We created a modular system that uses Gateway recombineering technology to combine the binarytetracycline-regulated expression components into a single vector. By using a targeting vector with a known genomic insertion site (Rosa26), we could introduce all of the components needed for tetracycline-regulated expression into a well-characterized and transcriptionally-active locus. This strategy allows us to use the same cell line for both the experimental (that is, Rs1 expressed) and control (that is, Rs1 expression suppressed) conditions, thus significantly decreasing the problem of clonal variability.
Although a variety of methods for combining the tetracycline-inducible expression components into a single vector have been described [14
], our modular system has the added advantage of facilitating the rapid interchange of the expression components for tetracycline-induced expression as well as the choice of viral and non-viral delivery backbones. We easily inserted the expression cassettes into two final destination vectors, including a new Gateway-enabled construct for targeting into the mouse Rosa26 locus. We then were able to integrate the combined tTA-TetO construct into the Rosa26 locus of pluripotent mouse cells in a single targeting step.
Our expression strategy also took advantage of a P2A ribosomal skip site to co-express the engineered GPCR Rs1 and the fluorescent marker mCherry. The P2A sequence has been used in a variety of systems to generate independent polypeptides from a single RNA transcript [27
]. In our experiments, combining the mCherry and Rs1 cistrons together allowed us to visualize the expression of mCherry and Rs1 and assess the effect of Rs1-induced Gs
signaling in both ES cell differentiation and mice. Our results indicate that the 2A strategy produces functional membrane-bound receptors and reporter proteins from a single RNA transcript. This strategy will be useful for expressing other GPCRs since it does not require the addition of a large fusion protein domain to the receptor.
In our study, expression of Rs1 and activation of Gs
signaling in mouse ES cells induced larger EB size. Although Gs
-induced cellular proliferation by cholera toxin can increase EB size [49
], our results indicated no detectable increase in cellular proliferation, decrease in apoptosis, or change in ES cell differentiation. We were unable to definitively assess whether a change in cell size could be contributing to the increased EB size; however, the difference in observed ES cell proliferation may be a result of lower Gs
activation from the weak EF1α
promoter activity in our system, lower expression of Rs1 in the ES-cell derived differentiated tissues (as indicated by our mouse results), or differences in how cholera toxin or GPCRs activate the Gs
or non-cannonical GPCR signaling pathways.
Although a single copy of the EF1α
promoter could drive doxycycline-dependent expression of our construct in ES cells, it was insufficient to drive high expression of the transgenes in differentiated mouse tissues. Our study examined a limited number of founder mice and did not assess whether EF1α-tTA expression varies between founder lines. However, several other reasons could result in low levels of tTA expression: the EF1α
promoter may not function as robustly in differentiated tissues as in pluripotent cells; the flanking insulator sequences could result in transgene silencing; or the reduced number of TetO repeats in our constructs results in a lower sensitivity to tTA activation. The latter two possibilities are less likely since our results from the mouse crosses indicate that the TetO-mCh-Rs1 portion of the transgene can respond to a separate transgene expressing tTA from the ColI(2.3) promoter, suggesting that the R26(EF1α-tTA/TetO-mCh-Rs1)-targeted locus is not silenced. In addition, prior studies show that constructs with as few as two TetO repeats can be induced by tTA [50
]. Finally, while we cannot exclude that the parallel positioning of the EF1α-tTA construct with the endogenous Rosa26 promoter may cause transcriptional interference as previously reported for the CMV promoter [51
], this should have been minimized by the use of a 5' insulator sequence.
The observed low levels of Rs1 expression in mouse tissues likely accounts for the absence of embryonic lethality and other pathology we would have expected if our model mimicked McCune-Albright syndrome [44
]. In addition, the differential activity of the EF1α
promoter, or different susceptibility of certain tissues to Gs
signaling, may be contributing to the heterogeneity of the Rs1 expression we observed in our R26(EF1α-tTA/TetO-mCh-Rs1) mice. Increasing the expression of the tTA transactivator by using an alternative ubiquitous promoter, such as the CMV-β actin (CAG) promoter [52
], may allow more robust expression of Rs1 and increased activation of the Gs
We believe that the modularly-designed single-vector tet system presented here provides an ideal system for the tissue specific expression of tTA or rtTA as well as the controlled expression of a transgene from the tetracycline response element. This system could be used for a variety of genetic studies where a single cassette is advantageous. Having both components of the tet-inducible system in the same cassette could facilitate the genetic modification of ES and induced pluripotent stem cells for regulatory studies as well as for making engineered tissues.
To facilitate these applications, we generated a series of improved vectors (Table S2 in Additional File 3
and Figures S2A-F in Additional File 4
). We created a new Rosa26 targeting vector that contains the yeast PI-SceI homing sequence (pR26 R1R2 RexNeo PI-SceI) to simplify linearization of large constructs. We also created pEntL1L3 and pEntR3L2 plasmids without the insulator sequences and adapter plasmids containing only a multiple cloning sequence (pEntL1L3-MCS and pEntR3L2-MCS) to allow use in lentiviral expression constructs (that is, to generate independent tet-regulator and -responder expression constructs in a Gateway-compatible lentivirus backbone such as pLenti6 (Invitrogen)). We believe that these constructs will help advance the generation of new inducible expression models. In addition, these constructs are compatible with new technologies such as transposon-mediated gene transfer, which can move large segments of DNA (> 20 kb) as a single cassette and can be easily mobilized into or excised from the genome [53
], as well as newer versions of the tet regulatory components such as Ptet [54