Manipulation of Transcription Factor Activity Using Fluorescein-Tagged Dumbbell Oligonucleotides

The functional analysis of transcription factors has contributed tremendously to the understanding of fundamental biological phenomena as well as disease mechanisms. Transcription factors exert their function indirectly by regulating their responding or target genes. The set of genes regulated by the same regulatory protein can vary significantly between different tissues or organs. Hence, the cellular context influences the physiological pathway that a given transcription factor will initiate. It is, therefore, conceivable that the knowledge of responding genes will provide profound insight into the biological role and function of the corresponding transcription factor. Consequently, the inhibition of specific regulatory proteins is a promising approach to identify responding genes in combination with powerful methods such as differential display polymerase chain reaction (PCR) (7) or PCR-Select (CLONTECH Laboratories, Palo Alto, CA, USA) (4). It has been shown that the transcriptional activity of DNA-binding proteins within the cell can be inhibited by introducing double-stranded oligonucleotides (1–3,8–10). The regulatory sequences of these oligomers compete specifically for transcription factor binding with their endogenous counterparts in the target gene promoters. Circular dumbbell oligonucleotides (dumbbells) provide increased stability towards nucleolytic degradation in a biological environment because of the absence of free termini and therefore are excellent tools for specifically inhibiting selected transcription factors (3). We describe the use of fluorescein-labeled dumbbells in combination with fluorescence-activated cell sorting (FACS) as a way to obtain well-defined pools of equally manipulated cells lacking the activity of a selected transcription factor. As a result, the homogeneity of the cellular material allows the analysis of differential gene expression and isolation of target gene candidates. RORα is a transcription factor that belongs to the superfamily of nuclear hormone receptors and is expressed in the brain, several peripheral tissues and immune cells. Recently, it has been shown that a deleted version of the RORα gene causes severe cerebellar ataxia in the natural mutant mouse staggerer (6). Highest expression of RORα is found in T and B cells, yet its function in these cells remains unclear. Therefore, we were interested in identifying RORα responding genes in human T cells. Since low transfection efficiency of T cells is notorious, selection of transfected cells is necessary to obtain a homogeneous dumbbell-containing cell population. For this purpose, we generated a fluorescein-labeled dumbbell (Figure 1) in the manner described below. Two 86-bp 5′-phosphorylated DNA oligomers were modified with a fluorescein phosphoramidite (CLONTECH) during synthesis on a Model 8909 Oligonucleotide Synthesizer (PerSeptive Biosystems, Cambridge, MA, USA). The modification of the oligonucleotides must not influence the interaction with RORα. For this reason, the fluorescein label was linked to position 68 located in the T-loop region and thus sufficiently distant from the binding site. After running the oligomer on a 12% polyacrylamide gel (7 M urea) that was analyzed by ultraviolet (UV) shadowing, a single band of the expected size was excised. The concentration of the purified oligomer was determined by UV absorption. To obtain a dumbbell product, the oligomer was ligated as described in Clusel et al. (3). The ligation efficiency was determined by incubating 1 pmol of the ligated or nonligated oligomer with 20 U of exonuclease III for 30 min at 37°C in 20 μL of the appropriate buffer. The enzyme exclusively digested the nucleotides 3′ of T loop T2 (Figure 1A) of non-ligated oligomers because of their free (non-ligated) 3′ OH group. The reaction was stopped by adding 0.2 μL 3 M NaOAc (pH 5.0), followed by incubation with 10 U mung bean nuclease, which digested all the single-stranded DNA, resulting in a 36-bp fragment of the ligated, but only a 12-bp fragment of the non-ligated, form of the original molecules. The analysis on a 4% MetaPhor® agarose gel (FMC BioProducts, Rockland, ME, USA) revealed a high efficiency of ligation because the smaller fragment was not visible in the ligated sample. The dumbbells contained either an optimal RORα response element (RORE: AATATGGGTCA for db.rore) or a scrambled element (ATCGTGGTAAA for db.c) embedded in an otherwise identical 36-bp doublestranded region flanked by a singlestranded loop of seven thymidines. DNA bandshift experiments demonstrated that db.rore was able to bind to RORα in vitro, while db.c did not (not shown). The dumbbells were then tested for their ability to compete with promoters containing a RORE in cellular reporter gene assays using the T-cell line Jurkat. For this purpose, we used a DNA reporter plasmid in which the expression of the gene encoding chloramphenicol acetyltransferase (CAT) is driven by a thymidine kinase (tk) minimal promoter fused to a RORE (RE.CAT). In each experiment, 0.4 μg of a pSG5 vector (Stratagene, La Jolla, CA, USA)-based RORα overexpressing plasmid (pRORα), 1.6 μg RE.CAT as reporter and 0.2 μg pCH110 (Pharmacia Biotech, Piscataway, NJ, USA) as control for transfection efficiency, as well as the indicated concentrations of db.c or db.rore, were added to 4% LIPOFECTAMINE in RPMI 1640 (both from Life Technologies, Gaithersburg, MD, USA). The amount of DNA was always about 2.2 μg in 200 μL solution. After 30 min incubation at room temperature (RT), the solution was added to 5 × 106 Jurkat cells in 0.8 mL RPMI 1640 medium, and the mixture was incubated at 37°C. After 4 h, the solution was replaced by 10 mL RPMI 1640/ 10% fetal calf serum (FCS) followed by an incubation for 16 h at 37°C. The CAT and β-galactosidase activities were measured using the CAT and βGalactosidase Enzyme Assay Systems (Promega, Madison, WI, USA) following the manufacturer’s protocol. Figure 1B shows that db.rore significantly reduces RORα activity in the Jurkat T cells. A reduction by approximately 50% of the tkRORE promoter activity was observed at 3 nM dumbbell