Indexing TNF-alpha gene expression using a gene-targeted reporter cell line

Abstract

Background: Current cell-based drug screening technologies utilize randomly integrated reporter genes to index transcriptional activity of an endogenous gene of interest. In this context, reporter expression is controlled by known genetic elements that may only partially capture gene regulation and by unknown features of chromatin specific to the integration site. As an alternative technology, we applied highly efficient gene-targeting with recombinant adeno-associated virus to precisely integrate a luciferase reporter gene into exon 1 of the HeLa cell tumor necrosis factor-alpha (TNF-alpha) gene. Drugs known to induce TNF-alpha expression were then used to compare the authenticity of gene-targeted and randomly integrated transcriptional reporters.

Results: TNF-alpha-targeted reporter activity reflected endogenous TNF-alpha mRNA expression, whereas randomly integrated TNF-alpha reporter lines gave variable expression in response to transcriptional and epigenetic regulators. 5,6-Dimethylxanthenone-4-acetic acid (DMXAA), currently used in cancer clinical trials to induce TNF-alpha gene transcription, was only effective at inducing reporter expression from TNF-alpha gene-targeted cells.

Conclusion: We conclude that gene-targeted reporter cell lines provide predictive indexing of gene transcription for drug discovery.

Figure 1

Adeno-associated virus-mediated TNF-α gene targeting in HeLa cells. (a) Genomic fragment containing the human TNF-α gene (top) and the recombinant adeno-associated virus (rAAV) TNF-α targeting vector (bottom). Arrows mark nested primers used for PCR screening of targeting events. Restriction sites and probes used for Southern blot confirmation of targeting events are also shown. The left and right homologous arms (Homologous Arm) in the AAV targeting vector contain endogenous sequence from the wild-type TNF-α allele. (b) Southern blot analysis of cells derived from the targeted intermediate clone Tg#28zeo^R, using restriction sites and the TNF-α probe indicated in panel (a). Parental HeLa cell genomic DNA is shown in lanes 1, 3 and 5 and targeted clone Tg#28zeo^R genomic DNA in lanes 2, 4 and 6. (c) Strategy used to generate the final TNF-α targeted R-Luc reporter cell line by LoxP/Cre-mediated excision of the Zeocin selection cassette. (d) Southern blot analysis of the zeocin-resistant targeted intermediate clone (Tg#28zeo^R, marked as Z⁺), and the Zeocin-sensitive targeted cell line (Tg#28zeo^-, marked as Z^-). Arrows to the left of the blots indicate the DNA fragments cut from non-targeted (solid) and gene-targeted (open) TNF-α alleles. (e) Renilla luciferase (R-Luc) reporter activity in the TNF-α targeted intermediate clone Tg#28zeo^R, the Tg-Zeo^- cell pool following excision of the selection marker, and from 5 individual Tg-Zeo^- clonal cell lines (#1–5) isolated from the Tg-Zeo^- cell pool. Results represent the mean (+/-SEM, N = 4). One-way ANOVA demonstrated no significant difference (p > 0.05) between the targeted cell pool and the five individual targeted clonal cell lines isolated from the pool.

Figure 2

Renilla luciferase activity and TNF-α mRNA expression in TNF-α targeted and non-targeted reporter cell lines. (a) Basal R-Luc reporter activity and TNF-α mRNA levels in the TNF-α targeted (Tg) cell line and in four non-targeted (nTg #1–4) cell lines. (b) Relative R-Luc activity in the targeted cell line (Tg) and in four non-targeted cell lines following treatment with the TNF-α transcriptional activator PMA (5 ng/ml), DOX (5 μM), TSA (100 ng/ml) or Aza-dC (5 mg/ml). (c) The drug-induction profiles of R-Luc activity (top), TNF-α mRNA expression (middle), and R-Luc mRNA expression (bottom) in the Tg cell line. Values represent the mean (+/-SEM, N = 4).

Figure 3

DMXAA treatment of TNF-α targeted and non-targeted R-Luc reporter cell lines. (a) Relative R-Luc activity in the targeted (Tg) and non-targeted (nTg #1–4) reporter cell lines in the absence or presence of DMXAA (100 μg/ml). (b) Dose-response profile for R-Luc activity following treatment of the Tg and nTg4 cell lines with DMXAA. (c) Dose-response profile for DMXAA-induced cytotoxicity in Tg and nTg4 cell lines. (d) Induction of TNF-α mRNA in the Tg and nTg4 cell lines following treatment with DMXAA. Despite the fact that DMXAA only induces R-Luc reporter activity in the Tg cell line, DMXAA induced TNF-α mRNA to equivalent levels in both cell lines. Values represent the mean (+/-SEM, N = 4).

Figure 4

Anthracycline treatment of TNF-α targeted and non-targeted R-Luc reporter cell lines. (a) Dose-response profile of R-Luc activity in the targeted (Tg) cell line treated with four closely related anthracycline antibiotics. Cells treated with Epirubicin were grown on a separate plate. Luciferase activity was assessed using biophotonic imaging on the IVIS. (b) Dose-response profile of cell viability in the Tg line following anthracycline treatment. The same culture plates that are shown in 'a' are also shown in 'b'. Fluorescence was assessed using biophotonic imaging on the IVIS. (c) Relative R-Luc activity in targeted (Tg) and non-targeted (nTg4) cell lines following treatment with a fixed concentration of the different anthracyclines (1 μM). (d) Anthracycline-induced cytotoxicity in the Tg and nTg4 cell lines was similar. Values represent the mean (+/-SEM, N = 4).