Constitutive Gs activation using a single-construct tetracycline-inducible expression system in embryonic stem cells and mice

Abstract

Introduction: The controlled expression of many genes, including G-protein coupled receptors (GPCRs), is important for delineating gene functions in complex model systems. Binary systems for inducible regulation of transgene expression are widely used in mice. One system is the tTA/TRE expression system, composed of a tetracycline-dependent DNA binding factor and a separate tetracycline operon. However, the requirement for two separate transgenes (one for each tTA or TRE component) makes this system less amenable to models requiring directed cell targeting, increases the risk of multiple transgene integration sites, and requires extensive screening for appropriately-functioning clones.

Methods: We developed a single, polycistronic tetracycline-inducible expression platform to control the expression of multiple cistrons in mammalian cells. This platform has three basic constructs: regulator, responder, and destination vectors. The modular platform is compatible with both the TetOff (tTA) and TetOn (rtTA) systems. The modular Gateway recombineering-compatible components facilitate rapidly generating vectors to genetically modify mammalian cells. We apply this system to use the elongation factor 1α (EF1α) promoter to drive doxycycline-regulated expression of both the fluorescent marker mCherry and an engineered Gs-coupled GPCR "Rs1" separated by a 2A ribosomal skip site.

Results: We show that our combined expression construct drives expression of both the mCherry and Rs1 transgenes in a doxycycline-dependent manner. We successfully target the expression construct into the Rosa26 locus of mouse embryonic stem (ES) cells. Rs1 expression in mouse ES cells increases cAMP accumulation via both basal and ligand-induced Gs mechanisms and is associated with increased embryoid body size. Heterozygous mice carrying the Rs1 expression construct showed normal growth and weight, and developed small increases in bone formation that could be observed in the calvaria.

Conclusions: Our results demonstrate the feasibility of a single-vector strategy that combines both the tTA and TRE tetracycline-regulated components for use in cells and mouse models. Although the EF1α promoter is useful for driving expression in pluripotent cells, a single copy of the EF1α promoter did not drive high levels of mCherry and Rs1 expression in the differentiated tissues of adult mice. These findings indicate that promoter selection is an important factor when developing transgene expression models.

Figure 1

A single-vector tetracycline construct allows doxycycline-regulated expression. (A) Overview showing the regulator plasmid containing the EF1α-tTA cassette (pEntL1L3-EF1α-tTA, left) and the responder plasmid containing the TetO-mCh-Rs1 cassette (pEntR3L2 TetO-mCh-Rs1, right). The mCherry and Rs1 cistrons are separated by a P2A ribosomal skip sequence to allow simultaneous expression of both peptides. The entry plasmids were recombined using Gateway technology into the desired destination vector containing the AttR1 and AttR2 Gateway sites. The TetO and EF1α-tTA portions are in opposite orientation (indicated by upside-down text) to minimize steric hindrance between the two promoters, as well as potential cross-activation of the TetO by the EF1α promoter. In addition, flanking insulator sequences are included to minimize any read-through activation of the constructs by surrounding promoters (such as Rosa26) that may lead to "leakiness" or steric interference from endogenous promoter activity. (B, C) HEK-293 cells carrying the Exp-pcDNA3.2(EF1α-tTA/TetO-mCh-Rs1) expression cassette and cultured in doxycycline (suppressed expression) or in the absence of doxycycline (transgene expression allowed) demonstrate doxycycline-dependent mCherry expression. (D) Schematic of targeted Rosa26 locus and Southern screening strategy. The Rosa26 locus in E14 ES cells was targeted by homologous recombination with the Exp-R26(EF1α-tTA/TetO-mCh-Rs1) construct. Regions in hatch marks indicate the 5' and 3' homology regions of the targeting vector and the endogenous Rosa26 locus (abbreviated R26 in the figure). The location of the 5' recombination Southern probe and HindIII restriction sites are indicated. (E) Southern blots of genomic DNA digested with HindIII and probed as in (D). Heterozygous ES cells at the Rosa26 locus are indicated by the two bands.

Figure 2

R26(EF1α-tTA/TetO-mCh-Rs1) function in mouse ES cells. (A) FACS analysis showing doxycycline-inducible mCherry expression in E14 mouse cells carrying the Exp-R26(EF1α-tTA/TetO-mCh-Rs1) construct. (B) Induction of Rs1 expression and treatment with the agonist RS67333 results in increased cAMP accumulation in mouse ES cells. Both basal and ligand-induced increases in cAMP are detectable. In addition, serotonin does not induce increased cAMP accumulation in either the wildtype or Rs1-expressing cells. (C) Schematic showing the differentiation protocol for making suspension EBs. (D) Expression and ligand activation of Rs1 during EB formation results in larger EB size. (E) Quantitation of EB size in the different culture conditions using ImageJ. A minimum of 114 EBs were measured for each condition. The analysis was performed on three separate EB differentiation experiments with similar results. Error bars represent average +/- 1 SD.

Figure 3

Rs1 expression in R26(EF1α-tTA/TetO-mCh-Rs1) mice show significant variability and degrees of induction among different tissues. Rs1 expression was allowed (off doxy) or suppressed (on doxy) in mice for four weeks starting from gestation. mRNA levels were assessed by qPCR for Rs1. Representative mice are shown. Error bars represent average +/- 1 SD of technical triplicates.

Figure 4

A single copy of the EF1α-tTA regulator region weakly drives expression of a TetO transgene in mice. (A) Areal bone mineral density by DEXA of nine-week-old mice shows that the ColI(2.3)-tTA x R26(EF1α-tTA/TetO-mCh-Rs1) mice have increased bone mass. N = 9 WT, 5 R26(EF1α-tTA/TetO-mCh-Rs1), and 9 ColI(2.3)-tTA x R26(EF1α-tTA/TetO-mCh-Rs1) mice. ***, P < 0.0001 vs. wildtype (WT). (B) RNA expression levels of Rs1, tTA, and mCherry in the humeri of nine-week-old ColI(2.3)-tTA x R26(EF1α-tTA/TetO-mCh-Rs1) and littermate controls, showing highest expression in the double-mutant mice. Error bars represent means of technical qPCR triplicates +/- 1 SD. (C) Fluorescence images of calviaria from representative 16-week-old littermate mice showing strongest mCherry expression [assessed in the Texas Red (TxR) channel] in ColI(2.3)-tTA x R26(EF1α-tTA/TetO-mCh-Rs1) but not wildtype littermates. Although mCherry was not clearly detected in the R26(EF1α-tTA/TetO-mCh-Rs1) calvaria, subtle increases in the mineralization of the skull bones (white patches, as indicated by the yellow arrow) are present indicating functional responses to Rs1. The level of bone formation is lower than that seen in the ColI(2.3)-tTA x R26(EF1α-tTA/TetO-mCh-Rs1) (green arrow), where additional activation of the Rs1 transgene is achieved by adding osteoblast-specific expression of the tTA element.