Self-inactivating lentivirus vector for safe and efficient in vivo gene delivery

Abstract

In vivo transduction of nondividing cells by human immunodeficiency virus type 1 (HIV-1)-based vectors results in transgene expression that is stable over several months. However, the use of HIV-1 vectors raises concerns about their safety. Here we describe a self-inactivating HIV-1 vector with a 400-nucleotide deletion in the 3' long terminal repeat (LTR). The deletion, which includes the TATA box, abolished the LTR promoter activity but did not affect vector titers or transgene expression in vitro. The self-inactivating vector transduced neurons in vivo as efficiently as a vector with full-length LTRs. The inactivation design achieved in this work improves significantly the biosafety of HIV-derived vectors, as it reduces the likelihood that replication-competent retroviruses will originate in the vector producer and target cells, and hampers recombination with wild-type HIV in an infected host. Moreover, it improves the potential performance of the vector by removing LTR sequences previously associated with transcriptional interference and suppression in vivo and by allowing the construction of more-stringent tissue-specific or regulatable vectors.

FIG. 1

Structure of SIN HIV-derived vectors. A schematic representation of an HIV-1 vector with enlarged 3′ LTR to show the binding sites for differents transcription factors on U3 is shown (not to scale). Although the 3′ LTR is depicted, the nucleotide numbering refers to the cap site at the beginning of R as +1 as for a 5′ LTR. Position −418 is the 5′ limit of all deletions; positions −78, −45, −36, and −18 indicate the 3′ limits of the different deletions described in the text. The deletion generating the SIN-18 vector created a novel BglII site. Details on the nuclear factors binding U3 can be found in references , , and and references therein. SD, splice donor; RRE, Rev-response element; SA, splice acceptor. The GenBank accession number for the wild-type 3′ LTR is M1991.

FIG. 2

Expression of a lacZ transgene delivered by SIN or full-length LTR vectors. 293T cells were transduced with equal volumes (200 μl) of two HR′CMVlacZSIN-18 or four HR′CMV lacZ vector stocks. Titers (TU per nanogram of p24) were similar for all stocks. β-Gal activity (in arbitrary units) at 48 h postinfection is plotted against the amount of p24 in vector stocks. Cells transduced with SIN-18 vectors express more than twice as much β-Gal per nanogram of p24 than cells transduced with full-length LTR vectors.

FIG. 3

Northern blot analysis of vector-derived transcripts in transduced HeLa cells. Total RNA was extracted from HeLa cells transduced with an RRLPGK-GFP vector (lane 1) or with its SIN versions SIN-18 (lane 2), SIN-36 (lane 3), SIN-45 (lane 4), and SIN-78 (lane 5). In lane 1, three bands with the sizes expected for the LTR-derived transcripts (unspliced and spliced) and the PGK-derived transcripts are visible. As expected for HeLa cells, transcription was initiated much more frequently at the internal PGK promoter than at the 5′ HIV-1 LTR. In lanes 2 to 5, transcripts derived from SIN vectors are 340 to 400 nucleotides smaller than the corresponding transcripts in lane 1. RNA initiated at the HIV-1 LTR is detectable in lane 5 but not in lanes 2 to 4. Positions of molecular size markers (in kilobases) are indicated on the right. Ψ, encapsidation signal; SD, splice donor; SA, splice acceptor.

FIG. 4

Activation pattern of HIV-1 vectors following infection of transduced SupT1 cells by HIV-1. Human lymphocytic SupT1 cells were transduced at a high multiplicity of infection by HIV-derived vectors carrying a PGK-GFP expression cassette and either a full-length LTR or the indicated U3 deletion construct. Six days later, the stably transduced cells were infected with VSV G-pseudotyped HIV-1 or were mock treated, and 48 h later they were analyzed by FACS for GFP fluorescence. Infection with HIV-1 strongly enhanced the expression of GFP in cells transduced by a vector with a full-length U3 LTR or the −78 deletion construct, while it had no effect on cells transduced with vectors having larger U3 deletions. The left and middle quadrants represent the fluorescence of cells not transduced and transduced by the GFP vector, respectively. The right quadrant includes cells with increased GFP expression upon infection by HIV-1. The increased expression of GFP indicates activation of vector transcription from the LTR and is due to translational readthrough of the PGK promoter sequence upstream of the GFP cDNA (see text). The increased in fluorescence intensity was 30-fold for cells transduced by the full-length LTR and 21-fold for those transduced by the SIN-78 vector. The HIV-1 had a deletion in the envelope gene and was thus limited to one round of infection. Similar patterns of Tat responsiveness were observed when HeLa-tat cells were transduced with the various vectors (not shown).

FIG. 5

In vivo transduction of GFP into neurons by SIN or full-length LTR vectors. HIV-1 vectors carrying a PGK-GFP expression cassette with the full-length U3 region (A and C) or the −18 deletion construct (B and D) were concentrated by ultracentrifugation and normalized for particle content prior to injection into the corpora striata of adult rats. One month after injection, brain sections were stained for immunoreactivity to the GFP protein. Both types of vectors transduced neurons very efficiently. The SIN vector often appeared to achieve a higher level of transgene expression. A representative section close to the injection site is shown for one of six injected striata per vector. Bars in panels A and C, 2 and 0.1 mm, respectively (magnifications are the same for panels B and D, respectively).