to BioTechniques free email alert service to receive content updates.
Off-on polyadenylation strategy as a supplemental mechanism for silencing toxic transgene expression during lentiviral vector production
Claude Bagnis, Gael Zwojsczyki, Jacques Chiaroni, and Pascal Bailly
Etablissement Français du Sang Alpes Méditerranée, UMR 7268 ADéS, Aix-Marseille Université-EFS-CNRS, Laboratoire Hématologie Moléculaire, Marseille, France
BioTechniques, Vol. 56, No. 6, June 2014, pp. 311–318
Full Text (PDF)

Many gene therapy strategies rely on lentiviral-mediated transfer and expression of genes coding for toxic proteins. Methods of controlling transgene expression in target cells have been extensively investigated, but comparatively little attention has been given to controlling toxic protein expression in viral vector-producing cells, despite its potential implications for viral production and transduction efficiency. In this work, we tested a new lentiviral vector with a backbone that inhibits transgene mRNA polyadenylation and subsequent transgene expression in vector-producing cells. Transgene mRNA polyadenylation was not affected in transduced cells. In a model using enhanced green fluorescent protein (EGFP) cDNA under the control of the human phosphoglycerate kinase (PGK) promoter, flow cytometry demonstrated that transgene expression was dramatically decreased in 293T cells transfected with this new vector in its plasmid configuration. Viral production was maintained, and expression was fully restored in transduced HuH7 and 293T cells. These results provide the basis for a new strategy to improve the production of lentiviral vectors expressing toxic transgenes.

Targeted expression of toxic proteins by lentiviral vectors is a strategy with great potential for gene therapy that has been particularly well studied for cancer treatment (1-6). Regulation of toxic protein expression is problematic during both the production and delivery of lentiviral vectors. A number of strategies have been developed to restrict expression of therapeutic transgenes to target cells during treatment. These include using tissue-specific promoters that are active in targeted diseased cells but inactive in neighboring cells showing a different histotype, placing transgenes under the control of conditionally active transcription, and targeting viral particles to selected cells (7, 8). Less study has been devoted to the regulation of toxic transgene expression during virus production.

Vector-producing cells used to produce lentiviral vectors show strong transgene expression. For example, 293T cells used to produce a vector that transfers and expresses enhanced green fluorescent protein (EGFP) under the control of a ubiquitous promoter show very strong EGFP expression (9). When producing a lentiviral vector containing a toxic protein expression unit, transgene-induced toxicity can profoundly alter vector production and reduce the ability to generate highly concentrated batches of viral vectors, which is a key parameter for efficient gene transfer in vivo. In addition, transgenic proteins can be packaged nonspecifically into lentiviral particles during production (10), and these particles can be engulfed by antigen-presenting cells (11), potentially causing off-target effects in those cells.


Optimizing transfer of transgenes coding for toxic proteins is a key requirement for gene therapies using lentiviral vectors. Taking advantage of U3 sequence duplication during the retroviral cycle, we designed a new vector backbone that markedly reduces transgene expression by vector-producing cells. No change was observed in the transfer and expression ability of the lentiviral vector. This strategy relies on turning the polyadenylation signal off in producing cells and on in transduced cells.

Controlling toxic transgene expression in vector-producing cells is a strategy that has received little attention. Two approaches have been proposed. The first consists of using inducible toxic proteins, such as caspase 8, whenever possible (12). The second approach, which applies specifically to inhibition of transgene expression during production oflentiviral vector particles, involves the use of a promoter that remains silent in the vector-producing cells. This has been done using an expression cassette containing a tissue-specific promoter, the coding sequence, and the polyadenylation signal placed in antisense orientation relative to the viral transcription unit (13-15). However, the use of tissue-specific promoters has two important shortcomings. The first is that it sometimes fails to achieve full transcriptional silencing in 293T cells (14), and the second is that viral titer may be dramatically decreased when using an internal polyadenylation signal placed in the antisense orientation (16).

This study describes our novel off-on strategy for silencing transgene expression during lentiviral vector production. This is achieved by cloning a transgene expression casset te devoid of a polyadenylation signal in the antisense orientation relative to the vector backbone, which prevents transgene expression in the vector-producing cell. In addition, in the vector under plasmid configuration, a polyadenylation signal in the sense orientation relative to the transgene casset te was inserted in the U3 element of the 3′LTR located upstream of the transgene cassette (relative to its sense orientation). During reverse transcription, that U3 element is duplicated to the 5′LTR, now placing its polyadenylation signal downstream of the transgene cassette (in the sense orientation relative to the transgene), thus allowing the transgene to be expressed in the transduced cell. Our experiments using this vector showed that EGFP expression in vector-producing 293T cells was dramatically reduced without detriment to viral production and that transgene expression was fully restored in transduced cells.

Materials and methods

Cells and medium

A human hepatocarcinoma cell line, HuH7 (JCRB403, JCRB, Osaka, Japan), and a human embryonic kidney cell line, 293T (HEK-293T, CRL-11268, ATCC, Manassas, VA), were cultured in DMEM supplemented with 10% FBS in a 5% humidified CO2 atmosphere.

  1    2    3    4