Upregulation of Protooncogene Ski by Thyroid Hormone in the Intestine and Tail during Xenopus Metamorphosis

Abstract

Thyroid hormone (T3) is important for adult organ function and vertebrate development, particularly during the postembryonic period when many organs develop/mature into their adult forms. Amphibian metamorphosis is totally dependent on T3 and can be easily manipulated, thus offering a unique opportunity for studying how T3 controls postembryonic development in vertebrates. Numerous early studies have demonstrated that T3 affects frog metamorphosis through T3 receptor (TR)-mediated regulation of T3 response genes, where TR forms a heterodimer with RXR (9-cis retinoic acid receptor) and binds to T3 response elements (TREs) in T3 response genes to regulate their expression. We have previously identified many candidate direct T3 response genes in Xenopus tropicalis tadpole intestine. Among them is the proto-oncogene Ski, which encodes a nuclear protein with complex function in regulating cell fate. We show here that Ski is upregulated in the intestine and tail of premetamorphic tadpoles upon T3 treatment and its expression peaks at stage 62, the climax of metamorphosis. We have further discovered a putative TRE in the first exon that can bind to TR/RXR in vitro and mediate T3 regulation of the promoter in vivo. These data demonstrate that Ski is activated by T3 through TR binding to a TRE in the first exon during Xenopus tropicalis metamorphosis, implicating a role of Ski in regulating cell fate during metamorphosis.

1. INTRODUCTION

Thyroid hormone (T3) plays important roles for organ development in vertebrates. Misregulation of T3 function during vertebrate development leads to moderate to severe pathological consequences such as cretinism in human, a developmental disease characteristic of growth retardation, developmental delay, and impaired mental development. T3 functions mainly during the postembryonic period when many organs mature into their adult forms. This period encompasses a few months prior to birth to several months after birth during human development when the level of T3 in the plasma gradually increases, peaks around birth, and then decreases to a steady adult level (Tata 1993). The studies on T3 function during mammalian development have been difficult, in part because the uterus-enclosed mammalian embryos are not easily manipulatable and the influence of maternal T3 makes it difficult to determine the in vivo T3 action in mammals even after birth. In contrast, the development of anurans, such as the highly related Xenopus laevis and Xenopus tropicalis, proceeds externally after fertilization and is independent of maternal hormone influence, thus providing an ideal animal model for studying developmental regulation of diverse processes such as organ maturation and tissue regeneration (Tata 1993; Gilbert et al. 1996; Shi 1999; Li et al. 2016; Marshall et al. 2019; Wang and Shi 2021).

Anuran metamorphosis is the postembryonic development when a tadpole undergoes extensive tissue remodeling to become a frog. Interestingly, anuran metamorphosis is totally controlled by T3 and can be easily manipulated via the availability of T3 to the tadpole, either by interfering with T3 synthesis in the tadpole or by adding physiological concentrations of T3 in the rearing water. During anuran metamorphosis, tadpoles undergo drastic changes in essentially every organ/tissue, including de novo development of adult organs, complete resorption of larval specific ones, and remodeling of others (Dodd and Dodd 1976; Shi 1999). The drastic tissue transformation is similar to that during mammalian postembryonic development, including the maturation and the remodeling of the brain, intestine, and the lung, etc. Such properties make anuran metamorphosis a unique model to study how T3 regulates postembryonic development.

Numerous studies have shown that T3 affects anuran metamorphosis through T3 receptor (TRs)-mediated gene regulation (Shi 1994; Puzianowska-Kuznicka et al. 1997; Sachs et al. 2000; Sachs and Shi 2000; Schreiber et al. 2001; Buchholz et al. 2003; Nakajima and Yaoita 2003; Buchholz et al. 2004; Buchholz et al. 2006; Brown and Cai 2007; Bagamasbad et al. 2008; Denver et al. 2009; Schreiber et al. 2009; Shi 2009; Shi et al. 2012; Grimaldi et al. 2013; Choi et al. 2015; Sachs 2015; Wen and Shi 2015; Yen 2015; Wen and Shi 2016; Choi et al. 2017; Wen et al. 2017; Buchholz and Shi 2018; Nakajima et al. 2018; Sakane et al. 2018; Nakajima et al. 2019; Shibata et al. 2020; Shi 2021). TRs form heterodimers with 9-cis retinoic acid receptors (RXRs) to regulate target gene transcription by recruiting transcriptional cofactors in a T3-dependent manner, i.e., recruiting corepressors in the absence of T3 to repress gene expression but coactivators to activate gene expression in the presence of T3. Thus TR/RXR has dual functions in regulating target gene expression depending on the availability of T3. There are two conserved TR subtypes in vertebrates, TRα and TRβ, that are encoded by separate genes but function similarly in forming TR/RXR heterodimers and mediating T3 regulation of T3 target genes (Laudet and Gronemeyer 2002). In Xenopus, TRα and TRβ expression is distinctly regulated during development. TRα expression rises in premetamorphic stages and remains at relatively high levels to the climax of metamorphosis and thus may play important functions in coordinating tadpole growth and developmental timing through T3-dependent dual functions (Wen et al. 2017). On the other hand, TRβ is expressed at very low levels before metamorphosis but activated by T3 at the onsite of metamorphosis with peak level expression at the climax of metamorphosis to participate in metamorphosis (Nakajima et al. 2018; Nakajima et al. 2019). Thus, TRα and TRβ likely function coordinately together to control developmental and metamorphic processes by regulating T3 target genes. In order to identify direct TR target genes during metamorphosis, we have previously performed a ChIP (chromatin immunoprecipitation)-on-chip analysis of TR binding in the intestine of premetamorphic tadpoles treated with or without T3 (Fu et al. 2017). Among the putative TR target genes thus identified is Ski, a proto-oncogene that was initially identified in chicken to be homologous to a gene inserted into the avian Slogan-Kettering retroviruses (v-Ski) (Li et al. 1986).

Ski was characterized as a proto-oncogene due to its characteristic capability of inducing cellular transformation in vitro (Stavnezer et al. 1981; Colmenares and Stavnezer 1989; Sun et al. 1999). Consistently, knockdown of Ski inhibits human melanoma tumor growth in vivo (Chen et al. 2009) and Ski expression is up-regulated in many types of cancer cells such as colorectal cancer (Buess et al. 2004; Bravou et al. 2009), esophageal squamous cell carcinoma (Fukuchi et al. 2004), pancreatic cancer (Heider et al. 2007), and breast cancer (Theohari et al. 2012; Wang et al. 2013), etc. Ski also plays important roles during embryogenesis and tissue homeostasis by functioning as a key transcriptional corepressor of Smad proteins to regulate the transforming growth factor-β(TGF-β)/Smad signaling pathway (Luo 2004; Bonnon and Atanasoski 2012; Tecalco-Cruz et al. 2018). While many studies have implicated that Ski plays important roles during vertebrate development and misregulation of its expression leads to moderate to serve pathological consequences such as cancers, much less is known about the regulation of Ski gene expression. We show here that Ski is activated in the intestine and tail during Xenopus tropicalis metamorphosis and that TR/RXR binds to a thyroid hormone response element (TRE) within the first exon of Ski gene to mediate transcriptional activation of the Ski promoter by T3.

2. MATERIALS AND METHODS

2.1. Experimental animals

Xenopus tropicalis tadpoles and Xenopus laevis frogs were purchased from Nasco (Fort Atkinson, MI) or reared in the laboratory. Tadpoles were staged according to Nieuwkoop and Faber (Nieuwkoop and Faber 1965). When indicated, premetamorphic tadpoles at stage 54 were treated with 10 nM T3 in rearing water at 25 °C for 2 days. All animal procedures were done as approved by NICHD Animal Use and Care Committee.

2.2. Quantitative RT-PCR

Genomic DNA-free total RNA was isolated from the intestine and tail of tadpoles at indicated stages during natural metamorphosis or premetamorphic tadpoles at stage 54 treated with or without 10 nM T3 for 2 days and 2.0 μg of each total RNA were subjected to cDNA synthesis as described (Fu et al. 2022). At least three tadpoles per sample were analyzed for each stage or day of T3 treatment and at least two samples were collected from pooled tissues for each stage or T3-treatment conditions. Quantitative PCR (qPCR) on the cDNA was carried out in technical replicates to evaluate gene expression by using SYBR Green PCR Master Mix on a StepOne Plus Real-Time PCR System (Applied Biosystems, Foster City, CA). The primers used were 5’- AAGCAGCCAGGATTTCCTCTG-3’ (forward, located in exon 1 & 2) and 5’- GGAAAGCAGACAGACGCTGA-3’ (reverse, located in exon 3) for Ski. EF1α (elongation factor 1α) was used as the normalization control as described (Das et al. 2009).

2.3. Bioinformatic identification of TREs

The sequence of X. tropicalis Ski gene was downloaded from ENSEMBL website (https://useast.ensembl.org/index.html) and the computational analysis tool NHR-Scan (http://nhrscan.genereg.net/cgi-bin/nhr_scan.cgi ) (Sandelin and Wasserman 2005) was used to identify putative TREs.

2.4. Gel mobility shift assay.

Gel mobility shift assay was performed essentially as described (Das et al. 2009). In brief, Xenopus tropicalis TRα and RXRβ proteins were made in vitro by using TnT T7 Quick Coupled Transcription/Translation System (Promega, Madison, WI) to mix together to form the TRE-binding TR/RXR heterodimers. They were mixed with infrared dye IR700 (LI-COR, Lincoln, NE)-labeled TRE of Xenopus laevis TRβA gene, a well-characterized T3 target gene (Ranjan et al. 1994), in the in vitro binding reactions in the presence or absence of indicated unlabeled competitors. The unlabeled competitors were made by annealing complementary oligonucleotides containing the putative TREs in the PCR buffer. The sequences of sense strand of oligonucleotides were: 5’- ATTTGTGACCTATGGTAACCATCCAT-3’ (Ski TRE1), 5’- ATTTGTGACCTGCGCGGAACTTCCAT-3’ (Ski TRE2), 5’- ATTTGTGATTTATGGTAACTATCCAT-3’ (Ski mTRE1) and 5’- ATTTGTGACTTGCGCGGATTTTCCAT-3’ (Ski mTRE2) (Bold letters indicate the TRE half sites with the mutated nucleotides underlined in the mutant TRE), respectively. Each binding reaction included 1 μl of 100 fmol of IR-700 labeled probe, 1 μl of each TR and RXR in vitro translated protein mixture, and 1 μl of the wild type or mutant TRE oligonucleotides at 400 fmol/μl, 2 pmol/μl, or 10 pmol/μl, respectively, to obtain 4×, 20×, or 100× unlabeled competitor oligonucleotides, respectively, in a total volume of 20 μl. The mixtures were incubated at room temperature for 20 min., electrophoresed on a 6% DNA retardation gel (Invitrogen, Carlesad, CA) and then scanned by using an Odyssey Infrared Scanner (LI-COR, Lincoln, NE).

2.5. Generation of luciferase reporter constructs for transcription assay.

The firefly luciferase reporter constructs containing Xenopus tropicalis Ski promoter with wild type or mutant TREs were made by using pGL4.10 firefly luciferase vector (Promega) through PCR cloning with high fidelity PrimeSTAR GXL DNA Polymerase (Takara, Mountain View, CA). A fragment encompassing the basic Ski promoter without any putative TRE was amplified by PCR on X. tropicalis genomic DNA with primers Ski_F (5’- CGCGCCGCTAGCAGCCTTATGTGGGGGAATAGT-3’ (restriction enzyme NheI recognition sequences underlined)) and Ski_R (5’-CGCGCCAAGCTTGCTCACGAGTTCCCCTTCAT-3’ (restriction enzyme HindIII recognition sequences underlined)), double digested with restriction enzymes NheI and HindIII, gel-purified and inserted into pre-digested pGL4.10 firefly luciferase vector bearing the same NheI and HindIII ends, to produce the Ski basic promoter (Ski-basic).

The promoter constructs containing the putative TRE1 of Ski (Ski-TRE1) was generated through PCR-mediated joining of a PCR fragment encompassing the putative Ski TRE1 to the 5’-end of the Ski basic promoter in the Ski-basic construct. In brief, the Ski basic promoter was amplified in PCR1 with primer Ski_F1 (5’-AGCCTTATGTGGGGGAATAGT-3’, binds to the same targets as Ski_F above) and Ski_R1 (5’- CGCGCCGATATCGCTCACGAGTTCCCCTTCAT-3’ (EcoRV recognition sequences underlined)), and the PCR fragment encompassing the putative Ski TRE1 was amplified in PCR2 with primer Ski-TRE1_F (5’-CGCGCCGCTAGCTGCTGCAAATTCATGCCTGG-3’ (NheI recognition sequences underlined) and Ski-TRE1_R (5’- ACTATTCCCCCACATAAGGCTCCCATAGCAACCAATCAGCG-3’, sequences in bold letters overlap with the primer F3 on the complementary strand). An aliquot of each gel-purified PCR1 and PCR2 products were mixed and used as templates in a third PCR reaction (PCR3) to amplify the recombinant Ski promoter containing the putative Ski TRE1 with Ski-TRE1_F and Ski_R1. The PCR3 products were gel-purified and digested with NheI and EcoRV, and inserted into pre-digested pGL4.10 firefly luciferase vector bearing the same NheI and EcoRV ends to make the Ski promoter construct with the putative TRE1 (Ski-TRE1). Its mutant version (Skim-TRE1) was made by using the Ski-TRE1 as the template for PCR with primer set 1: Ski-TRE1_F and Ski-mTRE1_R (5’-GATTAGTTACCATAAATCACTGC-3’, bold letters indicate the TRE half sites with the mutated nucleotides inactivating the TRE underlined) and primer set 2: Ski-mTRE1_F (5’- GCAGTGATTTATGGTAACTAATC -3’, bold letters indicate the TRE half sites with the mutated nucleotides inactivating the TRE underlined) and Ski_R1. The two PCR fragments were gel purified and then mixed together as the templates for another round of PCR with primers Ski-mTRE1_F and Ski_R1 to produce a ~0.8 kb fragment. The 0.8 kb fragment was then subjected to restriction enzyme digestion with NheI and EcoRV and gel-purified. The 794 bp restricted fragment was then subcloned into NheI - EcoRV digested pGL4.10 firefly luciferase vector to make Ski-mTRE1 construct.

The Ski promoter containing the putative Ski TRE2 was amplified through PCR with primers Ski-TRE2_F (5’-CTGGCCGGTACCACAGCCAAGTTTATTATCTACAAGG-3’(KpnI recognition sequences underlined) and Ski-TRE2_R (CTGGCCCCATGGTCTCCATGCCTGAAAGCTCC-3’(NcoI) and gel purified. This PCR product of ~2.5 kb was double digested with KpnI and NcoI, gel purified and inserted into pre-digested pGL4.10 firefly luciferase vector bearing the KpnI and NcoI ends to generate Ski promoter containing the Ski TRE2 (Ski-TRE2). Its mutant version (Ski-mTRE2) was made by using the Ski-TRE2 as template for PCR cloning essentially as described above for Ski-mTRE1. In brief, the first round PCRs were carried out in two parallel reactions with primer set 1: LZ631 (5’-tgttggatgctcatactcgtcc-3’, specific for pGL4.10 firefly luciferase vector sequences) and Ski-mTRE2_R (5’-GGCAAAATCCGCGCAAGTCAATCG-3’, bold letters indicate the TRE half sites with the mutated nucleotides inactivating the TRE underlined) and primer set 2: Ski-mTRE2_F (5’-TGACTTGCGCGGATTTTGCCCCAGT-3’, bold letters indicate the TRE half sites with the mutated nucleotides inactivating the TRE underlined) and LZ632 (5’- ggtaatgtccacctcgatatgtgc-3’, specific for pGL4.10 firefly luciferase vector sequences). The two PCR products were gel purified and mixed to serve as templates for another round of PCR with primers LZ631 and LZ632 to generate a product of ~2.9 kb fragment, which weas gel-purified and double digested with KpnI and NcoI for cloning into pre-digested pGL4.10 firefly luciferase vector bearing the KpnI and NcoI ends. All the promoter constructs were confirmed through DNA sequencing.

2.6. Transcription assay in Xenopus laevis oocyte

Oocyte transcription assay was performed as described (Wong and Shi 1995; Wang et al. 2008). Briefly, plasmid constructs containing GFP, Xenopus tropicalis TRα and RXRβ were linearized with EcoRI digestion and transcribed into mRNAs by using a mMESSAGE mMACHINE T7 Transcription Kit (Ambion, Grand Island, NY). The cytoplasm of stage VI Xenopus laevis oocytes was injected with 46 pg/oocyte of GFP mRNA or TR/RXR mRNA mixture. Two hour later, the firefly luciferase reporter constructs under the control of Ski basic promoter (Ski-basic), Ski promoter harboring the wild type Ski TRE1 (Ski-TRE1) or mutated Ski TRE1 (Ski-mTRE1), or Ski promoter harboring the wild type Ski TRE2 (Ski-TRE2) or mutated Ski TRE2 (Ski-mTRE2) was injected into the nuclei of these oocytes (33 pg/μl), along with the internal control Renilla luciferase reporter phRG-TK (3.3 pg/μl). After incubation at 18 °C overnight in the presence or absence of 100 nM T3, the oocytes were randomly sorted into groups for dual luciferase assay following the manufacturer’s instructions for the Dual-Luciferase-Reporter Assay kit (Promega, Madison, WI), with each group containing at least 5 oocytes and 5 groups per treatment were collected. The relative expression of firefly luciferase to Renilla luciferase was determined and presented as the average of at least three groups of repeats. The data shown were representative of a few independent experiments with similar results.

2.7. Statistical analysis

All quantitative data are presented as mean ± S.E.M. (standard error of the mean) and statistical analyses were done with Prism (GraphPad, San Diego, CA). The differences between two groups were evaluated by a paired t-test and the differences among multiple groups were evaluated with ordinary one-way ANOVA. Significance values were * P≤0.05.

3. RESULTS

3.1. Xenopus tropicalis Ski is upregulated in the intestine and tail during natural and T3-induced metamorphosis

In a previous study to identify direct TR target genes globally during Xenopus tropicalis metamorphosis, we discovered Ski as one of the putative TR-target genes (Fu et al. 2017). To investigate if and how Ski is upregulated by TR during natural metamorphosis, we performed qRT-PCR to analyze its expression in the intestine and tail of tadpoles at stages from premetamorphosis (stage 56), metamorphic climax (stage 58–64), to the end of metamorphosis (stage 66) (Fig. 1). The results indicated that Ski expression was significantly upregulated in the intestine during natural metamorphosis with a peak expression at stage 62, concurrent with the stage when T3 level in the plasm peaks and the intestine undergoes drastic remodeling, characterized by rapid larval intestinal epithelial degeneration through apoptosis and robust adult intestinal stem cell proliferation and differentiation (Fig. 1A) (Dodd and Dodd 1976; Shi and Ishizuya-Oka 1996). Ski expression in the tail was also significantly upregulated during metamorphosis and peaked when the tail undergoes rapid resorption around stage 62 to 64 (Fig. 1B) (Nieuwkoop and Faber 1965; Dodd and Dodd 1976). Thus, Ski is likely upregulated by T3 during frog metamorphosis to participate in the cell fate determination during intestinal remodeling and tail resorption.

Figure 1: Ski expression peaks at the climax of metamorphosis during natural frog metamorphosis.

Genomic DNA-free total RNAs were extracted from the intestine and tail of X. tropicalis tadpoles at indicated stages and subjected to qRT-PCR analysis with primers specific for Ski as well as the control gene EF1α, respectively. Ski expression was normalized to that of EF1α and presented as mean + standard error of the mean (S.E.M), relative to the level at stage 56 in the organs, which was set to 1. Statistical analysis was done with ordinary one-way ANOVA. *: p <0.05.

To investigate whether Ski expression is indeed regulated by T3, we treated premetamorphic tadpoles at stage 54 with or without 10 nM T3 for 2 days and analyzed Ski gene expression by qRT-PCR on total RNA isolated from the intestine and tail, respectively. The data indeed revealed significant activation of Ski expression by T3 in the intestine (Fig. 2A) and the tail (Fig. 2B), although more drastically in the intestine (Fig. 2A), suggesting tissue-dependent regulation of the Ski gene by T3 during frog metamorphosis. Thus, Ski is likely activated by T3 to participate in metamorphosis, particularly intestinal remodeling.

Figure 2: Ski expression is upregulated by T3 in X. tropicalis tadpoles.

Premetamorphic X. tropicalis tadpoles at stage 54 were treated with or without T3 for 2 days and subjected to total RNA extraction from the intestine and tail, respectively. cDNAs were made from the total RNAs and subjected to qRT-PCR analysis for Ski and EF1α expression, respectively. Ski expression was normalized to that of EF1α and presented as mean + standard error of the mean (S.E.M), relative to the level at stage 56 in the organs, which was set to 1. Student t-test was used to determine the significance of T3 regulation. *: p <0.05.

3.2. TR/RXR heterodimer can bind to a putative TRE in the first exon of Ski gene in vitro

To investigate if X. tropicalis Ski expression is directly regulated by TR as our earlier ChIP-on-chip analysis suggested, we carried bioinformatics analysis on the sequences from 15 kb upstream of the putative transcriptional start site to the end of its last exon (exon 7) of X. tropicalis Ski gene by using NHR-Scan (http://nhrscan.genereg.net/cgi-bin/nhr_scan.cgi ) and identified two putative TREs with the putative TRE1 about 3 kb upstream of the putative transcriptional start site and the putative TRE2 localizing in the first exon, about 400 bp downstream of the putative transcriptional start site (TRE1 – 2, Fig. 3A–B) and 59 bp upstream of its translational start site (data not shown). To investigate if TR/RXR heterodimer binds to these putative TREs, we carried out gel mobility shift assays by using IR700-labeled TRE of Xenopus laevis TRβA gene, a well characterized TRE consisting of two near perfect direct repeats of AGGTCA half sites separated by 4 bp, as the probe (Fig. 3B) (Ranjan et al. 1994; Wong et al. 1998), TR and RXR proteins made through in vitro translation, and unlabeled competitor TREs that included Xenopus laevis TRβA TRE and Xenopus tropicalis Ski TREs (Fig. 3B). The results indicated that the putative Ski TRE1 had little or weak ability to compete for binding against the labeled Xenopus laevis TRβA TRE to the TR/RXR heterodimer while the putative Ski TRE2 competed with similar effectiveness as the unlabeled Xenopus laevis TRβA TRE (Fig. 3C). To determine the specificity of the competition, we mutated the putative TREs at the positions known to abolish TRE binding to TR/RXR heterodimer (Ski mTRE1/2, Fig. 3B) and tested in the mobility shift assay under the same conditions as the wild type TREs. The results indicated that the mutations in the putative TREs of Ski gene abolished their ability to compete for binding to TR/RXR heterodimer. These findings indicate that putative TRE2 in the first exon of Xenopus tropicalis Ski genes binds to TR/RXR heterodimer strongly and specifically, and thus, likely mediates liganded TR activation on the Ski gene expression during frog metamorphosis.

Figure 3: TR binds to the putative TRE within the first exon of the Ski gene in vitro.

(A) Schematic diagram of putative TREs in X. tropicalis Ski gene with the first (E1) and the last (E7) exons shown as clear boxes.

(B) Sequences of the putative TREs in comparison with the sequences of consensus TRE and the TRE of Xenopus laevis TRβA gene (TRβA TRE). The numbers indicate the locations relative to the transcriptional start site of Ski gene (identified based on the 5’-end of cDNA sequence and denoted as +1). Mutant versions of Ski TREs (Ski mTRE1/2, mutated nucleotides are underlined) were used for studies below. Bold letters indicate TRE half sites. E1: exon 1; E7: exon 7.

(C)/(D) The putative Ski TRE2 compete efficiently against the TRβA TRE for binding to TR/RXR heterodimer in gel shift assay. The labeled double stranded TRβA TRE, used as the probe, was mixed with in vitro translated TR/RXR heterodimers in the presence or absence of 4x, 20x, or 100x of unlabeled wild type (C) or mutant (D) Ski TREs as indicated. The reaction mixtures were analyzed by Electrophoretic (or gel) Mobility Shift Assay (EMSA). The locations of TR/RXR bound probes (TR/RXR-probe complex) and unbound probes (Free probe) are indicated with arrows. Note that the labeled probe binding to the TR/RXR was competed away efficiently by the unlabeled TRβA TRE (C, lanes 3–5) and that Ski TRE2 also had strong competition (C, lanes 9–11) while Ski TRE1 had little competition (C, lanes 6–8). Neither one of the mutant Ski TREs had noticeable competition (D).

3.3. Xenopus tropicalis Ski TRE2 but not TRE1 conveys TR/RXR-mediated T3 regulation in vivo

To investigate whether the putative TREs of the Ski gene can render TR/RXR-mediated T3 regulation, we generated Ski promoter constructs containing either the wild type TRE1 of the Ski gene (Ski-TRE1) or its mutant version (Ski-mTRE1), respectively, to drive firefly luciferase gene expression, in comparison with a Ski promoter construct containing a basic Ski promoter region (Ski-basic) (Fig. 4A) and evaluated T3-regulation of their activities in Xenopus oocytes (Fig. 4B). As shown in Fig. 4B, none of these three promoter constructs responded to T3 treatment in oocytes pre-injected with either GFP mRNA (as a control) or mixture of TR/RXR mRNAs, suggesting that the Ski TRE1 have little capacity, if any, to mediate T3 regulation of the promoter in oocyte system under the experimental conditions. This is consistent with the weak binding of the Ski TRE1 to TR/RXR heterodimer as shown by gel mobility assay above (Fig. 3C).

Figure 4: The TRE2 but not TRE1 of X. tropicalis Ski gene mediates transcriptional regulation by T3 in vivo.

(A) Schematic diagram of firefly luciferase reporter constructs containing the wild type putative Ski TRE1 (Ski-TRE1) or a mutant TRE1 (Ski-mTRE1) generated through PCR-mediated cloning to bring the putative Ski TRE1 or its mutant closer to the transcription start site (This was done because TREs located far from the start site do not active the promoter effectively (Wong et al. 1997)). The Ski-basic has a truncated promoter region lacking the TRE1 or TRE2.

(B) The Ski-TRE1 promoter is not activated by liganded TR in frog oocytes. The Ski-basic, Ski-TRE1 or Ski-mTRE1 promoter construct was co-injected with the control Renilla luciferase construct phRG-TK into the nucleus of Xenopus laevis oocytes with prior cytoplasmic injection of GFP mRNA or Xenopus tropicalis TRα and RXRβ mRNAs. The oocytes were incubated at 18 °C overnight in the absence or presence of 100 nM T3 and then lysed for dual luciferase assay. The relative activities of the firefly luciferase to Renilla luciferase were plotted as a measure of the reporter promoter activity. Error bars indicate S.E.M. Note that none of the reporters, including the one with wild type TRE1, was regulated by TR in the oocytes.

(C) Schematic diagram of firefly luciferase reporter constructs containing the putative Ski TRE2 (Ski-TRE2) or a mutant TRE2 (Ski-mTRE2).

(D) The Ski-TRE2 promoter is strongly activated by liganded TR in frog oocytes. The Ski-TRE2 or Ski-mTRE2 promoter construct was co-injected with the control Renilla luciferase construct phRG-TK into the nucleus of Xenopus laevis oocytes with prior cytoplasmic injection of GFP mRNA or Xenopus tropicalis TRα and RXRβ mRNAs. Oocytes were treated and analyzed as in (B). Error bars indicate S.E.M. Note that the wild type TRE2-containing reporter was activated by T3 only in the presence of TR/RXR and that the mutations in the Ski TRE2 abolished this activation. Statistical analysis was done with ordinary one-way ANOVA. *: p <0.05.

To determine whether the putative TRE2 in the first exon of the Ski gene can render TR/RXR-mediated T3 regulation of the promoter, we generated Ski promoter constructs containing either the wild type TRE2 (Ski-TRE2) or its mutant version (Ski-mTRE2), respectively, to drive firefly luciferase gene expression (Fig. 4C) and evaluated T3-regulation of their activities in Xenopus oocytes (Fig. 4D). As shown in Fig. 4D, the Ski-TRE2 promoter did not respond to T3 treatment in oocytes preinjected with GFP mRNA, but was strongly activated by T3 in oocytes preinjected with a mixture of TR/RXR mRNAs, suggesting that TRE2 mediated T3 induction through the binding of the TRE to TR/RXR. However, the T3-regulation was absent when the Ski-mTRE2 construct was used, which produced similar firefly luciferase activities both in the presence or absence of T3 treatment in oocytes preinjected with either GFP mRNA or a mixture of TR/RXR mRNAs (Fig. 4D). Taken together, these findings indicate that the putative TRE2 but not TRE1 of X. tropicalis Ski gene can convey TR/RXR-mediated T3 regulation in Xenopus oocytes.

4. DISCUSSION

Anuran metamorphosis is an important developmental process for studying the molecular mechanism of vertebrate postembryonic development. Unlike the mammalian embryos that develop in the uterus and are under the maternal influence, anurans develop externally and are independent of maternal effects, thus allowing easy access and manipulation during development. Numerous studies have revealed that TR plays an essential role in mediating the effects of T3 during metamorphosis (Buchholz and Shi 2018; Shi 2021). TR binds to TRE constitutively and recruits cofactor complexes to the TRE regions of its target genes to activate or repress their expression depending on the presence or absence of T3. This makes it critical to identify and characterize T3 target genes during metamorphosis for exploring the molecular mechanisms underlying T3-regulation of anuran metamorphosis. Toward this end, we performed ChIP-on-chip analysis and found Ski among the 300 or so putative direct targets of TR during intestinal metamorphosis (Fu et al. 2017). Our studies here demonstrate that Ski is activated by TR directly via a TRE within the first exon during X. tropicalis metamorphosis and its expression peaks at the climax of metamorphosis in both the intestine and tail, implicating a critical role of Ski during intestinal remodeling and tail degeneration.

Ski is a conserved proto-oncogene identified in many vertebrates and numerous studies have shown that it plays important roles in human health and disease (Luo 2004; Bonnon and Atanasoski 2012; Tecalco-Cruz et al. 2018). However, how Ski expression is regulated has been poorly studied and the molecular mechanism underlining the regulation of Ski expression is largely unknown. Our earlier ChIP-on-chip analysis suggests that Ski is a direct target of TR and is regulated during X. tropicalis intestinal metamorphosis (Fu et al. 2017). Through qRT-PCR analysis, we have demonstrated that Ski is indeed up-regulated in the intestine as well as the tail during natural metamorphosis, with peak levels of expression at the climax of metamorphosis (Fig. 1). It is also up-regulated in both organs during T3-induced metamorphosis (Fig. 2). Furthermore, through bioinformatics analyses, we identified two putative TREs with TRE1 in the promoter and TRE2 in the exon 1. Through in vitro DNA-binding and in vivo transcriptional analyses, we have discovered the TRE2 in the first exon capable of strong binding to TR/RXR heterodimer and mediating T3-induction of the Ski promoter, although the TRE1 in the promoter had much weaker binding to TR/RXR heterodimer and failed to mediate T3-induction of the Ski promoter alone by itself. It is unclear why the putative TRE1 had much weaker binding capacity to TR/RXR heterodimer and if the TRE1 could still enhance the TRE2 mediated T3 regulation of Ski expression in the endogenous promoter context. Future studies to mutate the individual TREs or both through genome editing may reveal more clarity on this. Nevertheless, these findings indicate that T3 directly activate the transcription of the Ski gene via TR binding to the TRE in the first exon during X. tropicalis metamorphosis.

Anuran metamorphosis involves drastic changes in most, if not all organs, including the intestinal remodeling, where both larval cell death and adult stem cell development take place, and the tail resorption where the entire organ degenerate mainly through apoptosis. The upregulation of Ski gene expression by T3 in both the intestine and tail suggest that Ski participates in the metamorphosis of different organs/tissues. It is interesting to note that T3 regulation of Ski expression in the intestine during early stages of metamorphosis was more dramatic than that in the tail (Fig 1), similar to what observed when premetamorphic tadpoles at stage 54 were treated with T3 for 2 days (Fig.2), although it was upregulated in both organs. This may suggest organ-specific temporal involvement of SKI during metamorphosis since intestinal remodeling takes place earlier than tail resorption.

Earlier studies suggest that Ski plays important roles during embryogenesis and tissue homeostasis by functioning as a key transcriptional corepressor of Smad proteins to regulate the transforming growth factor-β (TGF-β)/Smad signaling pathway (Luo 2004; Bonnon and Atanasoski 2012; Tecalco-Cruz et al. 2018). Ski forms heterodimer with SnoN, a homologous protein of Ski, and interacts with Smad4 to inhibit TGF-β target genes in the absence of TGF-β through Smad4-mediated binding to their promoters. The canonical TGF-β signaling pathway exerts most of its biological functions through cross-talks with other pathways such as the Wnt, Notch, Hippo, PI3K-AKT, PKC, MAPKs, and JAK-STAT signal pathways and regulation of Ski gene expression plays important roles in the crosstalk (Tecalco-Cruz et al. 2018). Interestingly, most of these pathways, if not all, are directly or indirectly regulated by T3 during intestinal metamorphosis (Heimeier et al. 2010; Sun et al. 2013; Shibata et al. 2021; Tanizaki et al. 2021; Tanizaki et al. 2022a; Tanizaki et al. 2022b). Ski does not bind to target genes directly, it participates in transcriptional repression through association with DNA-binding transcriptional factors, such as Smad4, to target promoters (Nomura et al. 1999; Tokitou et al. 1999). Interestingly, one such transcriptional factor is TRβ (Nomura et al. 1999). This raises an interesting possibility that T3 induces the expression of Ski and Ski in turn feedback to regulate TR activity to affect amphibian metamorphosis. It would be interesting to test this model in the future.

Numerous studies have revealed mis-regulation of Ski gene expression in many types of cancer cells including colorectal cancer (Buess et al. 2004; Bravou et al. 2009) and esophageal squamous cell carcinoma (Fukuchi et al. 2004), suggesting that Ski plays a role in gastrointestinal tissue homeostasis. Our data have demonstrated that Ski expression is directly up-regulated during intestinal remodeling, which involves drastic cell fate changes including larval epithelial cells death through apoptosis and concurrent adult epithelial stem cell development, followed by their rapid proliferation and subsequent differentiation to form the adult intestinal epithelium (Dodd and Dodd 1976; Shi and Ishizuya-Oka 1996; Shi 1999). Thus, there is likely a conserved role of Ski in regulating gastrointestinal epithelial cell fate in vertebrates. Further studies on how the T3-regulated Ski expression participates in intestinal remodeling during anuran metamorphosis should not only improve our understanding of intestinal metamorphosis but also offer mechanistic insights on the involvement of Ski in human gastrointestinal epithelial homeostasis and pathogenesis. A promising approach toward understanding of Ski function during Xenopus intestinal remodeling is to knock out Ski through genome editing and investigate its impact on intestinal metamorphosis. Such experiments would also provide opportunities to further investigate how Ski participates in signal pathways to affect organ-specific changes during Xenopus metamorphosis.

Highlights:

Protooncogene Ski was upregulated by T3 in the tadpole intestine and tail

Ski expression peaked at the climax of intestine and tail metamorphosis

A TRE in exon 1 of the Ski gene could bind to TR/RXR heterodimer in vitro

The TR/RXR-binding to SKI TRE could mediate promoter activation by T3 in vivo

Ski was activated by T3 directly at the transcriptional level and might feedback to regulate TR activity during metamorphosis