Thyroid hormones (THs), thyroxine (T4), and 3,5,3′-triiodothyronine (T3), are important regulators of growth, development, and metabolism of all vertebrates. An extreme example of their importance during development is their absolute requirement during the process of anuran metamorphosis (reviewed in Shi,2000; Furlow and Neff,2006). Tadpole tail resorption, one of the last changes during metamorphosis, is mediated by apoptotic pathways (Veldhoen et al.,2006) and is entirely controlled by TH. This process can be precociously induced by exogenous administration of T3 (Wang and Brown,1993; Brown et al.,1996; Helbing et al.,2003) matching the onset and completion of natural metamorphosis (Wang and Brown,1993; Helbing et al.,2003). TH action at the cellular level is highly conserved from frogs to humans and is primarily mediated by mechanisms involving TH binding to specific nuclear thyroid hormone receptors (TRs) that regulate gene expression (Shi,2000; Yen,2001). TRα and TRβ are the two major TR isoforms that possess dual functions as transcription repressors and activators in the absence and presence of ligand (primarily T3), respectively (Sachs et al.,2000). Use of dominant-negative TRα (Schreiber et al.,2001; Buchholz et al.,2004) and dominant-positive TRα genes (Buchholz et al.,2003) in transgenic Xenopus laevis has shown that TRα is critical for T3-induced frog metamorphosis and for the regulation of known T3-response genes, such as TRβ, at the transcriptional level in the tail. TRα is present in the premetamorphic tadpole tail whereas TRβ levels are extremely low until TH induction (Eliceiri and Brown,1994; Veldhoen et al.,2002), suggesting that TRα is critical for the competence to respond to TH while TRβ is required for establishment of the metamorphic genetic program.
Protein phosphorylation plays important roles in modulating TH action. The pathways involved include those mediated by tyrosine kinases (Di Fulvio et al.,2000; Utoh et al.,2003; Martinez and Gomes,2005), protein kinase C (PKC) (Petcoff and Platt,1992; Phillips and Platt,1994; Wagner et al.,2001; Alisi et al.,2004), mitogen-activated protein kinase (MAPK)/extracellular signal-regulated kinase (ERK) (Lin et al.,1997,1998,1999a,b; Davis et al.,2000; Bergh et al.,2005), cyclin-dependent kinases (Barrera-Hernandez et al.,1999; Nguyen et al.,2003; Skirrow,2003; Skirrow and Helbing,2007), and phosphatidylinositol-3 kinase (PI3K) (Lei et al.,2004; Cao et al.,2005; Seo et al.,2005). However, the precise target substrates, the location of phosphorylation sites, their integration into the TH signaling pathway, and functional roles in specific tissues remain to be elucidated.
Since TRs are phosphoproteins, they could serve as key substrates for modulation by these pathways. Previous work in mammalian cell cultures has shown that phosphorylation of the TRβ protein increases its ability to bind to thyroid hormone response elements (TREs) (Bhat et al.,1994; Sugawara et al.,1994; Chen et al.,2003), thereby enhancing its transcriptional activity (Lin et al.,1992; Bhat et al.,1994; Chen et al.,2003). In contrast, phosphorylation of the TRα protein inhibits monomeric TRE binding, but not retinoid X receptor (RXR)/TRβ heterodimeric binding (Tzagarakis-Foster and Privalsky,1998) and enhances its nuclear localization (Nicoll et al.,2003). However, the functional role of TR phosphorylation in the context of apoptosis in a normal developmental process, such as T3-dependent tail resorption, is not clear.
Given that there is increasing evidence for the importance of protein phosphorylation cascades on TH action, it is plausible that this post-translational modification may be a potential target for endocrine disruptors, which could influence the establishment of gene expression programs by altering transcription factor stability, ability to interact with DNA response elements, and/or interaction with cofactors, thereby leading to deleterious effects (Tabb and Blumberg,2006). Genistein (4′, 5, 7–trihydoxyisoflavone) is a major component in soy foods and is an analog of estradiol (E2), which can exert either estrogenic or anti-estrogenic effects at low concentrations (10 nM–10 μM) (Dixon and Ferreira,2002) through binding to the estrogen receptor and sex hormone–binding proteins (Wang et al.,1996). At higher concentrations (≥30 μM), genistein is a potent tyrosine kinase inhibitor (Akiyama et al.,1987). Previous work indicates that genistein affects TH metabolism (Mori et al.,1996; Chang and Doerge,2000) and inhibits thyroid peroxidase activity (Doerge and Chang,2002; Doerge and Sheehan,2002), but no information exists as to its cellular effects on TH signaling.
Since genistein has been associated with goiter (Doerge and Sheehan,2002) and since there is crosstalk between estrogen and TH signaling pathways (Vasudevan et al.,2002), we examined the potential effect of this chemical on T3-responsive tail tissue in Rana catesbeiana tadpoles. Tail resorption is easy to manipulate in organ culture and serves as an ideal model system for investigation of T3-dependent apoptosis. The tail organ culture system allows one to answer questions about the involvement of particular signaling pathways through the use of specific inhibitors that would otherwise be toxic to the intact organism. Our work shows that genistein prevents both T3-induced tadpole tail regression in organ culture and up-regulation of TRβ transcript levels, suggesting that T3-induced tail regression is dependent on tyrosine kinase signaling likely mediated through PKC. This work provides the first evidence that changes in tyrosine phosphorylation are important in T3-dependent biological outcomes during postembryonic development and could be targets for endocrine disruption.
Inhibition of Tyrosine Kinase Signaling Prevents T3-Induced Tail Tip Regression
To determine whether genistein affects T3-dependent tail regression in Rana catesbeiana, 2-cm tail tips from premetamorphic tadpoles were cultured in serum-free medium for 72 hr and treated with vehicle control (DMSO), 100 nM T3, 0.1–100 μM genistein alone or T3/genistein in combination. Tails treated with either DMSO or genistein alone showed no regression (Fig. 1) while tails treated with 100 nM T3 showed a 55% reduction in tail area by 72 hr and 83% by 96 hr in culture (P < 0.01, T3 relative to DMSO for both; MWU) (Fig. 1). This regression was completely inhibited by the addition of 100 μM genistein to the culture medium (P = 0.023, T3 + 100G relative to T3; MWU) (Fig. 1). Lower concentrations (0.1, 1, or 10 μM) of genistein were not sufficient to significantly inhibit T3-induced tail regression although 10 μM showed a partial inhibition (Fig. 1).
In contrast to the above results, T3-induced tail regression was not altered by the addition of either PD098059 or Wortmannin (ERK and PI3K inhibitors, respectively) to the culture medium (Fig. 2). Neither inhibitor had any effect on tail tip area compared to the DMSO controls (Fig. 2). The effect of these inhibitors on ERK activity (an indicator of both pathways (Davis et al.,2000; Page et al.,2000) was assessed by immunoprecipitation of ERK complexes, from tail tips cultured with T3 for 48 hr, and subsequent measurement of kinase activity (Fig. 2C) PD098059 inhibited ERK1 activity by 30% in tail tips and ERK2 activity by 65%. Treatment of the tail tips with Wortmannin in conjunction with T3 also resulted in a substantial reduction in ERK1 activity (58%), although ERK2 was largely unaffected. These kinase activity assays indicated that both PD098059 and Wortmannin were functional even though T3-induced tail regression was not affected.
Genistein Reduces T3-Induced Tyrosine Phosphorylation and PKC Kinase Activity
Genistein inhibits the epidermal growth factor (EGF)-stimulated increase in phosphotyrosine levels in mammalian cells (Akiyama et al.,1987). To determine if genistein inhibition of T3-induced tail regression is occurring via inhibition of tyrosine phosphorylation, tadpoles were injected with either T3 or vehicle control (DMSO) and maintained in dechlorinated 25°C tap water for 24 hr prior to culturing their tail tips in serum-free medium in the presence of 100 μM genistein or vehicle control for a further 24 hr. After 24 hr, the tail tips were collected and homogenized for total protein isolation. Proteins were separated by SDS-PAGE, and phosphotyrosine levels were assessed by immunoblot analysis. Although, in any given experiment, several bands appeared to change in intensity, only two bands corresponding to ∼75 and 80 kDa reproducibly changed in each of four independent experiments (Fig. 3).
The ∼75-kDa (Fig. 3A and C) protein showed an apparent increase in tyrosine phosphorylation upon T3 treatment (P = 0.014, T3 alone relative to control; MWU), which was reduced in the presence of genistein (P = 0.043 T3-G relative to T3 alone; MWU). This highly abundant protein, which showed some staining with a general phosphoprotein stain and a strong signal with the anti-p-Tyr antibody (Fig. 3C), was identified as serum albumin (bullfrog) (NCBI accession number A37253) by mass spectrometry analysis from a 2D gel (data not shown).
A ∼80 kDa protein with increased tyrosine phosphorylation following T3 treatment was also identified (Fig. 3A and B). This increased phosphorylation was attenuated when the tail tips from T3-injected tadpoles were cultured in medium containing 100 μM genistein (P = 0.014 T3-G relative to D-D control and P = 0.021 T3-G compared to T3 alone; MWU). The 80-kDa phosphoprotein was detected by a general phosphoprotein stain (Fig. 3C) and co-migrated with PKC on both 1-D and 2-D Western blots (Fig. 3A and C) indicating that PKC may be a target of T3-dependent tyrosine phosphorylation and that genistein inhibits that phosphorylation. Immunoprecipitation of PKC and evaluation of its kinase activity showed that T3-induced PKC kinase acitivity was markedly inhibited by addition of genistein in culture (75% reduction T3-D vs. T3-G) (Fig. 3D).
Genistein Inhibition of T3-Induced Tail Regression Is Time-Dependent
Previous studies in Xenopus laevis showed that there is a critical timeframe for establishment of the tail regression genetic program (Wang and Brown,1993). Tail regression in tadpoles induced to undergo metamorphosis by T3 treatment was blocked by both transcriptional and translational inhibitors within the first 24 hr following T3 exposure. However, after 48 hr of exposure, these inhibitors were not as effective at inhibiting tail regression since the gene expression program was already initiated (Wang and Brown,1993). This important transition phase was referred to as the “commitment point” (Wang and Brown,1993). Based on this finding, we asked whether there was a similar window of opportunity for genistein to exert its inhibitory effect on tail regression. To determine this, premetamorphic tadpoles were injected with 3 × 10-10 mol/body weight T3 or an equal volume of vehicle alone (DMSO) and maintained in 25°C dechlorinated water for either 24 or 48 hr (Fig. 4). At each timepoint, 2-cm tail tips were collected and cultured in the presence of vehicle (DMSO) or 100 μM genistein and incubated at 25°C for 24, 48, or 72 hr. Tail tips from tadpoles injected with DMSO and subsequently cultured in either DMSO or 100 μM genistein showed no change in tail area (Fig. 5A), while tail tips from tadpoles injected with T3 and subsequently cultured with DMSO (T324-D and T348-D) showed at least 80% regression after 48 hr in culture (Fig. 5B). This is statistically significant compared to DMSO-cultured tail tips from DMSO-injected animals (D24-D and D48-D) (P < 0.001; MWU) (Fig. 5A). Tail tips cultured in the presence of 100 μM genistein 24 hr after tadpole T3 injection (T324-G) showed complete inhibition of tail regression (Fig. 5B), while those cultured in 100 μM genistein 48 hr after T3 injection (T348-G) showed only partial inhibition (32% regression) (Fig. 5B). This release from inhibition is statistically significant compared to the complete inhibition seen with the T324-G animals (P < 0.05; MWU) and suggests that the target(s) sensitive to genistein action are available up to 24 hr following T3 treatment, with more limited availability after 48 hr.
T3-Induced Up-Regulation of TRβ, But Not TRα, Transcripts Is Affected by Genistein
The previous experiment suggests that the effect of genistein may primarily be associated with initiation of the genetic program for tail regression. Key players in this transition are the TH receptors, TRα and TRβ (Schreiber et al.,2001; Buchholz et al.,2003,2004). To investigate whether genistein affects TRα and TRβ expression, tadpoles were injected with vehicle alone (DMSO) or T3 and tail tips were cultured after 24 hr (24-hr injection) or 48 hr (48-hr injection). At the indicated times of exposure to vehicle or genistein (G) in culture (Fig. 4), total RNA was isolated and analysed by QPCR.
Tail tips cultured in medium containing DMSO 24 hr post-T3 injection (T3-D) showed an increase in TRα mRNA levels at 24-, 48-, and 72-hr time points compared to tail tips from 24-hr DMSO-injected tadpoles (D-D) (P < 0.05; MWU) (Fig. 6A). However, when T3-induced tail regression was completely blocked in culture by treatment with 100 μM genistein, TRα transcript levels were still up-regulated by T3 compared to vehicle control (D-D) (P < 0.05; MWU) (Fig. 6A). Furthermore, there was no statistical difference in TRα transcript levels in tail tips cultured in 100 μM genistein compared to an equal volume of DMSO vehicle after T3 injection (Fig. 6A). This suggests that up-regulation of TRα transcripts is not affected by genistein. Tail tips cultured for 48 hr in DMSO-containing medium 48 hr after T3 injection (T3-D) showed no TRα up-regulation. This observation is consistent with previous observations (Zhang et al.,2006) that TRα levels decrease after 96 hr of T3 exposure. However, tail tips cultured in genistein-containing medium (D-G) did show a significant decrease in TRα levels compared to control (Fig. 6A).
Tail tips cultured in DMSO-containing medium 24 hr after T3 injection (T3-D) showed a significant increase in TRβ mRNA levels at 24-, 48-, and 72-hr time points compared to tail tips from DMSO-injected tadpoles (D-D) (P < 0.05; MWU) (Fig. 6B). This increase was attenuated when tail regression was completely blocked by 100 μM genistein (P < 0.05; MWU) (T3-D vs. T3-G). Tail tips cultured for 48 hr in DMSO-containing medium 48 hr after T3 injection showed an increase in TRβ mRNA levels compared to DMSO-injected tadpoles (P < 0.05; MWU) (T3-D vs. D-D; Fig. 6B). This T3-induced increase in TRβ transcript levels was attenuated in the presence of genistein (T3-D vs. T3-G) (P < 0.05; MWU), but the overall response was lower relative to the TRβ transcript levels measured in tail tips cultured 24 hr after T3-injection. This is in keeping with genistein's partial inhibition of T3-induced tail regression 48 hr after T3 exposure (compare Fig. 6B with 5B). In fact, the TRβ expression levels correlated strongly with tail regression (r = 0.368, P < 0.01; Spearman's correlation analysis). There was no correlation between TRα expression levels and tail regression (r = 0.086, P = 0.487; Spearman's correlation analysis).
TRα Phosphorylation Is Reduced by Genistein During T3-Induced Tail Regression
In vitro and mammalian cell culture studies showed that phosphorylation of TRs modulates TR-mediated regulation of target gene expression (Bhat et al.,1994; Sugawara et al.,1994), but there is no information about the functional role of TR phosphorylation in vivo. Our data showed that TRβ (Fig. 6B), but not TRα mRNA levels (Fig. 6A), are attenuated by genistein. To further understand the molecular mechanism underlying transcriptional regulation of TRβ and determine whether this regulation could occur via alteration of the TRα phosphorylation state, tadpoles were injected with either T3 or vehicle (DMSO) control. Tail tips were removed 24 hr later and incubated in serum-free medium for 24 hr in the presence of 100 μM genistein or DMSO. Tyrosine phosphorylation typically occurs as key regulatory events early in signal transduction cascades, whereas serine phosphorylation, such as that found on TRα, can occur as downstream events. Since TRs are mainly targeted for serine phosphorylation (Goldberg et al.,1988; Glineur et al.,1989; Ting et al.,1997; Davis et al.,2000), we examined the steady-state level of TRα serine phosphorylation. Tail tip protein homogenates were subjected to IP with a frog anti-TRα antibody and immunoblotted with an anti-human TRα (hTRα) that cross-reacts with frog TRα or an anti-phosphoserine antibody (Fig. 7A). T3 induces TRα accumulation and serine phosphorylation (Fig. 7A). Neither phosphorylation nor TRα accumulation are affected by genistein alone; however, an attentuation of phosphorylation was observed upon exposure to genistein in tail tips from T3-treated animals compared to T3-treated animals (Fig. 7A/B). This was accompanied by a marked increase in TRα levels compared to T3-treated tail tips incubated in vehicle alone (Fig. 7A, bottom panel). The ratio of serine phosphorylated TRα to total TRα was increased by 60% after T3 injection relative to vehicle-injected controls. Incubation of tail tips with genistein reduced this T3-mediated induction by 68%.
Protein Kinase C Inhibitor, H7, Inhibits T3-Induced Tail Tip Regression and Attenuates the Increase in TRβ Transcripts
Tail regression induced by T4 has previously been reported to be inhibited by the PKC inhibitor, H7 (Petcoff and Platt,1992; Phillips and Platt,1994), but no data have been published for T3-induced tail regression. Since there is increasing evidence for differential effects of T4 and T3 (Bergh et al.,2005; Zhang et al.,2006), we first examined the ability of H7 to inhibit T3-induced tail regression in organ culture. Tail tips from premetamorphic tadpoles were cultured in serum-free medium and treated with DMSO, 100 μM H7, 100 nM T3, or T3+H7 over 72 hr. Similar to the effects of genistein (Fig. 1), addition of H7 completely inhibited T3-induced tail regression (P = 0.004; T3+H7 relative to T3 alone; MWU) while the inhibitor alone had no effect (Fig. 8A).
Genistein and H7 have similar effects on T3-induced tail tip regression and genistein prevents the T3-induced increase of TRβ transcript levels (Fig. 6). If genistein is acting through PKC-mediated pathways, then inhibition of PKC activity through H7 should also affect TRβ transcript levels in a similar way. Tadpoles were injected with DMSO vehicle or T3 and tail tips were removed 24 hr later and incubated in serum-free medium for an additional 24 hr in the presence of DMSO or 100 μM H7. TRβ transcript levels were evaluated by QPCR and we observed a similar attenuation of the T3-mediated induction of TRβ transcript levels (P < 0.001 T3-D vs. T3-H7; MWU) (Fig. 8B) as was seen with genistein in Figure 6B. H7 alone (D-H7) also reduced TRβ transcript levels compared to DMSO controls (D-D; Fig. 8B).
Studies at the molecular level have demonstrated that TH regulates target gene expression through genomic effects by binding TRs (Yen,2001). Although there is increasing evidence of an important role for phosphorylation in TH signaling (Lin et al.,1997,1998,1999b; Schmidt et al.,2002; Wang et al.,2003) in in vitro and in cultured mammalian cell lines, few studies have addressed its relevance during development in vivo. In this study, by using a tadpole tail organ culture system as an experimental model, we demonstrate the possible involvement of T3-induced PKC tyrosine phosphorylation, which regulates TRα phosphorylation, thereby affecting tail regression.
Tadpole tail regression during anuran metamorphosis is one of the most conspicuous events at metamorphic climax. It requires a series of complex biochemical and gene expression changes (Wang and Brown,1993; Brown et al.,1995,1996; Shi,2000; Helbing et al.,2003). We chose this system because it has distinct advantages for study. The tail can be cultured in a serum-free medium for many days and it responds to exogenously administrated T3 in a fashion similar to the normal process of tail resorption that occurs at the climax of metamorphosis when endogenous TH is at its highest level (Furlow and Neff,2006). Using tail tip organ culture, we showed that genistein, but not MAPK or PI3K inhibitors, inhibited T3-induced tail resorption, even though the latter two inhibited ERK activity (Figs. 1 and 2). This suggests that MAPK and PI3K signaling is not important in the context of the T3-induced tail regression program. Protein analyses (Fig. 3) showed that T3-induced PKC phosphotyrosine levels and its kinase activities were reduced by genistein. Since tyrosine phosphorylation is important for PKC activity (Tapia et al.,2002; Pula et al.,2005), these data suggest that the inhibition of T3-induced tail regression may be through alteration of tyrosine kinase signaling. This is consistent with the observations that T3 or T4 enhances kinase activity of PKC in the chick embryo (Alisi et al.,2004) and that the PKC inhibitor, H7, inhibits T3 (Fig. 8A) or T4-induced tadpole tail regression (Petcoff and Platt,1992; Phillips and Platt,1994). PKC is a member of a ubiquitously expressed family of kinases that plays key roles in the regulation of diverse cellular activities. PKC is not only regulated by Ca2+ and diacylglycerol, but can also be regulated by phosphorylation on serine, threonine, and tyrosine residues making PKC catalytically active (Tapia et al.,2002; Newton,2003; Poole et al.,2004; Pula et al.,2005). Although enticing, it is still possible that other important targets for tyrosine phosphorylation exist, which could influence downstream target activity such as PDGF/PDGFR tyrosine kinase signaling (Utoh et al.,2003). Nevertheless, the up-stream and down-stream regulatory factors of PKC still need to be investigated.
An additional putative tyrosine phosphorylation target was identified as serum protein albumin. This protein increases during natural or precocious metamorphosis (Feldhoff,1971; Averyhart-Fullard and Jaffe,1990; Helbing et al.,2003) and acts as a carrier protein for a variety of hormones including THs (Shi,2000). As albumin is synthesized in the liver and is released into the blood as a serum protein, its presence in the tail tip is from the blood in these preparations. The protein is highly abundant, and the associated phospho-tyrosine signal may be an artifact of its abundance rather than albumin being a direct target for tyrosine phosphorylation. No tyrosine kinases targeting albumin have, to our knowledge, been identified as serum components. Alternatively, cellular components liberated through the process of tail regression may have included a tyrosine kinase that phosphorylated a subset of albumin proteins. It is, therefore, unlikely that albumin represents a key tyrosine phosphorylation target in the tail tip.
In the tadpole injection experiments, we showed that genistein was capable of inhibiting T3 action within 24 hr but that this chemical was less able to do so 48 hr after T3 injection (Fig. 5B). This is coincident with the time required for the establishment of the genetic program for tail resorption (Wang and Brown,1993). The timing of these effects suggests that tyrosine kinase signaling may be important in the genetic program transition and that phosphorylation events inhibited by genistein are important in T3-induced tail tip regression program. This would be consistent with genistein's action as a tyrosine kinase inhibitor, if critical tyrosine phosphorylation events occur early on in tail regression.
Using QPCR analysis, an induction of TRα and TRβ mRNA was observed upon T3 treatment (Fig. 6A and B) as was expected from previous studies (Crump et al.,2002; Veldhoen et al.,2002; Helbing et al.,2003; Zhang et al.,2006). The T3-induced up-regulation of TRβ, but not TRα, transcript levels was attenuated by genistein (Fig. 6B) in a manner that strongly correlated with the biological effects of this chemical on the inhibition of T3-induced tail regression (Figs. 1 and 5). This effect on TRβ transcripts was recapitulated using the PKC inhibitor, H7 (Fig. 8B), suggesting that PKC may be a modulator of this response. Previous work using a TRβ-specific antagonist, NH-3, and a TRβ-specific agonist, GC-1, has established that signaling through this receptor was critical for tail regression to occur in Xenopus laevis (Lim et al.,2002; Furlow et al.,2004). Our data suggest that tyrosine kinase signaling may play a role in the regulation of specific T3-responsive genes that are linked to biological outcome.
Although the TRs are ligand-activated receptors, they can also be functionally regulated by phosphorylation (Lin et al.,1992; Bhat et al.,1994; Sugawara et al.,1994; Ting et al.,1997; Tzagarakis-Foster and Privalsky,1998; Davis et al.,2000; Chen et al.,2003). Phosphorylation and dephosphorylation reactions are accomplished by multiple kinases and phosphatases. Since many signaling pathways contain tyrosine phosphorylation steps early on in the pathway that are amplified by subsequent serine/threonine phosphorylation events and since TRs have many serine/threonine phosphorylation sites with regulatory consequences (Goldberg et al.,1988; Glineur et al.,1989), we examined the steady-state levels of TRα serine phosphorylation to investigate a possible molecular mechanism of genistein's attenuation of the TRβ gene response to T3. We showed that T3 up-regulated serine-phosphorylation levels of TRα (Fig. 7), which is consistent with the previous observation that T4 enhances TRβ phosphorylation in mammalian cells (Davis et al.,2000). This T3-induced phosphorylation of TRα is reduced by genistein and suggests that T3 induces TRα phosphorylation contributing to its activation, which induces high expression of TRβ mRNA. Indeed, phosphorylation may serve as an important conduit in the transition of the TR from a transcription repressor to activator in the dual function model for the regulation of specific genes (Sachs et al.,2000). More importantly, this study provides the first biologically relevant in vivo evidence suggesting that TRα phosphorylation may be one of the mechanisms by which TRα achieves its diverse biological functions. A previous study suggested that TR phosphorylation may not be important during metamorphosis (Eliceiri and Brown,1994). However, they only examined Nieuwkoop and Faber (NF) stage 59 Xenopus laevis tadpoles in which most TRs might have already been phosphorylated and thus were not further available for labeling by [32P]orthophosphate; the method used for detection of TR phosphorylation in that study.
Phosphorylation targets proteins for degradation by the ubiquitin-proteosome pathway (Poizat et al.,2005). The T3-induced phosphorylation of TRα was greatly reduced in the presence of genistein in tail tissue but the steady-state levels of the protein increased (Fig. 7A). This implies that phosphorylation of TRα makes it more likely to be degraded and inhibition of that phosphorylation allows for the accumulation of the TRα protein. In CV-1 cells over-expressing TRα1, T3 treatment did promote TRα degradation (Chen et al.,2003). However, chemical enhancement of TRα1 phosphorylation reduced T3-induced TRα degradation (Chen et al.,2003) rather than increased it. The apparent discordance could be due to the nature of the experimental systems used or could reflect a difference in phosphorylation target sites. Since TRs have many phosphorylation sites with potentially different regulatory functions (Goldberg et al.,1988; Glineur et al.,1989), selective phosphorylation and dephosphorylation at specific TR residues may play a role in TR protein regulation and stability. It is also important to note that the T3-dependent induction of TRβ transcripts, although greatly attenuated, was not completely blocked even when tail regression was (Fig. 6B). This suggests that phosphorylation of TRα may serve an “enhancement” role to promote full transcriptional activity of the TRβ gene, and possibly other gene transcripts, leading to a biological consequence.
Previous work shows that protein kinase A (PKA) regulates TRα phosphorylation, which alters its TRE binding affinity in the context of T3 (Tzagarakis-Foster and Privalsky,1998); PKC increases TRα phosphorylation (Goldberg et al.,1988) and regulates TRα expression at protein and mRNA levels (Kenessey et al.,2006) in mammalian cell lines. Since PKA and PKC have a conserved kinase core (Newton,2003), it is likely that PKC regulates TRα phosphorylation, which in turn may regulate TRβ transcripts, which finally influences morphological metamorphic changes. TRα genes are constitutively expressed after the completion of embryogenesis and are required for the initial response to exogenous TH (Eliceiri and Brown,1994; Schreiber et al.,2001; Buchholz et al.,2003; Helbing et al.,2003). The TRβ genes have very low expression prior to metamorphosis and are up-regulated by exogenous T3, which is critical for the establishment of tissue-specific genetic programs for metamorphosis (Shi,2000; Helbing et al.,2003).
THs are critical for all phases of life in most vertebrates and, in humans, are particularly important during the perinatal period for proper brain development (reviewed in Shi,2000; Zoeller and Rovet,2004). Genistein, as a major isoflavone in soy products, is ingested in very high amounts by infants exclusively fed soy-based formulas. The plasma isoflavone concentration of these infants is maintained at ∼7 μM, which is ∼300-fold higher than milk formula- or breast-fed infants (Setchell et al.,1998; Badger et al.,2002). Our data suggest that concentrations similar to this could influence T3 action in frog tadpole tissues. However, we do not know what the effect such concentrations may have on TH signaling in human cells during chronic exposures. Given that our results suggest that genistein could directly disrupt TH signaling in target tissues important for proper development, longer term exposures are warranted to determine if TH signaling is affected at lower genistein concentrations.
In summary, this study demonstrates that phosphorylation is an important factor in the induction of T3-dependent tail regression in the frog tadpole. Our data strongly implicate PKC and TRα as targets for phosphorylation, which may selectively change the induction of specific T3-responsive target genes. This finding implies that environmental contaminant-induced changes in nuclear receptor activity can be through changes in phosphorylation status rather than by ligand competition, thus representing an important additional mechanism of endocrine disruption.
Taylor and Kollros (TK) (Taylor and Kollros,1946) stage VI–XV Rana catesbeiana tadpoles were locally caught (Victoria, BC) or purchased (Ward's Natural Science Ltd., St. Catharines, ON). The care and treatment of animals used in this study were in accordance with the guidelines of the Animal Care Committee, University of Victoria. Animals were housed in the University of Victoria aquatics facility and maintained in a 360-L all-glass flow-through aquarium containing recirculated water at 15°C with exposure to natural daylight. Tadpoles were fed daily with spirulina (Aquatic ELO-Systems, Inc., FL).
Tail Organ Culture
The procedure used for tail culture was adapted from that described previously (Helbing et al.,1992). Tadpoles were euthanized in 0.1% tricaine methane sulfonate (MS-222) (Syndel Laboratories, Vancouver, BC). The animals were then rinsed in 100 ml sterile distilled water for 10 sec, followed by a 5-sec immersion in 100 ml 70% ethanol and two subsequent successive 10-sec immersions in 100 ml sterile distilled water. Then 2-cm tail tips were aseptically removed and placed into individual 6-cm dishes (1 tip/dish) containing 5 ml Tadpole Minimal Essential Medium (TMEM). TMEM consists of a 55% strength solution of Minimal Essential Medium (Invitrogen), 25 mM HEPES (Sigma), 3 mM NaHCO3, 1.2 mM Na2HPO4, 1.2 mM NaH2PO4, 20 mM NaCl, 2 mM L-glutamine, 50 U/ml penicillin, 50 μg/ml streptomycin, and 50 μg/ml neomycin, and the final pH was adjusted to 7.1.
Two types of experiments were done. In the first, tail tips were isolated from tadpoles and allowed to recover in culture for 24 h at 25°C prior to the addition of chemical treatments with equal volumes of dimethyl sulfoxide (DMSO) vehicle solution (ACP Chemicals Inc., Montréal, QC), T3 (Sigma-Aldrich, Canada Ltd.), genistein (Sigma), 1-(5-isoquinolinesulfonyl)-2-methylpiperazine (H7; Sigma), PD098059 (Sigma), or Wortmannin (Calbiochem). Media and chemicals were replenished every 24 hr for the duration of the experiment. The second type of experiment is described in detail below.
In Vivo T3 Tadpole Treatment Followed by Chemical Treatment of Tail Tips in Organ Culture
Prior to T3 exposure, tadpoles were collected and maintained at 25°C in dechlorinated tap water (1 per 2 L water), using an automatic aquarium heater (Warnock Hersey), for 2 days to acclimate to laboratory conditions. After 2 days, the animals were immobilized on ice and injected intraperitoneally through the tail muscle with either DMSO or T3-containing solution at 3 × 10-10 mol per gram body weight, which equals 3 μl per gram body weight of 10-4 M T3 (Fig. 4). Animals were continuously maintained at 25°C in dechlorinated tap water for the duration of the experiment, with daily water changes. During the acclimatization and in vivo T3 exposure periods, animals were not fed. A subset of animals was anaesthetized in 0.1% MS-222 at 24 hr (24-hr injection set) and 48 hr (48-hr injection set), and their tail tips were aseptically removed for tail organ culture as described above. Cultured tail tips were treated for 24, 48, or 72 hr with equal volumes of DMSO (solvent control), 100 μM genistein, or 100 μM H7. Media and chemical reagents were replenished every 24 hr. At each time point, tail tissue was collected and preserved in RNAlater (Ambion Inc., Austin, TX) at 4°C for later RNA isolation.
Photographs of cultured tails were taken every 24 hr using a CCD digital camera (DVC Company, Austin, TX). Tail area was measured using Northern Eclipse version 5.0 (Empix Imaging Inc., Mississauga, ON) software.
Tail tips were homogenized on ice using a Barnant Mixer in a buffer containing 50 mM HEPES, 150 mM NaCl, 2.5 mM EGTA, 1 mM EDTA, 0.1% Tween 20, 100 μM phenylmethylsulfonyl fluoride (PMSF), 20 U/ml aprotinin, 10 mM β-glycerophosphate, 0.1 mM Na3VO4, 1 mM NaF, 1 mM dithiothreitol (DTT), 53 U/ml benzonase (Sigma), using 3 ml of buffer per gram of tissue (Minshull et al.,1994). After tissue homogenization, the mixture was sonicated 4 times for 15 sec each on ice and then incubated on ice for 30 min. Homogenates were centrifuged at 12,000g for 10 min at 4°C and the collected supernatant was stored at −70°C. Protein concentration was determined using the Bio-Rad Protein Assay (Bio-Rad).
Tissue homogenates (500 μg) were diluted to a 1-ml volume in IP buffer containing 25 mM Tris-HCl (pH 8.0), 400 mM KCl, 10 mM EDTA, 0.1% Triton X-100, 1 mM DTT, 1 mM β-glycerophosphate, 2.5 mM sodium pyrophosphate, 1 mM Na3VO4, 1 μg/ml leupeptin, and 1 mM PMSF. Homogenates were precleared by rotation with 20 μl protein G-sepharose beads (Amersham) per 1 ml of diluted tissue homogenate for 20 min at 4°C. Following preclearing, the mixture was centrifuged at 3,000g for 10 min at 4°C and the supernatant transferred to microfuge tubes containing 20 μl of fresh beads and 10 μl of anti-PKC (H-300), 5 μl of anti-ERK1 (K-23), anti-ERK2 (C-14) (Santa Cruz Biotechnology, Santa Cruz, CA), or anti-frog TRα rabbit polyclonal antibody (Eliceiri and Brown,1994) (a generous gift from D. Brown, Carnegie Institute). The antibody-bead-homogenate mixture was mixed by rotation at 4°C for 3 hr and the beads were washed 3 times with 1 ml IP buffer. The IP complexes were used for kinase assays or boiled for 3 min in 30 μl 3 × sodium dodecyl sulfate (SDS) sample buffer (187.5 mM Tris-HCl, 6% SDS, 30% glycerol, 150 mM DTT, 0.03% bromophenol blue, pH 6.8) for immunoblotting.
The IP complexes from the PKC, ERK1, or ERK2 IPs were further washed in 1 ml kinase reaction buffer containing 50 mM HEPES (pH 8.0), 10 mM MgCl2, 1 mM EDTA, 10 mM β-glycerophosphate, 0.1 mM Na3VO4, 1 mM NaF, 1 mM DTT, and 20 μM reduced glutathione. The immunoprecipitates were then incubated in 25 μl kinase reaction buffer containing 1.2 μM cold ATP (Invitrogen Canada Inc., Burlington, ON), 10 μCi γ32P-ATP (Amersham Biosciences) and 1 μg histone H1 (Roche) or myelin basic protein (MBP) (Sigma) for 40 min at 30°C. The reactions were terminated by boiling for 3 min in SDS sample buffer. MBP reactions were separated by 15% SDS-polyacrylamide gel electrophoresis (SDS-PAGE) while the proteins in the histone H1 reactions were separated by 12% SDS-PAGE. The gels were dried and exposed to phosphor imaging screens (Amersham Biosciences) for 16 hr on average. Radiographic data was obtained using a Storm 820 optical scanner phosphorimaging system at 50 μm resolution (Amersham Biosciences). The density of individual bands was analyzed using Northern Eclipse v5.0 and normalized to the DMSO control.
Equal quantities of tail tissue homogenates [64 μg (30 μl)/lane] or the proteins immunoprecipitated with the anti-TRα antibody were separated on an 8% or 10% SDS-PAGE gel, respectively. Protein loading and transfer quality were verified by membrane staining with 0.1% Ponceau S (Sigma) in 5% acetic acid. Membranes were blocked with 5% nonfat milk in 140 mM NaCl, 20 mM Tris-base (TBS), pH 7.6, at 4°C overnight and probed with a primary antibody solution diluted in 5% nonfat milk/TBS and 0.1% Tween (TBST) with gentle agitation for 1 hr at room temperature. The primary antibodies were a mouse monoclonal anti-phosphotyrosine antibody (anti-p-Tyr-100, 1:2,000 dilution, Cell Signaling Technologies, Inc., MA), a rabbit polyclonal anti-phosphoserine antibody (1:125 dilution, Stressgen Bioreagents, BC), or a mouse monoclonal anti-human TRα whose epitope has 60% amino acid sequence identity to Rana catesbeiana TR° (1:1,000 dilution, LabVision, Corp., CA). We have previously verified that this antibody recognizes bacterially expressed frog TR° protein (data not shown). The blots were washed with TBST for 1 hr. The membranes were incubated with secondary antibody solution diluted (1:2,000) in 5% nonfat milk/TBST for 1 hr at room temperature. The secondary antibody is an IRDye 800CW conjugated anti-mouse IgM or anti-rabbit IgG antibody depending upon the primary antibody used (Rockland Inc., Gilbertsville, PA). The blots were washed with TBST for 1 hr. The prepared blots were scanned by an Odyssey infrared imaging system (LI-COR Biosciences, Lincoln, NE) at 42 μm resolution. The band intensities were measured using Northern Eclipse version 5.0 software. The densitometric values of the bands of interest were normalized to the total protein loading detected by amido black 10B stain (Link,1999).
Two-Dimensional (2D) Polyacrylamide Gel Electrophoresis
A T3-DMSO treatment (tadpoles injected with T3 and tail tips removed at 24 hr and cultured in the presence of DMSO for a further 24 hr) tail-tip homogenate sample was desalted using a Sephadex G-25M column in 60 mM ammonium bicarbonate. Three 300-μg samples were then aliquoted, lyophilized, and solubilized in isoelectric focusing (IEF) buffer for 6 hr and subsequently separated by 2D gel electrophoresis as described previously (Helbing et al.,2003). Proteins on one of the three 2D gels were stained sequentially with both a colloidal Coomassie stain (Sigma) and a ProQ Diamond phosphoprotein stain (Molecular Probes, Eugene, OR) to examine the total proteins and phosphoproteins, respectively. Proteins from the other two 2D gels were transferred onto nitrocellulose membrane and probed as described in the Immunoblotting section. One of the 2D blots was incubated with anti-PKC antibody at a 1:1,000 dilution at room temperature for 1 hr. The other 2D blot was incubated with anti-phosphotyrosine antibody at a 1:2,000 dilution at 4°C overnight. Secondary antibodies and visualization were as described above. Protein spots of interest were excised from the Coomassie-stained gel and identified by peptide mass fingerprinting (PMF) of their tryptic fragments, using matrix- assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) as described previously (Helbing et al.,2003).
Preparation of RNA
Total RNA was isolated from RNAlater-preserved tail tips using TRIzol reagent according to the manufacturer (Invitrogen). The procedure was adapted from that described previously (Crump et al.,2002; Veldhoen et al.,2002; Helbing et al.,2003). Individual tail tips were homogenized in 1 ml TRIzol per 100 mg tail tissue using a MM 300 Mixer Mill (Retsch GmbH & Co. KG, Haan, Germany) at 20 Hz for 6 min. After tissue homogenization, chloroform (200 μl chloroform/1 ml TRIzol) was added for phase separation. RNA was precipitated from the aqueous phase using isopropyl alcohol and washed with 75% ethanol. The isolated RNA was resuspended in RNase-free water and stored at −70°C until cDNA was prepared. The concentration of total RNA for each sample was determined by spectrophotometry at 260 nm.
Preparation of cDNA
One microgram total RNA was annealed with 500 ng random hexamer oligonucleotides and cDNA was generated using 200 U Superscript II Rnase H− reverse transcriptase as described by the manufacturer (Invitrogen). The 20-μl reaction was incubated at 42°C for 2 hr to generate cDNA and then diluted 20-fold for real-time quantitative polymerase chain reaction (QPCR) examination.
A MX4000 real-time quantitative polymerase chain reaction system (Stratagene, La Jolla, CA) was used to examine the expression of several gene transcripts as described previously (Veldhoen et al.,2006). Each 15-μl DNA amplification reaction contained 5 mM Tris HCl, 5 mM Tris Base, 50 mM KCl, 3 mM MgCl2, 0.01% Tween 20, 0.8% glycerol, 40,000-fold dilution of SYBR Green I (Molecular Probes Inc., Eugene, OR), 200 μM dNTPs, 83.3 nM ROX reference dye (Stratagene), 10 pmol of each primer, 2 μl of 20-fold diluted cDNA, and 1.0 U platinum Taq DNA polymerase (Invitrogen). The primers used were: TRα-forward 5′-GGACCAGAATCTTAGCGG-3′, TRα-reverse 5′-CATTGGTGCTTCGGTGAG-3′; TRβ-forward 5′-AGCAGCATGTCAGGGTAC-3′, TRβ-reverse 5′–TGAAGGCTTCTAAGTCCA-3′; L8-forward 5′-CAGGGGACAGAGAAAAGGTG-3′, L8-reverse 5′-TGAGCTTTCTTGCCACAG-3′. For each set of QPCRs, we included two controls, one without cDNA template and the other without Taq DNA polymerase, to establish the specificity of target cDNA amplification. The specificity of the appropriate products was verified by electrophoresis on 2% agarose gels and sequence confirmation. Standard plots for each target sequence were generated using known amounts of plasmid containing the amplicon of interest. The thermocycle started at 95°C (9 min) followed by 40 cycles of 95°C (15 sec), 55°C (30 sec) annealing, and 72°C (45 sec) extension. Cycle threshold (Ct) values obtained were converted into copy numbers. The copy numbers were determined using standard plots of Ct versus log copy number. Quadruplicate (Veldhoen et al.,2006) data for each target cDNA amplification were averaged and normalized to an invariant mRNA encoding ribosomal L8 protein control to generate relative copy numbers.
Statistical analysis was conducted using SPSS version 12.0 software (Chicago, IL). The Shapiro-Wilk test and Levene's test was used for examination of normality of data distribution and homogeneity of variances, respectively. As the data were not normally distributed, a non-parametric Kruskal-Wallis test (for K-independent samples) followed by Mann-Whitney U (MWU) two-tailed test (for two-independent samples) was applied. Spearman 2-tailed test was applied for correlation analysis. Statistical significance was at P < 0.05.
We thank D. Brown for the generous gift of TR antibodies. This work was funded by a NSERC operating grant to C.C.H. C.C.H. is also the recipient of a NSERC University Faculty Award, a SETAC early career investigator award for applied ecological research, and a Michael Smith Foundation for Health Research Scholar Award. D.D. is a recipient of NSERC graduate scholarships (PGS-M and PGS-D).