| Nucleic Acids Research | Pages |
Use of the human EF-1[alpha] promoter for expression can significantly increase success in establishing stable cell lines with consistent expression: a study using the tetracycline-inducible system in human cancer cells
Introduction
Materials And Methods
Construction of plasmids
Cell culture and derivation of stable cell lines
RNA blot and luciferase activity analysis
Nude mouse tumorigenesis studies
Results
Construction and initial testing of the EF-1[alpha] promoter-based rtTA expression vector
Analysis of stable cell lines expressing the rtTA cDNA under regulation of the EF-1[alpha] promoter
Functional analysis of stable clones expressing cDNAs under inducible regulation of Tc using the EF1prtTA construct
Determination of temporal stability of clones expressing cDNAs under inducible regulation of EF1prtTA
Discussion
Acknowledgements
References
Use of the human EF-1[alpha] promoter for expression can significantly increase success in establishing stable cell lines with consistent expression: a study using the tetracycline-inducible system in human cancer cells
Received August 2, 1999; Revised and Accepted October 27, 1999
ABSTRACT Establishing cells with an exogenously introduced gene of interest under the inducible control of tetracycline (Tc) initially requires clonal cell lines stably expressing the tetracycline activator (tTA or rtTA). The originally described plasmid vectors expressing tTA/rtTA are driven by the cytomegalovirus (CMV) immediate early (IE) promoter-enhancer, known for its robust activity in a wide spectrum of cell types. While many reports testify to the utility and efficacy of this construct, instances of inexplicable failure to establish cell lines having inducible expression of the cDNA under study are encountered. Spontaneous extinction of CMV promoter activity in cells has been observed in a temporal and cell type-dependent manner. This could be a contributing factor in the failure to establish Tc-responsive cell lines. We here report that a change of the expression cassette to the human elongation factor-1[alpha] (EF-1[alpha]) promoter has permitted successful establishment of several inducible cell lines from diverse human tumor tissue origins. We interpret these results to imply that extinction of rtTA (or tTA) expression might be a significant factor in the lack of success in establishing Tc-inducible cell lines. Moreover, the present findings have general relevance to experiments requiring the use of stable cell lines.
INTRODUCTION
Since the first report by Gossen and Bujard (1) and subsequent documentation of a variant form (2), the tetracycline (Tc)-regulated system has been broadly adopted and is widely acknowledged as the method of choice in experiments requiring inducible expression of genes of interest. In its originally reported form, the system employs two plasmids. One plasmid expressing the tTA or rtTA cDNA (jointly referred to as TA), a fusion protein of the bacterial Tc repressor fused to the C-terminal acidic activation domain of the Herpes simplex virus (HSV) VP16 transcriptional transactivator. The second plasmid enables cloning of a cDNA of interest downstream of a synthetic heptamerized Tc operator transcription regulatory DNA sequence. This sequence is fused to a DNA element providing basal promoter activity in mammalian cells derived either from the CMV IE or HSV thymidine kinase promoters. Establishing a cell line having Tc-regulatable expression of the gene of interest involves a two-step process. In the first, a cell line stably expressing the TA cDNA is established and identified by clonal selection and expression analysis through transient transfection with a Tc-responsive reporter. In the second step, the gene of interest cloned under control of the Tc-responsive element is introduced into the cell line made in the previous step and a second round of selection is performed to identify clones displaying Tc-responsive inducibility of the cDNA (1,2). The Tc-regulated system has effectively overcome several drawbacks seen in earlier systems which showed high basal levels of expression, poor responsiveness and toxicity of the inducing agent. In addition, the Tc-inducible system is able to achieve induction over a range of several orders of magnitude in a graded manner responsive to varying levels of inducer. Furthermore, the system is extremely versatile and amenable to several types of modifications permitting the study of the role of a particular gene, or combinations thereof, in a wide variety of cell types of interest. The potential to use this system in medical applications including gene therapy protocols and pharmacological small molecule screening are areas of active investigation. Its versatility has enabled adaptation to situations requiring inducible gene expression in a tissue-specific or generalized manner in animal or plant models, thus opening up new avenues to study gene function in vivo.
The Tc-inducible expression system has been modified in several ways, in attempts to improve performance or tailor it to specific needs. Autoregulatory control was achieved by placing both the tTA and exogenous cDNA under control of the Tc-operator sequence (3), which reportedly permitted regulation of available tTA levels only on induction and thereby increased overall performance in terms of inducibility and frequency of positive clones obtained. Single plasmid vectors containing the tTA sequence and the gene of interest in opposite orientations have been developed to obviate the need for multiple rounds of clonal selection (4-6). The development of retroviral vectors for delivery of both components of the system either in a single or in a combination of two separate viruses has overcome the barrier of introduction of DNA into transfection recalcitrant cells (7-11). Several promoters have been used to enable generalized or tissue-specific expression of tTA in plants (5) or animals (12-15). Modification of the Tc operator-containing plasmid to reduce leaky expression or reduce the effects of the integration site has been attempted. Strategies toward this end include using Epstein-Barr virus (EBV) replication origin-based vectors that are maintained episomally (16), modified basal promoters to reduce uninduced expression (17) and incorporation of sequences that prevent interference from adjoining elements at the site of integration (14,18,19).
The original report and several other studies have documented potential pitfalls and have provided troubleshooting strategies using the Tc-regulated system (reviewed in 20-22). However, anecdotal evidence or documented reports of failure to establish inducible cell lines do exist (23,24). At least one specific report showing a high degree of variability of expression of the target gene using the CMV IE promoter exists (15). Analysis of expression of the gene of interest was highly variable even in the same cell type of a given tissue, based on histochemical analysis (15). Our own initial attempts at using the classical Tc-regulated system failed in establishing inducible cell lines in several cell types, principally the HO-1 human melanoma model system. Drawing upon previous experiences using expression constructs with strong viral promoters based on CMV or simian virus 40 (SV40) derived sequences, we surmised that extinction of expression of the transactivator function could be a potentially significant factor in the inability to establish Tc-responsive cell lines. This might be of special relevance in cells having a relatively slow growth rate and/or the potential to differentiate, making them particularly sensitive to this phenomenon, since changes in cell physiology could affect the activity of exogenously introduced promoter constructs. Often the time involved in identifying and selecting the initial TA-expressing clone is fairly extensive. It is during this time period that the host cell possibly stops supporting CMV (or other viral) promoter enhancer expression, resulting in the shutdown of TA expression, making attempts to identify inducible clones impossible. Despite the recent introduction of retroviral vectors that enable single step and therefore relatively quick selection of positive clones, several of these also depend on viral promoters for expression of one or more elements and are also prone to similar problems in the long term. The construction of a specific retrovirus is in itself time consuming and is far from a routine procedure in many laboratories when compared to transfection or electroporation of plasmid DNA into cells. Based on these factors we decided to modify the existing construct for rtTA cDNA expression by placing it under control of the promoter regulating the human EF-1[alpha] gene. This gene has a housekeeping function in all cells and is expressed to high levels. Most importantly, due to its indispensable housekeeping function in all cells, EF-1[alpha] promoter expression is consistent from a temporal viewpoint, relatively insulated from changes in cell physiology and is cell type independent (25-27). Utilization of this construct in cells derived from diverse human cancers enabled the successful construction of Tc-regulatable lines in every case attempted so far. We believe this modified vector will not only be of general utility, but will be especially useful in cases where difficulties have been previously experienced in successfully establishing Tc-responsive clones.
MATERIALS AND METHODS
Construction of plasmids
An EcoRI-BamHI fragment containing the rtTA open reading frame was isolated from pUHD 17-1neo (2). This fragment was cloned directionally into the mammalian expression vector pCDEF3 (25) at the 5[prime] EcoRI and 3[prime] XbaI sites of the vector multiple cloning site to generate the final construct, termed EF1prtTA. Ligation of the 3[prime] XbaI site of pCDEF3 and the BamHI site of the fragment was possible after Klenow filling the overhangs to make them blunt-ended. This modified vector places the rtTA gene under direct transcriptional control of the human EF-1[alpha] promoter. Plasmids expressing the Mda-7 and JunB cDNAs were constructed in pUHD 10-3 (1) by blunt cloning of isolated cDNA fragments into Klenow filled blunt vector followed by sequence analysis for confirmation.
Cell culture and derivation of stable cell lines
All cell lines used in this study were grown and maintained under standard conditions as previously described (28). Selection of stable clones expressing the rtTA cDNA using EF1prtTA was carried out in the presence of 500-1000 µg/ml G418 (Life Technologies Inc.) depending on the individual cell line. After the selection period, macroscopic visible colonies were picked, expanded and analyzed for activity by assaying for luciferase activity, for rtTA expression or by northern blot analysis of inducible cDNA, such as Mda-7 or JunB, respectively.
RNA blot and luciferase activity analysis
Total cellular RNA was resolved by denaturing formaldehyde-agarose gel electrophoresis after isolation of RNA using a RNAeasy Kit (Qiagen). Transfer was onto Hybond nylon membranes (Amersham) which were probed with appropriately labeled cDNA probes for Mda-7 and JunB. Luciferase assays were performed using a Luciferase Assay Kit (Promega) and quantitation was performed on a Turner Design TD 20/20 luminometer. Equal quantities of RNA were loaded on each gel following spectrophotometric estimation at 260 nm. Normalization of RNA levels between samples was confirmed by visualizing RNA on ethidium bromide stained gels. Normalization of luciferase activity was achieved by quantitating protein and adjusting the amount of extract to a fixed amount of protein.
Nude mouse tumorigenesis studies
HeLa cells expressing the cDNA under Tc regulation were injected subcutaneously (1 × 106 cells/animal, suspended in sterile phosphate-buffered saline in 400 µl final volume) into nude mice (Taconic Farms). Four weeks after injection, animals were sacrificed and parts of the tumors explanted under aseptic conditions, the remainder being used to isolate RNA as described above. Tumor tissue was minced and pieces plated for growth and maintained under standard conditions as previously described (28). Under these conditions only the originally injected HeLa cells survived after culture for ~14 days. These cells were used for subsequent analysis of doxycycline (Dox)-inducible expression of Mda-7.
RESULTS
Construction and initial testing of the EF-1[alpha] promoter-based rtTA expression vector
Details of the cloning steps for construction of the EF-1[alpha] promoter rtTA (EF1prtTA) expression vector are described in Materials and Methods and Figure 1. The protein expressed by this cDNA, a mutant form of the original bacterial Tc repressor (2), binds to and activates transcription of genes downstream of Tc operator binding elements only when Tc is present. EF1prtTA was transiently co-transfected with the Tc-responsive luciferase reporter plasmid, pUHC 13-3 (2), into HO-1 human melanoma cells to determine if the construct was active. A parallel set of transfections was performed with the original CMV IE-based construct, pUHD 17-1neo (2), in the absence or presence of 1 µg/ml Dox. Cells were harvested 48 h after transfection and luciferase activity was determined (Fig. 2) using a luminometric luciferase assay system (Promega). The initial experiments clearly demonstrated that the EF1prtTA expression vector was functional at comparable levels to the original pUHD 17-1neo construct in transient assays. Based on the positive activity obtained, the EF1prtTA construct was utilized to establish stable lines expressing rtTA in HeLa (human cervical carcinoma), HO-1 (human melanoma), MCF-7 (human breast carcinoma) and PC3 and DU-145 (human prostate carcinoma) cell lines.
Figure 1. Plasmid map of the EF1prtTA expression construct. The map shows individual component elements of the vector including the rtTA ORF, human EF-1[alpha] promoter, bovine growth hormone (BGH) polyadenylation [poly(A)] signal and partial multiple cloning site retained from the vector pCDEF3 (25) after cloning. The neomycin resistance marker (NeoR) flanked by the SV40 promoter and poly(A) signal, Ampicillin resistance marker (AmpR) for bacterial propagation and selection and some reference restriction sites are also shown.
Figure 2. Luciferase assay to test activity of the EF1prtTA vector. Extracts from human HO-1 melanoma cells transiently co-transfected with the original (pUHD 17-1neo) or modified (EF1p Tet on) rtTA expression vectors and the Tc luciferase reporter pUHC 13-3 were quantitated for luciferase activity. These extracts were prepared from cells treated without (-Dox) or with (+Dox) the inducer. Treatment with inducer was for 48 h as described in Materials and Methods.
Analysis of stable cell lines expressing the rtTA cDNA under regulation of the EF-1[alpha] promoter
Cells were transfected with the EF1prtTA construct using Superfect transfection reagent (Qiagen) based on standard conditions recommended in the manufacturer's usage protocol. The efficiency of transfection, reflected in the number of clones obtained at the end of the selection period, varied with each cell line. Colonies were selected using neomycin resistance conferred by the marker present within the construct. For every cell line, 24 neomycin-resistant colonies were isolated for further analysis. These individually selected clones were transiently transfected with the Tc-responsive luciferase reporter pUHC 13-3 (2) to determine the presence and level of rtTA activity. It is emphasized here that several previous attempts to establish Tc- or ecdysone-inducible cell lines in HO-1 cells had failed (Table 1). In these cases inducible expression was only obtained in transient assays and no stable lines could be established from antibiotic-selected clonal populations of cells. We show no data for essentially negative results but the experience of collaborators with the original construct utilizing the CMV IE-based pUHD 17-1neo (2) and our modified vector have independently corroborated the findings reported here (unpublished results; L.Ossowski and S.Grant, personal communication).
Table 1. Comparative efficiency of expression-positive stable clones obtained using CMV versus EF-l[alpha] based expression systems
| Construct | Promoter | No. of expression-positive stable clones/total analyzed | No. of independent transfections |
| pUHD17-1 Neo | CMV | 0/96a | 3 |
| EGFPC2 (Clontech) | CMV | 0/48a,b | 1 |
| pVgRXR (Invitrogen) | CMV + SV40 | 0/96a,c | 2 |
| pEF1/HisA with different cDNAs (Invitrogen) | EF-1[alpha] | 5/24 | 4 |
| pCDEF3 with different cDNAs | EF1[alpha] | 6/24 | 2 |
| pEF1prtTA | EF-1[alpha] | 25/96 | 4 |
bWeak expression observed after initial clonal selection was lost on expansion of the colony as a function of time, based on fluorescence microscopy and western blotting.
clntegration of the construct into genomic DNA was confirmed by Southern analysis.
Results obtained in a screen to identify Tc-responsive clonal cell lines in HO-1, MCF-7, PC3 and DU-145 cells (Fig. 3) indicated that an average of at least two isolates of 24 showed Tc-responsiveness. This frequency of positive clones ([le]1:12) is comparable to that previously reported (1,2). Transient assays poorly reflect the level of inducibility actually obtainable after final selection of stable clones, since basal levels of expression change dramatically once plasmid DNA is integrated into chromatin (1,2,20). Therefore, the fold induction observed in the presence of Tc, though relatively low, is likely to be a reflection of uninduced leaky expression due to the transient transfection conditions used in this initial screen. Despite this leakiness, clones with high or low relative levels of inducibility were identified in every case and potentially usable cell lines were selected with relative ease at frequencies described previously (1,2). As described below, the level of inducibility (average 5-fold) in transient conditions ultimately results in very effective TA-expressing cell lines when used to stably express exogenous cDNAs, validating the relevance of a 5-fold inducibility cut-off value as an acceptable level for further studies. This screening assay has enabled us to rapidly identify suitable clones expressing the rtTA cDNA from a small number of cells in a reproducible manner with relatively high throughput.
Figure 3. Luciferase assay to select Tc-inducible clones. The panels show quantitation of luciferase assays from individual neomycin-resistant clonally isolated cell lines of human prostate (DU-145 and PC3), cervical (HeLa), breast (MCF-7) and melanoma (HO-1) tumor origin. Each stable clone was transiently transfected with Tc luciferase reporter pUHC 13-3 in the absence (-Dox) or presence (+Dox) of inducer. Extracts prepared from these cells were assayed for luciferase activity to identify clones showing adequate levels of inducibility for each cell type as described in Material and Methods.
Functional analysis of stable clones expressing cDNAs under inducible regulation of Tc using the EF1prtTA construct
A major focus of our research involves the identification and analysis of the role of specific known and novel genes in melanoma differentiation (29-32). To further this aim, stable cells expressing differentiation-associated genes including the transcription factor JunB (33 and references therein) and the cancer growth suppressing gene Mda-7 (28-30) under Tc regulation were established in HO-1 melanoma cells. This human melanoma cell line has the ability to terminally differentiate in the combined presence of fibroblast interferon (interferon-[beta]) and the protein kinase C (PKC) activator mezerein (29). These cells are sensitive to culture conditions due to their capacity to differentiate and are difficult to transfect. They also take a relatively long time during selection (35-45 days) to form visible colonies suitable for clonal isolation. HO-1 therefore presents an ideal experimental model in which to test the efficacy of the EF-1[alpha] promoter-based vector. Using a suitable rtTA-expressing cell line identified in the previous screen described above (Fig. 3, HO-1 panel, clone 2), transfections were performed with Tc-regulatable JunB and Mda-7 cDNAs cloned into the vector pUHD 10-3 (1) in the sense orientation. Colonies were isolated and individual clones were analyzed for expression and induction of JunB and Mda-7 by northern blotting. To determine the level of inducibility of individual clones, RNAs were isolated from each clone grown in the absence or presence of Dox. Northern blots, probed with JunB and Mda-7 cDNA probes (Fig. 4) indicated that several positive clones had been obtained for each cDNA. As anticipated, varying degrees of clone-dependent (i.e. genomic integration site-dependent) basal and inducible levels of expression were observed, as summarized in Table 2. It may be noted that the parental EF1prtTA cell line chosen from the initial screen (Fig 3, HO-1 panel, clone 2) had not exhibited a very high level of fold inducibility in transient assays. However, upon introduction of a Tc operator-regulatable construct in stably integrated form, high levels of Tc-dependent induction were observed in individual clones (Fig. 4A, compare lanes 1, 3, 9, 10, 14, 16 and 17, induced and uninduced level, and similarly Fig. 4B, lanes 4, 6 and 9; summarized in Table 2). Overall, the frequency, variability and basal to induced levels obtained in various clones can be attributed to integration site dependence, as reported in the original Tc system (1,2,20). Quantitation by densitometric scanning of fold activation observed in the stable clones gave values ranging between 0 and >100-fold (Table 2). Therefore, in addition to the relative enhanced efficiency in establishing stable expression lines, the modification performed at comparable levels to CMV-based vectors. As described below, the expression levels were maintained over an extended period of time with no apparent loss of inducibility at the time of submission (i.e. ~12 months).
Figure 4. Northern blot analysis of individual Tc-responsive clones expressing regulatable Mda-7 or JunB cDNAs. Autoradiographic detection of levels of induced RNA message levels expressed in clonally selected cells stably transfected with the Mda-7 (A) or JunB (B) cDNAs under regulation of Tc, probed with the respective radiolabeled cDNA probes after transfer to nylon membranes. Each similarly numbered sample was derived from the same clone without induction [1-17 (A) and 1-9 (B)] or after addition of inducer [1[prime]-17[prime] (A) and 1[prime]-9[prime] (B)].
Table 2. Quantitation of range of inducibility of Mda-7 and JunB mRNA levels obtained in a gross screen of individual clones
| Clone no. (Fig. 4) | Total no. of clones | Fold induction in the presence of Doxa |
| Mda-7 clones 6, 8, 11-13, 15 | 6 | No inducibility |
| Mda-7 clones 2, 4 | 2 | <10-fold induction |
| Mda-7 clones 1, 7, 9, 10, 14, 16, 17 | 7 | Between 10- and >100-fold induction |
| JunB clones 1, 3, 10 | 3 | No inducibility |
| JunB clones 2, 7, 8 | 3 | <10-fold induction |
| JunB clones 4-6, 9 | 4 | Between 10- and 50-fold induction |
Determination of temporal stability of clones expressing cDNAs under inducible regulation of EF1prtTA
To systematically document that expression and inducibility of clones continued over an extended period of time, two HO-1 cell lines were analyzed over a period of 60 days. Continued inducibility of gene expression was determined at time points of 30, 45 and 60 days after start of the experiment by northern blot analysis (Fig. 5). It may be noted that these clones selected from the screen shown in Figure 4A were actually isolated ~180 days before this series of experiments was initiated. Clone 1 showed very high levels of expression and inducibility but also demonstrated leaky expression. On the other hand, clone 14 showed lower levels of peak expression, but low basal activity. Both lines retained essentially the same expression characteristics over the entire period of analysis, which were similar to the levels shown when they were initially isolated (compare Fig. 4A, lanes 1 and 14, with Fig. 5A-C; quantified in Table 3). This is in contrast to attempts at establishing constitutively expressing cell lines for Mda-7 driven by a CMV promoter in HO-1 cells. Expression of a green fluorescent protein epitope-tagged CMV-Mda-7 construct showed a gradual time-dependent reduction in exogenous gene expression, ending in complete extinction within a few weeks, as monitored by fluorescence microscopy. These experiments were performed in parallel with the Tc-regulatable cells described above, highlighting the relevance of using promoters other than virally derived ones to obtain stable expression in certain cell types.
Figure 5. Time course northern blot analysis of Tc-regulatable Mda-7 cDNA expression. (Upper) Autoradiographic detection of Mda-7 transcripts in wild type HO-1 cells (WT HO-1) and Tc-regulated stable HO-1 Mda-7 clones 1 and 14 grown in the absence (-) or presence (+) of Dox, probed with the 32P-labeled Mda-7 cDNA probe. (Lower) Ethidium bromide stained gel for normalization of RNA amounts showing 18S and 28S rRNA bands corresponding to samples in the upper panel. Samples were isolated after continuous exposure of cells to Dox for 30 (A), 45 (B) and 60 (C) days, respectively. Cells were trypsinized, counted and re-plated before attaining confluence to maintain growth of the line in the presence or absence of Dox, as applicable, during the course of the experiment.
Table 3. Quantitation of average fold inducibility of Mda-7 mRNA levels in selected HO-1 and HeLa clones
| Clone no. | Fold induction in the presence of Doxa |
| HO-1 clone 1 | 100 |
| HO-1 clone 14 | 1000 |
| HeLa clone 1 | >100 |
| HeLa clone 2 | >200 |
| HeLa clone 4 | >100 |
To add another level of complexity to the study, a HeLa cell line expressing Mda-7 in a Tc-inducible manner was injected subcutaneously into nude mice. The animals were treated with Dox in their drinking water. Palpable tumors formed after 4 weeks, at which point the animals were sacrificed. Part of the tumor was explanted to re-isolate the injected cells and RNA was also isolated directly from tumors to determine if induction occurred in vivo. Of the five tumors isolated from nude mice, four showed expression of Mda-7 (Fig. 6A, compare lane 1 to lanes 2-5; quantified in Table 3). Microscopic analysis of surviving cells from the explants after 3 weeks in culture (see Materials and Methods) demonstrated that they were identical to HeLa cells, with no evidence of cells derived from surrounding mouse tissue. The explant-derived cells from three separate mice were re-examined for their ability to respond to Dox. Inducibility of the Mda-7 cDNA was determined by northern blot analysis. It was apparent that inducible expression of the Mda-7 cDNA was obtained in the animals, continued after passage of the HeLa cells through mice and persisted through further passages in vitro for a period extending beyond 14 days (Fig. 6B and C) from the start of analysis. Overall continued expression and inducibility of the exogenously introduced cDNA in these HeLa clones had therefore been maintained in excess of 90 days (since initially established) and persisted subsequent to passage through nude mice. A detailed analysis of the phenotypic effects of inducible expression of Mda-7 and JunB on the relevant cell lines or in nude mice tumorigenesis will be reported elsewhere. We plan to establish transgenic mouse lines using the EF1prtTA construct in the near future since inducibility of the rtTA driven Mda-7 cDNA was positive in vivo (nude mouse tumors) in mice fed with Dox in their drinking water (Fig. 6A, lanes 2-5).
Figure 6. Time course northern blot analysis of Tc-regulatable Mda-7 cDNA expression of HeLa cells explanted from nude mouse tumors. (Upper) Autoradiographic detection of Mda-7 transcripts from independent nude mouse tumors (A) or tumor explants (B and C) of HeLa cells following growth in the absence (-) or presence (+) of Dox. Clones 1, 2 and 4 in (B) and (C) were derived from independent tumors in different mice. (Lower) Ethidium bromide stained gel for RNA normalization, corresponding to RNA samples from the upper panels. (A) RNA from nude mice with Dox present continuously in their drinking water. (B and C) Samples isolated after continuous exposure of cells to Dox for 3 and 17 days, respectively. Cells were trypsinized, counted and re-plated before attaining confluence to maintain growth of the line in the presence or absence of Dox, as applicable, during the course of the experiment.
DISCUSSION
Inability to support continual strong expression from a given type of promoter, specifically those of viral origin, has been documented for certain cell types (34-38). Other than anecdotal information from colleagues and our own experience with premature reduction or ablation of promoter activity, rigorous studies that address these issues, perhaps understandably, are not well documented. The primary goal of this work was to reduce what we believe is a significant and hitherto unaddressed variable in successfully establishing Tc-inducible cells and also to demonstrate that choice of the appropriate promoter is critical for stable expression. We reasoned that the Tc operator expression construct, pUHD 10-3 (1) or its derivatives, into which the cDNA of interest is usually cloned is entirely dependent for regulatable expression on the TA gene product. Success rates in establishing a cell line can be achieved by preventing or avoiding TA cDNA expression shutdown, a variable that is likely to be cell type related, among other factors (23,24). Our own experience using pUHD 17-1neo (2) indicated that activity and inducibility in transient studies with sensitive luciferase assays worked reasonably well. However, further propagation of these cells after clonal selection of individual lines failed, despite the presence of expression construct DNA in the genome as detected by Southern analysis. To achieve steady and adequate levels of TA cDNA expression, relatively independent of temporal factors, cell type, cell physiology status and cell passage number, we replaced the CMV IE promoter-enhancer with the cellular EF-1[alpha] promoter (25-27).
Numerous modifications of the basic Tc-regulatable system have been reported in the literature, directed toward enhancing its performance. Alternative promoters have been utilized to drive expression of the TA cDNA. Many of these are based on the requirement for tissue- or species-specific expression in a plant (5), Drosophila (39) or mouse (11,14,17,34,37,40-44). Another modification of the TA-expressing construct involves use of bi- or multi-cistronic plasmids which drive expression through oppositely oriented promoters, of both TA cDNA and Tc operator-regulated cDNAs, mainly to circumvent two rounds of transfection of separate plasmids (4-6,41,45). However, they are based on one or a combination of viral promoters. These are unfortunately accompanied by the drawbacks mentioned above. A generally applicable modification to the original TA expression construct involved expression of both TA cDNA and exogenous cDNA under regulation of Tc operator sequences (3,41), the rationale being that exquisite regulation with very high inducibility could be built into a system when both the activator molecule and the regulatable gene of interest are under control of the same inducer through an autoregulatory loop. Unfortunately, it appears that the high levels of tTA protein produced as a result of induction result in toxic side-effects in cells (1,46), most likely due to interference in cellular metabolism by the acid activation domain of the HSV VP16 protein present in TA proteins. This could be an additional reason why certain cell types apparently shut down expression of TA cDNA after extended periods of time. Alternatively, cells strongly expressing TA proteins might be at a selective disadvantage, particularly those with a long doubling time, due to accumulation of toxic levels of TA protein.
Multi-cistronic retroviruses or combinations of two or more retroviruses expressing different components have also been constructed (7-11). These overcome the barrier of gene delivery into cells. However, expression is often based on viral promoter sequences, prone to possible shutdown in some cell types. The relatively complex steps involved in making a virus for a given cDNA of interest, including complicated cloning strategies due to large vector size, somewhat offset the advantages they present over classical DNA transfection approaches. Making retroviral vectors is presently restricted to a relatively small number of laboratories and safety concerns impose limitations of use in several setups. Therefore, while these vectors hold considerable promise, the likelihood of a major shift to their usage from widespread DNA transfection approaches may only be in the long term. The relevance of improved plasmid vectors is therefore still strong.
From this study it appears that switching over to the EF-1[alpha] expression cassette is able to overcome inconsistent or apparent loss of TA expression for reasons that could range from promoter activity extinction to toxicity of TA protein. We base this conclusion on our observations over an extended period of time, in the case of certain EF1prtTA cells lines, such as those established in HO-1 melanoma, over a period of 12 months. The parental HO-1 EF1prtTA cell line was made and initially analyzed over a 60 day period before being expanded and frozen for future use. These parental cells when used to establish inducible JunB and Mda-7 expression (Fig. 4A and B) showed functional levels of TA expression and inducible properties after being thawed out several months and passage numbers subsequent to when the line had been initially established and frozen. This cell line and others (Figs 3-6) continue to retain Tc-responsive properties and are routinely maintained in the absence of antibiotic selection. Expression of the rtTA cDNA continued irrespective of lack of a positive selective pressure, passage number, passage through nude mice and time elapsed between introduction and integration of the plasmid DNA and final usage of the cell line. This contrasts with the inability to establish TA-expressing lines in the cell types of interest using the original CMV IE-driven constructs. Overall, following modification of the rtTA cDNA expression vector, we have obtained significant enhancement of the likelihood of establishing cell lines that are Tc-regulatable compared to initial failures with the original vectors. It appeared that positive clone frequency was at least comparable to that previously reported (~1:12). Consistent inducible expression levels comparable to levels from the original constructs (i.e. 100- to 1000-fold inducibility at the RNA level) were obtained and clonal stability over an extended period of time was accomplished. From this we conclude that the modified EF1prtTA is a useful reagent with broad applicability in establishing Tc-regulatable cells. In addition, this study demonstrates the importance of examining alternatives to widely used expression constructs based on strong viral promoters since they are more likely to be suited for short-term transient studies, particularly in a given cell type-specific context.
ACKNOWLEDGEMENTS
All original Tc-regulated plasmid constructs were a kind gift from Prof. H. Bujard. The EF-1[alpha] promoter vector was a gift from J. A. Langer. The authors would like to acknowledge T. Kinzy for advice on usage of pCDEF3 and M. T. Maddireddi and D. C. Kang for the Mda-7 and JunB constructs, respectively. This research was supported in part by grants CA35675 and CA74468 from the National Institutes of Health, National Cancer Institute, the Samuel Waxman Cancer Foundation and the Chernow Endowment. P.B.F. is the Michael and Stella Chernow Urological Cancer Research Scientist.
REFERENCES
*To whom correspondence should be addressed at: Department of Urology, Herbert Irving Comprehensive Cancer Center, Columbia University, College of Physicians and Surgeons, 630 West 168th Street, New York, NY 10032, USA. Tel: +1 212 305 3441; Fax: +1 212 305 8177; Email: pbfl{at}columbia.edu
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: jnl.info{at}oup.co.uk
Last modification:
Copyright© Oxford University Press, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  B. M. Taylor, J. S. Foulke, R. Flinko, A. Heredia, A. DeVico, and M. Reitz An Alteration of Human Immunodeficiency Virus gp41 Leads to Reduced CCR5 Dependence and CD4 Independence J. Virol., June 1, 2008; 82(11): 5460 - 5471. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Weber, W. Bacchus, M. Daoud-El Baba, and M. Fussenegger Vitamin H-regulated transgene expression in mammalian cells Nucleic Acids Res., September 27, 2007; 35(17): e116 - e116. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. P. Turner, O. Moussa, M. Sauane, P. B. Fisher, and D. K. Watson Prostate-Derived ETS Factor Is a Mediator of Metastatic Potential through the Inhibition of Migration and Invasion in Breast Cancer Cancer Res., February 15, 2007; 67(4): 1618 - 1625. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N.A. Yamada, A. Castro, and R.A. Farber Variation in the extent of microsatellite instability in human cell lines with defects in different mismatch repair genes Mutagenesis, May 1, 2003; 18(3): 277 - 282. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. A. Yamada and R. A. Farber Induction of a Low Level of Microsatellite Instability by Overexpression of DNA Polymerase {beta} Cancer Res., November 1, 2002; 62(21): 6061 - 6064. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. Su, Y. Shi, R. Friedman, L. Qiao, R. McKinstry, D. Hinman, P. Dent, and P. B. Fisher PEA3 sites within the progression elevated gene-3 (PEG-3) promoter and mitogen-activated protein kinase contribute to differential PEG-3 expression in Ha-ras and v-raf oncogene transformed rat embryo cells Nucleic Acids Res., April 15, 2001; 29(8): 1661 - 1671. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||