Highly efficient site-specific transgenesis in cancer cell lines

Abstract

Background: Transgenes introduced into cancer cell lines serve as powerful tools for identification of genes involved in cancer. However, the random nature of genomic integration site of a transgene highly influences the fidelity, reliability and level of its expression. In order to alleviate this bottleneck, we characterized the potential utility of a novel PhiC31 integrase-mediated site-specific insertion system (PhiC31-IMSI) for introduction of transgenes into a pre-inserted docking site in the genome of cancer cells.

Methods: According to this system, a "docking-site" was first randomly inserted into human cancer cell lines and clones with a single copy were selected. Subsequently, an "incoming" vector containing the gene of interest was specifically inserted in the docking-site using PhiC31.

Results: Using the Pc-3 and SKOV-3 cancer cell lines, we showed that transgene insertion is reproducible and reliable. Furthermore, the selection system ensured that all surviving stable transgenic lines harbored the correct integration site. We demonstrated that the expression levels of reporter genes, such as green fluorescent protein and luciferase, from the same locus were comparable among sister, isogenic clones. Using in vivo xenograft studies, we showed that the genetically altered cancer cell lines retain the properties of the parental line. To achieve temporal control of transgene expression, we coupled our insertion strategy with the doxycycline inducible system and demonstrated tight regulation of the expression of the antiangiogenic molecule sFlt-1-Fc in Pc-3 cells. Furthermore, we introduced the luciferase gene into the insertion cassette allowing for possible live imaging of cancer cells in transplantation assays. We also generated a series of Gateway cloning-compatible intermediate cassettes ready for high-throughput cloning of transgenes and demonstrated that PhiC31-IMSI can be achieved in a high throughput 96-well plate format.

Conclusions: The novel PhiC31-IMSI system described in this study represents a powerful tool that can facilitate the characterization of cancer-related genes.

Figure 1

Docking-incoming system. (A) The basic docking site (DockZ) was modified to contain luciferase as a fusion transgene with puromycin N-acetyl-transferase (PAC), deriving DZL. (B) PhiC31 integrase-mediated site-specific insertion of the incoming vector. Correct integrants are selected based on resistance to neomycin. (C) A series of incoming vectors with different promoters and reporter genes. IncBasic is promoter-less and contains a multiple cloning site (MCS). IncCAP contains the pCAGGs promoter allowing for constitutive expression of the transgene as well as the Gateway cassette with reading frame A (RfA). IncTAP and IncTAG contain the second-generation tetracycline-regulated promoter (TRE) allowing for inducible expression. Inc-TAG allows for indirect monitoring of the expression of the transgene through a bicistronic arrangement with an IRES followed by EGFP.

Figure 2

Characterization of luciferase containing docking site (DZL). Normalized luciferase levels of single-copy DZL-containing clones of SKOV-3 (A, n=3) and DU145 lines (B, n=3).

Figure 3

Fidelity of the docking-incoming system. (A) Position of the primer pairs used for screening of integration (IntF and IntR) as well as the tk probe used for southern analysis. (B) PCR amplification of the integration junction using primers recognizing the attL site and the neo^R probe. Sixty-six colonies (60 colonies for Pc-3-A7 and 6 for SKOV-3-13) were screened and all had the correct integration site. Analysis was done using a multicomp agarose gel. (C) Southern blot analysis of subclones derived by integration of two different incoming vectors (IncCAG-transgene; lanes 2–7, and IncTRE-transgene; lanes 9–13) into line Pc-3-A7 (lanes 2–7) and SKOV-3-13 (lanes 9–13). Genomic DNA was digested with BamHI and the tk probe was used. Original Pc-3-A7 and SKOV3-13 DockZ lines were also included in lanes 8 and 14, respectively. 1-Kb marker is shown in the first lane.

Figure 4

Expression of reporter genes from isogenic clones. (A) Normalized luciferase levels of nine isogenic clones derived after integration of an incoming luciferase containing plasmid under the control of pCAGGs (IncCAP-luciferase) into line Pc-3-A7. There was no significant difference in luciferase activity in isogenic clones (P=0.509, n=3). Error bars show the standard deviation of the mean. (B) Histogram plots of EGFP expression of ten isogenic clones derived after integration of (i) IncCAP-EGFP and (ii) IncCAP- tCD4-IRES-EGFP into the Pc-3-A7. The GFP levels were an order of magnitude higher (Table 1) when EGFP was expressed as a single gene. (C) Fluorescent images of xenografts derived from Pc-3-A7 lines stably integrated with (i) IncCAP-EGFP and (ii) IncCAP-tCD4-IRES-EGFP. The relative ratio of EGFP levels of the two vectors was maintained in vivo as well.

Figure 5

Generation of an inducible system. (A) Plasmid encoding the second generation of rtTA under the control of CMV. (B) EGFP expression after integration of IncTRE-EGFP into Pc-3-A7 rtTA sublines #23 and #25, before and after induction with dox for 24 h and 48 h. bf; brightfield, gfp; fluorescent. (C) Histogram plots of EGFP expression of three isogenic clones derived after integration of IncTRE-Flt1-Fc-IRES-EGFP into Pc-3-A7-23. (D) Expression levels of (i) Flt1-Fc and (ii) VEGF in the supernatant of Pc-3-A7-23 stably transfected with IncTRE-EGFP or IncTRE-Flt1Fc-IRES-EGFP, before (white columns) and 48h after (grey columns) dox induction (n=3). (E) Tumour volume of Pc-3-A7-23 xenografts stably transfected with IncTRE-EGFP (empty grey squares, n=4) or IncTRE-Flt1Fc-IRES-EGFP (filled blue squares, n=4). Arrow indicates the time point at which the mice were switched from normal chow to doxycycline containing food pellets.