Introduction

Production of genetically engineered animals by pronuclear injection or retroviral delivery systems has been a successful strategy for generating animal models to better understand the functionality of genes. In transgenic animals created by embryo microinjection, the site of integration of the transgene within the genome is a random event. Thus, when multiple embryos have been injected or infected with the same DNA, the integration site will be different in each founder animal. When the integration events have occurred at the one-cell stage, they should exhibit germline transmission with the potential to be inherited by the founder's offspring. If the integration occurs at a later stage, the resulting mosaic founder may or may not exhibit germline transmission of the transgene. In the case of pronuclear injection, there is typically one insertion site, although multiple transgene copies are often found in a tandem array at that integration site (1).

Lentivirus transgenesis is becoming an increasingly attractive alternative to pronuclear injection because it is more efficient in terms of successful transgene incorporation into the host genome, less invasive to the embryo, and technically less demanding to perform (2). Lentiviral delivery systems have been used successfully to generate transgenic mice, rats, pigs, and cattle (2,3,4,5,6,7). The disadvantage of lentivirus is that there are often multiple integration events with random transgene insertions on several chromosomes.

Independent of the method of transgene delivery, the insertion site can have profound effects on transgene expression. This can lead to phenotypic effects in the transgenic animal that are not due to the transgene per se, but are a consequence of the integration site, a phenomenon referred to as position effect (8). It is critical to correlate phenotype with genotype, particularly in animals created via lentivirus transgenesis, since not all copies of the transgene may be contributing equally to the phenotype.

Determining transgene integration sites is challenging. A number of PCR-based methods, often referred to as chromosome walking techniques, have been developed to isolate DNA fragments adjacent to known sequences, including inverse PCR (9), ligation-mediated PCR (LMPCR) (10), randomly primed PCR (RP-PCR) (11,12), and T-linker PCR (13). The method described in this paper incorporates several elements of these techniques in a unique way that allows the capture of DNA fragments containing the chromosomal region flanking the transgene. Our method enables quick and inexpensive determination of multiple independent transgene integration sites in founder animals and their offspring. Here, we demonstrate that this method is useful for identifying and monitoring multiple transgene integration sites in transgenic animals created using lentivirus.

Materials and Methods

Animals

Lewis rat lines carrying an enhanced green fluorescent protein (EGFP) transgene were created by using the EGFP DNA construct and experimental protocol described by Lois et al. (2). Transgene positive founder animals were identified using an EGFP PCR assay (14). To assess GFP expression, tail biopsies were examined for fluorescence under a Nikon SMZ1500 UV dissecting scope (Nikon Instruments, Melville, NY, USA). Founders (F0) were bred to wild-type Lewis rats obtained from Harlan Sprague Dawley (Indianapolis, IN, USA) to generate the N1. N1 animals were genotyped using integration site-specific PCR assays. Selected N1 animals were mated to wild-type Lewis rats to generate the N2. N2 animals were genotyped using integration site-specific genotyping assays, and GFP expression was confirmed. Several of these lines (F455.5, F456.9, F458.7, F463.1, and F463.5) demonstrated stable transmission of the integration site-specific transgene coupled with robust GFP expression and were donated to the Rat Research and Resource Center (RRRC) at the University of Missouri. All other rat lines and mouse strains used are available either through the RRRC (www.nrrrc.missouri.edu) for rats or the University of Missouri/Harlan Mutant Mouse Regional Resource Center (MMMRC; www.mmrrc.org) for mice.

Preparation of DNA

DNA was isolated from tail biopsies using the DNeasy® Tissue kit (Qiagen, Valencia, CA, USA). Restriction endonuclease digestion was performed with PstI and HhaI (Figure 1, step 1). These enzymes were chosen because they created 3′ overhangs and their recognition sites were not present within the transgene sequence. Three micrograms genomic DNA were digested with 20 U enzyme in a total reaction volume of 30 µL as recommended by the manufacturer. Reactions were incubated for 2 h for partial digestion.