TBP2 is a substitute for TBP in Xenopus oocyte transcription
© Akhtar and Veenstra. 2009
Received: 9 March 2009
Accepted: 3 August 2009
Published: 3 August 2009
Skip to main content
© Akhtar and Veenstra. 2009
Received: 9 March 2009
Accepted: 3 August 2009
Published: 3 August 2009
TATA-box-binding protein 2 (TBP2/TRF3) is a vertebrate-specific paralog of TBP that shares with TBP a highly conserved carboxy-terminal domain and the ability to bind the TATA box. TBP2 is highly expressed in oocytes whereas TBP is more abundant in embryos.
We find that TBP2 is proteolytically degraded upon meiotic maturation; after germinal vesicle breakdown relatively low levels of TBP2 expression persist. Furthermore, TBP2 localizes to the transcriptionally active loops of lampbrush chromosomes and is recruited to a number of injected promoters in oocyte nuclei. Using an altered binding specificity mutant reporter system we show that TBP2 promotes RNA polymerase II transcription in vivo. Intriguingly, TBP, which in oocytes is undetectable at the protein level, can functionally replace TBP2 when ectopically expressed in oocytes, showing that switching of initiation factors can be driven by changes in their expression. Proteolytic degradation of TBP2 is not required for repression of transcription during meiotic maturation, suggesting a redundant role in this repression or a role in initiation factor switching between oocytes and embryos.
The expression and transcriptional activity of TBP2 in oocytes show that TBP2 is the predominant initiation factor in oocytes, which is substituted by TBP on a subset of promoters in embryos as a result of proteolytic degradation of TBP2 during meiotic maturation.
A key regulatory step in eukaryotic transcription initiation is the assembly of basal transcription apparatus at the core promoter. This apparatus includes RNA polymerase II (pol-II) and a set of basal transcription factors. For a long time the basal transcriptional machinery was thought to be universal and mainly invariant across different promoters. However, a growing body of evidence points towards a more dynamic and regulatory role for this machinery (reviewed in [1–4]). TATA-box binding protein (TBP) was once thought to be the general transcription factor involved in all transcription in eukaryotic cells. In higher eukaryotes, however, a number of TBP paralogs are present and there is clear evidence for TBP-independent transcription in a variety of model organisms [5–12]. So far three TBP paralogs have been described in metazoans; the insect-specific TBP-related factor 1 (TRF1) , the metazoan-specific TBP-like factor (TLF, also known as TRF2, TLP or TBPL1)  and the vertebrate-specific factor TBP2 (also known as TRF3 or TBPL2) [8, 12, 14].
TBP2 is the most closely related TBP paralog, sharing 95% sequence identity in the core domain [8, 12, 14]. It can bind to TATA box, interacts with TFIIA and TFIIB and promotes basal transcription in vitro. In addition, knockdown studies in fish and frogs showed that TBP2 is indispensable for embryonic development [8, 12] and is preferentially required for the transcription of embryonic vertebrate-specific genes and those involved in ventral specification during gastrulation in Xenopus . TBP2 is also essential in differentiation pathways in zebrafish and mouse [10, 11]. Although some TBP2 is present and required in early Xenopus embryos, it is most abundant in oocytes, suggesting that TBP2 has an important role in oocyte transcription . Here we provide functional evidence for the role of TBP2 in pol-II transcription by examining its binding to oocyte chromosomes and pol-II promoters, and by employing in vivo transcriptional assays involving an altered binding specificity mutant reporter system. We show that TBP2 is localized to active promoters in oocytes and can promote pol-II transcription. TBP2 is degraded during meiotic maturation. Surprisingly, TBP, when exogenously expressed in oocytes, can substitute for TBP2, which is indicative of dynamic and rapidly adaptable nature of core transcription machinery. Together, these observations establish the involvement of TBP2 in transcription initiation of oocytes. Moreover the proteolytic degradation of TBP2 during meiotic maturation is relevant for the initiation factor switching that occurs during the course of early development.
The Xenopus TBP (pSP64A-xTBP)  and TBP2 (pT7TSA-xTBP2)  constructs were utilized to generate altered binding specificity mutant versions of these proteins by introducing three point mutations in the carboxy-terminal domain of xTBP (I250F, V259T and L261V) and xTBP2 (I273F, V282T and L282V) using the Quick Change site-directed mutagenesis kit (Stratagene). Capped mRNAs were transcribed by an in vitro RNA synthesis kit (Ambion, Austin, TX). pCMV-CAT and pG4-hsp70-CAT have already been described  whereas pG4-BLCAT- ZFP36L2, pG4-BLCAT-TLF, pG4-BLCAT-polr2h and pG4-BLCAT-Cyr61 and pG4-BLCAT-intron (containing an intronic fragment of the WD repeat domain 42A (wdr42a) gene) were generated by cloning these promoters amplified from X. tropicalis genomic DNA, into pG4-BLCAT2  at the BamHI and XhoI sites (WA and GJCV, manuscript in preparation).
Germinal vesicle (GV) spreads were prepared as described in  with minor modifications. Defolliculated stage VI oocytes were transferred one at a time to isolation medium (83 mM KCl, 17 mM NaCl, 6.5 mM Na2HPO4, 3.5 mM KH2PO4, 1 mM MgCl2, 1 mM dithiothreitol, pH 7.0). The GV was isolated, the nuclear membrane was removed and the contents of GV quickly transferred to 20 μl of spreading medium containing 1% paraformaldehyde and 0.15% Triton X-100 on a glass cover slip. The cover slips were kept at room temperature for 1 hr before being stored in phosphate-buffered saline (PBS). GV spreads were rinsed in PBS containing 0.2% Tween and blocked with 0.5% bovine serum albumin and 0.4% gelatin followed by 1 hr incubations with primary and secondary antibodies at room temperature. TBP2 antibody 1G6 has already been described  whereas 3E6 is a mouse monoclonal antibody raised against the amino-terminal domain of Xenopus TBP2. The secondary antibody was Alexa 488-labelled goat anti-mouse immunoglobulin G (IgG) in a 1:100 dilution also containing 4 μg/ml 4',6-diamidino-2-phenylindole (DAPI). Images were taken with Zeiss Axiophot2 Fluorescence microscope.
Defolliculated oocytes were injected with 1 ng of in vitro-transcribed TBP/2-abs RNA in the cytoplasm using a Drummond Nanoject microinjection apparatus (Drummond Scientific, Broomall, PA). After a delay of 4 hr, GVs of these oocytes were injected with different amounts of plasmid constructs as indicated in the text and figure legends. The next day the oocytes were collected and RNA was isolated using Trizol (Invitrogen) extraction and LiCl precipitation. The RNA equivalent to four oocytes was analyzed by primer extension as described in .
Stage VI oocytes were incubated at 18°C in MBSH buffer containing 1 μg/ml progesterone (Sigma). Maturing oocytes were synchronized at germinal vesicle break down (GVBD) and then collected at different time points as indicated in the figure legends. Extracts from collected oocytes were prepared as previously described . The antibodies used were anti-TBP 58C9, anti-TFIIB C-18, anti-TFIIF RAP 74 C-18, anti-Mos C237 (Santa Cruz Biotechnology), anti-phospho-p44/42 MAP Kinase (Cell Signalling Technology), anti-TBP SL27, SL30 and SL33  and anti-H3 core and anti-H3Ser10P (Abcam).
Stage VI oocytes were injected with promoter constructs (for amounts see the figure legends) in the nucleus. After overnight incubation at 18°C, oocytes were cross-linked and processed for chromatin immunoprecipitation (ChIP) as described previously . For ChIP of HA-tagged TBP, 2 ng mRNA encoding HA-TBP2 was injected in the cytoplasm 4 hr prior to nuclear injections. The antibodies used were anti TBP-2 1G6 , anti-RNA polymerase II (Diagenode) and anti-HA (Covance).
TBP2 on the other hand is abundant in oocytes, whereas in eggs and early embryos very low levels of TBP2 are present. This was also observed when oocyte GV and embryonic nuclear extracts were probed with a TBP2-specific monoclonal antibody 3E6 (Figure 1B). The mechanism of the decrease in TBP2 expression, however, is unknown. To investigate this issue we followed the TBP2 levels during the course of oocyte maturation. This can be done in vitro by treating fully grown stage VI oocytes with progesterone. The maturation is marked by the appearance of a spot on the animal pole of the oocyte as a result of GVBD. TBP2 levels start to drop within 30 min after GVBD and by 4 hr the protein is hardly detectable (Figure 1C). This was further assessed by looking at some markers of meiotic maturation in oocytes, including degradation of cytoplasmic polyadenylation element binding protein and phosphorylation of MAPK and histone H3 serine10. TBP2 downregulation coincides with the accumulation of Xenopus Mos protein (Figure 1C). The Mos protein is a kinase that plays an important role in oocyte maturation. In response to progesterone treatment, the translation of Mos mRNA and the stability of the Mos protein increase, resulting in the gradual accumulation of the protein .
To examine the kinetics of overexpressed TBP and TBP2 in maturing oocytes, oocytes were injected with mRNA encoding TBP and HA-TBP2 (which can be distinguished from endogenous TBP2) and treated with progesterone. The injected synthetic TBP mRNA lacks its native 5' and 3'-UTR and is translated in oocytes, in contrast to the endogenous maternal TBP mRNA which is translationally masked . The kinetics of overexpressed TBP and TBP2 are similar, although the levels of endogenous TBP2 tend to decrease more rapidly (Figure 1D), which may reflect differences in expression level at the start of the experiment.
To assess whether the decrease in TBP2 levels is due to changes in translation or degradation, oocytes were treated with the translation inhibitor cycloheximide. To check translation efficiency under these conditions oocytes were injected with TBP mRNA in the presence and absence of this inhibitor. After 16 hr of incubation with cycloheximide there was hardly any effect on TBP2 levels, showing that TBP2 protein is quite stable in oocytes, whereas in progesterone-treated cells TBP2 expression is lost (Figure 1E). Therefore, proteolytic degradation of TBP2 is markedly increased during meiotic maturation. TBP, when ectopically expressed in oocytes, is also degraded during maturation although residual levels after maturation are substantially higher, similar to the situation when TBP2 is overexpressed (Figures 1D and 1E). Intriguingly, even well after GVBD, TBP2 is not completely degraded, which is consistent with the observation that some TBP2 persists in early Xenopus embryos (Figure 1A, see ).
To further confirm the association of TBP2 with promoters in oocytes, we expressed an HA-tagged version of TBP2 in oocytes and tested the recruitment of TBP2 on these promoters with an anti-HA antibody. Oocytes injected with promoter constructs but lacking HA-TBP2 were used as a negative control. A similar pattern of recruitment was obtained (Figure 3B). Together, these ChIP results show that TBP2 is recruited to pol-II and pol-III promoters in oocytes.
Next, we analyzed the behaviour of these proteins in oocyte transcription, for which we made use of the hsp70 promoter, which is known to be active in Xenopus oocytes and has a canonical TATA box upstream of its transcription start site . The TATA box of this promoter was mutated to TGTA. Wild-type and the mutant promoters were injected in the oocytes with and without mRNA encoding TBP-abs or TBP2-abs (Figure 4B). Mutation of the TATA box reduces transcription from this promoter, although some transcription persists (Figure 4C). In the presence of TBP2-abs, transcription from the TGTA promoter is rescued. In contrast, wt TBP2 is unable to promote transcription from the TGTA promoter, confirming that the rescue is specifically carried out by the exogenously provided abs-mutant proteins. Interestingly, TBP-abs, when expressed in oocytes, also rescues the transcription to the same extent (Figure 4C) whereas the levels of exogenous abs proteins in both cases are comparable (Figure 4D).
We also tested the X. tropicalis ZFP36L2 promoter, which has a canonical TATA box positioned 27 bp upstream of the transcription start site. We generated the TGTA mutant of this promoter and used it to test the activity of TBP and TBP2 on this promoter. The TGTA mutation significantly reduces transcription from this promoter (Figure 4E). Both TBP-abs and TBP2-abs can rescue transcription, with TBP2-abs being slightly more efficient than TBP in this regard (Figure 4E).
Collectively, our findings show that TBP can substitute for TBP2 if provided artificially to the oocyte core transcription machinery. On the hsp70 and ZFP36L2 promoters, the relative expression levels of TBP and TBP2 appear to be a major determinant of initiation factor substitution. To analyze the behaviour of TBP and TBP2 on promoters lacking a TATA box, we examined the recruitment of exogenously provided HA-TBP and HA-TBP2 on two TATA-less promoters (TLF, polr2h) and two TATA box-containing promoters (hsp70, ZFP36L2). Oocytes injected with promoter constructs but lacking HA-TBP/2 were used as a negative control. Furthermore, a background control construct (intron) was injected. The results substantiate the notion that TBP, if ectopically expressed, can be recruited to these promoters in oocytes (Figure 4G).
In the present study we show that Xenopus oocytes lack any detectable TBP, whereas TBP2 is abundant in oocytes. In eggs and early embryos relatively low levels of TBP2 are present. TBP starts to accumulate after meiotic maturation and during cleavage stages of development (Figure 1A). A similar expression profile of TBP and TBP2 has been observed in mouse oocytes , which suggests a major role for TBP2 in oocyte transcription. Indeed, TBP2 is recruited to the transcriptionally active loops of lampbrush chromosomes (Figure 2). In addition, TBP2 also appears to be a part of CBs and BSs, which are considered to be involved in preassembly of transcription complexes . The association of TBP2 with active promoters was further confirmed by ChIP on GV-injected promoters and on endogenous 5S rRNA (Figure 3). Surrounding follicle cells outnumber the oocytes; therefore it is not possible to perform ChIP on single copy loci in oocytes. Due to this limitation, injected promoter constructs were used to model oocyte transcription. This system has been used effectively to study the biochemistry of chromatin assembly and transcription during oogenesis and early embryogenesis . In order to examine the activity of TBP2 in oocytes, we utilized an altered binding specificity (abs) mutant reporter system . It cannot be ruled out that endogenous TBP2 behaves differently compared with its overexpressed abs mutant. However, the involvement of TBP2 in pol-II mediated transcription in oocytes may be inferred from the ability of abs-TBP2 to rescue the deleterious effects of TGTA promoter mutations, probably due to binding of abs-TBP2 to the TGTA sequence to which the endogenous TBP2 cannot bind (see Figure 4). These findings are congruous with the observation that only TBP2 is present in frog and mouse oocytes with no detectable TBP, which reinforces the notion that germ cell transcription is different from that of somatic cells .
We have found that during later stages of meiotic maturation TBP2 protein is actively degraded (Figure 1C). This could potentially achieve two biological goals. First, it might be essential for proper shutdown of global transcription. In early embryos, transcriptional repression can be perturbed by premature accumulation of TBP together with disruption of chromatin assembly [15, 27]. Second, this degradation of TBP2 in combination with the translational upregulation of TBP in embryos might be a mechanism to achieve switching of TATA binding factors between the oocyte and embryonic and/or somatic transcription machineries. The results of our experiments do not support a role for TBP2 degradation in the shutdown of transcription as it occurs even in the presence of abundant TBP2 (Figure 5), although a redundant role in transcriptional repression cannot be excluded. Furthermore, transcriptional repression is already established before TBP2 is degraded (Figure 6). TBP2 degradation is likely to be a mechanism that brings about the switching of TATA binding core transcription factors, facilitating the transition from a germ cell transcription apparatus to a somatic transcription machinery. However, this switching is not yet complete in early embryos because of low levels of TBP2, as discussed below.
The altered binding specificity mutant of TBP, when expressed in the oocytes, could also promote transcription from the hsp-70 and ZFP36L2 promoters (Figure 4). Earlier observations indicated that TBP and TBP2 can replace each other in initiating basal transcription in vitro as well as in complexes with TFIIA/ALF and DNA [8, 30, 31]. Intriguingly, many promoters that exclusively require TBP in embryos are also maternally expressed . TBP protein, however, is not present in oocytes where these maternal transcripts were initially made, implying an initiation factor switching on these promoters between oocytes and early embryos. Furthermore, TBP2 overexpression in TBP knockdown embryos significantly, although not completely, rescued the phenotype of TBP ablation . On a similar note, it has been proposed that in cells heterozygous for TBP (tbp +/-), elevated levels of TBP2 compensate abnormally low levels of TBP at some but not all promoters . This shows that on many promoters the two TATA binding proteins can act redundantly.
A complicating matter is that both TBP and TBP2 are required for early development in fish and frog [5, 6, 8, 12], and promoters that strictly require either TBP or TBP2 in embryos have been identified [9, 10]. The molecular basis of TBP-TBP2 selectivity remains to be elucidated. The embryonic transcription machinery (abundant TBP, low levels of TBP2) is not yet fully switched to the common somatic machinery (abundant TBP, no TBP2), effectively leaving TBP and TBP2 to compete for interactions with embryonic activators and promoter elements. Paradoxically, despite the relatively low levels of TBP2 in embryos, many genes depend on this factor, whereas ablation of TBP causes relatively moderate effects on gene expression . TBP2 may be competitive in embryonic gene regulation because of preferential interactions with embryonic activators or other maternal factors that – like TBP2 – still persist in early development; such preferential interactions would be analogous to the preference of the Caudal activator for a promoter nucleoprotein architecture with the downstream promoter element . For example, TBP2 preferentially accommodates activation of vertebrate-specific embryonic and ventral-specific genes . During later development, core factor switching may be completed when TBP2 levels further decline .
Our observations not only establish an important role of TBP2 in transcription of a highly specialized cell type which lacks any detectable TBP, but also provide evidence that the basal transcription machinery is highly flexible and can rapidly exchange factors depending on their expression. This leads to initiation factor switching on a subset of promoters that are active both in oocytes and embryos, whereas other promoters show factor selectivity, which accounts for their non-redundant function in early embryos .
This work was supported by Higher Education Commission of Pakistan with a PhD fellowship to WA and by NWO-ALW (Netherlands Organization for Scientific Research-Research Council for Earth and Life Sciences) with grant number 864.03.002 to GJVC. We thank B Scheijen and M Koeppel for critical reading of the manuscript, and RC Akkers, U Jacobi, E Jansen, N van Bakel and R Engels for help with frogs and animal care.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.