Tetracycline (Tet)-controlled transcriptional activators allow for the inducible expression of protein-coding genes or shRNAs, and are frequently used to decipher gene function in cultured cells and in vivo [1]. The system is based on the bacterial Tet operon that mediates resistance to Tet derivatives, such as doxycycline (Dox). In the Tet-off system, fusion of a Tet repressor protein with the transcriptional activation domain of the herpes simplex virus protein VP16 results in a Tet-controlled transcriptional activator (tTA), which constitutively binds to Tet-responsive elements (TRE), but is inhibited in its transcriptional activity by Dox [1]. In contrast, in the more frequently used Tet-on system, a mutant tTA-version generates a reverse tTA (rtTA) that is only recruited to the TRE upon Dox binding and then mediates transgene expression [1]. Since the Tet-system allows transgene expression in a tissue-specific and temporal manner, it has been extensively used for regulating expression of oncogenes or silencing of tumor suppressor genes [1,2,3]. Despite its enormous success, however, we report here that, irrespective of the transgene, the Tet-system might compromise activated T-cells, thereby providing a cautionary note for its use in immunological studies.
To investigate T-cell responses, we used transgenic mice carrying an advanced Tet-on transactivator, driving expression of a miR30-based shRNA in frame with a turboGFP reporter. To reduce effects associated with random integration, the Dox-inducible shRNA-reporter transgene was delivered by a recombinase-mediated cassette exchange approach into the type-I collagen (Col1a1) locus. During an in-depth analysis of the transgenic mice, we discovered that specific splenic T-cell populations were absent not only in those mice expressing shRNAs targeting particular mRNAs, but surprisingly also in mice carrying a control shRNA for Renilla luciferase. These effects were independent of the promoter for rtTA expression, and observed in independent mouse lineages of both ROSA26 and CAG promoter-driven systems. Although total CD4+ T-cells in the spleen were only marginally reduced, antigen-experienced CD4+ T-cells, which we characterized as CD44+ and CD62Llow, were nearly absent in the GFP-positive fractions (Fig. 1). We confirmed the reduction of activated T-cells in the GFP-positive population by CD25 and CD69 staining (Fig. 1). Importantly, disappearance of antigen-experienced T-cells was not mediated by Dox itself, since Dox-treated mice lacking rtTA expression exhibited normal numbers of activated T-cells (Fig. 1). Already 6–10 days of Dox treatment were sufficient to trigger a profound reduction in CD25+/CD44+/CD4+ T-cells, which characterize effector memory or regulatory T-cells. Thus, their rapid disappearance suggests that Dox-activated rtTA/turboGFP expression induces toxicity in these cells (Fig. 1). Interestingly, the frequency of double-negative thymocytes expressing CD25 and CD44 was not altered (Suppl. Fig. S1). However, antigen-experienced GFP-positive CD8+ T-cells, characterized by CD44 expression, were reduced similarly to the CD4+ lineage (Suppl. Fig. S2).
