Adeno-associated virus vector-mediated transgene integration into neurons and other nondividing cell targets

Abstract

The site-specific integration of wild-type adeno-associated virus (wtAAV) into the human genome is a very attractive feature for the development of AAV-based gene therapy vectors. However, knowledge about integration of wtAAV, as well as currently configured recombinant AAV (rAAV) vectors, is limited. By using a modified Alu-PCR technique to amplify and sequence the vector-cellular junctions, we provide the first direct evidence both in vitro and in vivo of rAAV-mediated transgene integration in several types of nondividing cells, including neurons. This novel technique will be highly useful for further delineating the mechanisms underlying AAV-mediated integration, including issues of frequency, site preference, and DNA rearrangement in human as well as animal cells. Results from these studies should be beneficial for the development of the next generation of gene delivery vectors.

FIG. 1

Schematic of Alu-PCR. Human genomic DNA from CMVβ-gal-transduced cells was amplified by using the first set of primers: Alu5, an Alu-specific primer containing a tag; and C1, a CMV-specific primer. After an initial 10 cycles of PCR, the first set of primers was destroyed by uracil DNA glycosylase-induced nicks at dUTP (filled box). cTag5 (open box) represents a sequence complementary to the tag sequence on the newly synthesized strand. These PCR products were further amplified by using an internal primer, C2, for the CMV sequence plus TagA5, containing 16 nucleotides of Tag sequence and 6 nucleotides of Alu5 sequence. C3 was used as a specific CMV oligonucleotide probe for Southern hybridization. C4 and TagA5 were used to further amplify the above Alu-PCR products in order to obtain discrete bands for direct sequencing.

FIG. 2

5-Bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal) histochemical staining of 293 cells 2 days after transduction with the CMVβ-gal vector at an MOI of 15.

FIG. 3

Alu-PCR analysis on 293 cell samples. (a and b) Cells were either untreated or transduced with a crude preparation of CMVβ-gal. (a) Aliquots of 25 μl of PCR products were loaded in a 1% agarose gel containing ethidium bromide. Lane 1, control sample without a genomic DNA template; lanes 2 to 6, 100 ng of genomic DNA isolated 8 days after transduction of the CMVβ-gal vector into 293 cells; lanes 7 to 11, 100 ng DNA from control 293 cells; lanes 2 and 7, 10 pmol of Alu5 primer alone for the first 10 cycles; lanes 3 and 8, 10 pmol of C1 primer alone; lanes 1, 4, and 9, Alu5-C1 at 10 pmol: 10 pmol; lanes 5 and 10, Alu5-C1 at 100:10; lanes 6 and 11, Alu5-C1 at 10:100. (b) The gel shown in panel a subjected to Southern blot analysis. PCR-amplified products on a nylon membrane were hybridized with a ³²P-end-labeled specific CMV oligonucleotide probe (C3). (c and d) Alu-PCR amplification from 100 ng of genomic DNA isolated 8 days after transduction with CsCl gradient-purified CMVβ-gal. (c) Aliquots of 25 μl of PCR products loaded in an agarose gel. Lane 1, Alu5-C1 at 10 pmol: 10 pmol; lane 2, C1 primer alone. (d) The gel shown in panel c subjected to Southern blot analysis.

FIG. 4

Sequence of a vector-cellular junction identified by Alu-PCR from 293 cells transduced with the CMVβ-gal rAAV vector. (a) The vector portion of the junction includes the 5′ end of the CMV IE promoter (open box) and some AAV sequences (underlined) containing a partial AAV ITR. The cellular sequence contains a fragment with 64% homology to a human DNA sequence on chromosome Xq26.1 (dashed box), which is flanked by unidentified sequences not present in the CMVβ-gal vector (unmarked). The actual sequence in the shaded box is shown in panel b. The arrows in panels a and b point to the site of the vector-cellular junction.

FIG. 5

Alu-PCR analysis using an Alu5 primer. (a) Aliquots of 33 μl of Alu-PCR products from 100 ng of genomic DNA were loaded on a 1% agarose gel containing ethidium bromide. Lane 1, control human NT neurons; lane 2, NT neurons transduced with CMVβ-gal for 7 days; lane 3, control human alveolar macrophages; lane 4, transduced alveolar macrophages. (b) The gel shown in panel a subjected to Southern blot analysis with a ³²P-end-labeled specific CMV oligonucleotide probe (C3).

FIG. 6

Sequence of a vector-cellular junction identified by Alu-PCR from human NT neurons transduced with the CMVβ-gal rAAV vector. (a) The vector portion of the junction includes the 5′ end of the CMV IE promoter (open box) and some AAV sequences (underlined) containing a partial AAV ITR. The cellular sequence includes a fragment with 98% homology to a human DNA sequence on chromosome 9q34 (dashed underlined). The actual sequence in the shaded box is shown in panel b. The arrows in panels a and b point to the site of the vector-cellular junction.

FIG. 7

Alu-PCR analysis using an Alu3 primer. (a) Aliquots of 33 μl of Alu-PCR products from 100 ng of genomic DNA were loaded on a 1% agarose gel containing ethidium bromide. Lane 1, control human NT neurons; lane 2, NT neurons transduced with CMVβ-gal for 7 days; lane 3, control human alveolar macrophages; lane 4, transduced alveolar macrophages. (b) The gel shown in panel a subjected to Southern blot analysis with a ³²P-end-labeled specific CMV oligonucleotide probe (C3).

FIG. 8

B1-PCR analysis of rat brain samples. (a) Aliquots of 30 μl of B1-PCR products from 100 ng of genomic DNA were loaded on a 1% agarose gel containing ethidium bromide. Lanes 1 to 5, brain samples from a rat 2 weeks after injection with artificial CSF; lanes 6 to 10, samples from a rat injected with 5 × 10⁵ IU of a crude preparation of the CMVβ-gal AAV vector; lanes 1 and 6, cerebral cortex; lanes 2 and 7, cerebellum; lanes 3 and 8, striatum; lanes 4 and 9, brainstem; lanes 5 and 10, hypothalamus. (b) The gel shown in panel a subjected to Southern blot analysis with a ³²P-end-labeled specific CMV oligonucleotide probe (C3). (c and d) Alu-PCR amplification from 100 ng of hypothalamic genomic DNA isolated 2 weeks after injection with the CsCl gradient-purified CMVβ-gal. (c) Aliquots of 40 μl of PCR products were loaded in an agarose gel. Lane 1, B1-5-C1 at 10 pmol: 10 pmol; lane 2, C1 primer alone. (d) The gel shown in panel c subjected to Southern blot analysis.