Ribozyme suppression of endogenous thyroid hormone receptor activity in Xenopus laevis cells

Abstract

Xenopus laevis is an excellent model for thyroid hormone (T3)-regulated gene expression. T3 initiates two drastically different pathways during metamorphosis: death of larval tissues and growth of adult tissues. The role that each T3 receptor (TR) isotype, α and β, plays in metamorphosis is uncertain. The X.laevis tetraploid genome limits experiments to overexpression, misexpression and dominant negative studies. Ribozymes offer an alternative by suppressing gene activity through specific mRNA reduction. It has been suggested that ribozymes will not work in X.laevis because of the organism’s intracellular environment and body temperature. In this study, we show that hammerhead ribozymes are active in vitro against transcribed TRβ message and in vivo against a TRβ–luciferase fusion protein. We next show that TRβ-targeted ribozymes can inhibit T3-induced transcription of a reporter gene in cultured X.laevis cells, using T3 response elements from two T3-responsive transcription factor genes. One has early expression kinetics in response to T3 and is proposed to be TRα regulated whereas the other has intermediate induction kinetics and thus may be partially TRβ regulated. Therefore, ribozymes are a potentially valuable tool for overcoming the limitations in this system for examining gene function in X.laevis.

INTRODUCTION

In Xenopus laevis thyroid hormone (T3) initiates metamorphosis causing the death of larval tissue and the proliferation and differentiation of adult tissue. These two drastically different pathways are controlled by two thyroid hormone receptor (TR) isotypes, α and β (1). The specific role of each receptor in Xenopus metamorphosis is not completely understood because the animal’s tetraploid genotype, lack of a suitable stem cell line and long life cycle prevents the use of gene knockout approaches.

TR gene selectivity is currently predicted using TR overexpression studies or through correlation with spatial and temporal patterns of gene up-regulation (2–5). TRα and TRβ are nearly 100% conserved in the DNA-binding domain, thus, when overexpressed, they may artificially bind identical DNA sequences (6). TRα is expressed in the tadpole before the development of an active thyroid gland (7,8). One model holds that this early TRα expression is important for inhibition of T3 response genes prior to metamorphosis. Increasing TRβ mRNA levels coincide with rising thyroid hormone levels and reach maximal levels at the climax of metamorphosis (7). Early T3 response genes, such as the basic transcription element-binding protein (BTEB) and TRβ genes, may be largely controlled by TRα (9,10). Genes induced with intermediate kinetics, such as basic region leucine zipper transcription factor (TH/bZIP), or late kinetics, such as various protease genes, may be controlled by TRβ (9). During metamorphosis the growing limbs have high TRα levels but maintain low TRβ expression (11). The dying tail has the opposite profile, with low initial TRα and highly inducible TRβ that becomes the predominant TR isotype in the tail at climax (11,12). Therefore, genes induced in the limb are presumed to be controlled by TRα with little if any contribution from TRβ.

Pharmacological experiments with the synthetic TRβ preferential thyroid hormone analog GC-1 have furthered our knowledge of TRβ function (13–15). TRβ and TRα have 87% amino acid homology in the ligand-binding domain, thus preventing GC-1 from exclusively inducing TRβ without affecting TRα. GC-1 binds TRα with 10-fold lower affinity and induces transcription 100 times less effectively than T3 (J.D.Furlow, M.Hsu, H.Y.Yang, D.J.Ermio, W.Lim, G.Chiellini and T.S.Scanlan, unpublished results) (15). GC-1 binds and activates TRβ nearly equally to T3. However, as long as both isotypes are expressed no definitive distinctions can be made between TRα and TRβ.

Ribozymes, RNA with enzymatic activity to specifically cleave RNA, provide an attractive alternative method of reducing specific endogenous mRNAs, suppressing or even eliminating gene activity (16). Ribozymes have the advantage of differentiating between the two isotypes at the nucleic acid level, where they have 75% sequence homology. Rather than introducing exogenous receptor and reporter by transient transfection, ribozymes can specifically suppress one endogenous receptor and determine the effects on reporter gene activity. Ribozymes have been used against multiple targets, including cancer, inherited diseases, and viral infections. It has been suggested that ribozymes are ineffective in the X.laevis embryo because of incompatible salt and pH conditions (17). Previous studies have injected ribozymes against co-injected exogenous targets into Xenopus oocytes. However, these studies subjected the oocytes to non-physiological conditions and transcribed the ribozymes in vitro (18–20). This study examines endogenous TRβ regulation of two T3-responsive genes, BTEB and TH/bZIP (9,21). TRβ selectivity for the two thyroid hormone response elements (TREs) was examined through the use of minimized hammerhead ribozymes, optimized to cleave with a high efficiency compared to the wild-type hammerhead ribozyme (18). We also created and investigated a twinzyme, tethered minimized hammerhead ribozymes with two active domains designed for increased activity.

MATERIALS AND METHODS

Transient transfection assays

Xenopus laevis XLA (kidney) and XTC (fibroblast) cells were maintained and transfected according to Furlow and Brown (9) with a few modifications. Lipofectamine 2000 reagent (Invitrogen) was used at 2 µl/well. In co-transfection experiments with luciferase reporter and ribozyme expression vectors in XLA cells, 0.1 µg TH/bZIP TRE ΔMTV-luciferase or BTEB TRE ΔMTV-luciferase (9) was mixed with 0.8 µg pCS+ribozyme or pCS+GFP3 (a gift of Enrique Amaya, Wellcome/CRC Institute, Cambridge, UK) and 0.1 µg pCS2+ βGalactosidase (a gift of Dave Turner, Fred Hutchinson Cancer Research Center, Seattle, WA). Experiments with TR–luciferase fusion proteins utilized 0.1 µg pCS2+ βGalactosidase, 0.8 µg pCS+ribozyme and 0.1 µg pTR/luc reporter (see below).

In XTC transient transfection experiments using TR and ribozyme expression vectors, cells were transfected in Corning 75 cm² cell culture flasks using 5 µl of Lipofect amine 2000 reagent in 500 µl of OptiMEM (Invitrogen) and 0.5 µg pCS2TRβ (a gift of Haochu Huang and Don Brown, Carnegie Institution of Washington, Baltimore, MD) or miwTRα (9) and 4.5 µg pCS+GFP3 or pCSTRβ ribozyme in a total of 500 µl of OptiMEM. The transfection was done overnight in a total of 5 ml of OptiMEM. The medium was then replaced with 70% L-15 plus stripped fetal bovine serum (FBS) and allowed to recover. Twelve hours later the medium was replaced with 70% L-15 plus stripped FBS. Cells were harvested after 48 h and assayed as described previously (9).

Plasmid constructs

βMiniHH was created by sequential rounds of oligonucleotide elongation. Primers CGTAATACGACTCACTAGTGGGA GATCTGGCGACTTCGG and TTCAGTTTCGTTACCTC ATCAGTTGATGATACCGAAGTGGC were heated to 94°C for 20 s, annealed at 55°C for 20 s and elongated at 72°C in Promega PCR buffer, 3 mM MgCl₂ and 0.2 mM dNTPs. This was repeated for 10 cycles. The product was amplified and elongated with primers CGTAATACGACTCACTAGTG GGAGATCTGGCGACTTCGG and ATCTATCTCGAGGG GTATCGCTTTCTTCATTCAGTTTCGTTACC using standard PCR conditions. The PCR product was inserted into pCS+GFP3 under the control of the cytomegalovirus promoter using XhoI and BglII restriction enzymes. βΔHH was created in a similar fashion using a single round of 10 cycle primer elongation of primers ATGCGGAGATCTCGGTATCAT CTTCTGATGAGTCCGTG and ATCTATCTCGAGTCTT CATTCAGTTTCGTCCTCACGGACTCATC.

βTwin was created by annealing and elongating primers ATCTATCTCGAGGGGTATCGCTTTCTTCATTCAGTTTCGTTACC and TTCGGTATCATCAACTGATGAGGTAA CGAAACTGAA for 10 cycles. Separately, primers TTGA TGATACCGAAGTTTCGTTACCTCATCAGGCCT and ATGCGGAGATCTCAGCCTGCAACAAGGCCTGATC AGGTA were annealed and elongated for 10 cycles. The two cycle products were mixed, denatured at 94°C, annealed at 63°C and finally extended at 72°C for 25 cycles. The final product was inserted into pCS+GFP3 with BglII and XhoI and checked by sequencing.

αMiniHH was created by annealing primers GATCTTC TGGCTCTCTGATGAGGTAACGAAACAAGCAGTC and TCGAGACTGCTTGTTTCGTTACCTCATCAGAGAGCC AGAA and insertion into pCS+GFP3 at the BglII and XhoI sites. The clone was verified by sequencing.

Construction of pTRβ/luc was by first amplifying TRβA1 using primers ATGCGGAGATCTATGGAAGGGTATATA CCC and AGGCCCGGGCTGCAGCTAGTCCTCAAACA CTTC. Amplified TRβA1 coding sequence was then inserted into pCS+GFP3 using BglII and PstI. Luciferase coding sequence was amplifed from pGL-2 Basic (Promega) using primers CTAGGAGCTAGCATGGAAGACGCCAAAAAC and TCTAGACTCGAGGTCATCGCTGAATACAGTTAC ATT. The amplified luciferase coding sequence was inserted in-frame downstream of TRβ using NheI and XhoI, creating pTRβ/luc.

Plasmid pTRα/luc was created by amplyifying TRα using primers ATGCGGAGATCTATGGACCAGAATCTCAGC and CTAGGAGCTAGCGTCTGGATCGTAGCGTACAGC. The amplified product was inserted into pTRβ/luc using NheI and XhoI, creating pTRα/luc.

In vitro cleavage assay

The ribozymes were generated as described above except that a T7 promoter site was added to βMiniHH, βΔHH and αMiniHH using primers CGTAATACGACTCACTATAG GGGGCGACTTCGG, CGTAATACGACTCACTATAGGG CGGTATCATCTTCTG and CGTAATACGACTCACTAT AGGGTCTGGCTCTCTGATG, respectively. The PCR products were transcribed with Promega T7 polymerase at 37°C for 1 h as described by the manufacturer. DNase I was added after transcription and incubated at 37°C for 30 min. Ribozyme transcription was visualized by acrylamide gel electrophoresis and ethidium bromide staining. EcoRI- linearized pSP64A-TRβ (12) was used to transcribe the target using Amersham SP6 polymerase at 40°C for 1 h as described by the manufacturer. DNase I was added after transcription and incubated at 37°C for 30 min.

Equal volumes of transcribed target and ribozyme were incubated together for 3 h (20 mM Tris–HCl pH 7.9, 3 mM MgCl₂, 1 mM spermidine and 5 mM DTT) at room temperature. Samples were briefly heated at 94°C before loading on 6.5% acrylamide–8 M urea gels. Dried gels were exposed to Molecular Dynamics PhosphorImager Screens overnight before scanning using a Storm 680 (Amersham Pharmacia). Quantitation was performed using ImageQuant software.

Quantitative PCR

Total RNA was extracted from transfected cells using Trizol Reagent (Invitrogen) according to the manufacturer’s specifications. Synthesis of cDNA was performed using pd(N)₆ (Amersham Pharmacia), Anti-RNase (Ambion) and Superscript II (Life Technologies) according to the manufacturer’s specifications. Quantitative PCR was performed using a GeneAmp 5700 sequence detection system and SYBR green PCR core reagents (Applied Biosystems). TRβ was detected using primers TGGTGAGATGGCAGTGACAAG and GGCGACTTCGGTATCATCAAG designed with Primer Express (Applied Biosystems). Samples were normalized using oligonucleotides specific to RNA encoding the X.laevis ribosomal protein subunit L8, GCTGAGCTTTCTTGC CACAGT and GGTTGCATTCCGTGATCCTTA. cDNA was diluted 1:500 before PCR. The conditions for PCR were 94°C for 10 min followed by 40 cycles of 94°C for 20 s, 58°C for 30 s, 72°C for 30 s. Since SYBR green non-specifically binds double-stranded DNA, we ensured specificity of the amplified products through analysis by gel electrophoresis and through a dissociation protocol at the end of the quantitative PCR run as described in the manufacturer’s recommendations. All primers produced a single product.

RESULTS

Ribozyme-mediated TRβ mRNA cleavage in vitro and in cultured cells

Synthetically transcribed TRβ-targeted ribozymes (Fig. 1) and TRβ mRNA were incubated together in transcription buffer at room temperature before gel electrophoresis. The GUX TRβ target site was selected based on an MFOLD predicted RNA structure and proximity to a neighboring GUX site, allowing the creation of a tethered ribozyme against both cleavage sites in subsequent experiments. The βMiniHH ribozyme cleaved TRβ in a dose-dependent fashion (Fig. 2). The 300 bp band is not well defined because of early transcriptional termination of the 1100 bp product. This results in inconsistent 3′ ends whose minor differences are not evident in the full-length transcript, but very evident in the cleaved 300 bp product. The negative control ribozymes, βΔHH and αMiniHH, showed no detectable cleavage under these conditions (data not shown). Therefore, βMiniHH can hybridize and cleave its target efficiently at room temperature.

Figure 1.

Schematics of TRβ ribozymes and their secondary structure predictions. The colored sequences are either TRβ (blue) or TRα (red) mRNA sequences with the number and arrow corresponding to the position of the predicted ribozyme cleavage site in the open reading frame. βMiniHH is a minimized hammerhead ribozyme lacking helix II. βΔHH has a nucleotide mismatch (circled) opening helix III near the cleavage site that is predicted to disrupt ribozyme activity. βTwin is comprised of a pair of minimized hammerhead ribozymes with two target sites at nucleotides 815 and 832 in TRβ. αMiniHH is a TRα-targeted ribozyme with a minimized hammerhead active domain.

Figure 2.

In vitro cleavage of a synthetic TRβ transcript. TRβ RNA and increasing βMiniHH ribozyme molar ratios were incubated for 3 h at room temperature before separation on a 6.5% denaturing acrylamide gel. Full-length transcripts and the cleaved products are indicated by arrows. Ribozyme βMiniHH cleaved TRβ transcript in a dose-dependent fashion. The listed concentrations are approximate molar ratios. The band labeled N.S. is a non-specific product of in vitro transcription.

We next tested the ability of ribozymes to cleave endogenous target mRNA in living X.laevis cells. We find that XTC fibroblasts have low basal and uninducible TR mRNA, making it an optimal cell line for measuring transfected TR mRNA reduction by expressed ribozymes (Fig. 3). Untransfected cells showed <1% endogenous TRβ RNA compared to cells transfected with a TRβ expression vector. The βMiniHH ribozyme reduced the expressed TRβ mRNA by 40% in XTC cells (Fig. 3). Ribozyme βΔHH did not affect TRβ RNA levels. In addition, TRα levels were unaffected by βMiniHH expression (data not shown).

Figure 3.

TRβ mRNA cleavage in X.laevis XTC fibroblast cells. Cells were transiently transfected with TRβ and ribozyme or parent control expression vectors. TRβ mRNA levels were measured using quantitative PCR. The amplified product spanned the cleaved site, limiting detection to uncleaved RNA. The untransfected cells had <1% endogenous TRβ compared to transfected cells.

TRβ-targeted ribozyme activity reduces the levels of a TRβ–luciferase fusion protein as well as transcription of TRE-containing reporter genes

The ability of a ribozyme to reduce TRβ RNA levels was measured by two different methods. One method measured ribozyme cleavage of mRNA encoding a TRβ–luciferase fusion protein. If the mRNA encoding the fusion protein is cleaved the downstream luciferase sequence will not be translated, thus leading to decreased luciferase activity. Compared to the control vector, co-transfection of the βMiniHH ribozyme expression vector decreased luciferase activity in a dose-dependent manner (Fig. 4). The highest ribozyme concentrations showed the greatest luciferase reduction of 40%. In an attempt to increase the efficiency of cleavage, a paired βminiHH was created that targeted adjacent sites in the TRβ message (βTwin). The βTwin ribozyme decreased luciferase activity of the fusion protein by 54% (Fig. 5). The βMiniHH and βTwin ribozymes were statistically different from control vector transfection, but were not significantly different from each other. Negative control ribozymes βΔHH and αMiniHH showed no significant cleavage activity (Fig. 5). In addition, expression of a TRα–luciferase fusion was not affected by the ribozyme βMiniHH.

Figure 4.

βMiniHH ribozyme cleaves TRβ–luciferase in a dose-dependent fashion in XLA cells. Luciferase activity of the TRβ–luciferase fusion protein decreased in response to increasing βMiniHH ribozyme. The asterisks above the 1:4 and 1:8 molar ratios denote statistical significance compared to the no ribozyme control.

Figure 5.

TRβ–luciferase fusion RNA is cleaved in vivo but not TRα–luciferase fusion RNA. Xenopus laevis XLA kidney cells were transiently transfected with constructs bearing CMV promoters driving TRβ– or TRα–luciferase fusion and ribozyme or parent expression vector control. Ribozyme activity was measured through inhibition of TR–luciferase translation and expression. The asterisks denote a significant reduction in luciferase activity compared to control based on an unpaired t-test with a P value of 0.05.

The second method measured ribozyme effects on TRβ transcriptional activation of T3-responsive reporter genes. XLA cells that contain functional endogenous TRα and inducible TRβ were transfected with T3-responsive TH/bZIP or BTEB TRE upstream of a minimal promoter driving the firefly luciferase gene (9). The βMiniHH ribozyme reduced TH/bZIP and xBTEB TRE induction by 23 and 45%, respectively (Fig. 6). The effect of the βMiniHH ribozyme was statistically different from the control. βTwin ribozyme expression reduced TH/bZIP and xBTEB TRE induction by 23 and 51%, respectively (Fig. 6). Ribozyme βΔHH did not affect TRβ induction of the TH/bZIP TRE, which is consistent with the in vitro cleavage assay. This result demonstrates that the ribozyme is responsible for the reduction in TRβ activity and not RNA interference by the antisense portion of the ribozymes. In these experiments we could not use αMiniHH as a negative control because it is designed to cleave TRα mRNA. TRα is known to bind and induce activity of these TREs (9) (J.D.Furlow and A.Kanamori, unpublished results). We conclude that the minimized TRβ ribozymes described here can successfully reduce endogenous TRβ and thereby selectively reduce its activity in living Xenopus cells.

Figure 6.

Expression of the βMiniHH ribozyme inhibits transcription of a T3-responsive promoter in XLA cells. Cells were transiently co-transfected with a TRE-containing luciferase reporter and ribozyme or control parent expression vectors. Closed bars indicate cells treated with 10 nM T3 and open bars represent untreated cells. The single asterisk denotes a significant reduction in luciferase activity compared to the control. βTwin ribozyme further reduced BTEB TRE ΔMTV luciferase activity compared to βMiniHH ribozyme and is therefore marked with a double asterisk. Expression of the βΔHH ribozyme showed no cleavage compared to the control parent vector.

DISCUSSION

Ribozymes have been used as a tool in multiple organisms in an attempt to reduce endogenous mRNA targets. Previous studies in the Xenopus system synthesized the ribozyme in vitro and injected the ribozyme into oocytes. This is the first study that synthesized the ribozymes within an amphibian cell against an endogenous mRNA target, namely TRβ. This study focused on two different hammerhead ribozyme strategies, a single minimized hammerhead riboyzme and a twinzyme. Minimized hammerhead ribozymes were used in both cases because free magnesium ions are not abundant in living cells and activity of the full ribozyme is dependent on its availability (22).

Twinzymes were originally designed with hairpin ribozymes as a tool for cleaving multiple targets (23). We attempted to utilize this strategy to increase ribozyme activity in Xenopus cells because we reasoned that there might be an increase in targeting errors or decreased activity of the ribozyme at ambient temperatures. The presence of two active domains could ensure that at least one ribozyme may align properly and cleave the target. βTwin showed increased activity in the TRβ–luciferase fusion and BTEB TRE ΔMTV reporter assays. However, only the BTEB TRE experiment showed a clear statistical difference between the two ribozymes. Thus, βTwin may have only slightly more activity in this system.

One practical application of ribozyme-mediated RNA suppression is to selectively decrease the expression of a specific member of a gene family. Transient transfection experiments examining TR isotype-specific transcriptional regulation typically express receptor protein at levels far exceeding physiological concentrations in cells that typically do not express the receptor (9). The two TR isotypes share nearly 100% amino acid homology in the DNA-binding domain, therefore receptor overexpression could allow for artificial binding of TRβ to a TRα response element. Ribozymes can specifically reduce one endogenous TR isotype and examine selective induction at physiological TR expression levels. TH/bZIP and BTEB are T3-responsive genes with different induction kinetics. TH/bZIP was suspected to be TRβ regulated based on minimal cycloheximide resistance and intermediate T3 response kinetics (9). BTEB was suspected to be a TRα-regulated gene based on greater cycloheximide resistance and earlier T3 response kinetics. In this study, both TRE were found to be susceptible to a reduction in TRβ expression. Therefore, both genes may be primarily TRα regulated with some additional TRβ dependence. This model follows from previous experiments with the TH/bZIP TRE showing that co-transfection of TRα or TRβ with a wild-type TH/bZIP promoter resulted in higher levels of induction but did not change the expression kinetics per se. Properties of the promoter other than TR binding to the TRE may delay TH/bZIP expression (9). In addition, TRβ represses transcription in the absence of T3 (24). Neither TRE ΔMTV reporter showed any relief of repression in the presence of ribozymes. Since XLA cells have no detectable TRβ in the absence of T3, transcriptional repression of these TREs is likely mediated only through TRα (25).

In summary, we demonstrate that minimized hammerhead ribozymes can cleave endogenous mRNA targets in X.laevis cells. Ribozymes have multiple potential uses in Xenopus. Using the recently described transgenesis approach in this organism, ribozymes can be used to make functional knockouts of specific genes. In particular, they can be designed to target only selected members of a closely related gene family. At first glance, 40–50% inhibition of expression may not seem sufficient. While it may be difficult to further reduce mRNA levels, if the protein made from that message is limiting in cells, like many transcription factors, this level of mRNA reduction may be enough to produce a scorable phenotype. Therefore, ribozymes are a potentially powerful tool for functional analysis of X.laevis genes that circumvent the tetraploid genotype and long life cycle of this important model organism.