Mechanisms controlling titer and expression of bidirectional lentiviral and gammaretroviral vectors


Bidirectional lentiviral vectors mediate expression of two or more cDNAs from a single internal promoter. In this study, we examined mechanisms that control titer and expression properties of this vector system. To address whether the bidirectional design depends on lentiviral (LV) backbone components, especially the Rev/Rev responsive element (RRE) system, we constructed similar expression cassettes for LV and gammaretroviral (GV) vectors. Bidirectional expression levels could be adjusted by the use of different internal promoters. Furthermore, removal of the constitutive RNA transport element of Mason-Pfizer monkey virus, used in first generation bidirectional LV vectors, improved gene expression. Titers of bidirectional vectors were 10-fold reduced in comparison to unidirectional vectors, independent of the Rev/RRE interaction. We reasoned that titer reductions were due to the formation of interfering double-stranded RNA in packaging cells. Indeed, cotransfection of Nodamuravirus B2 protein, an RNA interference suppressor, increased bidirectional vector titers at least fivefold. We validated the potential of high titer bidirectional vectors by coexpressing a fluorescent marker with O6-methylguanine-DNA methyltransferase from integrating, or with Cre recombinase from integrating and non-integrating GV and LV backbones. This allowed for the tracking of chemoprotected and recombined cells by fluorescence marker expression.


Gammaretroviral (GV) and lentiviral (LV) gene transfer with monocistronic vectors has shown a promising potential for the treatment of monogenetic and acquired diseases of the hematopoietic system.1, 2, 3, 4 A number of applications, such as the cure of polygenetic diseases as well as adoptive T-cell transfer and generation of induced pluripotent stem cells require more complex vector architectures capable of coexpressing two or more transgenes from the same vector backbone.5, 6

The most commonly used coexpression strategies are based on 2A cleavage or internal ribosome entry sites (IRES). In both cases, a bicistronic RNA has to be generated from a single promoter and coexpression depends on the function of the interspersed element. In the case of picornavirus-derived self-processing 2A cleavage sites, cotranslational separation of both transgenes will be at the expense of protein alterations, risk of incomplete cleavage and immunogenicity of the engineered gene products.7 In contrast, IRES suffer from their relatively large size (500 nt), their cell-type dependency8 and a potentially reduced gene expression.9, 10 Retroviral vectors harboring two consecutive promoters have also been described, but frequently suffered from promoter interference and transcriptional suppression.11, 12 This could not be observed when the minimal core element of the human cytomegalovirus promoter (mCMV) was fused back-to-back to a strong activating promoter thereby yielding an artificial bidirectional promoter/expression cassette; this bidirectional promoter design was described for tet-inducible non-viral vectors13 and constitutively active LV self-inactivating (SIN) vectors.14

However, the message initiated from the mCMV will, in part, be complementary to the genomic viral RNA in the packaging cell, which in turn might induce a double-stranded RNA (dsRNA) response. In general, dsRNA longer than 30 bp is a potent inducer of interferon-dependent and -independent antiviral pathways in mammals.15, 16 The latter include the destruction of dsRNA by RNase L and a block of translation by protein kinase R-mediated inhibition of eukaryotic initiation factor 2α.17, 18 Although it has remained unclear whether long dsRNA can also be processed by mammalian Dicer into small interfering RNA with antiviral activity,19 this is one of the main features of innate immunity against viral RNA in plants, fungi and invertebrates.20 To circumvent these defense mechanisms, a multitude of viruses encode suppressors of RNA interference (RNAi) with some of them being functional across kingdom borders. Interestingly, Nodamuravirus B2 protein (NovB2), important to establish lethal infections in insects and mammals, was shown to bind to long dsRNA and interfere with Dicer-mediated cleavage as well as post Dicer processes.21 Yet another RNAi suppressor is the adenoviral VA1-RNA that counteracts phosphorylation of eukaryotic initiation factor 2α, inhibits Dicer22 and was shown to increase LV vector titers.23

Besides cytosolic destruction of dsRNA, the latter canalso be the substrate for nuclear adenosine deaminasesthat act on RNA (ADAR). ADAR induce adenosine to inosine (A → I) hypermutations subsequently resulting in adenosine to guanine (A → G) changes in reverse transcribed cDNA. Interestingly, strongly deaminized dsRNA will be retained within the nucleus unless additional export systems, such as the human immunodeficiency virus-1 (HIV-1)-derived Rev/Rev responsive element (RRE), are used.24 Rev/RRE is normally responsible for the export of incompletely spliced HIV RNA25 by hijacking the cellular Crm-1-dependent protein export machinery.26 Rev also increases RNA stability and translation.27 In contrast, the constitutive transport element (CTE) of Mason-Pfizer monkey virus mediates the export of unspliced RNA through the cellular RNA export factor Tap.28

We hypothesized that the export of partially double-stranded genomic bidirectional vector RNA depends on Rev's function to overcome nuclear quality control restrictions. Therefore, HIV-Rev dependency would restrict the usage of bidirectional expression cassettes to LV vectors. By constructing similar LV and GV bidirectional vectors, we could show that Rev is not required for bidirectional vector production, and that the low titer observed for these vectors is mainly because of the formation of dsRNA per se. This could be overcome by interfering with the dsRNA response pathways of the packaging cell. Finally, we show the utility of bidirectional vectors for the coexpression of Cre recombinase as well as the chemoprotective selection marker O6-methylguanine-DNA methyltransferase (MGMT) in concert with a fluorescent marker.


Minor restriction of LV gene transfer in murine cells

To obtain a fair comparison of GV and LV gene transfer, we first examined potential species-specific transduction restrictions in the murine and human cellular backgrounds. Therefore, we produced identical integrase proficient (IP) and integrase defective (ID) LV and GV SIN vectors expressing eGFP from an internal spleen focus-forming virus (SFFV) promoter, and transduced human (HT1080) and murine (SC-1) fibroblast cells. The usage of ID retroviral vectors allows for transient gene expression29, 30 and may reveal whether a potential restriction targets the integration step.

Titers of VSVg pseudotyped IP and ID LV vectors were approximately 5–6.3 times lower when determined on SC-1 cells than on HT1080 cells (Figure 1a). Thus, the same supernatant transduced only 0.16–0.2 murine cells per transduced human cell with ID and IP LV vectors, respectively (Figure 1b). In contrast, this difference was less pronounced when using GV IP and ID vectors (Figure 1b). These experiments revealed that murine cells imposed a mild block to LV transduction that occurred independent of the function of the integrase protein, and potentially even before nuclear import. When murine and HLA-2-positive human cells were mixed before transduction, IP LV again transduced murine cells fourfold less effective than human cells. For IP GV this difference was only twofold (Figures 1c and d).

Figure 1

Integrating and non-integrating LV gene transfer is restricted in murine cells. (a) Titers of integrase proficient (IP) and integrase defective (ID) lentiviral (LV) and gammaretroviral (GV) VSVg pseudotyped particles determined on murine (SC-1) and human (HT1080) fibroblast cells. Vector titers of three independent experiments are shown as mean values with standard deviation and significance was determined by Student's unpaired two-sided t-test with ***P<0.001; NS, not significant. (b) Data from (a) plotted as titers on SC-1 divided by titers on HT1080 cells for IP and ID LV and GV particles. (c) Murine and HLA-2-positive human cells were mixed in a 1:1 ratio and transduced with IP LV and GV SIN vectors expressing GFP under the control of the SFFV promoter. FACS analysis 9 days post transduction revealed better transduction of human cells than murine HLA-negative cells with GV and LV vectors, respectively. (d) Ratio of murine and human transduced cells for LV (n=5) and GV (n=6) IP vectors as determined 9 days post transduction. Data are shown as mean values with standard deviation, and significance according to Student's unpaired two-sided t-test with *P<0.05.

As a recent report indicated a transcriptional block of LV in murine T-cells,31 we wondered whether this would also apply to LV vectors harboring the strong SFFV promoter. Vector copy numbers per transduced eGFP expressing SC-1 and HT1080 cell were 1 as determined by flow cytometry and quantitative PCR indicating that every vector integrate was transcriptionally active (Supplementary Figures S1a and b).

Together, these initial experiments indicated a minor (twofold) restriction of LV as compared with GV gene transfer in murine cells, which could depend on capsid processing.32 On the basis of these data, a fair comparison of both vector families is possible in murine models.

The modular bidirectional vector design

We next cloned identical GV and LV bidirectional vectors to study their Rev/RRE dependency. The Rev/RRE system might be needed for bidirectional vector production to overcome nuclear retention caused by edited dsRNA.24

Bidirectional GV SIN vectors (Figure 2a) consist of two independent transcription units with the sense-oriented unit resembling a normal SIN transcript driven from an internal enhancer/promoter element with message termination at the 3′ LTR. The second transcription unit was cloned in antisense orientation to the vector backbone, and combines an mCMV promoter and a 3′ untranslated region consisting of an optional CTE of Mason-Pfizer monkey virus and an SV40 polyadenylation signal. The same elements were used in the original description of bidirectional LV vectors.14 Having introduced appropriate restriction sites, the whole transgene cassette could be exchanged between GV and LV vectors (with pRRL.PPT. backbone yielding the pLBid vector series).

Figure 2

Expression from GV bidirectional backbones depends on the choice of the internal promoter. (a) General design of GV (pRBid) and LV (pLBid) bidirectional SIN vectors with and without (nC, no CTE) the constitutive transport element (CTE) of Mason-Pfizer monkey virus. The GV vector backbone is formed by 5′ and 3′ long terminal repeats (LTRs) (ΔU3, R, U5) with SIN deletion in the U3 region, splice donor (SD), packaging signal (Ψ) and posttranscriptional regulatory element (wPRE) of woodchuck hepatitis virus. LV backbones contain an additional splice acceptor (SA), the central polypurine tract (cPPT), the Rev responsible element (RRE) and an extended encapsidation signal (Ψ) including the 5′ region of gag (ΔGA). The bidirectional expression cassette consists of an SV40 polyadenylation signal (pA), the CTE, first gene of interest (GOI-1), and the minimal CMV (mCMV) promoter forming the antisense expression unit. The sense unit is formed by a strong enhancer/promoter (E/P) element and a second gene of interest (GOI-2). The expression cassette contains restriction sites for the modular exchange of its components and its transfer between GV and LV vector backbones. (b) GV expression cassette with eGFP in antisense and mCherry in sense orientation and variable strong enhancer/promoter (E/P) elements; short form of the eukaryotic elongation factor 1α (EFS); human phosphoglycerate kinase (PGK); or SFFV promoter. Murine SC-1 cells were transduced with ecotropic particles and analyzed by flow cytometry 11 days later.

Bidirectional particle formation is independent of Rev/RRE

To test whether GV bidirectional vectors naturally lacking RRE are able to produce infectious particles, we introduced eGFP and mCherry as gene of interest 1 and 2, respectively, and coupled their expression to therapeutically relevant promoters such as the short form of the eukaryotic elongation factor 1α (EFS), the human phosphoglycerate kinase promoter (PGK) and the viral SFFV promoter (Figure 2a).

We initially assessed vector performance by transducing SC-1 cells at low MOI (resulting in <30% transduced cells). This ensures the establishment of only one vector copy in the majority of cells,33 which is a prerequisite for a fair analysis of coexpression strategies. Although all three vectors generated comparable amounts of double-positive cells (Figure 2b), the PGK and SFFV promoters led to a small number of single-positive cells (2%). Cells transduced with the EPS promoter construct showed almost equal numbers of single eGFP and double-positive cells (8% vs 10%). However, we cannot exclude that more sensitive methods would have detected a majority of double-positive cells in this setting too. This may therefore reflect problems of working with weak promoters and a pair of transgenes that differ in the sensitivity of detection.

In contrary to our initial hypothesis, double-positive cells could be seen with all three GV vector constructs indicating independence of the Rev/RRE system. If significant editing of the genomic viral RNA would have occured, a major fraction of the transduced cells should only express the marker driven from the sense message (mCherry in Figure 2). However, our data indicate that editing of the bidirectional genome does not impose a major limitation to the development of bidirectional retrovirus-based vectors.

Hence, bidirectional vector RNA can be packaged independently of accessory mRNA export factors (for example, HIV-Rev). Furthermore, the expression levels of bidirectional vectors can be adjusted dependent on the type of internal promoter used.

NovB2 increases GV and LV vector titers

An important limitation of bidirectional vectors is the 10-fold reduction in titers as compared with their unidirectional counterparts. As we observed the same severe reduction in both GV and LV packaging systems (data not shown), this suggests that the drop in titer cannot be overcome by Rev's function on RNA export or stabilization and must therefore depend on another mechanism. We hypothesized that the expression of two complementary RNA molecules (the viral genomic RNA and the internal antisense transcript) in the packaging cell could be responsible for low titers, most likely through the induction of cellular defense mechanisms against double-stranded RNA. We, therefore, focused on strategies preventing either the expression of antisense message or the protection of dsRNA in packaging cells. Importantly, incorporating a T-cell-specific murine CD3δ promoter34 into a bidirectional GV vector did not prevent expression of the antisense message and thus did not overcome titer restrictions possibly due to interaction of the internal promoter with enhancer sequences of the 5′ promoter as described before (Supplementary Figure S2).35 Consequently, we focused on the screening of RNAi suppressor proteins by cotransfecting either empty vector (ctrl), HIV Tat,36 human foamy virus Tas,37 tomato bushy stunt virus p19,38 betanodamuravirus B239 or NovB221, 40 expression plasmid to a normal transfection mixture for the production of GV bidirectional vectors. These experiments identified NovB2 as most potent in enhancing bidirectional vector titers in a dose-dependent manner. Titers obtained with 5 μg of NovB2 (1:1 ratio with transfer vector) were 5.3-fold better than with the control plasmid (Figure 3a). This strong titer increase mediated by NovB2 fits to its proposed mechanism of action, which is the unspecific binding to long dsRNA and its subsequent protection against Dicer-mediated cleavage.21 Although we did not directly determine RNA levels in packaging cells, we observed an increase in eGFP and mCherry fluorescence of the NovB2 as compared with the control cotransfected cells (Supplementary Figure S3). This most likely reflects an accumulation of viral RNA mediated by NovB2's ability to protect dsRNA against cellular RNases, resulting in enhanced translation and improved packaging.

Figure 3

Characterization of NovB2's effect on increasing bidirectional vector titers. Titrations of bidirectional vectors harboring an eGFP-SFFV-mCherry expression cassette were performed with ecotropic GV or VSVg pseudotyped LV particles on SC-1 cells and subsequent analysis by flow cytometry 4–5 days post transduction. All titers are shown as mean values of transducing units per milliliter (t.u. ml−1) and standard deviation derived from three independent experiments. Test for significant differences between samples was performed by ‘one-way ANOVA’ and post-hoc ‘Bonferroni Multiple Comparison’ test. Significance was reached when *P<0.05, **P<0.01 and ***P<0.001, respectively. (a) Titers of GV bidirectional vectors generated by cotransfection of 1, 3, 5 or 15 μg of NovB2- or control (ctrl) plasmid. No significant differences between control samples; highly significant (***P<0.001) differences between each control and NovB2 pair. (b) GV vectors with standard configuration (pRBid), without CTE (pRBid nC), with enhanced formation of genomic RNA containing (pREBid) and lacking (pREBid nC) the CTE were transfected together with 5 μg of NovB2 or control plasmid. No significant differences between control samples; highly significant (***P<0.001) differences between each control and NovB2 pair. (c) Competition of NovB2 and VA1 to increase bidirectional vector titers. pREBid vectors were produced by cotransfection of VA1 plasmids in 1:10–1:1 ratios to 5 μg of NovB2 or control plasmids. Significance between 1:10 VA1 and 1:1 VA1 was reached (*P<P<0.05). Cotransfection of NovB2 to various amounts of VA1 always gave highly significant differences (***P<0.001) in titer. (d) Production and titration of LV particles generated with either 5 μg of control, NovB2, VA1 or VA1+NovB2 plasmids. NovB2 samples were highly significant different (***P<0.001) from control and VA1 samples.

Next, we addressed the impact of bidirectional vector cis-elements on titers. We thus used vectors with the standard configuration (pRBid), vectors depleted of the CTE (nC) and GV vectors with enhanced production of genomic RNA (pREBid) (pSERS backbone)35 containing and lacking the CTE (Figure 3b). Again, in all configurations, NovB2 significantly increased vector titers up to 6 × 106 t.u. ml−1. In addition, these experiments showed that the CTE had no impact on vector titers.

Furthermore, we compared NovB2 with a commercially available suppressor of dsRNA response, adenoviral VA1-RNA, which counteracts phosphorylation of eukaryotic initiation factor 2α and subsequent translational inhibition17 as well as Dicer function.22, 41 When increasing amounts of VA1 were cotransfected with pREBid vectors only a mild titer increase could be observed. In contrast, even smallest amounts of VA1 supplemented with 5 μg of NovB2 increased titers to saturating levels (Figure 3c). In line with the above results obtained using GV bidirectional vectors, NovB2 also led to a significant increase in LV bidirectional vector titers (up to 6.8-fold, Figure 3d) and was clearly superior to adenoviral VA1.

We conclude that NovB2 alone, most likely by protecting long dsRNA from degradation, is sufficient to increase bidirectional vector titers irrespective of the retroviral backbone (GV and LV), and thus irrespective of the pathway used for mRNA export (Rev dependent for LV, conventional mRNA export for GV).

Coexpression efficiency is enhanced in CTE-deprived GV and LV vectors

Next, we aimed at further characterizing our bidirectional expression system focusing on promoter-independent vector modifications with the potential to enhance transgene expression. One of these elements was the CTE of Mason-Pfizer monkey virus, which was initially introduced to the antisense message14 to support the export of (unspliced) RNA from the nucleus. Of note, new generations of LV bidirectional vectors42 lacked the CTE, as it was found to reduce the expression of a number of transgenes (Luigi Naldini, CONSERT Annual Meeting, 2008).

We constructed similar GV and LV vectors containing and lacking the CTE element to compare its effect on the transgene expression from both vector backbones (Figure 2a). When eGFP and mCherry were used as transgene pair (Figure 4a, left), the removal of the CTE resulted in a threefold (GV) to fivefold (LV) increase in the mean fluorescence intensity (MFI) of the antisense gene (eGFP). Surprisingly, also the MFI of the sense transcript (mCherry) increased twofold in both GV and LV vectors. CTE-deprived LV bidirectional vectors could be shown to efficiently transduce human CD34+ cells (Supplementary Figure S4).

Figure 4

Removal of the CTE and comparison of bidirectional, 2A cleavage and internal ribosome entry site-mediated coexpression strategies. (a) Direct comparison of GV and LV vectors with and without CTE (Figure 2a). Expression of eGFP and mCherry (left) or mCherry and truncated CD34 (tCD34) (right) was driven by the SFFV promoter and analyzed by flow cytometry 4 days post transduction. MFI is indicated by x and y values. tCD34 stained by an FITC-labeled antibody. (b) Vector design of GV and LV vectors harboring a bidirectional expression cassette lacking the CTE (Bid nC). Bidirectional transgene expression of the antisense-oriented eGFP and the sense-oriented mCherry was driven from the internal SFFV promoter. Accordingly, unidirectional gene expression was driven from the SFFV promoter but mCherry and eGFP transgenes were coexpressed through the 2A sequence of porcine teschovirus (P2A) or the internal ribosome entry site (IRES) of encephalomyocarditis virus. (c) FACS plots of SC-1 cells 5 days post transduction. MFI is indicated by x and y values. (d) SC-1 cells transduced with GV and LV Bid nC, P2A and IRES vectors at less than 30% have been expanded for 11 days before determination of mean vector copy numbers per expressing cell by quantitative PCR (n=3) using primers directed against the wPRE and the cellular PTBP-2 allele. Data are shown as mean values with standard deviation. (e) SC-1 cells from two different experiments (n=6) have been analyzed by FACS for eGFP and mCherry MFI in the double-positive cell fraction (<40%) 5 days post transduction with VSVg pseudotyped vectors (as depicted in b). (f) Jurkat cells (n=3) have been analyzed by FACS 5 days post transduction for eGFP and mCherry MFI in the double-positive cell fraction (<40%). Vectors as in (b). Statistical analysis for (e, f) was performed by ‘one-way ANOVA’ and post-hoc ‘Bonferroni Multiple Comparison’ test. Significance was reached with **P<0.01; ***P<0.001; NS, not significant.

To further investigate the issue of increased gene expression in CTE-deprived constructs, we cloned bidirectional vectors with mCherry in antisense and truncated CD34 receptor (tCD34),43 a transmembrane protein, in sense orientation (Figure 4a, right). Again, we could detect an increase in the MFI of the antisense protein if GV or LV vectors were lacking the CTE but the accompanying increase in the expression of the sense-oriented transcription unit was only modest for tCD34. This could be because of saturating expression of tCD34 at the plasma membrane.

Comparison of bidirectional to 2A and IRES-mediated coexpression

To assess how bidirectional vectors behave in comparison with unidirectional coexpression vectors, a panel of GV and LV SIN vectors was cloned in which coexpression of mCherry and eGFP was governed by the SFFV promoter (Figure 4b). Coexpression was either mediated by the 2A cleavage site of porcine teschovirus (P2A) or by the encephalomyocarditis virus IRES and compared with our best bidirectional vectors (lacking the CTE element; Bid nC vector series). Murine SC-1 cells were transduced with VSVg pseudotyped particles and analyzed for eGFP and mCherry expression 4 days later (Figure 4c). As exemplified in Figure 4d, cell populations transduced to less than 30% contained single vector integrants that were transcriptionally active.

In the double-positive SC-1 cell fraction, bidirectional GV vectors expressed 1.8-fold more eGFP and mCherry than comparable LV vectors resulting in significantly (P<0.001) higher eGFP expression from GV bidirectional vectors in comparison to all other vector configurations (Figure 4e). In contrast, bidirectional LV vectors showed comparable eGFP expression as their unidirectional counterparts, and significantly reduced mCherry expression as compared with bidirectional (P<0.01) and unidirectional (P<0.05) vectors (Figure 4e).

When we performed the same experiments on a human T-cell line (Jurkat) (Figure 4f), we observed a 1.4-fold higher eGFP expression from GV than from LV bidirectional vectors (P<0.001), which was also significantly higher than from all other vector constructs (P<0.001). In contrast, mCherry expression from GV and LV bidirectional vectors did not differ but was significantly lower than from GV 2A (pR P2A) vectors (P<0.001). Hence, efficient coexpression by bidirectional, 2A and IRES-based strategies depends not only on the cell type but also on the vector backbone and possibly its integration preferences.

Bidirectional vectors mediate protection against BCNU-induced cell death

Furthermore, we wanted to test the potential of bidirectional vectors for therapeutical and biotechnological applications focusing on the simultaneous expression of a fluorescent marker together with either Cre recombinase or the P140K mutant of O6-methylguanine-DNA methyltransferase (MGMT). The latter is a clinically relevant selection marker used to protect hematopoietic cells against DNA damage and subsequent apoptosis induced by alkylating agents like bis-chloronitrosourea (BCNU).44

GV and LV bidirectional vectors were constructed harboring MGMT-P140K and eGFP in either orientation. We chose the modest human PGK promoter as high MGMT-P140K expression was shown to inhibit proliferation of hematopoietic stem/progenitor cells.44 Pilot experiments performed with GV and LV vectors showed that only the expression of MGMT-P140K in sense orientation was strong enough to yield a large amount of double-positive cells (data not shown). When a murine myeloid progenitor cell line, 32D, was transduced with these vectors (Figure 5a) and FACS analyzed for eGFP and MGMT-P140K expression, we observed 18–39% double-positive cells (control). The same cells treated with 35 μM of BCNU showed an accumulation of 80% double-positive cells (Figure 5b) showing that eGFP expression correlated with MGMT-P140K positivity. Normalizing to the number of cells residing within the lower left quadrant (untransduced cells), GV and LV transduced cells could be enriched more than 25-and 13-fold, respectively (Figure 5b).

Figure 5

MGMT-P140K-mediated protection of hematopoietic cells against alkylating agents. (a) CTE containing GV and LV bidirectional vectors (pRBid GPM/pLBid GPM) expressing eGFP in antisense and MGMT-P140K (MGMT) in sense orientation both under the control of the PGK promoter. (b) 32D cells were transduced at low multiplicity of infection with VSVg pseudotyped pRBid and pLBid GPM vectors. Cells were expanded and aliquots (n=3) were incubated with 25 μM O6-benzylguanine before selection with 0 (control), 25, 35 or 50 μM BCNU. Representative samples treated with 35 μM BCNU were fixed, permeabilized and incubated with MGMT-P140K biotinylated primary and phycoerythrin-labeled streptavidin secondary antibody. Samples were analyzed for MGMT-P140K and eGFP expression by flow cytometry. (c) Same cells (n=3) as in (b) but only analyzed for eGFP expression by flow cytometry. Statistical significance was tested with Student's unpaired, two-sided t-test, and significance was reached when ***P<0.001.

According to the coexpressed eGFP (Figure 5c), GV and LV transduced cells could be stepwise enriched from 20 to 93% or 43 to 93% by increasing BCNU concentrations from 0 to 35 μM, respectively. Bidirectional vectors can thus be used to mediate protection against alkylating drugs in conjunction with a coexpressed protein.

Cre-mediated recombination by integrating and non-integrating bidirectional vectors

Cre recombinase derived from bacteriophage P1 mediates excision of transgenes flanked by loxP sites as well as the site-specific integration of a transgene into a single loxP site engineered into eukaryotic genomes. Although Cre recombinase belongs to a family of site-specific recombinases, overexpression may be cytotoxic.45 Cre-mediated recombination can be monitored by expressing a selectable marker.

Bidirectional GV and LV vectors were constructed (Figure 6a) with Cre in antisense and mCherry in sense, both under the control of the SFFV enhancer/promoter. The Cre message contained the CTE as it does not compromise titers but reduces gene expression, which should avoid potential cytotoxicity of overexpressed Cre. The usage of non-integrating vectors would further reduce the potential of genotoxic events induced by ectopic Cre expression. Thus, IP and ID vectors were used to transduce murine and human Cre indicator cell lines.46 Indicator cells contained a reporter allele consisting of the coding sequences of DsRed and a start codon-deprived eGFP. Cre-mediated recombination of loxP sites located after the start and stop codons of DsRed would thus restore the eGFP open reading frame accompanied by a shift from red to green fluorescence.

Figure 6

Cre-mediated recombination of Cre indicator cells by integrating and non-integrating GV and LV vectors. (a) GV and LV bidirectional vectors (pR/LBid) harboring Cre recombinase in antisense and mCherry in sense with transgene expression driven from the SFFV enhancer/promoter element. pA, polyadenylation signal of SV40 virus; CTE, constitutive transport element of Mason-Pfizer monkey virus. (b) Cre indicator cells (SC-1) 2 days post transduction with integrase proficient (IP) or integrase defective (ID) GV (pRBid) or LV (pLBid) bidirectional vectors. Cells positive for eGFP underwent recombination by vector delivered Cre and stably or transiently coexpressed mCherry. (c) Analysis of transiently expressing vectors over time. Data show the percentage of recombined (left y axis) and mCherry (right y axis) expressing cells. The mCherry signal got gradually lost during 13 days of observation in SC-1 and HT1080 indicator cells whereas eGFP signal from recombined cells was permanently detectable.

Two days post transduction of murine indicator cells, the majority of Cre-mCherry expressing cells had recombined the reporter allele to express eGFP, and no differences between IP and ID vectors were observed (Figure 6b). Of note, transiently expressing ID vectors were lost within 13 days post transduction, as indicated by a loss of mCherry fluorescence, whereas recombined cells stably expressed eGFP during the time of observation (Figures 6c and d).

In conclusion, we could readily develop efficient fluorescence-tagged bidirectional expression vectors for Cre-mediated recombination with similar efficiency for GV and LV backbones, and showed their utility in the form of IP and ID vectors.


In this study, we addressed mechanisms underlying the production and expression characteristics of bidirectional retrovirus-based vectors (both GV and LV). In the course of this study, we developed a new GV bidirectional expression system that produces high titer particles in a Rev/RRE independent manner. The system is suitable for the stable and transient coexpression of fluorescent markers alone or in combination with therapeutic transgenes like the chemoselection marker MGMT-P140K or site-specific recombinases, and displays comparable characteristics like its LV counterpart. On the basis of the results achieved in this study, we were also able to improve titer and expression of LV bidirectional vectors.

Initially, we speculated that the formation of bidirectional vector particles depends on the presence of an active RNA export element to enhance nucleo-cytoplasmic export. This hypothesis was supported by the report that long dsRNA is subjected to A → I editing, which not only can lead to nuclear retention but also to hypermutations within the dsRNA stretch.24, 47 In this case, nucleo-cytoplasmic export can be induced by incorporation of the RRE but not the CTE into the message.24 As bidirectional vectors ultimately produce two complementary RNAs (the genomic RNA and the message initiated from the mCMV) within the packaging cell this would be a perfect target for nuclear retention. As we could produce GV particles with bidirectional genomes in similar quantities as corresponding LV particles, it is tempting to speculate that A → I editing does not occur on bidirectional transcripts most likely because intermolecular dsRNA formation is less likely to occur within the nucleus than intramolecular hybridization.48

Although bidirectional particle formation was generally possible, titers were reduced at least 10-fold in comparison to unidirectional vectors. We therefore hypothesized that the detection of cytoplasmic rather than nuclear dsRNA is a limiting factor for high titer vector production. This concept was supported by the observation that cotransfection of an RNAi suppressor protein, NovB2, increased low bidirectional vectors titers more than fivefold. Additionally, titers of LV vectors encoding short hairpin RNAs could be rescued by cotransfecting NovB2 to the packaging cells.21, 40, 49 As NovB2's mode of action is the unspecific binding of long dsRNA and subsequent protection against cytoplasmic Dicer-mediated cleavage,21 it is likely that the gain in titer is the result of increased amounts of packageable genomic RNA (Supplementary Figure S3). Of note, the titer increasing effect was not only observed for GV and LV vectors but also for a multitude of transgenes supporting the hypothesis of rather unspecific protection of dsRNA against degradation. Besides Dicer, RNase L as part of the dsRNA response pathway could also help to degrade bidirectional dsRNA in the packaging cell. It remains debatable whether NovB2 would also inhibit this pathway and is able to suppress interferon induction. Viral VA1-RNA, known to counteract a dsRNA and interferon-induced block of translation initiation17 as well as Dicer,22 mediated only a minor increase in bidirectional vector titers, and did not show additive effects in the presence of NovB2. We therefore conclude that VA1 and NovB2 both act on the same cellular dsRNA response pathway with NovB2 being more potent than VA1. The exact target, potentially Dicer or suppression of interferon response, remains to be defined.

A retroviral transactivation assay could show that the internal SFFV but not PGK or EF1α promoter was able to enhance expression from a distantly located mCMV promoter.50 In line with these results, we could show that the SFFV promoter was also capable of efficiently activating the mCMV in a bidirectional setting. This also applied to the PGK and the CD3 promoters supporting the hypothesis that the activation of a neighboring promoter depends on enhancer strength and distance. Accordingly, even internal cell-type-specific promoters may thus not be sufficient to improve bidirectional vector titers.51 Interestingly, we could show that not only the promoter but also elements on the antisense message control gene expression. The gain in expression observed for CTE-deprived antisense-oriented genes remains unexplained, especially in the context of earlier studies that showed no effect of the CTE when fused to an eGFP lacking 5′ splice sites in the context of GV vectors.52 Nevertheless, we observed an increase in MFI from both GV and LV vectors, indicating that suppressive effects were mediated by the CTE itself, possibly in a cell type and transgene-dependent manner.

The choice of the viral backbone became more important when we compared the expression intensities of otherwise identical bidirectional vectors. SC-1 cells transduced with GV vectors had a 1.7-fold higher MFI for eGFP and mCherry than their LV counterparts. As GV tend to integrate in the vicinity of promoters, DNaseI hypersensitive sites as well as CpG islands and LV within active transcription units,53 the different expression levels observed for both vector families could reflect a better accessibility of transcription factors to the GV promoter/enhancer complex. The integration of bidirectional vectors within actively transcribed genes might be problematic considering that the cellular RNA running over the viral LTR boundaries will ultimately result in the formation of dsRNA that can trigger nuclear and cytoplasmic dsRNA responses and subsequent A → I editing of the message, reduced cap-dependent translation or cell death.24, 54, 55

Finally, we showed a proof of principle for the use of bidirectional vectors in human CD34+ cells, for the efficient Cre-mediated recombination of indicator cells as well as MGMT-P140K-mediated protection of hematopoietic cells along with efficient cell tracking by fluorescence marker expression. The bidirectional vector system benefits from the generation of two independent messages with coordinated expression. This makes it an attractive alternative to commonly used 2A and IRES-based coexpression strategies, and should also support combinatory designs to express more than two cDNAs.

Materials and methods

Quantitative PCR

Quantitative PCR was performed on an Applied Biosystems StepOnePlus Real-Time PCR System (Foster City, CA, USA) using the Quantitect or QuantiFast SYBR Green PCR kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. Primer pairs for the amplification of a 94 bp fragment of the woodchuck hepatitis virus posttranscriptional regulatory element (wPRE-FW: 5′-IndexTermGAGGAGTTGTGGCCCGTTGT-3′ and wPRE-RV: 5′-IndexTermTGACAGGTGGTGGCAATGCC-3′) and for the normalization with a 107 bp fragment of the murine flk-1 intron (flk_FW: 5′-IndexTermGTGAATTGCAGAGCTGTGTGTTG-3′ and flk_RV: 5′-IndexTermATTCATTGTATAAAGGTGGGATTG-3′) have been described.56

Alternatively, a primer pair amplifying a 142 bp fragment of a conserved intron (between exons 9 and 10) of the polypyrimidine tract-binding protein 2 (PTBP-2, PTBP2_FW 5′-IndexTermGTCTCCATTCCCTATGTTCATGCT-3′; PTBP2_RV: 5′-IndexTermATGAACAGATCATCCATAACGA-3′) was used to normalized the amount of input DNA and determine vector copy numbers in murine and human samples, and specificity of gene products was controlled by melting temperature determination.

Results were quantified according to Pfaffl57 by making use of primer pair-specific real-time PCR efficiencies and by comparing sample CT values to a standard curve generated with a hematopoietic cell clone with a known number of insertions. Genomic DNA was extracted with the QIAamp DNA Blood Mini kit (Qiagen) earliest 5 days post transduction to avoid detection of plasmid contamination or episomal DNA.

Production of viral supernatants and cultivation of cells

GV cell-free supernatants were generated by transient transfection of 293T cells seeded in 10 cm cell culture dishes, using plasmids encoding the viral vector, gag/pol and an ecotropic58 or VSV-G59 envelope proteins. LV supernatants were produced accordingly by cotransfection of LV vector, gag/pol, envelope and a Rev encoding plasmid. To increase bidirectional vector titers, transfection mixtures contained, if not otherwise stated, additional 5 μg NovB2 expression plasmid. Viral titers were determined on murine SC-1 fibroblasts or human fibrosarcoma HT1080 cells by transducing 1 × 105 cells in 500 μl medium supplemented with 4 μg ml−1 protamine sulfate (Sigma, Seelze, Germany) and spin occulation at 2000 r.p.m. for 1 h at 37 °C. SC-1 and HT1080-derived Cre indicator cell lines have been described earlier;46 293T, SC-1 and HT1080 (and their derivatives) cells were maintained in Dulbecco's modified Eagle's medium (high glucose) (Biochrom, Berlin, Germany) supplemented with 10% fetal calf serum, 100 U ml−1 penicillin/streptomycin and 0.1 mg ml−1 sodium pyruvate (all PAA, Pasching, Austria); Jurkat cells were cultivated in RPMI 1640 supplemented with 10% fetal calf serum and 100 U ml−1 penicillin/streptomycin (all PAA); 32D cells were maintained in Iscove's modified Dulbecco's medium (Biochrom) supplemented with 10% fetal calf serum, 100 U ml−1 penicillin/streptomycin, 0.1 mg ml−1 sodium pyruvate (all PAA) and 2 ng ml−1 murine IL3 (Peprotech, Hamburg, Germany).

MGMT-P140K expressing 32D cells were selected by incubation for 2 h with 25 μM O6-benzylguanine (Sigma), and subsequent addition of BCNU to a final concentration of 25, 35 or 50 μM. Cells were analyzed for simultaneous MGMT-P140K and eGFP expression by antibody staining against MGMT-P140K proteins.


Detailed cloning strategies of GV and LV bidirectional vectors are available on request. In brief, a 452 bp fragment consisting of an SV40 polyadenylation signal and a Mason-Pfizer monkey virus CTE was PCR amplified using primers 5′CTEpA_as MluI (5′-IndexTermGCTACGCGTGATCATAATCAGCCATACCACAT-3′) and 3′CTEpA_as Not Eco47III (5′-IndexTermGCTAGCGCTGCGGCCGCAGACCACCTCCCCTGCGAGCTAA-3′) and a pSF2-c(NMCG) template (restriction sites in bold). The PCR product was subcloned into pSRS11.PGK.eGFP.pre*,35 additionally equipped with unique MluI and Eco47III restriction sites. Sequence identity of the subcloned PCR product was confirmed by sequencing. The resulting vector (pSRS11.pA.CTE.PGK.eGFP.pre*) was equipped with a DsRedExpress (DsRex) fluorescence marker through Eco47III and NotI restriction yielding pSRS11.pA.CTE.DsRex.PGK.eGFP.pre*. The minimal CMV promoter was PCR amplified from a pSRS11.CMV.eGFP.pre* 35 template using 5′ CMVmin (5′-IndexTermGCTATTCGAATATCGATTGCTAGCGCTGGGTAGGCGTGTACGGTGGGAGG-3′) and 3′ CMVmin (5′-IndexTermCGATCCATGGTGGCCGCGGAGGCTGGATCGGTCCCGGT-3′) primers harboring Eco47III and NcoI restriction sites, respectively. The PCR product was sequence verified and cloned into pSRS11.pA.CTE.DsRex.PGK.eGFP.pre* through Eco47III and NcoI restriction sites. The resulting vector pRBid.DsRex.PGK.eGFP.pre* was used as a template for all subsequent cloning steps. LV bidirectional vectors (pLBid) were generated by the transfer of the GV bidirectional expression cassette through MluI/BsrGI digestion.

The ID MLV gag/pol construct (MLVg/p D184A) was cloned by PCR amplification of a 283 bp fragment introducing a single D184A point mutation into the coding region of the retroviral integrase. PCR was performed with forward primer SphI_FW (5′-IndexTermCGGCATGCCTCAGGTATTGGGAACTGcCAATGGGCCTGCCTTCGT-3′) introducing the point mutation (in small letter), and the reverse primer NdeI_RV (5′-IndexTermATATAAGATCTCATATGGGGTGAGGCCATGG-3′). PCR products were subcloned into pCR2.1 vector (Invitrogen, Carlsbad, CA, USA), sequence verified and introduced into pcDNA3.MLVgag/pol through SphI/NdeI restriction sites.

The LV ID gag/pol construct (pcDNA3.GP.4 × CTE with D64V mutation) was kindly provided by Mick Milsom (Boston, MA, USA). The plasmid pcDNA3.1.Puro.NovB2 (NovB2) was kindly provided by Christopher S Sullivan and Don Ganem,21 and pAdVantage, encoding for adenoviral VAI and VAII RNA, purchased from Promega (Madison, WI, USA).

Flow cytometry

Cells transduced with integrating GV or LV vectors were analyzed earliest 4 days after transduction. Cells transduced with ID particles were analyzed 36–48 h post transduction if not otherwise mentioned. Staining of human MGMT protein with mouse monoclonal antibodies (Milipore, Billerica, MA, USA) was performed on fixed and permeabilized cells (BD Cytofix/Cytoperm, BD Biosciences, Heidelberg, Germany) in the presence of 0.1% (w/v) saponine (Riedel-de-Haen, Seelze, Germany) in phosphate-buffered saline with mouse anti-human MGMT and secondary phycoerythrin-labeled goat anti-mouse antibodies (BD Pharmingen, Franklin Lakes, NJ, USA). Staining of human truncated CD34 was performed with FITC-labeled antibodies (8G12, BD Bioscience, Franklin Lakes, NJ). Samples were analyzed for eGFP, mCherry, DsRed, human MGMT, HLA-2 or human truncated CD34 expression on a FACSCalibur or LSRII using Cell Quest (BD) and FlowJo (Treestar, Ashland, OR, USA) software; 15 000 cells per sample were monitored and a gate was set on a homogeneous cell population by scatter characteristics.

Conflict of interest

The authors declare no conflict of interest.


  1. 1

    Aiuti A, Slavin S, Aker M, Ficara F, Deola S, Mortellaro A et al. Correction of ADA-SCID by stem cell gene therapy combined with nonmyeloablative conditioning. Science 2002; 296: 2410–2413.

    CAS  Article  Google Scholar 

  2. 2

    Hacein-Bey-Abina S, Le Deist F, Carlier F, Bouneaud C, Hue C, De Villartay JP et al. Sustained correction of X-linked severe combined immunodeficiency by ex vivo gene therapy. N Engl J Med 2002; 346: 1185–1193.

    CAS  Article  Google Scholar 

  3. 3

    Ott MG, Schmidt M, Schwarzwaelder K, Stein S, Siler U, Koehl U et al. Correction of X-linked chronic granulomatous disease by gene therapy, augmented by insertional activation of MDS1-EVI1, PRDM16 or SETBP1. Nat Med 2006; 12: 401–409.

    CAS  Article  Google Scholar 

  4. 4

    Levine BL, Humeau LM, Boyer J, MacGregor RR, Rebello T, Lu X et al. Gene transfer in humans using a conditionally replicating lentiviral vector. Proc Natl Acad Sci USA 2006; 103: 17372–17377.

    CAS  Article  Google Scholar 

  5. 5

    Morgan RA, Dudley ME, Wunderlich JR, Hughes MS, Yang JC, Sherry RM et al. Cancer regression in patients after transfer of genetically engineered lymphocytes. Science 2006; 314: 126–129.

    CAS  Article  Google Scholar 

  6. 6

    Sommer CA, Stadtfeld M, Murphy GJ, Hochedlinger K, Kotton DN, Mostoslavsky G . iPS cell generation using a single lentiviral stem cell cassette. Stem Cells 2009; 27: 543–549.

    CAS  Article  Google Scholar 

  7. 7

    de Felipe P . Polycistronic viral vectors. Curr Gene Ther 2002; 2: 355–378.

    CAS  Article  Google Scholar 

  8. 8

    Borman AM, Le Mercier P, Girard M, Kean KM . Comparison of picornaviral IRES-driven internal initiation of translation in cultured cells of different origins. Nucleic Acids Res 1997; 25: 925–932.

    CAS  Article  Google Scholar 

  9. 9

    Mizuguchi H, Xu Z, Ishii-Watabe A, Uchida E, Hayakawa T . IRES-dependent second gene expression is significantly lower than cap-dependent first gene expression in a bicistronic vector. Mol Ther 2000; 1: 376–382.

    CAS  Article  Google Scholar 

  10. 10

    Kaufman RJ, Davies MV, Wasley LC, Michnick D . Improved vectors for stable expression of foreign genes in mammalian cells by use of the untranslated leader sequence from EMC virus. Nucleic Acids Res 1991; 19: 4485–4490.

    CAS  Article  Google Scholar 

  11. 11

    Proudfoot NJ . Transcriptional interference and termination between duplicated alpha-globin gene constructs suggests a novel mechanism for gene regulation. Nature 1986; 322: 562–565.

    CAS  Article  Google Scholar 

  12. 12

    Greger IH, Demarchi F, Giacca M, Proudfoot NJ . Transcriptional interference perturbs the binding of Sp1 to the HIV-1 promoter. Nucleic Acids Res 1998; 26: 1294–1301.

    CAS  Article  Google Scholar 

  13. 13

    Baron U, Freundlieb S, Gossen M, Bujard H . Co-regulation of two gene activities by tetracycline via a bidirectional promoter. Nucleic Acids Res 1995; 23: 3605–3606.

    CAS  Article  Google Scholar 

  14. 14

    Amendola M, Venneri MA, Biffi A, Vigna E, Naldini L . Coordinate dual-gene transgenesis by lentiviral vectors carrying synthetic bidirectional promoters. Nat Biotechnol 2005; 23: 108–116.

    CAS  Article  Google Scholar 

  15. 15

    Gantier MP, Williams BR . The response of mammalian cells to double-stranded RNA. Cytokine Growth Factor Rev 2007; 18: 363–371.

    CAS  Article  Google Scholar 

  16. 16

    Sen GC, Peters GA . Viral stress-inducible genes. Adv Virus Res 2007; 70: 233–263.

    CAS  Article  Google Scholar 

  17. 17

    Kitajewski J, Schneider RJ, Safer B, Munemitsu SM, Samuel CE, Thimmappaya B et al. Adenovirus VAI RNA antagonizes the antiviral action of interferon by preventing activation of the interferon-induced eIF-2 alpha kinase. Cell 1986; 45: 195–200.

    CAS  Article  Google Scholar 

  18. 18

    Thomis DC, Samuel CE . Mechanism of interferon action: evidence for intermolecular autophosphorylation and autoactivation of the interferon-induced, RNA-dependent protein kinase PKR. J Virol 1993; 67: 7695–7700.

    CAS  PubMed  PubMed Central  Google Scholar 

  19. 19

    Yang S, Tutton S, Pierce E, Yoon K . Specific double-stranded RNA interference in undifferentiated mouse embryonic stem cells. Mol Cell Biol 2001; 21: 7807–7816.

    CAS  Article  Google Scholar 

  20. 20

    Ding SW, Voinnet O . Antiviral immunity directed by small RNAs. Cell 2007; 130: 413–426.

    CAS  Article  Google Scholar 

  21. 21

    Sullivan CS, Ganem D . A virus-encoded inhibitor that blocks RNA interference in mammalian cells. J Virol 2005; 79: 7371–7379.

    CAS  Article  Google Scholar 

  22. 22

    Lu S, Cullen BR . Adenovirus VA1 noncoding RNA can inhibit small interfering RNA and MicroRNA biogenesis. J Virol 2004; 78: 12868–12876.

    CAS  Article  Google Scholar 

  23. 23

    Pernod G, Fish R, Liu JW, Kruithof EK . Increasing lentiviral vector titer using inhibitors of protein kinase R. Biotechniques 2004; 36: 576–578; 580.

    CAS  Article  Google Scholar 

  24. 24

    Zhang Z, Carmichael GG . The fate of dsRNA in the nucleus: a p54(nrb)-containing complex mediates the nuclear retention of promiscuously A-to-I edited RNAs. Cell 2001; 106: 465–475.

    CAS  Article  Google Scholar 

  25. 25

    Malim MH, Hauber J, Le SY, Maizel JV, Cullen BR . The HIV-1 rev trans-activator acts through a structured target sequence to activate nuclear export of unspliced viral mRNA. Nature 1989; 338: 254–257.

    CAS  Article  Google Scholar 

  26. 26

    Bogerd HP, Echarri A, Ross TM, Cullen BR . Inhibition of human immunodeficiency virus Rev and human T-cell leukemia virus Rex function, but not Mason-Pfizer monkey virus constitutive transport element activity, by a mutant human nucleoporin targeted to Crm1. J Virol 1998; 72: 8627–8635.

    CAS  PubMed  PubMed Central  Google Scholar 

  27. 27

    Felber BK, Hadzopoulou-Cladaras M, Cladaras C, Copeland T, Pavlakis GN . rev protein of human immunodeficiency virus type 1 affects the stability and transport of the viral mRNA. Proc Natl Acad Sci USA 1989; 86: 1495–1499.

    CAS  Article  Google Scholar 

  28. 28

    Gruter P, Tabernero C, von Kobbe C, Schmitt C, Saavedra C, Bachi A et al. TAP, the human homolog of Mex67p, mediates CTE-dependent RNA export from the nucleus. Mol Cell 1998; 1: 649–659.

    CAS  Article  Google Scholar 

  29. 29

    Yu SS, Dan K, Chono H, Chatani E, Mineno J, Kato I . Transient gene expression mediated by integrase-defective retroviral vectors. Biochem Biophys Res Commun 2008; 368: 942–947.

    CAS  Article  Google Scholar 

  30. 30

    Apolonia L, Waddington SN, Fernandes C, Ward NJ, Bouma G, Blundell MP et al. Stable gene transfer to muscle using non-integrating lentiviral vectors. Mol Ther 2007; 15: 1947–1954.

    CAS  Article  Google Scholar 

  31. 31

    Tervo HM, Goffinet C, Keppler OT . Mouse T-cells restrict replication of human immunodeficiency virus at the level of integration. Retrovirology 2008; 5: 58.

    Article  Google Scholar 

  32. 32

    Hatziioannou T, Cowan S, Bieniasz PD . Capsid-dependent and -independent postentry restriction of primate lentivirus tropism in rodent cells. J Virol 2004; 78: 1006–1011.

    CAS  Article  Google Scholar 

  33. 33

    Fehse B, Kustikova OS, Bubenheim M, Baum C . Pois(s)on—it's a question of dose? Gene Therapy 2004; 11: 879–881.

    CAS  Article  Google Scholar 

  34. 34

    Ji HB, Gupta A, Okamoto S, Blum MD, Tan L, Goldring MB et al. T cell-specific expression of the murine CD3delta promoter. J Biol Chem 2002; 277: 47898–47906.

    CAS  Article  Google Scholar 

  35. 35

    Schambach A, Mueller D, Galla M, Verstegen MM, Wagemaker G, Loew R et al. Overcoming promoter competition in packaging cells improves production of self-inactivating retroviral vectors. Gene Therapy 2006; 13: 1524–1533.

    CAS  Article  Google Scholar 

  36. 36

    Bennasser Y, Le SY, Benkirane M, Jeang KT . Evidence that HIV-1 encodes an siRNA and a suppressor of RNA silencing. Immunity 2005; 22: 607–619.

    CAS  Article  Google Scholar 

  37. 37

    Lecellier CH, Dunoyer P, Arar K, Lehmann-Che J, Eyquem S, Himber C et al. A cellular microRNA mediates antiviral defense in human cells. Science 2005; 308: 557–560.

    CAS  Article  Google Scholar 

  38. 38

    Voinnet O, Pinto YM, Baulcombe DC . Suppression of gene silencing: a general strategy used by diverse DNA and RNA viruses of plants. Proc Natl Acad Sci USA 1999; 96: 14147–14152.

    CAS  Article  Google Scholar 

  39. 39

    Fenner BJ, Thiagarajan R, Chua HK, Kwang J . Betanodavirus B2 is an RNA interference antagonist that facilitates intracellular viral RNA accumulation. J Virol 2006; 80: 85–94.

    CAS  Article  Google Scholar 

  40. 40

    Johnson KL, Price BD, Eckerle LD, Ball LA . Nodamura virus nonstructural protein B2 can enhance viral RNA accumulation in both mammalian and insect cells. J Virol 2004; 78: 6698–6704.

    CAS  Article  Google Scholar 

  41. 41

    Andersson MG, Haasnoot PC, Xu N, Berenjian S, Berkhout B, Akusjarvi G . Suppression of RNA interference by adenovirus virus-associated RNA. J Virol 2005; 79: 9556–9565.

    CAS  Article  Google Scholar 

  42. 42

    Brown BD, Gentner B, Cantore A, Colleoni S, Amendola M, Zingale A et al. Endogenous microRNA can be broadly exploited to regulate transgene expression according to tissue, lineage and differentiation state. Nat Biotechnol 2007; 25: 1457–1467.

    CAS  Article  Google Scholar 

  43. 43

    Fehse B, Richters A, Putimtseva_Scharf K, Klump H, Li Z, Ostertag W et al. CD34 splice variant: an attractive marker for selection of gene-modified cells. Mol Ther 2000; 1: 448–456.

    CAS  Article  Google Scholar 

  44. 44

    Milsom MD, Jerabek-Willemsen M, Harris CE, Schambach A, Broun E, Bailey J et al. Reciprocal relationship between O6-methylguanine-DNA methyltransferase P140K expression level and chemoprotection of hematopoietic stem cells. Cancer Res 2008; 68: 6171–6180.

    CAS  Article  Google Scholar 

  45. 45

    Pfeifer A, Brandon EP, Kootstra N, Gage FH, Verma IM . Delivery of the Cre recombinase by a self-deleting lentiviral vector: efficient gene targeting in vivo. Proc Natl Acad Sci USA 2001; 98: 11450–11455.

    CAS  Article  Google Scholar 

  46. 46

    Galla M, Will E, Kraunus J, Chen L, Baum C . Retroviral pseudotransduction for targeted cell manipulation. Mol Cell 2004; 16: 309–315.

    CAS  Article  Google Scholar 

  47. 47

    Prasanth KV, Prasanth SG, Xuan Z, Hearn S, Freier SM, Bennett CF et al. Regulating gene expression through RNA nuclear retention. Cell 2005; 123: 249–263.

    CAS  Article  Google Scholar 

  48. 48

    Kawahara Y, Nishikura K . Extensive adenosine-to-inosine editing detected in Alu repeats of antisense RNAs reveals scarcity of sense-antisense duplex formation. FEBS Lett 2006; 580: 2301–2305.

    CAS  Article  Google Scholar 

  49. 49

    Poluri A, Sutton RE . Titers of HIV-based vectors encoding shRNAs are reduced by a dicer-dependent mechanism. Mol Ther 2008; 16: 378–386.

    CAS  Article  Google Scholar 

  50. 50

    Zychlinski D, Schambach A, Modlich U, Maetzig T, Meyer J, Grassman E et al. Physiological promoters reduce the genotoxic risk of integrating gene vectors. Mol Ther 2008; 16: 718–725.

    CAS  Article  Google Scholar 

  51. 51

    Liu B, Paton JF, Kasparov S . Viral vectors based on bidirectional cell-specific mammalian promoters and transcriptional amplification strategy for use in vitro and in vivo. BMC Biotechnol 2008; 8: 49.

    Article  Google Scholar 

  52. 52

    Schambach A, Wodrich H, Hildinger M, Bohne J, Krausslich HG, Baum C . Context dependence of different modules for posttranscriptional enhancement of gene expression from retroviral vectors. Mol Ther 2000; 2: 435–445.

    CAS  Article  Google Scholar 

  53. 53

    Lewinski MK, Yamashita M, Emerman M, Ciuffi A, Marshall H, Crawford G et al. Retroviral DNA integration: viral and cellular determinants of target-site selection. PLoS Pathog 2006; 2: e60.

    Article  Google Scholar 

  54. 54

    Tanaka N, Sato M, Lamphier MS, Nozawa H, Oda E, Noguchi S et al. Type I interferons are essential mediators of apoptotic death in virally infected cells. Genes Cells 1998; 3: 29–37.

    CAS  Article  Google Scholar 

  55. 55

    Der SD, Yang YL, Weissmann C, Williams BR . A double-stranded RNA-activated protein kinase-dependent pathway mediating stress-induced apoptosis. Proc Natl Acad Sci USA 1997; 94: 3279–3283.

    CAS  Article  Google Scholar 

  56. 56

    Modlich U, Bohne J, Schmidt M, von Kalle C, Knoss S, Schambach A et al. Cell-culture assays reveal the importance of retroviral vector design for insertional genotoxicity. Blood 2006; 108: 2545–2553.

    CAS  Article  Google Scholar 

  57. 57

    Pfaffl MW . A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res 2001; 29: e45.

    CAS  Article  Google Scholar 

  58. 58

    Morita S, Kojima T, Kitamura T . Plat-E: an efficient and stable system for transient packaging of retroviruses. Gene Therapy 2000; 7: 1063–1070.

    CAS  Article  Google Scholar 

  59. 59

    Burns JC, Friedmann T, Driever W, Burrascano M, Yee JK . Vesicular stomatitis virus G glycoprotein pseudotyped retroviral vectors: concentration to very high titer and efficient gene transfer into mammalian and nonmammalian cells. Proc Natl Acad Sci USA 1993; 90: 8033–8037.

    CAS  Article  Google Scholar 

Download references


This study was supported by grants of the German Ministry for Research and Education (TreatID, Pidnet), the DAAD (German-Chinese junior research groups), the Deutsche Forschungsgemeinschaft (DFG, that is SFB738 and Excellence Cluster REBIRTH) and the European Union (CONSERT, LSHB-CT-2004-005242; Clinigene, LSHB-CT-2006-018933, Persist), the German National Merit foundation (stipend to TM) and the Else Kröner Fresenius Stiftung (fellowship to AS). We thank Bernhard Gentner and Luigi Naldini for helpful comments and suggestions. We are grateful to Ute Modlich for sharing the qPCR standard, to Don Ganem and Christopher Sullivan for kindly providing the Nodamuravirus B2 expression plasmid, to Dirk Lindemann providing the HFV tas cDNA, to Beau Fenner and Jimmy Kwang for providing the Betanodavirus B2 expression plasmid, to Kajohn Boonrod for providing p19 RNA, and to Roger Y Tsien for providing the mCherry cDNA.

Author information



Corresponding author

Correspondence to A Schambach.

Additional information

Supplementary Information accompanies the paper on Gene Therapy website (

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Maetzig, T., Galla, M., Brugman, M. et al. Mechanisms controlling titer and expression of bidirectional lentiviral and gammaretroviral vectors. Gene Ther 17, 400–411 (2010).

Download citation


  • bidirectional
  • lentivirus
  • gammaretrovirus
  • Nodamuravirus
  • O6-methylguanine-DNA methyltransferase
  • Cre

Further reading