-IE minimal promoter. In the presence of tetracycline at subtoxic levels, tTA cannot bind to tetO sequences and the gene of interest under control of the CMV-IE promoter is transcribed at low to non-detectable levels. In the absence of tetracycline, however, the TetR portion of the protein encoded by tTA enables high affinity binding to the tetO sites, positioning VP16 to activate the CMV-IE promoter. In order to generate stable cell lines with controlled gene expression two labor-intensive steps are required: (i) introduction of the transactivator plasmid ptTA followed by functional testing of clonally selected cells using transient transfection with pTetP containing a reporter gene; (ii) a second round of transfection with the pTetP plasmid containing the gene of interest, followed again by clonal selection and functional testing of individual clones for regulated gene expression (4 ).
To shorten the time-consuming process of clonal selection following transfection with the pTetP plasmid, we explored whether an episomal vector containing the tetO/CMV-IE promoter (TetP) could be substituted for the pTetP plasmid vector originally used. For this purpose we used pCEP4, carrying OriP [Epstein Barr virus (EBV) replication origin] as well as the gene encoding EBNA-1 (EBV nuclear antigen 1), both of which control episomal replication of the plasmid in most primate cells (3 ). A derivative of this vector (pCEP4-Luc) that contains the luciferase reporter gene was modified by replacing the CMV promoter with the tetO/CMV-IE promoter. The resulting plasmid, pCEPTetP-Luc, was then transfected into two tTA-expressing clones derived from human melanoma cell line WM793 established in our laboratory, as well as into tTA-expressing epithelial cells (HeLa) originally established by Gossen and Bujard (1 ). After selection cells were maintained as mass cultures and tested for their ability to transiently and stably express the reporter luciferase gene in a Tc-dependent fashion.
MATERIALS AND METHODS
Construction of plasmids ptTA-neo and pCEPTetP
The plasmids of the original Tc-regulatable expression system, pUHD15-1 (ptTA), pUHD10-3 (pTetP) and pUHC13-3 (pTetP-Luc), were provided by Dr H.Bujard (1 ). The ptTA-neo plasmid was derived by replacing the PvuI-BamHI fragment of pUHD15-1 with the PvuI-BamHI fragment from MT-CB6+ carrying a neomycin resistance expression cassette (10 ). Thus, ptTA-neo contains the gene for tTA-VP16 protein driven by the hCMV promoter followed by the human growth hormone poly(A)+ signal and the neomycin resistance gene driven by the SV40 promoter and followed by the SV40 poly(A)+ signal. To generate a plasmid (pCEPTetP-Luc) that expressed the luciferase gene under control of the tetP hybrid promoter, the CMV-IE enhancer promoter of pCEP4-Luc, a derivative of pCEP4 (Invitrogen, San Diego, CA) containing the luciferase gene, was replaced by tetP derived from pTetP-hygro, a derivative of pUHD10-3 (Fig. 1 ). This plasmid also contains the hygromycin resistance gene under separate promoter control. To obtain the control plasmid pCEPTetP, the luciferase gene was removed.
Transfections and luciferase assays
The primary melanoma cell line WM793 was maintained as described (11 ). HeLa cells stably expressing tTA (1 ) were obtained from Clontech (Palo Alto, CA) and maintained according to the manufacturer's instructions. Transfections were performed using either the calcium phosphate technique (20 [mu]g total DNA/10 cm dish) or by using Lipofectamine (Gibco BRL, Gaithersburg, MD) with 1.5 [mu]g DNA/35 mm dish following the manufacturer's instructions. To generate stable WM793 clones expressing tTA, cells transfected with ptTA-neo were selected with G418 (400 [mu]g/ml; Gibco BRL). Clones were isolated using cloning cylinders and tested for their ability to activate transcription from tetP by transient transfection with pUHC13-3 in the presence and absence of tetracycline (1 [mu]g/ml; Sigma, St Louis, MO), followed by determination of luciferase activity.
For luciferase assays cells were harvested either by trypsinization, followed by two washes with ice-cold PBS, and lysis in 1* lysis buffer (Analytical Luminescence Labs, Ann Arbor, MI) for 20 min at 4oC or by direct lysis in tissue culture dishes. Luciferase activity was determined using a Luminometer Monolight 2010, series 1036 (Analytical Luminescence Labs). Protein concentrations of cell lysates were determined using the micro-assay method (BioRad, Hercules, CA or Pierce, Rockford, IL) following the protocols provided. To control for transfection efficiency of transient transfections, a plasmid carrying LacZ [pAdCMV-LacZ (12 )] was used. [beta]-Galactosidase activity in co-transfected cells was quantitatively assessed using the Galacto Light Assay (Tropix, Bedford, MA) according to the manufacturer's protocol. Stable, double transfectants of WM793 were selected and maintained in growth medium containing 200 [mu]g/ml G418, 100 [mu]g/ml hygromycin B and 2 [mu]g/ml tetracycline.
RESULTS
Tc-regulated expression of luciferase using the episomal pCEPTetP vector in transiently transfected cells
To compare Tc-dependent regulation of gene expression of the new episomal construct (pCEPTetP) with the original Tc-regulatable construct pTetP, we performed transient expression experiments. First, we co-transfected ptTA with either pCEPTetP-Luc or pTetP-Luc into the WM793 melanoma cell line. Second, a tTA-expressing clone 793-tTA6, derived from WM793, was transiently transfected with either pCEPTetP-Luc or pTetP-Luc. In both cell lines tested, luciferase expression was mediated by the transactivator tTA in a Tc-dependent manner. Tetracycline at a concentration of 1 [mu]g/ml suppressed expression almost completely. The level of activation, measured as luciferase activity in the absence of tetracycline, was 150-350-fold and slightly higher in both cell lines using the episomal vector when compared with the original pTetP plasmid (Table 1 ). This was probably not due to a higher plasmid copy number per cell since the same amount of pCEPTetP-Luc plasmid corresponds to only ~50% of the number of pTetP-Luc molecules.
. Tetracycline-dependent luciferase expression in melanoma cells transiently transfected with pCEPTetP-Luc or pTetP-Luc
Cell line
Plasmid
Luciferase-activity (normalized RLU)
Activation factor
With Tc
Without Tc
WM 793
pCEPTetP
0.7
0.9
1.3
pTetP-Luc
2.1
5.0
2.4
pCEPTetP-Luc
4.2
5.0
1.2
ptTA+pTetP-Luc
18.0
4900
270
ptTA+pCEPTetP-Luc
15.0
5200
350
793-tTA6
pCEPTetP
0.6
0.7
1.1
pTetP-Luc
6.8
1000
150
pCEPTetP-Luc
4.7
1600
340
Tetracycline-dependent luciferase expression in melanoma cells transiently transfected with pCEPTetP-Luc or pTetP-Luc. Cells were transfected in 10 cm dishes with calcium phosphate-precipitated mixtures of: 5 [mu]g expression plasmid (pTetP-Luc = pUHC13-3, pCEPTetP-Luc or pCEPTetP as mock vector), 5 [mu]g control plasmid (pADCMV-LacZ), 10 [mu]g effector plasmid (ptTA = pUHD15-1) or 10 [mu]g non-specific plasmid (Puc19) per plate. After 6 h the medium was changed to normal growth medium in the presence (1 [mu]g/ml) or absence of tetracycline. Cells were harvested 36 h after start and analyzed as described. To account for differences in transfection efficiency, luciferase activity was normalized to [beta]-galactosidase activity. Values given are relative light units (RLU) of luciferase activity per RLU [beta]-galactosidase activity/[mu]g protein * 1000.
. Tetracycline-dependent luciferase expression in melanoma and HeLa cells stably transfected with pCEPTetP-Luc
Cell line
Plasmid
Luciferase-activity (RLU/[mu]g protein)
Activation factor
With Tc
Without Tc
793-tTA6
pCEPTetP
22 +- 6
21 +- 2
0.9
pCEPTetP-Luc
3110 +- 964
302 581 +- 10 221
97
793-TAN4
pCEPTetP
18 +- 3
24 +- 2
1.3
pCEPTetP-Luc
2772 +- 637
762 128 +- 123 616
275
WM793
pCEPTetP
41 +- 21
32 +- 3
0.8
pCEPTetP-Luc
880 +- 314
1096 +- 146
1.2
HeLa-tTA
pCEPTetP
300 +- 53
239 +- 21
0.8
pCEPTetP-Luc
1488 +- 91
1 927 597 +- 313 506
1300
HeLa
pCEPTetP
175 +- 17
165 +- 14
0.9
pCEPTetP-Luc
2809 +- 1461
2580 +- 1240
0.9
Tetracycline-dependent luciferase expression in cells stably transfected with pCEPTetP vectors. Cells were seeded in triplicate in 24-well plates and incubated in the presence (1 [mu]g/ml) or absence of tetracycline until reaching 90% confluency. After cell lysis luciferase activity was determined as described and normalized to total cellular protein content. Data given are the mean and standard deviation of triplicates.
Regulation of luciferase expression in stably transfected mass cultures
Next, we generated stable double transfectants by transfecting two different WM793-derived tTA-expressing clones, 793-tTA6 and 793-TAN4, with pCEPTetP-Luc and the mock vector pCEPTetP as control, followed by selection in mass culture for hygromycin resistance. To account for background transcription levels of pCEPTetP-Luc we also transfected the parental WM793 with pCEPTetP-Luc and pCEPTetP and selected in the same manner. The resulting cell lines were seeded in the presence and absence of tetracycline (1 [mu]g/ml) and, after reaching ~90% confluence, luciferase activity was determined. As shown in Table 2 , luciferase expression was activated only in the absence of tetracycline in the two WM793 clones expressing tTA but not in parental WM793 cells. The levels of induced luciferase expression were of the same order of magnitude, i.e. 100-200-fold, as observed for transient transfection (see Table 1 ).
Similar results were obtained when tTA-expressing HeLa cells and parental HeLa cells were transfected with pCEPTetP-Luc and control pCEPTetP (Table 2 ). The levels of induced luciferase expression in HeLa cells for both types of vectors were at least one order of magnitude higher than those observed for the melanoma cell lines. Since the amplitude of luciferase expression is likely to reflect the amount of ptTA molecules expressed per cell, we assume that the melanoma cell lines used in this study express tTA at a comparatively lower level. Several tTA-expressing cell lines derived from the WM793 melanoma cell line consistently showed lower luciferase expression levels than HeLa cells when transiently transfected with pTetP-Luc (data not shown). These results are consistent with the conclusion that expression of higher tTA levels might be toxic to WM793 cells.
Effective inhibitory tetracycline concentration
In order to determine the effective inhibitory tetracycline concentration for the episomal vector, we seeded 793-tTA6 and 793-TAN4 cells at different tetracycline concentrations ranging from 10 pg/ml to 10 [mu]g/ml (Fig. 2 ). Inhibition of luciferase expression was detectable at tetracycline concentrations as low as 1 ng/ml and was maximal at 0.1 [mu]g/ml. These results were quantitatively comparable with those described for the original Tc-regulatable system in stable double transfected HeLa cells (1 ).
DISCUSSION
This study demonstrates that an EBNA-based episomal vector can be used for Tc-regulatable stable expression of transgenes in human melanoma and HeLa cells. A previously described episomal vector system carrying both the tTA gene and the tetP element on one plasmid proved to be poorly regulatable with tetracycline (15 ). Although high expression of a reporter gene was observed in the absence of tetracycline, gene expression could be reduced by only ~75% in the presence of 1 [mu]g/ml tetracycline. To obtain more efficient repression of gene expression, higher concentrations of tetracycline (10 [mu]g/ml) were required, which are potentially toxic. In contrast, the expression system described here provided tighter (>99%) regulation at 1 [mu]g/ml tetracycline. As long as stably transfected cells were maintained in the presence of hygromycin, regulated expression of the reporter gene remained unchanged over a period of 3 months.
This modified Tc-regulatable expression system has several advantages over the original system. Most importantly, stably transfected cells can be selected with relative ease while reproducing the high efficiency of induced gene expression (two to three orders of magnitude) described for the original system. Specifically, selection and expansion of mass cultures with Tc-regulatable gene expression was complete within 3-4 weeks after transfection of both cell types used in this study with pCEPTetP vectors. In contrast, clonal selection and expansion using derivatives of the original pUHC13-3 plasmid typically required at least 6 weeks, followed by screening of multiple clones. The comparably short time required to obtain stable cell lines after pCEPTetP transfection is likely to reflect the high efficiency of transfection of episomal vectors as compared with plasmids which require chromosomal integration for stable maintenance. This is most probably due to the nuclear retaining and replication promoting capability of EBNA-1 attached to OriP (3 ). A major advantage of the episomal system over the recently described Tc-regulatable retroviral expression system (16 ) is that the episomal system has no size limitation for the gene of interest to be expressed. This is particularly important if future modifications such as bidirectional Tc-regulatable promoter use are considered.
The episomal expression system also provides the opportunity for further modifications, including those published recently (17 -19 ). Furthermore, the ongoing development of tTA expressing cell lines in different laboratories will facilitate application of this episomal Tc-regulated expression system in different experimental systems. In this context it should be noted that use of this expression system is restricted to primate cells.
ACKNOWLEDGEMENTS
We are very grateful to Dr H.Bujard for providing plasmids pUHD15-1, pUHD10-3 and pUHC13-3. We also thank Dr P.Curtis for pCEP4-Luc, Dr G.Prendergast for pTetP-hygro, Dr J.Wilson for pAdDCMV-LacZ and Dr H.Riethman for a critical review of the paper. This work was supported by the National Institute of Health (CA25874-16).
REFERENCES
1 Gossen,M. and Bujard,H. (1992) Proc. Natl. Acad. Sci. USA, 89, 5547-5551.MEDLINE Abstract
