Abstract：Homologous recombination (gene targeting) is an important technique that is used to modify mammalian genome. We constructed an efficient common gene targeting vector pGT-C1 based on the plasmid pEGFP-C1. The vector consisted two different multiple cloning sites(MCS), two locus of X-over P1 (LoxP) and positive and negative selection markers. Two positive selection markers, neomycin resistance gene (Neo) and aequorea coerulescens green fluorescent protein gene (AcGFP) flanked by two LoxP sites. The two synthesized multiple cloning sequences MCS1 and MCS2 were placed in outside of LoxP sites. Additionally, a negative selection marker HSV-tk (herpes simplex virus thymidine kinase) gene was located adjacent to MCS2. Each element was detected by enzyme digestion, and positive and negative selection markers were verified by adding geneticin(G418) or ganciclovir (GANC) to cell culture media, respectively. The cutting efficiency of Cre-LoxP system was measured by semi-quantitative PCR after transfected Cre expression vector. Finally, we built the cow (Bos taurus) beta casein targeting vector pGT-L-R, and tested the targeting efficiency after transfection and drug screening. Targeting vector pGT-C1 had the following unique features: it contains two multiple cloning sites (MCS), which facilitated the directional insertion of homologous arms and then enhanced the versatility of this vector. The green fluorescent tag AcGFP between two LoxP was used to monitor instantly the transfection efficiency and integration efficiency, and the selection markers was able to be removed from the recipient cells by Cre-LoxP system to avoid the potential damage of selection markers to the recipient cells and increase the efficiency of gene knockout. Collectively, we constructed a gene targeting vector pGT-C1 with high efficiency and flexibility, and subsequently produced the transgenic bovine fibroblasts via constructing the vector pGT-L-R target the β-casein gene. This work provides basic tool for the development of transgenic animal models.
