Skip to main content
NCBI home page
The Journal of General Virology logoLink to The Journal of General Virology
. 2014 Jan;95(Pt 1):142–152. doi: 10.1099/vir.0.056838-0

The adenovirus 55 residue E1A protein is a transcriptional activator and binds the unliganded thyroid hormone receptor

Vishnuka D Arulsundaram 1, Paul Webb 2, Ahmed F Yousef 3, Peter Pelka 4, Greg J Fonseca 1, John D Baxter 2,†, Paul G Walfish 5, Joe S Mymryk 1,✉
PMCID: PMC3917064  PMID: 24136366

Abstract

The early region 1A (E1A) of human adenovirus types 2 and 5 is differentially spliced to yield five distinct mRNAs that encode different proteins. The smallest E1A RNA transcript encodes a 55 residue (55R) protein that shares only 28 amino acid residues with the other E1A proteins. Even though it is the most abundant E1A transcript at late times post-infection, little is known about the functions of this E1A isoform. In this study, we show that the E1A 55R protein interacts with, and modulates the activity of the unliganded thyroid hormone receptor (TR). We demonstrate that E1A 55R contains a signature motif known as the CoRNR box that confers interaction with the unliganded TR; this motif was originally identified in cellular corepressors. Using a system reconstituted in the yeast Saccharomyces cerevisiae, which lack endogenous TR and TR coregulators, we show that E1A 55R nonetheless differs from cellular corepressors as it functions as a strong co-activator of TR-dependent transcription and that it possesses an intrinsic transcriptional activation domain. These data indicate that the E1A 55R protein functions as a transcriptional regulator.

Introduction

Early region 1A (E1A) is the first viral gene expressed after human adenovirus (HAdV) infection (Nevins et al., 1979). It has been most extensively characterized in species C HAdV, specifically types 2 (HAdV-2) and 5 (HAdV-5), which are virtually identical in sequence and are typically treated interchangeably (Berk, 2005; Frisch & Mymryk, 2002; Pelka et al., 2008; Yousef et al., 2012). The primary E1A transcript is extensively spliced to produce the 13S, 12S, 11S, 10S and 9S mRNAs, which encode proteins of 289, 243, 217, 171 and 55 residues (R), respectively, in species C (Perricaudet et al., 1979; Stephens & Harlow, 1987) (Fig. 1a). Splicing preserves the translational reading frame in all cases except the 9S mRNA, which utilizes an alternative reading frame in exon 2. As a result, the 9S-encoded E1A 55R protein shares 26 amino acids also present in the other four E1A proteins, with the remaining 29 amino acid residues derived from the alternative reading frame in exon 2. Interestingly, the first two of the amino acid residues encoded by exon 2 in the 9S are identical to residues 27 and 28 in the 13S- and 12S-encoded E1A proteins, but not the 11S and 10S products (Fig. 1b). During the early stages of infection, splicing favours the production of the 13S and 12S E1A mRNAs. However, splicing at late times during infection favours the production of the 9S E1A mRNA, which then predominates. Despite the fact that the 9S mRNA is the major E1A product at late times and was originally identified in 1978 (Berk & Sharp, 1978), little is known about the 55R E1A protein other than that it can promote viral gene expression and replication (Miller et al., 2012).

Fig. 1.

Fig. 1.

Open in a new tab

Schematic depiction of HAdV-2 E1A products and their ability to function as transcriptional co-activators of unliganded thyroid hormone receptor (TR). (a) The primary HAdV-2 E1A RNA transcript is alternatively spliced to yield five mRNA encoding proteins ranging in size from 289 to 55 amino acid residues. Note that the translational reading frame is preserved in all cases except the 9S mRNA. (b) Sequence alignment of the N-terminal portions of the five HAdV-2 E1A proteins. The sequences comprising the putative CoRNR box motif (CBM) are indicated in bold. Shading indicates residues common to most products, and the intron/exon boundaries are indicated. (c) The 289R, 243R and 55R E1A proteins activate transcription via the unliganded TR. The panel shows results of β-galactosidase assays performed with a yeast strain containing a single copy of the chicken lysozyme (F2) T3 response element (TRE) and expressing hTRβ1, the various wt E1A proteins fused to the Gal4 DBD, or the Gal4 DBD empty vector. Cultures were treated with either vehicle (grey bars) or 10−6 M T3 (black bars). β-Galactosidase activities were expressed as Miller units (mg protein)−1. Data shown were pooled from three independent experiments and calculated as mean±se.

Much more is known about the larger E1A proteins, which are responsible for activating expression of other viral early genes and reprogramming cell growth to provide an optimal environment for viral replication (Berk, 2005; Frisch & Mymryk, 2002; Pelka et al., 2008; Yousef et al., 2012). These changes result from the ability of these E1A proteins to interact with a variety of cellular regulatory proteins. Interaction of E1A with many of these proteins requires sequences that are highly conserved throughout the different serotypes, which are referred to as conserved regions 1–4 (Avvakumov et al., 2004). The E1A 55R protein does not contain any of these conserved regions and is therefore unlikely to be able to target most of the cellular regulators bound by the larger E1A isoforms.

The nuclear receptor (NR) superfamily is a diverse group of transcriptional regulators (McEwan, 2009). Many NRs are targeted by viruses to regulate expression of their own genes and to optimize cellular gene expression to facilitate viral propagation (Miller & Mymryk, 2011). The thyroid hormone receptors (TRs) are hormone-regulated NRs that control growth, development and homeostasis by binding to thyroid hormone response elements (TREs) in target promoters and altering transcription (Cheng, 2000; Mangelsdorf et al., 1995; Yen, 2001). Unliganded TRs typically repress thyroid hormone activated genes by recruiting corepressors, such as NCoR (NR corepressor) and SMRT (silencing mediator for retinoid and thyroid hormone receptors) (Chen & Evans, 1995; Hörlein et al., 1995; Sande & Privalsky, 1996). NCoR and SMRT contact the TR directly via C-terminal NR interacting domains, called CoRNR box motifs (CBMs) (Glass & Rosenfeld, 2000). CBMs are composed of a consensus sequence I/LXXI/H/LIXXX/L, and adopt a three-turn α-helical structure that docks on a hydrophobic surface of the unliganded TR ligand binding domain (Glass & Rosenfeld, 2000; Hu & Lazar, 1999; Marimuthu et al., 2002; Nagy et al., 1999; Perissi et al., 1999; Webb et al., 2000; Xu et al., 2002). Thyroid hormone, predominantly tri-iodothyronine (T3), activates transcription by enhancing packing of TR C-terminal helix 12 over the lower part of the corepressor’s binding surface, simultaneously promoting corepressor release and creating a binding site for co-activators (Glass & Rosenfeld, 2000).

We previously determined that E1A 289R and 243R proteins behave as co-activators of unliganded TRs that bind in the typical corepressor mode (Meng et al., 2003, 2005; Sato et al., 2009). We used the budding yeast Saccharomyces cerevisiae to reconstruct individual TR transcription complexes and analyse their activities as yeast contains conserved eukaryotic general transcription factors and chromatin remodelling enzymes, but lacks endogenous NRs and NR coregulators that can complicate the interpretation of experiments. We found that the HAdV-2/5 E1A 289R and 243R proteins enhanced activity of unliganded TRs in yeast (Meng et al., 2003, 2005). Activation was suppressed by T3 and was dependent upon a consensus CBM spanning E1A residues 20–28 that binds TR similarly to corepressors (Fig. 1b).

In this study, we report that the CBM is preserved in the E1A 55R proteins of HAdV-2 and -5, despite the presence of an intron/exon boundary in the portion of the transcript encoding this motif, and bind and co-activate unliganded TR. Further, the E1A 55R proteins function as transcriptional activation domains when recruited directly to a reporter gene construct via fusion to a heterologous DNA-binding domain. These results further support a role for the E1A 55R proteins in transcriptional regulation.

Results

The HAdV-2 E1A 55R protein can activate transcription by the unliganded TR

We previously determined that HAdV-2 E1A 289R and 243R proteins function as co-activators of unliganded TRs in yeast (Meng et al., 2003, 2005). This activity requires the CBM between residues 20 and 28. Alternative splicing of the primary HAdV-2 E1A transcript occurs within the region encoding the CBM to generate the 11S, 10S and 9S mRNA products (Fig. 1a, b). Interestingly, while the signature motif of the CBM is destroyed in the 11S- and 10S-encoded proteins, it is regenerated in the 9S-encoded E1A 55R protein by the use of an alternative reading frame in exon 2 (Fig. 1b). To assess whether the E1A 55R protein could function as a co-activator of TR-dependent transcription, each of the five HAdV-2 E1A products was tested in a yeast model system.

In accordance with previous results, E1A 289R and 243R functioned as co-activators for the unliganded TR, and this effect was downregulated by the addition of T3 (Fig. 1c). The E1A 217R and 171R proteins, which do not contain a complete CBM, failed to activate transcription by the unliganded TR. In contrast, the E1A 55R protein functioned as a strong co-activator of the TR and this was downregulated by T3. Thus, alternative splicing appears to regenerate a functional CBM in the E1A 55R, allowing it to function as a co-activator of unliganded TRs similar to the 289R and 243R proteins.

The 55R E1A proteins of HAdV-2 and -5 contain a CBM and activate transcription by the unliganded TR

Alignment of the sequences of the HAdV-2 and the closely related HAdV-5 55R E1A proteins revealed that both contain a consensus CBM, but differ by three amino acids located near the C terminus (Fig. 2a). The HAdV-5 E1A 55R E1A protein also functioned as a co-activator of the unliganded TR, albeit at a lower level than the closely related HAdV-2 E1A 55R (Fig. 2b). This suggests that the three amino acid residues that differ between these two proteins affect function, despite the fact that they are not part of the CBM (Fig. 2a). Inspection of the amino acid sequences of the smallest E1A proteins expressed by the more distantly related HAdV-3 (species B) and 12 (species A) indicates that neither contains a consensus CBM (Fig. 2a). Not surprisingly, neither functioned as a co-activator (Fig. 2b).

Fig. 2.

Fig. 2.

Open in a new tab

Comparison of the activities of the smallest E1A proteins of subgroups A, B and C HAdV. (a) Sequence alignment of the 55R E1A proteins of HAdV-2 and -5 with the 57R E1A protein of HAdV-3 and 53R protein of HAdV-12. Darker shading depicts a higher degree of conservation; asterisks indicate non-conserved residues in HAdV-2 and -5 E1A 55R. The location of the putative CBM in HAdV-2 is indicated. (b) The E1A 55R proteins of HAdV-2 and -5 transcriptionally co-activate unliganded TR, but not the E1A 57R or 53R proteins of HAdV-3 and -12. Other details as for Fig. 1(c). (c) TR interacts with the HAdV-2 and -5 E1A 55R isoforms in vitro. Panels show an exposure of a 10 % SDS-polyacrylamide gel loaded with input labelled hTRβ1 or hTRβ1 retained on GST beads in the presence of the indicated GST-E1A 55R proteins with or without T3 (10−6 M).

The presence of the CBM in the HAdV-2 and -5 E1A 55R proteins suggested that their ability to co-activate TRs in the absence, but not in the presence, of T3 results from their ability to only bind unliganded TR. To confirm this, we examined interactions of TRβ with bacterially expressed E1A 55R proteins in glutathione S-transferase (GST) pulldown assays in vitro. TRβ bound to HAdV-2 and -5 E1A 55R in the absence of T3, and these interactions were decreased in the presence of T3 (Fig. 2c). This T3-dependent decrease in interaction by E1A parallels what is observed in TR–corepressor interactions, confirming that the E1A 55R proteins bind the TR like a cellular corepressor (Glass & Rosenfeld, 2000). GST pulldowns using recombinant HAdV-3 and -12 E1A did not detect binding of TRβ, as expected based on their lack of a consensus CBM and lack of function in the yeast co-activation assay (data not shown).

Mapping the regions of the E1A 55R protein needed for co-activation

To determine the regions of the HAdV-2 E1A 55R protein necessary for co-activation of transcription by unliganded TR, we constructed and tested a panel of mutants in the yeast system for their ability to activate TR-dependent gene expression (Fig. 3a). L23A/I24A, a double point mutation within the CBM or deletion of most of the CBM (Δ4–25) abrogated the ability of the E1A 55R protein to activate transcription, as did an N-terminal deletion of amino acids 2–9 and C-terminal truncation at residue 28. Thus, activation requires the CBM and flanking regions from residue 5 to residue 32.

Fig. 3.

Fig. 3.

Open in a new tab

Mutational analysis of the regions of the HAdV-2 and -5 E1A 55R proteins required for co-activation of unliganded TR. (a) The CBM is required for co-activation and the C terminus of E1A negatively regulates function. The panel on the right shows results of β-galactosidase assays performed as in Fig. 1(c), with yeast strains expressing wt HAdV-2 E1A 55R, wt HAdV-5 E1A 55R or the indicated 55R mutants in the presence or absence of T3 (10−6 M). (b) TR interaction requires the E1A 55R CBM in vitro. Other details as for Fig. 2(c). The L23A, L24A mutant contains two point mutations that disrupt the CBM.

Interestingly, the N- and C-terminal residues that flank the region of E1A 55R that is required for TR co-activation seemed to inhibit E1A activity. Indeed, removal of amino acids 1–6 enhanced activity, and truncations removing the C-terminal region greatly improved co-activation. Furthermore, the Δ1–4/Δ33–55R mutant, lacking both N- and C-terminal residues, functioned as a strong co-activator for the unliganded TR.

As mentioned above, the HAdV-2 and HAdV-5 E1A 55R proteins differ only by three amino acid residues located in the C-terminal region (Fig. 2a), and HAdV-5 E1A 55R is a weaker activator of the TR than HAdV-2 E1A 55R (Fig. 2b). Replacement of two of these residues with those found in HAdV-2 E1A 55R (HAdV-5 55R E1A K50N/R52C) increased activity substantially (Fig. 3a). Notably, C-terminal truncations of the E1A 55R protein that delete this region also increase activity, suggesting that the C terminus negatively regulates the ability to activate TR-dependent transcription (Fig. 3a). We noted that two of the three changes in HAdV-5 55R E1A introduce basic amino acids, which are generally rare in the acidic E1A proteins. Conversion of these residues to aspartic acid (K50D/R52D) increased function dramatically (Fig. 3a).

In GST pulldown experiments, the Δ33–55 and Δ1–4/Δ33–55 E1A mutants exhibited enhanced interaction with the TR in the absence of T3 (Fig. 3b). In contrast, the L23A/I24A double point mutant, which is predicted to compromise binding of the CBM to the TR, no longer bound the TR under any conditions. Thus, amino acids 33–55 of the E1A 55R protein function to negatively regulate TR co-activation in yeast and reduce TR binding in vitro.

E1A 55R binds the TR in mammalian cells

Our results suggest that the species C E1A 55R proteins bind the TR in vitro and in vivo in yeast (Figs 2 and 3). We tested whether such an interaction could occur in mammalian cells using a mammalian two-hybrid test in human U2OS osteosarcoma cells. U2OS cells were transfected with a Gal4-responsive luciferase reporter, a vector expressing the hTRβ1 fused to the VP16 transcriptional activation domain, and vectors expressing the Gal4 DBD fused to the wt HAdV-2 or -5 55R E1A proteins or the C-terminal mutants that exhibit improved binding and activity in vivo in yeast and in vitro (Fig. 4). The results of these experiments exactly reproduce our observations in yeast and in vitro. Virtually no interaction was seen between the wt HAdV-5 55R E1A and the TR, a weak interaction was seen with the wt HAdV-2 55R E1A protein, and the C-terminal mutants interacted very strongly with the TR in the absence of T3. As observed in vitro and in yeast, these interactions in mammalian cells were abolished by the addition of T3. These results confirm that the HAdV-2 55R E1A protein can interact with the TR in the context of mammalian cells and also suggest that the C-terminal region inhibits this interaction.

Fig. 4.

Fig. 4.

Open in a new tab

E1A 55R interacts with the TR in mammalian two-hybrid analyses. U2OS cells were transfected with a Gal4-responsive luciferase reporter, the indicated Gal4 DBD fusions (bait) and TR-VP16 transcriptional activation domain fusions (prey) in the presence or absence of hormone (T3, 10−6 M). Luciferase activity was normalized to β-galactosidase activity and expressed as relative light units (RLU).

Mutations in the TR corepressor binding surface influence activation by E1A 55R

We examined the requirement for the TRβ corepressor binding surface in co-activation by HAdV-2 E1A 55R (Fig. 5a) using a panel of TRβ mutants previously characterized for their effects upon NCoR interaction (Marimuthu et al., 2002). We compared this with our previous data generated using a C-terminal fragment of NCoR (NCoRC), which also functions as a co-activator of unliganded TR (Meng et al., 2005). A number of these TRβ mutants also inhibited co-activation by E1A 55R in yeast. Most notably, mutations within residues that lie under the usual position of helix 12 in liganded TRβ and form an essential part of the corepressor binding surface (I280K, C309W) impaired TR co-activation by both E1A and NCoRC. However, there are three subtle differences. Specifically, (i) E1A 55R co-activation was not as significantly affected by the C309W mutation as was NCoRC; (ii) co-activation by E1A 55R was strongly reduced using the V284 mutant, which does not substantially affect co-activation by NCoRC; and (iii) the I302A mutant functioned better than wt hTRβ1 with E1A 55R, whereas it was slightly impaired with NCoRC. This is consistent with previous results (Hu et al., 2001; Marimuthu et al., 2002), which suggest that different NCoR and SMRT CBM sequences contact the same TR surface but exhibit subtle variations in this binding mode. Thus, E1A 55R interacts with an hTRβ1 surface that is similar, but not identical, to the one that binds NCoR.

Fig. 5.

Fig. 5.

Open in a new tab

E1A 55R co-activation requires the TR corepressor binding surface. (a) Mutations in the TR CoR binding surface disrupt co-activation by E1A and NCoR truncations. The panel shows results of β-galactosidase assays performed as in Fig. 1(c), with yeast strains expressing wt E1A 55R or a C-terminal fragment of NCoR (NCoRC) along with wt TR or the indicated TR mutants. (b) HAdV-2 55R E1A co-immunoprecipitates with unliganded TR. U2OS cells were transfected with plasmids expressing the indicated FLAG tagged TRs, and EGFP fused to the HAdV-2 55R E1A. Cells were treated or not treated with T3, lysates were immunoprecipitated (IP) with anti-FLAG antibody, and immunoblots (IB) were performed using an anti-EGFP antibody to detect E1A 55R. (c) CBM peptide blocks TR binding to E1A 55R. The sequence of the NCoR CBM 3 peptide is shown at the bottom. Key hydrophobic residues that define the CBM are shaded. The panels show an exposure of a 10 % SDS-polyacrylamide gel loaded with TRs retained on GST-E1A 55R columns plus increasing amounts of CBM peptide (0.3–3 µg). Experiments were performed in the absence of T3 except where indicated.

Selective TR mutants were also tested for their ability to bind the 55R E1A protein in mammalian cells using co-immunoprecipitation assays (Fig. 5b). U2OS cells were transfected with plasmids expressing FLAG epitope tagged wt TR or selected TR mutants and the HAdV-2 55R E1A protein fused to EGFP. The HAdV-2 E1A 55R protein was co-immunoprecipitated with wt TR in the absence, but not in the presence of T3 (Fig. 5b). The F451X TR mutant, which contains a C-terminal truncation removing helix 12, is unable to bind ligand and associated with the E1A 55R in both the presence and absence of T3. The I280K mutation, which disrupts corepressor binding by the TR (Marimuthu et al., 2002), also greatly reduced interaction of the TR with the HAdV-2 E1A 55R protein, further suggesting that the CBM in E1A mediates interactions with the TR that are similar to that observed for various NR corepressors. These data agree fully with the yeast co-activation data obtained in Fig. 5(a).

To further confirm that HAdV-2 55R E1A interacts with hTRβ1 in the typical corepressor mode, we tested if a short peptide that overlaps NCoR CBM 3 could interfere with the ability of E1A 55R to bind unliganded TR in GST pulldowns (Fig. 5c). As expected, this peptide functioned in a dose-dependent fashion to inhibit E1A 55R wt as well as the Δ1–4/Δ33–55 mutant from interacting with the TR. Together, our observations indicate that E1A 55R is a TR co-activator that binds the TR in a manner similar to cellular NR corepressors.

HAdV-2 E1A 55R contains an intrinsic transcriptional activation domain

Given that HAdV-2 E1A 55R binds the unliganded TR and functions as a co-activator, it seemed likely that this E1A protein functions in some way to recruit components of the transcriptional apparatus to the transcriptional template. We tested the E1A 55R protein to determine if it contains a transcriptional activation domain by fusing it to the heterologous LexA DNA-binding domain (DBD) and expressing it in yeast containing a β-galactosidase reporter gene under the control of a minimal promoter containing LexA binding sites. If the E1A 55R protein contained a transcriptional activation domain, it would be expected to activate expression of the reporter gene when tethered by the LexA DBD. Intriguingly, the wt E1A 55R protein did not activate transcription in this assay. However, removal of the C-terminal 11 residues resulted in a strong activation of reporter gene expression, as did the mutation in which the two basic residues in the HAdV-5 55R E1A were replaced with acidic residues (Fig. 6a). Similar results were seen in transient transfection assays in human U2OS osteosarcoma cells using Gal4 DBD fusions and a Gal4-dependent luciferase reporter construct (Fig. 6b, top). The higher level of transcriptional activation by the E1A 55R mutants seen in U2OS cells did not result from large changes in protein expression as determined by Western blot analysis (Fig. 6b, bottom). These results demonstrate that both the HAdV-2 and -5 E1A 55R proteins contain a transcriptional activation domain, but that it is somehow masked by the C-terminal region. Interestingly, mutations that unmask this transcriptional activity also improve co-activation of unliganded TR (Fig. 3). This correlation suggests that these two activities are indeed related and also that interaction of the E1A 55R protein with the TR results in a conformational change that exposes this otherwise hidden transcriptional activation domain.

Fig. 6.

Fig. 6.

Open in a new tab

E1A 55R protein functions as a transcriptional activator. (a) Results of yeast one-hybrid analysis performed in S. cerevisiae wt yeast BY4741 and the indicated gene deletion strains. Yeast were cotransformed with plasmids expressing the indicated LexA DBD fused E1A bait proteins and β-galactosidase reporter gene under the control of a minimal promoter containing LexA binding sites. Both the HAdV-2 E1A 55R Δ45–55 and the HAdV-5 E1A 55R R52D/K50D mutants contain an intrinsic activation domain, whose function relies on components of the SNF/SWI and SAGA complex. (b) E1A 55R functions as a transcriptional activator in human cells. Human U2OS osteosarcoma cells were transiently transfected with a Gal4-dependent luciferase reporter construct and plasmids expressing the indicated E1A 55R E1A products fused to Gal4 DBD. Extracts were analysed for luciferase activity (top) and expression of Gal4 fusions (bottom). Removal or mutation of the C-terminal region of the E1A 55R protein unmasks a transcriptional activation function in mammalian cells. Error bars in (a) and (b) indicate ±sd.

Transcriptional activation by E1A 55R requires SAGA and SWI/SNF

The yeast Gcn5 acetyltransferase serves as a highly conserved catalytic subunit of the multi-component SAGA (SPT, ADA, GCN5, Acetylation) transcriptional adaptor complex (Timmers & Tora, 2005). In addition, the yeast SWI/SNF chromatin remodelling complex is a highly conserved eukaryotic multi-component adaptor complex that provides an important ATP-dependent function that is required for transcription of a subset of yeast genes (Martens & Winston, 2003).

We assessed the effect of disrupting the genes encoding the Gcn5 or Ada2 component of the SAGA complex and the Swi2 component of the SWI/SNF complex on the ability of HAdV-2 E1A 55R to activate transcription when tethered to a reporter gene by the LexA DBD (Fig. 6a). Disruption of genes encoding SAGA components, as well as SWI2, reduced E1A 55R-dependent activation. These transcriptional defects appear specific as our previous studies indicate that these yeast strains retain the ability to support activation by other portions of E1A or p53 (Yousef et al., 2008, 2009). These results indicate that E1A 55R requires both SAGA and SWI/SNF for activation of transcription.

Discussion

Despite its original identification as the major mRNA product of the HAdV E1A gene during the late phase of virus infection, little is known about the function of the 9S-encoded E1A 55R protein. In large part, this lack of information is due to the fact that this protein lacks the four regions of highly conserved sequence present in the E1A proteins of all the different HAdV types (Avvakumov et al., 2004). Our previous analysis of regions of the E1A 289R protein required to co-activate the unliganded TR implicated a CBM consensus motif (LDQLIEEVL) spanning residues 20–28 (Meng et al., 2003, 2005). We also determined that the same motif is conserved in E1A 55R. Therefore, we investigated whether E1A 55R and other E1A proteins produced by alternative splicing could potentiate transcriptional activation by unliganded TRs.

Our data confirm that the E1A 55R protein of species C, like the E1A 289R and 243R proteins, functions as a transcriptional coregulator of the TR. This appears to be a function not present in other HAdV species, indicating that this may be of secondary importance for infection. Mutational analysis has defined distinct regions of the E1A 55R that contribute to this process (Fig. 7a). The first of these is the CBM motif, which is reconstituted in the E1A 55R protein in spite of the presence of an intron/exon boundary within the region encoding this sequence (Fig. 1b). The CBM in the E1A 55R protein is required for co-activation of unliganded TR (Fig. 3a) and binding (Fig. 3b). The E1A 55R CBM binds the TR similarly to CBM containing cellular corepressors. Firstly, the E1A 55R–TR interaction was blocked by T3, a characteristic feature of corepressor binding. Secondly, a CoRNR box peptide derived from NCoR could compete in vitro with E1A 55R for interaction with the TR. Thirdly, mutations in the TR corepressor binding surface greatly impair the ability of the E1A 55R protein to co-activate unliganded TR in a yeast model system. Thus, the E1A 55R isoform is a viral TR interacting protein that binds in the corepressor mode.

Fig. 7.

Fig. 7.

Open in a new tab

Summary of E1A 55R activity. (a) E1A 55R contains a CBM-like sequence spanning residues 20–28 that confers interaction with the TR. The minimal region required for activation of TR-dependent gene expression spans residues 7–32 and flanking regions appear to repress the activity of the minimal activation domain. (b) Model of E1A-dependent activation of TR-dependent gene expression. E1A 55R interacts with the TR and requires the SWI/SNF and SAGA complexes for activation of gene expression.

In addition to the CBM, our data also indicate that the E1A 55R protein contains an intrinsic transcriptional activation domain and that the SAGA and SWI/SNF complexes are required for both the intrinsic activation by E1A 55R and TR-dependent co-activation (Fig. 6). Interestingly, co-activation by both the HAdV-2 and HAdV-5 E1A 55R proteins appears to be negatively regulated by the C terminus, as mutants lacking this region bind and co-activate TR better than their wt counterparts (Figs 3 and 4). Removal or mutation of the C-terminal negative regulatory region also exposes the activation function of the E1A 55R protein when recruited directly to DNA by the Gal4 or LexA DBD in yeast and mammalian systems (Fig. 6). In addition, the C-terminal region of the HAdV-5 E1A 55R appears to be a more effective inhibitor. In the HAdV-5 E1A 55R protein, two of the three amino acid differences in the C-terminal region are basic residues which could potentially interact with the acidic residues present in the CBM or surrounding N-terminal region, stabilizing an inactive state. Mutation of these basic residues to acidic residues produces a hyperactive co-activator with improved ability to bind the TR, supporting this conjecture (Figs 3 and 4). Intriguingly, there are a number of serines located in the C-terminal region of the 55R E1A proteins. Phosphorylation of these residues and the resulting negative charges could similarly regulate 55R E1A function by affecting potential N- and C-terminal interactions. Binding of the E1A CBM to the unliganded TR may also destabilize the interaction of the E1A 55R C terminus with its N terminus, exposing its intrinsic transcriptional activation domain.

Taken together, we propose a model in which the E1A 55R protein enhances unliganded TR activity by acting as a bridge that permits TR to recruit and or stabilize chromatin modification and remodelling complexes, enhancing transcription (Fig. 7b). Most importantly, these data suggest that the E1A 55R protein has the potential to act as a transcriptional coregulator, much like the larger E1A splice variants. There are analogies between the properties of alternatively spliced HAdV E1A and another differentially spliced factor encoded by a small DNA tumour virus. Alternative splicing of the SV40 early transcript generates Large T and Small t antigens that retain a common 82 N-terminal amino acid segment before diverging as a result of splicing. This common region contains a J-domain that directs interaction with the host-encoded Hsp70 chaperones (Ahuja et al., 2005). Thus, the use of alternative splicing to generate proteins with novel properties appears to be a common theme in these viruses. However, we are unaware of any other example where alternative splicing into a different reading frame reconstitutes an identical protein binding motif, as is so elegantly engineered by HAdV E1A.

The exact role of the interaction of the E1A 55R proteins with the TR is not clear, although it is presumably related to transcriptional regulation occurring at later stages of infection, as this is when the E1A 55R protein is expressed (Miller et al., 2012). Given similarities between nuclear hormone receptors, it is also possible that E1A 55R could engage in functionally important interactions with corepressor domains of other members of this family of transcription factors. Nevertheless, there are precedents for a role for thyroid hormone in the viral life cycle. SV40 late gene expression is directly modulated by the TR through a TRE located at the transcription initiation site (Desvergne & Favez, 1997; Zuo et al., 1997). Other viruses also utilize the TR to aid in controlling their transcription (Desai-Yajnik et al., 1995; Park et al., 1993).

Our results provide further evidence for a role of the E1A 55R proteins in transcriptional regulation and demonstrate that it can interact with unliganded TRs via a CBM motif that is preserved despite differential splicing.

Methods

Yeast strains and media.

S. cerevisiae wt strain YPH499 (MATα ura3 lys2 ade2 trp1 his3 leu2) was used for most transcriptional activation assays. Wild-type yeast BY4741 (MATα ura3Δ0 his3Δ1 leu2Δ0 met15Δ0) and congenic haploid strains containing disruptions of ADA2, GCN5 or SWI2 were obtained from Dr C. Boone (University of Toronto, Canada). Yeast transformants were grown under standard conditions (Adams et al., 1998).

Assay of TR-dependent transcriptional activation in yeast.

hTRβ1-YEp56, hTRβ1-YEp46 and the β-galactosidase reporter plasmids TRE-F2×1 were described previously (Anafi et al., 2000; Meng et al., 2003; Walfish et al., 1997). Wild-type and mutant E1As were expressed as Gal4 DBD fusions from the vector pAS1 with either TRP1 or LEU2 selectable marker (Shuen et al., 2002). Selected hTRβ1 mutants were expressed using the Yep46 yeast expression vector as previously described (Meng et al., 2005).

GST pulldown assays and synthetic peptide competition analyses.

The wt E1A 55R and various mutants were expressed as GST fusions in Escherichia coli BL21 cells and prepared by standard methods. Full-length 35S-labelled hTRβ1 was synthesized by in vitro transcription and translation (TNT; Promega). Equivalent amounts of GST or GST fusion protein were used for in vitro binding assays as previously described (Webb et al., 2000). The sequence of the NCoR CBM 3 competition peptide corresponds to residues 1938–1961.

Yeast one-hybrid assays.

Mutant and wt E1As were expressed as LexA DBD fusions in L40 strain yeast (Clontech) and grown overnight in 2 % galactose and minimal media containing all essential amino acids except leucine. Intrinsic transcriptional activation abilities were monitored using the lacZ reporter gene and assayed as Miller units (mg protein)−1 (Meng et al., 2003).

Mammalian two-hybrid and one-hybrid assays.

Human U2OS osteosarcoma cells were cultured at 37 °C with 5 % CO2 in Dulbecco’s modified Eagle’s medium (DMEM; Invitrogen) supplemented with 10 % FBS (Sigma), penicillin (50 U ml−1) and 100 µg ml−1 streptomycin (Invitrogen). Cells were transfected using Lipofectamine 2000 (Invitrogen). Following transfections, all samples were grown in DMEM containing no antibiotics and 5 % charcoal/dextran-treated FBS (Hyclone). For the one-hybrid studies, 800 ng Gal4 DBD fusions of bait plasmid were prepared with 100 ng luciferase reporter plasmid and 100 ng of a plasmid constitutively expressing β-galactosidase. For the two-hybrid studies, cells were co-transfected with 300 ng Gal4 DBD bait plasmid, 500 ng VP16 prey plasmid, 100 ng luciferase reporter plasmid and 100 ng of the β-galactosidase expressing plasmid. Forty-eight hours after transfection, cells were harvested for luciferase assay (Promega). Photon production was detected as relative light units (RLU) and values were normalized to β-galactosidase activity. The A420 β-galactosidase reading was also normalized with respect to total protein present in extract, which was determined using the Bradford DC Protein Assay kit (Bio-Rad).

Immunoprecipitation and Western blotting.

U2OS cells were transfected with plasmids expressing either wt FLAG tagged hTRβ1 or the indicated mutants, and HAdV-2 E1A 55R with EGFP fused to the C terminus of E1A, using X-tremeGENE transfection reagent (Roche Applied Science). Transfected cells were grown in medium containing 5 % stripped FBS for 3 days. Cells were treated or not treated with T3, and lysates were prepared 2 h later. TR and associated proteins were precipitated with anti-FLAG antibody and Western blots were performed using an anti-EGFP antibody to detect E1A. Levels of E1A 55R and TR expression were determined by Western blot of the lysate with anti-EGFP and anti-FLAG antibodies, respectively.

Acknowledgements

We thank C. Boone for yeast strains and R. Y. Wang and J. Curran for technical assistance. This work was supported by a CIHR grant MOP-111173 (to J. S. M.), the Mount Sinai Hospital Foundation of Toronto Annual Da Vinci Gala Fundraiser and the Mount Sinai Hospital Department of Medicine Research Fund (to P. G. W.) and NIH grant 41428 (to P. W.).

References

  1. Adams A., Gottschling D. E., Kaiser C. A., Stearns T. (1998). Methods in Yeast Genetics. Plainview, N.Y.: Cold Spring Harbor Laboratory Press [Google Scholar]
  2. Ahuja D., Sáenz-Robles M. T., Pipas J. M. (2005). SV40 large T antigen targets multiple cellular pathways to elicit cellular transformation. Oncogene 24, 7729–7745 10.1038/sj.onc.1209046 [DOI] [PubMed] [Google Scholar]
  3. Anafi M., Yang Y. F., Barlev N. A., Govindan M. V., Berger S. L., Butt T. R., Walfish P. G. (2000). GCN5 and ADA adaptor proteins regulate triiodothyronine/GRIP1 and SRC-1 coactivator-dependent gene activation by the human thyroid hormone receptor. Mol Endocrinol 14, 718–732 10.1210/me.14.5.718 [DOI] [PubMed] [Google Scholar]
  4. Avvakumov N., Kajon A. E., Hoeben R. C., Mymryk J. S. (2004). Comprehensive sequence analysis of the E1A proteins of human and simian adenoviruses. Virology 329, 477–492 10.1016/j.virol.2004.08.007 [DOI] [PubMed] [Google Scholar]
  5. Berk A. J. (2005). Recent lessons in gene expression, cell cycle control, and cell biology from adenovirus. Oncogene 24, 7673–7685 10.1038/sj.onc.1209040 [DOI] [PubMed] [Google Scholar]
  6. Berk A. J., Sharp P. A. (1978). Structure of the adenovirus 2 early mRNAs. Cell 14, 695–711 10.1016/0092-8674(78)90252-0 [DOI] [PubMed] [Google Scholar]
  7. Chen J. D., Evans R. M. (1995). A transcriptional co-repressor that interacts with nuclear hormone receptors. Nature 377, 454–457 10.1038/377454a0 [DOI] [PubMed] [Google Scholar]
  8. Cheng S. Y. (2000). Multiple mechanisms for regulation of the transcriptional activity of thyroid hormone receptors. Rev Endocr Metab Disord 1, 9–18 10.1023/A:1010052101214 [DOI] [PubMed] [Google Scholar]
  9. Desai-Yajnik V., Hadzic E., Modlinger P., Malhotra S., Gechlik G., Samuels H. H. (1995). Interactions of thyroid hormone receptor with the human immunodeficiency virus type 1 (HIV-1) long terminal repeat and the HIV-1 Tat transactivator. J Virol 69, 5103–5112 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Desvergne B., Favez T. (1997). The major transcription initiation site of the SV40 late promoter is a potent thyroid hormone response element. Nucleic Acids Res 25, 1774–1781 10.1093/nar/25.9.1774 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Frisch S. M., Mymryk J. S. (2002). Adenovirus-5 E1A: paradox and paradigm. Nat Rev Mol Cell Biol 3, 441–452 10.1038/nrm827 [DOI] [PubMed] [Google Scholar]
  12. Glass C. K., Rosenfeld M. G. (2000). The coregulator exchange in transcriptional functions of nuclear receptors. Genes Dev 14, 121–141 [PubMed] [Google Scholar]
  13. Hörlein A. J., Näär A. M., Heinzel T., Torchia J., Gloss B., Kurokawa R., Ryan A., Kamei Y., Söderström M. & other authors (1995). Ligand-independent repression by the thyroid hormone receptor mediated by a nuclear receptor co-repressor. Nature 377, 397–404 10.1038/377397a0 [DOI] [PubMed] [Google Scholar]
  14. Hu X., Lazar M. A. (1999). The CoRNR motif controls the recruitment of corepressors by nuclear hormone receptors. Nature 402, 93–96 10.1038/47069 [DOI] [PubMed] [Google Scholar]
  15. Hu X., Li Y., Lazar M. A. (2001). Determinants of CoRNR-dependent repression complex assembly on nuclear hormone receptors. Mol Cell Biol 21, 1747–1758 10.1128/MCB.21.5.1747-1758.2001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Mangelsdorf D. J., Thummel C., Beato M., Herrlich P., Schütz G., Umesono K., Blumberg B., Kastner P., Mark M. & other authors (1995). The nuclear receptor superfamily: the second decade. Cell 83, 835–839 10.1016/0092-8674(95)90199-X [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Marimuthu A., Feng W., Tagami T., Nguyen H., Jameson J. L., Fletterick R. J., Baxter J. D., West B. L. (2002). TR surfaces and conformations required to bind nuclear receptor corepressor. Mol Endocrinol 16, 271–286 10.1210/me.16.2.271 [DOI] [PubMed] [Google Scholar]
  18. Martens J. A., Winston F. (2003). Recent advances in understanding chromatin remodeling by Swi/Snf complexes. Curr Opin Genet Dev 13, 136–142 10.1016/S0959-437X(03)00022-4 [DOI] [PubMed] [Google Scholar]
  19. McEwan I. J. (2009). Nuclear receptors: one big family. Methods Mol Biol 505, 3–18 10.1007/978-1-60327-575-0_1 [DOI] [PubMed] [Google Scholar]
  20. Meng X., Yang Y. F., Cao X., Govindan M. V., Shuen M., Hollenberg A. N., Mymryk J. S., Walfish P. G. (2003). Cellular context of coregulator and adaptor proteins regulates human adenovirus 5 early region 1A-dependent gene activation by the thyroid hormone receptor. Mol Endocrinol 17, 1095–1105 10.1210/me.2002-0294 [DOI] [PubMed] [Google Scholar]
  21. Meng X., Webb P., Yang Y. F., Shuen M., Yousef A. F., Baxter J. D., Mymryk J. S., Walfish P. G. (2005). E1A and a nuclear receptor corepressor splice variant (N-CoRI) are thyroid hormone receptor coactivators that bind in the corepressor mode. Proc Natl Acad Sci U S A 102, 6267–6272 10.1073/pnas.0501491102 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Miller M. S., Mymryk J. S. (2011). An unhealthy relationship: viral manipulation of the nuclear receptor superfamily. Future Microbiol 6, 999–1019 10.2217/fmb.11.80 [DOI] [PubMed] [Google Scholar]
  23. Miller M. S., Pelka P., Fonseca G. J., Cohen M. J., Kelly J. N., Barr S. D., Grand R. J., Turnell A. S., Whyte P., Mymryk J. S. (2012). Characterization of the 55-residue protein encoded by the 9S E1A mRNA of species C adenovirus. J Virol 86, 4222–4233 10.1128/JVI.06399-11 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Nagy L., Kao H. Y., Love J. D., Li C., Banayo E., Gooch J. T., Krishna V., Chatterjee K., Evans R. M., Schwabe J. W. (1999). Mechanism of corepressor binding and release from nuclear hormone receptors. Genes Dev 13, 3209–3216 10.1101/gad.13.24.3209 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Nevins J. R., Ginsberg H. S., Blanchard J. M., Wilson M. C., Darnell J. E., Jr (1979). Regulation of the primary expression of the early adenovirus transcription units. J Virol 32, 727–733 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Park H. Y., Davidson D., Raaka B. M., Samuels H. H. (1993). The herpes simplex virus thymidine kinase gene promoter contains a novel thyroid hormone response element. Mol Endocrinol 7, 319–330 10.1210/me.7.3.319 [DOI] [PubMed] [Google Scholar]
  27. Pelka P., Ablack J. N., Fonseca G. J., Yousef A. F., Mymryk J. S. (2008). Intrinsic structural disorder in adenovirus E1A: a viral molecular hub linking multiple diverse processes. J Virol 82, 7252–7263 10.1128/JVI.00104-08 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Perissi V., Staszewski L. M., McInerney E. M., Kurokawa R., Krones A., Rose D. W., Lambert M. H., Milburn M. V., Glass C. K., Rosenfeld M. G. (1999). Molecular determinants of nuclear receptor–corepressor interaction. Genes Dev 13, 3198–3208 10.1101/gad.13.24.3198 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Perricaudet M., Akusjärvi G., Virtanen A., Pettersson U. (1979). Structure of two spliced mRNAs from the transforming region of human subgroup C adenoviruses. Nature 281, 694–696 10.1038/281694a0 [DOI] [PubMed] [Google Scholar]
  30. Sande S., Privalsky M. L. (1996). Identification of TRACs (T3 receptor-associating cofactors), a family of cofactors that associate with, and modulate the activity of, nuclear hormone receptors. Mol Endocrinol 10, 813–825 10.1210/me.10.7.813 [DOI] [PubMed] [Google Scholar]
  31. Sato Y., Ding A., Heimeier R. A., Yousef A. F., Mymryk J. S., Walfish P. G., Shi Y. B. (2009). The adenoviral E1A protein displaces corepressors and relieves gene repression by unliganded thyroid hormone receptors in vivo. Cell Res 19, 783–792 10.1038/cr.2009.55 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Shuen M., Avvakumov N., Walfish P. G., Brandl C. J., Mymryk J. S. (2002). The adenovirus E1A protein targets the SAGA but not the ADA transcriptional regulatory complex through multiple independent domains. J Biol Chem 277, 30844–30851 10.1074/jbc.M201877200 [DOI] [PubMed] [Google Scholar]
  33. Stephens C., Harlow E. (1987). Differential splicing yields novel adenovirus 5 E1A mRNAs that encode 30 kD and 35 kD proteins. EMBO J 6, 2027–2035 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Timmers H. T., Tora L. (2005). SAGA unveiled. Trends Biochem Sci 30, 7–10 10.1016/j.tibs.2004.11.007 [DOI] [PubMed] [Google Scholar]
  35. Walfish P. G., Yoganathan T., Yang Y. F., Hong H., Butt T. R., Stallcup M. R. (1997). Yeast hormone response element assays detect and characterize GRIP1 coactivator-dependent activation of transcription by thyroid and retinoid nuclear receptors. Proc Natl Acad Sci U S A 94, 3697–3702 10.1073/pnas.94.8.3697 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Webb P., Anderson C. M., Valentine C., Nguyen P., Marimuthu A., West B. L., Baxter J. D., Kushner P. J. (2000). The nuclear receptor corepressor (N-CoR) contains three isoleucine motifs (I/LXXII) that serve as receptor interaction domains (IDs). Mol Endocrinol 14, 1976–1985 10.1210/me.14.12.1976 [DOI] [PubMed] [Google Scholar]
  37. Xu H. E., Stanley T. B., Montana V. G., Lambert M. H., Shearer B. G., Cobb J. E., McKee D. D., Galardi C. M., Plunket K. D. & other authors (2002). Structural basis for antagonist-mediated recruitment of nuclear co-repressors by PPARalpha. Nature 415, 813–817 10.1038/415813a [DOI] [PubMed] [Google Scholar]
  38. Yen P. M. (2001). Physiological and molecular basis of thyroid hormone action. Physiol Rev 81, 1097–1142 [DOI] [PubMed] [Google Scholar]
  39. Yousef A. F., Xu G. W., Mendez M., Brandl C. J., Mymryk J. S. (2008). Coactivator requirements for p53-dependent transcription in the yeast Saccharomyces cerevisiae. Int J Cancer 122, 942–946 10.1002/ijc.23174 [DOI] [PubMed] [Google Scholar]
  40. Yousef A. F., Brandl C. J., Mymryk J. S. (2009). Requirements for E1A dependent transcription in the yeast Saccharomyces cerevisiae. BMC Mol Biol 10, 32 10.1186/1471-2199-10-32 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Yousef A. F., Fonseca G. J., Cohen M. J., Mymryk J. S. (2012). The C-terminal region of E1A: a molecular tool for cellular cartography. Biochem Cell Biol 90, 153–163 10.1139/o11-080 [DOI] [PubMed] [Google Scholar]
  42. Zuo F., Kraus R. J., Gulick T., Moore D. D., Mertz J. E. (1997). Direct modulation of simian virus 40 late gene expression by thyroid hormone and its receptor. J Virol 71, 427–436 [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from The Journal of General Virology are provided here courtesy of Microbiology Society

RESOURCES