A Negative Regulatory Element (Base Pairs −204 to −177) of the EICP0 Promoter of Equine Herpesvirus 1 Abolishes the EICP0 Protein's trans-Activation of Its Own Promoter

Abstract

The early EICP0 protein is a powerful trans-activator that activates all classes of equine herpesvirus 1 (EHV-1) promoters but, unexpectedly, trans-activates its own promoter very weakly. Transient transfection assays that employed constructs harboring deletions within the EICP0 promoter indicated that EICP0 cis-acting sequences within bp −224 to −158 relative to the first ATG abolished the EICP0 protein's trans-activation of its own promoter. When inserted into the promoters of other EHV-1 genes, this sequence also downregulated activation of the immediate-early IE(−169/+73), early thymidine kinase TK(−215/+97), and late glycoprotein K gK(−83/+14) promoters, indicating that the cis-acting sequence (−224 to −158) downregulated expression of representative promoters of all classes of EHV-1 genes and contains a negative regulatory element (NRE). To define the cis-acting element(s), three synthetic oligonucleotides (Na [bp −224 to −195], Nb [bp −204 to −177], and Nc [bp −185 to −156]) were synthesized and cloned upstream of the EICP0(−157/−21) promoter. Of the three synthetic sequences, only the Nb oligonucleotide caused the downregulation of the EICP0 promoter. The NRE was identified as a 28-bp element to lie at −204 to −177 that encompassed the sequence of ([−204]AGATACAGATGTTCGATAAATTGGAACC[−177]). Gel shift assays performed with mouse L-M, rabbit RK-13, and human HeLa cell nuclear extracts and γ-³²P-labeled wild-type and mutant NREs demonstrated that a ubiquitous nuclear protein(s) (NRE-binding protein, NREBP) binds specifically to a sequence (bp −193 to −183) in the NRE. The NREBP is also present in the nucleus of EHV-1-infected cells; however, the amount of NREBP in EHV-1-infected L-M cells that bound to the Nb oligonucleotide was reduced compared to that in uninfected L-M cells. Transient transfection assays showed that deletions or mutations within the NREBP-binding site abolished the NRE activity of the EICP0 promoter. These results suggested that the NREBP may mediate the NRE activity of the EICP0 promoter and may function in the coordinate expression of EHV-1 genes.

The EICP0 gene of the equine herpesvirus 1 (EHV-1) KyA virus encodes an early nuclear phosphoprotein of 419 amino acids that trans-activates all classes of EHV-1 promoters (3, 4). The EICP0 protein is not a DNA-binding protein (13) and does not bind to its own promoter (S. K. Kim and D. J. O'Callaghan, unpublished data). The EHV-1 EICP0 protein and its homologs in herpes simplex virus type 1 (HSV-1) (10, 15), varicella-zoster virus (44, 50), bovine herpes virus 1 (54), and pseudorabies virus 1 (9, 53) contain a conserved cysteine-rich zinc RING finger (C₃HC₄ type) near the N terminus that contributes to their regulatory functions (8, 13, 44, 47). trans-Activation assays with the EICP0 mutants d8-46 and d19-30, which lack the RING finger motif (amino acids 8 to 46), revealed that this motif is essential for activation of the E and L (γ1 and γ2) promoters (3). The EICP0 protein's serine-rich region (amino acids 210 to 217) may serve as a potential site for phosphorylation and is necessary for the maximal trans-activation functions of this protein, as indicated by the observation that deletion of this domain reduced the ability of the EICP0 protein to activate representative late promoters by more than 70% (3). The regulatory functions of the EICP0 protein are severely antagonized by the immediate-early (IE) protein (35), and thus the EHV-1 EICP0 protein differs from HSV-1 ICP0, which functions synergistically with ICP4 to activate expression of the HSV-1 E and L promoters (8, 15). An interaction of the EICP0 protein with the IE protein and with the basal transcription factors TFIIB and TATA box-binding protein (TBP) may explain the antagonism between the IE and EICP0 proteins (35).

The EHV-1 immediate-early (IE) gene encodes a 1,487-amino-acid polypeptide (22) that is essential for replication (18). The IE protein trans-activates EHV-1 and heterologous viral promoters and trans-represses its own expression (48, 49). Residues 422 to 597 of the IE protein are sufficient for its sequence-specific DNA binding to the consensus binding sequence 5′-ATCGT-3′, which overlaps the transcription initiation site of the IE promoter, and to degenerate versions of this cognate cis-acting sequence in the E and L promoters (37). The IE protein represses transcription of the true late glycoprotein K (gK) gene by binding to the transcription initiation site of the gK promoter (33). In transient transfection assays, the EICP0 protein is able to release the γ2 L gK promoter from repression mediated by the IE protein (33). Although the EICP0 protein is a promiscuous trans-activator of all classes of EHV-1 promoters, our recent studies surprisingly showed that the functions of the EICP0 protein are dispensable for EHV-1 lytic replication in vitro (55). However, the EICP0 protein may be important for efficient virus egress, cell-to-cell spread, and efficient expression of late viral genes (55).

The experimental data presented in this paper concern the intriguing findings that the EICP0 protein unexpectedly trans-activates its own promoter very weakly. We present findings that EICP0 cis-acting sequences within bp −204 to −177 relative to the first ATG abolished the EICP0P's trans-activation of its own promoter and also downregulated the IE(−169/+73), late gK(−83/+14), and early thymidine kinase TK(−215/+97) promoters. Gel shift assays demonstrated that a ubiquitous nuclear protein (NRE-binding protein, NREBP) binds specifically to a sequence (bp −193 to −183) in the NRE. These findings suggest that the NRE abolishes EICP0's trans-activation of its own promoter by a mechanism in which a cellular regulatory factor binds to the NRE and thereby blocks the normal regulatory function of the EICP0 protein at its own promoter.

MATERIALS AND METHODS

Cell culture.

The mouse fibroblast L-M, rabbit kidney RK-13, and human ovarian HeLa cells were maintained at 37°C in complete Eagle's minimal essential medium supplemented with 100 U of penicillin per ml, 100 μg of streptomycin per ml, nonessential amino acids, and 5% fetal bovine serum.

Plasmids.

Plasmids were constructed and maintained in Escherichia coli HB101 by standard methods (46). Plasmids pIE(−169/+73)-CAT, pEICP0-CAT, pCETIF, pcDR4, pSVICP0K (pSVEICP0), pgK(−83/+14)-CAT, pSVIE, pEICP22-CAT, and pTK-CAT were described previously (4, 27, 33-35, 48). To generate pEICP0(−335/−21)-CAT, the XbaI and MscI fragment of pEICP0(−761/−21)-CAT (48) was cloned into pCAT-basic (Promega, Madison, Wis.). pCAT-basic was previously digested with SalI and treated with Klenow fragment to create a blunt end, and then the DNA was cut a second time with XbaI. To generate pEICP0(−258/−21)-CAT, pEICP0(−761/−21)-CAT was digested with SmaI and HindIII, filled with Klenow fragment, and self-ligated. To generate pEICP0(−224/−21)-CAT, pEICP0(−761/−21)-CAT was digested with PpuMI and HindIII, filled with Klenow fragment, and self-ligated. To generate pEICP0(−157/−21)-CAT, the XbaI-SphI fragment of pEICP0(−761/−21)-CAT was cloned into the XbaI and SphI sites of pCAT-basic.

To generate pNa-, pNb-, and pNc-EICP0(−157/−21)-CAT, the three synthetic oligonucleotides Na, Nb, and Nc (see Fig. 3) were cloned into the HindIII and SphI sites of pEICP0(−224/−21)-CAT, respectively. The HindIII fragment of pIE(−169/+73)-CAT was cloned into the HindIII site of pSVSPORT1 (Gibco-BRL, Rockville, Md.) to generate pSPORT-IE169. The XbaI-SphI fragment of pSPORT-IE169 was cloned into the XbaI and SphI sites of pEICP0(−224/−21)-CAT to generate pNRE-IE(−169/+73)-CAT. The HindIII-XbaI fragment of pgK(−83/+14)-CAT was cloned into the HindIII and XbaI sites of pSVSPORT1 to generate pSPORT-gK83. The XbaI-SphI fragment of pSPORT-gk83 was cloned into the XbaI and SphI sites of pEICP0(−224/−21)-CAT to generate pNRE-gK(−83/+14)-CAT. The XbaI-SphI fragment of pTKCAT (48) was cloned into the XbaI and SphI sites of pEICP0(−224/−21)-CAT to generate pNRE-TK(−215/+97)-CAT.

FIG. 3.

Negative regulatory element (NRE) of the EICP0 promoter. The amino acid sequence of EICP0 is shown in the single-letter code beneath the DNA sequence. Potential cis-acting elements (Sp1, octamer factor 1, and TATA box) are underlined. The regions of three synthetic oligonucleotides (Na, Nb, and Nc) are depicted. IEBS, putative binding site for the EHV-1 IE protein. The NRE is in boldface.

To generate pEICP0(−335N1)-CAT, the 136-bp DNA fragment of the EICP0 gene promoter from pEICP0(−761/−21)-CAT was amplified by PCR with primers EP-bm14 (5′-ACGCATGCCAAATTTGGCATGGGGACCCGTTTGAATGCGATTGGTG-3′) and EP-1 (5′-GGAAGCTTAACTATGCAACCCCAAAAAGCA-3′), which contain SphI and HindIII sites, respectively, and the fragment was cloned into the HindIII and SphI sites of pEICP0(−761/−21)-CAT. To generate pEICP0(−335N2)-CAT, the 146-bp DNA fragment of the EICP0 gene promoter from pEICP0(−761/−21)-CAT was amplified by PCR with primers EP-bm12 (5′-ACGCATGCCAAATTTGCAGGGGTATAGCATGGGGACCCGTTTGAAT-3′) and EP-1, which contain SphI and HindIII sites, respectively, and the fragment was cloned into the HindIII and SphI sites of pEICP0(−761/−21)-CAT.

To generate pEICP0(−335N3)-CAT, the 146-bp DNA fragment of the EICP0 gene promoter from pEICP0(−761/−21)-CAT was amplified by PCR with primers EP-bm13 (5′-ACGCATGCCAAATTTGCACGGGATACGCATGGGGACCCGTTTGAATGCGATT GGTG-3′) and EP-1, which contain SphI and HindIII sites, respectively, and the fragment was cloned into the HindIII and SphI sites of pEICP0(−761/−21)-CAT. To generate pEICP0(−335N4)-CAT, the 156-bp DNA fragment of the EICP0 gene promoter from pEICP0(−761/−21)-CAT was amplified by PCR with primers EP-bm9 (5′-ACGCATGCCAAATTTGCACGGGATACCAGGGGTATAGCATGGGGACCCGTTTGAAT-3′) and EP-1, which contain SphI and HindIII sites, respectively, and the fragment was cloned into the HindIII and SphI sites of pEICP0(−761/−21)-CAT.

To generate pEICP0(−335N5)-CAT, the 169-bp DNA fragment of the EICP0 gene promoter from pEICP0(−761/−21)-CAT was amplified by PCR with primers EP-bm10 (5′-ACGCATGCCAAATTTGCACGGGATACAACATCTGTATCTCAGGGGTATAGCATGGG-3′) and EP-1, which contain SphI and HindIII sites, respectively, and the fragment was cloned into the HindIII and SphI sites of pEICP0(−761/−21)-CAT. To generate pEICP0(−335N6)-CAT, the 175-bp DNA fragment of the EICP0 gene promoter from pEICP0(−761/−21)-CAT was amplified by PCR with primers EP-bm11 (5′-ACGCATGCCAAATTTGCACGGGATACGGTTCCAATTTATCGAACACAGGGGTATAGCATGGGGACCCGTTTGAAT-3′) and EP-1, which contain SphI and HindIII sites, respectively, and the fragment was cloned into the HindIII and SphI sites of pEICP0(−761/−21)-CAT.

The 130-bp DNA fragment of the EICP0 gene promoter from pEICP0(−761/−21)-CAT was amplified by PCR with primers EP-1 and EP-2 (5′-GGTCTAGAGGGTATAGCATGGGGACCCGTT-3′), which contain HindIII and SphI sites, respectively, and the fragment was cloned into the HindIII and SphI sites of pUC19 and designated pU-PCR1. The 180-bp DNA fragment of the EICP0 gene promoter from pEICP0(−761/−21)-CAT was amplified by PCR with primers EP-bm0 (5′-GGTCTAGATACAGATGTTCGATAAATTGGA-3′) and EP-3 (5′-GGGAATTCCAAATGAAAAGGCTGTATCAGC-3′), which contain XbaI and EcoRI sites, respectively, and the fragment was cloned into the XbaI and EcoRI sites of pU-PCR1 and designated pU-NREM0. The HindIII-SphI DNA fragment of pU-NREM0 was cloned into the HindIII and SphI sites of pEICP0(−761/−21)-CAT to generate pEICP0(−335M0)-CAT. The 180-bp DNA fragment of the EICP0 gene promoter from pEICP0(−761/−21)-CAT was amplified by PCR with primers EP-bm1 (5′-GGTCTAGAGGCAGATGTTCGATAAATTGGAACCGT-3′) and EP-3, which contain XbaI and EcoRI sites, respectively, and the fragment was cloned into the XbaI and EcoRI sites of pU-PCR1 and designated pU-NREM1.

The HindIII-SphI DNA fragment of pU-NREM1 was cloned into the HindIII and SphI sites of pEICP0(−761/−21)-CAT to generate pEICP0(−335M1)-CAT. The 180-bp DNA fragment of the EICP0 gene promoter from pEICP0(−761/−21)-CAT was amplified by PCR with primers EP-bm2 (5′-GGTCTAGATATCGCTGTTCGATAAATTGGAACCGTATCC-3′) and EP-3, which contain XbaI and EcoRI sites, respectively, and the fragment was cloned into the XbaI and EcoRI sites of pU-PCR1 and designated pU-NREM2. The HindIII-SphI DNA fragment of pU-NREM2 was cloned into the HindIII and SphI sites of pEICP0(−761/−21)-CAT to generate pEICP0(−335M2)-CAT. The 180-bp DNA fragment of the EICP0 gene promoter from pEICP0(−761/−21)-CAT was amplified by PCR with primers EP-bm3 (5′-GGTCTAGATACATACGTTCGATAAATTGGAACCGTATCCC-3′) and EP-3, which contain XbaI and EcoRI sites, respectively, and the fragment was cloned into the XbaI and EcoRI sites of pU-PCR1 and designated pU-NREM3.

The HindIII-SphI DNA fragment of pU-NREM3 was cloned into the HindIII and SphI sites of pEICP0(−761/−21)-CAT to generate pEICP0(−335M3)-CAT. The 180-bp DNA fragment of the EICP0 gene promoter from pEICP0(−761/−21)-CAT was amplified by PCR with primers EP-bm4 (5′-GGTCTAGATACAGATGCTACATAAATTGGAACCGTATCCCGTGCA-3′) and EP-3, which contain XbaI and EcoRI sites, respectively, and the fragment was cloned into the XbaI and EcoRI sites of pU-PCR1 and designated pU-NREM4. The HindIII-SphI DNA fragment of pU-NREM4 was cloned into the HindIII and SphI sites of pEICP0(−761/−21)-CAT to generate pEICP0(−335M4)-CAT. The 180-bp DNA fragment of the EICP0 gene promoter from pEICP0(−761/−21)-CAT was amplified by PCR with primers EP-bm5 (5′-GGTCTAGATACAGATGTTCGACAGCTTGGAACCGTATCCCGTGCAAATTT-3′) and EP-3, which contain XbaI and EcoRI sites, respectively, and the fragment was cloned into the XbaI and EcoRI sites of pU-PCR1; the resulting construct was designated pU-NREM5.

The HindIII-SphI DNA fragment of pU-NREM5 was cloned into the HindIII and SphI sites of pEICP0(−761/−21)-CAT to generate pEICP0(−335M5)-CAT. To generate pEICP0(−335M6)-CAT, the 184-bp DNA fragment of the EICP0 gene promoter from pEICP0(−761/−21)-CAT was amplified by PCR with primers EP-1 and EP-bm6 (5′-ACGCATGCCAAATTTGCACGGGATACGGTTCAGCTTTATCGAACATCTGTATCTCAGGG-3′), which contain HindIII and SphI sites, respectively, and the fragment was cloned into the HindIII and SphI sites of pEICP0(−761/−21)-CAT. To generate pEICP0(−335M7)-CAT, the 184-bp DNA fragment of the EICP0 gene promoter from pEICP0(−761/−21)-CAT was amplified by PCR with primers EP-1 and EP-bm6 (5′-ACGCATGCCAAATTTGCACGGGATACGGCGTCAATTTATCGAACATCTGTATCTCA-3′), which contain HindIII and SphI sites, respectively, and the fragment was cloned into the HindIII and SphI sites of pEICP0(−761/−21)-CAT.

Gel shift assays.

The synthetic oligonucleotides IE1 (38), Na, Nb (NRE), Nc, NREm1, NREm2, NREm3, NREm4, NREm5, NREm6, and NREm7 (see Fig. 5 and 7) were end labeled with [γ-³²P]ATP as previously detailed (37). DNA binding assays were conducted as previously described (37). The standard DNA binding reaction mixture contained 0.5 to 1 ng of radiolabeled DNA fragments (2 × 10⁴ cpm/ng), 0.1 μg of poly(dI-dC) as a nonspecific competitor, and 1 to 10 μl of nuclear or cytoplasmic extract (36) in 20 μl of DNA binding buffer (20 mM HEPES-KOH [pH 7.9], 0.5 mM dithiothreitol, 10% glycerol, 0.1 mM EDTA, 0.025% NP-40, 25 mM KCl, 2 mM MgCl₂). After 20 min at 22°C, 5 μl of loading buffer (200 mM HEPES-KOH [pH 7.5], 50% [vol/vol] glycerol, 0.02% bromophenol blue) was added, and the sample was subjected to electrophoresis at 22°C for 1.5 h at 200 V on a 3% polyacrylamide gel with 0.5× Tris-borate-EDTA running buffer.

FIG. 5.

NRE (bp −204 to −177) of the EICP0 promoter downregulates EHV-1 promoters. (A) Schematic diagram of pNRE-IE-CAT, pNRE-TK-CAT, and pNRE-gK-CAT. The cis-acting sequences (bp −224 to −157) containing the NRE were cloned upstream of the EHV-1 IE(−169/+73), early TK(−215/+97), and late gK(−83/+14) promoters. (B) Transient transfection assays with the EICP0 reporter plasmids. L-M cells were transfected with 0.5 pmol of the reporter plasmid. Each transfection was performed in triplicate. Data are averages and are representative of several independent experiments. Error bars show standard deviations.

FIG. 7.

NREBP binds to the nucleotide −193 to −183. (A) Mutant NREs. Mutated nucleotides are underlined. NREBP-binding sequences are in boldface. (B) Gel shift assays with mutant NREs. NE, nuclear extract; −, absent. (C) Relative NREBP binding activity. +++, strong; ++, moderate; +, weak; −, no binding. (D) Competition assays. Radiolabeled wild-type NRE competed with unlabeled excess competitors NREm1 to NREm7.

DNA transfection and chloramphenicol acetyltransferase (CAT) assays.

L-M cells seeded at 3 × 10⁶ cells per 60-mm tissue culture dish in Eagle's minimal essential medium with 5% fetal bovine serum were transfected by the liposome-mediated DNA transfection method at 24 h as described elsewhere (6). The reporter and effector plasmids were transfected in the amounts indicated in the figure legends. The total amount of DNA per transfection was adjusted to 8 μg by the addition of pSVSPORT1. After a further 5 h, the cells were washed and refed with fresh medium. At 45 h, total cell extracts were prepared, and CAT activities were assayed as described previously (48).

RESULTS

EICP0 protein very weakly trans-activates its own promoter.

Previous studies in our laboratory showed that the EICP0 protein is a powerful trans-activator that activates all classes of EHV-1 promoters (3, 4, 34). To determine whether the EICP0 protein trans-activates the EICP0 promoter, the pEICP0-CAT reporter plasmid was used for transient transfection assays. The pEICP0-CAT reporter construct contains the EICP0 promoter region located at bp −761 to −21 relative to the first ATG of the EICP0 open reading frame (6). The IE protein strongly trans-activated the pEICP0-CAT reporter plasmid (Fig. 1E, bar 3), a finding that was consistent with our previous results (6). The EICP27 protein also trans-activated the pEICP0-CAT reporter (Fig. 1E, bar 6). As expected, the ETIF protein was not able to trans-activate expression of the early EICP0 promoter (Fig. 1E, bar 4), which was consistent with our previous results (34). Unexpectedly, the EICP0 protein very weakly trans-activated the pEICP0-CAT reporter (Fig. 1E, bar 2), indicating that the EICP0 protein very weakly trans-activates its own promoter.

FIG. 1.

EICP0 protein very weakly trans-activates its own promoter. Transient transfection assays were performed with the EICP0 promoter (−335 to −21)-CAT reporter plasmid. Transient transfection assays were performed as described in Materials and Methods. L-M cells were transfected with 0.5 pmol of the reporter plasmid and 0.2 pmol of effector plasmids (pSVIE, pSVEICP0, pcDR4 [EICP22 expression vector], pSVICP27, and pCETIF). Each transfection was performed in triplicate. Data are averages and are representative of several independent experiments. Error bars show standard deviations.

cis-Acting sequences (mucleotides −224 to −158) of the EICP0 promoter may abolish the EICP0 protein's trans-activation on its own promoter.

To investigate the inability of the EICP0 protein to activate its own promoter, four upstream deletion constructs of the EICP0 promoter were generated (Fig. 2A) and used for CAT assays. These experiments revealed that the EICP0 protein trans-activated the deletion constructs pEICP0(−335/−21)-CAT, pEICP0(−258/−21)-CAT, and pEICP0(−224/−21)-CAT as weakly as the full-length pEICP0(−761/−21)-CAT was trans-activated by the EICP0 protein (Fig. 2B, bars 2, 5, 8, 11, and 14). Interestingly, the EICP0 protein strongly trans-activated the pEICP0(−157/−21)-CAT construct, which contains only the core promoter elements (i.e., TATA box and cap site) of the EICP0 promoter (Fig. 2B, bar 14). However, the IE protein strongly trans-activated all of the pEICP0-CAT deletion constructs (Fig. 2B, bars 3, 6, 9, 12, and 15). These results also showed that the inability of the EICP0 protein to trans-activate the EICP0-CAT constructs was not due to an inherent defect of the CAT reporters. These results indicated that the cis-acting sequences (nucleotides −224 to −158 relative to the first ATG) may harbor a negative regulatory element (NRE) that abolishes EICP0's trans-activation of its own promoter.

FIG. 2.

cis-Acting sequences (bp −224 to −157) of the EICP0 promoter may abolish EICP0's trans-activation on its own promoter. (A) Schematic diagram of upstream deletion constructs of the EICP0 promoter. The top diagram represents the EHV-1 EICP0 promoter. TATA, TATA box. (B) Transient transfection assays with the EICP0 reporter plasmids. L-M cells were transfected with 0.5 pmol of the reporter plasmid and 0.2 pmol of effector plasmids (pSVIE and pSVEICP0). Each transfection was performed in triplicate. Data are averages and are representative of several independent experiments. Error bars show standard deviations.

EICP0 promoter sequence bp −204 to −177 contains the NRE.

To define the NRE, three synthetic oligonucleotides (Na, Nb, and Nc) (Fig. 3), which cover the cis-acting sequence (nucleotides −224 to −158) of the EICP0 promoter, were synthesized, cloned upstream of the EICP0(−157/−21) promoter (Fig. 4A), and employed in CAT assays. When the pNb-EICP0(−157/−21)-CAT reporter was examined, a two- to threefold decrease in CAT activity was observed (Fig. 4B, bar 5). The EICP0 protein also weakly trans-activated the pNb-EICP0(−157/−21)-CAT reporter (Fig. 4B, bar 6). However, when the pNa-EICP0(−157/−21)-CAT and pNc-EICP0(−157/−21)-CAT reporter plasmids were transfected in the absence of the EICP0 expression construct, a slight decrease in CAT activity was observed (Fig. 4B, bars 3 and 7, respectively). Contrary to the results of the activation of the pNb-EICP0-CAT by the EICP0 protein, the EICP0 protein strongly trans-activated these two reporters (Fig. 4B, bars 4 and 8). Deletions of six nucleotide sequences that were either 5′ ([−204]AGATAC[199]) or 3′ ([−182]GGAACC[−177]) of the NRE (Nb) sequence reduced the repression of the EICP0 promoter activity by 75 and 25%, respectively (data not shown). These results demonstrated that the Nb oligonucleotide ([−204]AGATACAGATGTTCGATAAATTGGAACC[−177]) contains the negative regulatory element.

FIG. 4.

cis-Acting sequences (Nb, bp −204 to −177) of the EICP0 promoter may contain the negative regulatory element. (A) Schematic diagram of pNa-, pNb-, and pNc-EICP0-CAT. Three synthetic oligonucleotides (Na [bp −224 to −195], Nb [bp −204 to −177], and Nc [bp −185 to −156]) were synthesized and cloned upstream of the EICP0 (bp −157 to −21) promoter. EICP0 Pro, EICP0 promoter region, bp −157 to −21. (B) Transient transfection assays with the EICP0 reporter plasmids. L-M cells were transfected with 0.5 pmol of the reporter plasmid and 0.3 pmol of effector plasmid (pSVEICP0). Each transfection was performed in triplicate. Data are averages and are representative of several independent experiments. Error bars show standard deviations.

NRE within bp −224 to −158 of the EICP0 promoter downregulates other EHV-1 promoters.

To investigate whether the cis-acting sequences (nucleotides −224 to −158) of the EICP0 promoter downregulated basal activity of other EHV-1 promoters by other regulation proteins, the cis-acting sequence (bp −224 to −158) was cloned upstream of the IE(−169/+73), early TK(−215/+97), and late gK(−83/+14) promoters (Fig. 5A). In CAT assays, the cis-acting sequence containing the identified NRE downregulated the basal activities of the IE(−169/+73), TK(−215/+97), and gK(−83/+14) promoters compared to the activities of these promoters in the absence of the NRE (Fig. 5B, bars 2, 4, and 6, respectively). These observations confirm that the 67-bp DNA fragment (bp −224 to −157) of the EICP0 promoter contains the negative regulatory element. When the 28-bp Nb oligonucleotide (bp −204 to −177) was cloned upstream of the IE(−169/+73), TK(−215/+97), and gK(−83/+14) promoters, the Nb sequence also weakly downregulated these EHV-1 promoters (data not shown). When the Nb oligonucleotide was cloned upstream of the simian virus 40 promoter region (no enhancer), Nb caused its downregulation as well (data not shown). These results suggested that the 28-bp Nb (bp −204 to −177) of the EICP0 promoter contains the NRE and that additional elements within the sequence from −224 to −158 contribute to the NRE activity.

A ubiquitous nuclear protein (NREBP) binds specifically to the NRE.

From these results, we can speculate that a cellular protein(s) may mediate NRE activity by binding to the NRE (Nb). To investigate whether cellular or viral proteins interact with the Nb oligonucleotide, gel shift assays were performed with L-M cell nuclear extracts. A nuclear protein(s) bound to the Na and Nb sequences, as revealed by the presence of shifted bands (indicated with an arrow) (Fig. 6A, lanes 2 and 6, respectively); however, no proteins bound to the Nc sequence (Fig. 6A, lane 10). There were two nonspecific bands in the reactions with the Nc oligonucleotide, which were also detected in the samples containing the Na and Nb probes (indicated as a) (Fig. 6A). The formation of the DNA-protein complexes present at the Na or Nb sequence was completely blocked by the addition of unlabeled excess specific competitor Na or Nb (Fig. 6A, lanes 3 and 7, respectively), but was not eliminated by the nonspecific competitors IE1 (Fig. 6A, lanes 4 and 8), Na (Fig. 6B, lane 3), or Nc (Fig. 6B, lane 4). These results showed that this cellular factor(s) binds specifically to the Nb sequence.

FIG. 6.

A ubiquitous nuclear protein (NREBP) binds specifically to Nb. (A) Gel shift assays with the Na, Nb, and Nc oligonucleotides. One microgram of nuclear extracts from L-M cells was incubated with radiolabeled probes. (B) Additional competition assays. (C) Gel shift assays with L-M cell nuclear (NE) and cytoplasmic (CE) extracts. (D) Gel shift assays with mouse L-M, rabbit RK-13, and human HeLa cell nuclear extracts. (E) NREBP in nuclear extracts of EHV-1-infected cells binds weakly to the Nb oligonucleotide. NE, nuclear extract; CE, cytoplasmic extract; +, present; −, absent; comp, competitor; U, uninfected L-M cell nuclear extract; 5 h, nuclear extracts of EHV-1-infected L-M cells at 5 h postinfection; 8 h, nuclear extracts of EHV-1-infected L-M cells at 8 h postinfection. The Nb and NREBP complexes are indicated by an arrow on the left. The a on the left indicates nonspecific bands.

To confirm that a nuclear cellular protein interacted with the Nb sequence, an additional gel shift assay showed that nuclear proteins bound to the Nb sequence (Fig. 6C, lane 1), whereas cytoplasmic proteins failed in this regard. These results indicated that a nuclear protein (NRE-binding protein, NREBP) binds specifically to the Nb region. The formation of the Na-protein complexes was not eliminated by the nonspecific competitor Nb or Nc (data not shown), indicating that a nuclear protein also binds specifically to the Na region. To investigate whether the NREBP is a ubiquitous protein of all cell types or is present only in L-M cells, gel shift assays were performed with nuclear extracts prepared from rabbit RK-13 cells and human HeLa cells. The results of these gel shift assays were similar to those observed with nuclear extracts prepared from mouse L-M cells in that a cellular protein(s) within the RK and HeLa nuclear extracts interacted with the Nb sequence (Fig. 6D, lanes 4 to 9), indicating that the NREBP is a cellular protein(s) that is present in diverse mammalian cell lines.

To investigate whether the NREBP is present in the nucleus of EHV-1-infected cells, gel shift assays were performed with EHV-1-infected L-M cell nuclear extracts. Nuclear extracts of EHV-1-infected L-M cells were prepared at 5 h (early in infection) and 8 h (late in infection) postinfection. Interestingly, the amount of NREBP in the EHV-1-infected L-M cells bound to the Nb sequence was reduced (Fig. 6E, lanes 3 and 5) compared to that of the uninfected cells (Fig. 6E, lane 1). At the late stage of infection, less NREBP bound to the Nb sequence (Fig. 6E, lane 5) compared to the early stage of infection (Fig. 6E, lane 3). These results indicated that NREBP is still present in the nucleus of EHV-1-infected cells, but its binding to Nb is weaker than its binding to Nb in uninfected cells.

NREBP binds to the NRE sequence ([−193]TTCGATAAATT[−183]).

To determine the nucleotide sequence that the NREBP recognizes within the NRE, seven mutant NRE oligonucleotides (Fig. 7A; mutated sequences are underlined) were employed in gel shift assays. In these experiments, the NREBP bound to the NRE m1, m2, m3, and m7 oligonucleotides (Fig. 7B, lanes 4, 8, 12, and 28, respectively), indicating that the mutations within these sequences did not disrupt the binding site of the NREBP. The formation of these DNA-NREBP complexes was completely blocked by the addition of unlabeled excess specific competitors NRE m1, m2, m3, and m7, respectively (Fig. 7B, lanes 5, 9, 13, and 29, respectively) and with wild-type NRE (Fig. 7B, lanes 7, 11, 15, and 31), but was not prevented by inclusion of the nonspecific competitor Na (Fig. 7B, lanes 6, 10, 14, and 30). However, the NREBP failed to bind to the NREm4 and NREm5 oligonucleotides (Fig. 7B, lanes 16 to 23), and it bound weakly to the NREm6 oligonucleotide (Fig. 7B, lanes 24 to 27). In competition assays, the formation of the DNA-NREBP complexes was completely blocked by the addition of unlabeled excess competitors NREm1, NREm2, NREm3, and NREm7 and was partially blocked by NREm6 (Fig. 7D, lanes 5 to 7, 11, and 10, respectively), but was very weakly reduced by NREm4 and NREm5 (Fig. 7D, lanes 8 and 9, respectively). Collectively, these results indicated that the NREBP binds specifically to the sequence ([−193]TTCGATAAATT[−183]) of the EICP0 promoter's NRE.

NREBP may mediate NRE activity.

To determine which nucleotide sequences in the NRE are important for NRE activity, six internal deletion mutants of the NRE region were generated in the context of the pEICP0(−335/−21)-CAT reporter plasmid (Fig. 8A) and used in CAT reporter assays. Four of the deletion mutants, pEICP0(−335N1)-CAT, pEICP0(−335N2)-CAT, pEICP0(−335N3)-CAT, and pEICP0(−335N4)-CAT, which harbor deletions of the entire 28-bp NRE, lost all (>100%) of the NRE activity (Fig. 8B, bars 7, 10, 13, and 16, respectively). The pEICP0(−335N5)-CAT reporter, which harbors the minimal deletion that spans the NREBP-binding sequence, lost most (>80%) of the NRE activity (Fig. 8B, bar 19). However, the pEICP0(−335N6)-CAT reporter, in which a sequence (−204 to −196) upstream of the NREBP-binding sequence was deleted, retained most (80%) of the NRE activity (Fig. 8B, bar 22).

FIG. 8.

Nucleotide sequences important for NRE activity. (A) Construction of the pEICP0(−335/−21)-CAT reporter containing the deleted NRE. The top diagram represents the EHV-1 EICP0 promoter. Mutated nucleotides are underlined. NREBP-binding sequences are underlined. TATA, TATA box. (B) Transient transfection assays. L-M cells were transfected with 0.5 pmol of the reporter plasmid and 0.2 pmol of effector plasmids (pSVIE and pSVEICP0). Each transfection was performed in triplicate. Data are averages and are representative of several independent experiments. Error bars show standard deviations. −157, pEICP0(−157/−21)-CAT; −335, pEICP0(−335/−21)-CAT; −335N1, pEICP0(−335N1)-CAT; −335N6, pEICP0(−335N6)-CAT. (C) Summary of relative CAT activities of the reporter plasmids.

Consistent with the observation that −335N6 retained most of the NRE activity, the EICP0 protein weakly trans-activated the pEICP0(−335N6)-CAT reporter (Fig. 8B, bar 23). However, the IE protein strongly trans-activated all of the pEICP0-CAT deletion constructs (Fig. 8B, bars 3, 6, 9, 12, 15, 18, and 24), indicating that the low basal activity of these reporter constructs was not attributable to inherent defects in the activity of the promoters. The data presented in Fig. 8C show a summary of the relative CAT activities obtained for each reporter by itself. These results demonstrated that the nucleotide sequence (bp −213 to −166) that harbors the NREBP-binding site is important for NRE activity.

To investigate further which nucleotide sequences in the NRE are important for NRE activity, mutations in the 28-bp NRE of the EICP0 promoter were introduced by PCR (Fig. 9A; mutated sequences are underlined). These additional EICP0 promoter mutants were then examined in CAT reporter assays. The EICP0(−335M4)-CAT and EICP0(−335M5)-CAT reporter plasmids lost 37 to 50% of the NRE activity (Fig. 9B, bars 19 and 22, respectively), indicating that the sequence ([−193]TTCGATAAA[−185]) is important for NRE activity. The sequence ([−193]TTCGATAAA[−185]) overlapped the NREBP-binding sequence ([−193]TTCGATAAATT[−183]) in the NRE. Figure 9C shows a summary of the relative CAT activities obtained with each reporter in the absence of any effector protein. These results suggested that the NREBP mediates NRE activity by binding to the sequence (bp −193 to −183) within the EICP0 promoter.

FIG. 9.

Nucleotide sequences important for NRE activity. (A) Construction of the pEICP0(−335/−21)-CAT reporter containing the mutated NRE. The top diagram represents the EHV-1 EICP0 promoter. Mutated nucleotides are underlined. NREBP-binding sequences are in boldface. TATA, TATA box. (B) Transient transfection assays. L-M cells were transfected with 0.5 pmol of the reporter plasmid and 0.2 pmol of effector plasmids (pSVIE and pSVEICP0). Each transfection was performed in triplicate. Data are averages and are representative of several independent experiments. Error bars show standard deviations. −157, pEICP0(−157/−21)-CAT; −335, pEICP0(−335/−21)-CAT; −335M0, pEICP0(−335M0)-CAT; −335M7, pEICP0(−335M7)-CAT. (C) Summary of relative CAT activities of the reporter plasmids.

Summary of the matched sequences in the NRE of the EICP0 promoter.

The 28-bp NRE sequence was analyzed with a BLAST program (www.ncbi.nlm.nih.gov/BLAST) to elucidate potential homology and conservation of this sequence within other viral, cellular, and bacterial genes. One complete match was identified only in the EICP0 promoter of the EHV-1 genome; however, no sequences from other herpesvirus genomes matched the NRE sequence contained within the EHV-1 EICP0 promoter.

To elucidate potential transcription factors that bind to the NRE of the EICP0 promoter, the NRE sequence was examined in a transcription factor motif search program (25; www.genome.ad.jp). Four transcription factors, CdxA (42), Lmo2 (52), GATA-2 and -3 (41), and Tal-1/E47 (28), were selected based on matrix scores that exceeded the cutoff threshold of 0.85. The chicken homeodomain protein CdxA binds to the consensus sequence A(A/T)T(A/T)AT(A/G) and can activate transcription in cells (42). The CdxA consensus binding sequence (GATAAAT) was the only sequence that fully overlapped the NREBP-binding sequence ([−193]TTCGATAAATT[−183]) (Fig. 10), while the consensus binding sequence for Tal-1 beta/E47 overlapped half of the NRE sequence.

FIG. 10.

Summary of sequences that match the NRE of the EICP0 promoter. Transcription factors that may bind to the NRE of the EICP0 promoter were identified with a motif program. The binding region and direction of four transcription factors are depicted. Lmo2, a key factor in hematopoiesis; CdxA, chicken homeodomain protein; GATA-2 and -3, GATA-binding factors 2 and 3; Tal-1 beta/E47, Tal-1 beta/E47 (basic helix-loop-helix transcription factor) heterodimer.

DISCUSSION

The EICP0 protein, encoded by an early gene, is a powerful trans-activator of all classes of EHV-1 promoters (3, 4), but, unexpectedly, this protein failed to substantially trans-activate its own promoter. The studies presented in this paper suggest that the EICP0 promoter harbors a negative regulatory element (NRE) that selectively abolishes the trans-activation of the EICP0 promoter by the EICP0 protein, but trans-activation of this promoter by the IE protein is unaffected by the presence of this NRE. The mechanism by which the NRE negatively regulates the EICP0 promoter is thought to involve the recruitment of a cellular protein(s) which represses the activation of this promoter by the EICP0 protein to this sequence. When cloned upstream of other EHV-1 promoters (IE, early TK, and late gK promoters), the NRE ([−204]AGATAGATGTTCGATAAATTGGAACC[−177]) downregulated expression of these promoters, which are representative of all classes of EHV-1 genes. However, this downregulation was only moderate if a construct expressing the EICP0 protein was included in the transfection (data not shown). Analysis of the NRE with a BLAST program revealed one complete match of the 28-bp NRE sequence with a corresponding sequence in the EICP0 promoter of the EHV-1 genome. However, no other herpesvirus genomes contain the NRE sequence, indicating that the NRE of the EHV-1 EICP0 promoter is unique in the Herpesviridae family.

Gel shift assays showed that a ubiquitous nuclear protein (NREBP) binds specifically to a sequence ([−193]TTCGATAAATT[−183]) in the NRE. Four transcription factors, CdxA (42), Lmo2 (52), GATA-2 and 3 (41), and Tal-1/E47 (28), were selected by the transcription factor search motif program as being putative NRE-binding proteins, among which the sequence recognized by CdxA completely overlapped the NREBP-binding sequence (Fig. 10). The chicken homeobox gene cdxA is expressed in all enterocytes during embryonic and posthatch development (17) and may have a role in posthatch enterocyte maturation (19). The CdxA protein binds the consensus sequence A(A/T)T(A/T)AT(A/G) and can activate transcription in cells (42). Gel shift assays showed that a nuclear protein also binds specifically to the Na region. As shown in Fig. 3, the Na and Nb sequences overlap by 10 nucleotides which contain binding sites for the transcription factors GATA-2 and -3 and Lmo2 (Fig. 10). However, the Na fragment does not contain any potential Na-specific transcription factor binding site.

Our findings in this study show that the NREBP is present in the nucleus of EHV-1-infected cells. However, the amount of NREBP from infected cells that bound to Nb was reduced compared to that from uninfected cells. There are two possibilities to explain this effect. One possible mechanism is that the expression level of the cellular protein NREBP is reduced by EHV-1 infection. No virion-associated inhibition of cellular protein synthesis or herpesvirus-induced cellular mRNA degradation was detected in cells infected with any of three EHV-1 strains (Ab4, KyA, and KyD) at multiplicities of infection at which HSV-1 strain F exhibited maximal activity (16). EHV-1's low virion-associated host shutoff protein (vhs) activity in infected cells is not a reflection of the ORF19 protein's intrinsic vhs activity but may be due instead to the amount of ORF19 protein associated with viral particles or to modulation of ORF19 protein's intrinsic activity by another viral component(s) (16). Another possibility is that some viral proteins block or squelch NREBP's interaction with the NRE of the EICP0 promoter. The NREBP in the late stage of infection bound weakly to the Nb sequence compared to the early stage of infection. These results suggest that decreases in NREBP activity are more likely due to its interplay with other viral proteins than to its inhibition by vhs activity.

The HSV-1 ICP0 protein is required for efficient reactivation of the HSV-1 lytic gene expression program from neuronal latency (24). Transgenic mice containing HSV-1 immediate-early (ICP0 and ICP27) gene promoters fused to the E. coli β-galactosidase coding sequence were generated (40). Experiments with these transgenic mice showed that factors present in the neurons of trigeminal ganglia differentially activate the immediate-early ICP0 and ICP27 promoters in neurons (40). It is possible that repression of EICP0 gene expression by specific transcription factors is important in establishment of EHV-1 latency in equine neurons.

Our findings showed that the NRE of the EICP0 promoter specifically abolished the EICP0 protein's trans-activation of its own promoter; however, the IE and EICP27 proteins still strongly trans-activated the EICP0 promoter. It should be noted that the IE protein of EHV-1 is a DNA-binding protein (37) and that four potential binding sites for the IE protein are present in the EICP0 promoter (see Fig. 3). Also, the IE protein binds to factors in the basal transcriptional machinery such as TFIIB and TBP (1, 2, 29) and in addition binds to other EHV-1 regulatory proteins (11, 12, 35).

Our previous studies showed that the EICP0 protein interacts directly with basal transcription factors TFIIB and TBP as well as the IE protein (35) and contains a cysteine-rich zinc RING finger near the N terminus that is essential for activation of the E and L promoters (3). Of the many possibilities that may explain how NREBP abolishes the EICP0 protein's trans-activation of its own promoter, one possible mechanism is that NREBP abolished the EICP0 protein's trans-activation by blocking or squelching its interaction with TFIIB and TBP. HSV-1 ICP0, which is a powerful and promiscuous trans-activator, interacts with viral regulatory molecules (ICP4 and ICP27), cellular proteins (EF-1δ, cyclin D3, proteasomes, etc.), and the cellular transcription factor BMAL1 (5, 14, 26, 30-32, 39, 51, 56). HSV-1 ICP0 dynamically interacts with the proteasome and possesses in vitro ubiquitin E3 ligase activity (23, 51), which may direct the degradation of specific cellular proteins.

Another possible explanation of the NREBP mechanism is that NREBP may abolish the EICP0 protein's trans-activation by interacting directly with the EICP0 protein. The EICP0 protein is not a DNA-binding protein (13) and does not bind to its own promoter (S. K. Kim and D. J. O'Callaghan, unpublished data). Once the NREBP is identified and its gene is cloned into a suitable expression vector, the potential interaction of these two factors will be examined by in vitro and in vivo binding assays that will employ the NREBP and the NRE of the EICP0 promoter. A third possibility is that the NREBP may bind to the EICP0 promoter and cause or induce a conformational change of the promoter that prevents binding by viral and cellular transcription factors. This possibility will be explored by future DNase I footprint assays.

In transient transfection assays, the EICP0 protein weakly trans-activated its own promoter at lower concentrations (<0.1 pmol); however, the EICP0 protein was not able to trans-activate its own promoter at higher concentrations (>0.3 pmol) (data not shown), indicating that the EICP0 protein efficiently trans-activates its own promoter at a lower concentration. These results suggest that overexpression of the EICP0 gene may block EICP0's trans-activation by squelching its interaction with limiting amounts of cellular transcription factors.

During a productive lytic infection, the genes of EHV-1 are coordinately expressed and temporally regulated in an immediate-early (IE), early (E), and late (L) fashion (5, 20, 21). The sole IE protein turns on the expression of the E genes (27, 43, 45, 48, 57), but by itself is not able to activate the expression of some of the γ1 L genes and γ2 true L genes (27, 48). The early EICP0 protein is the only regulatory protein capable of independently activating γ2 L gene expression (3, 4, 33), whereas the IE protein acts in concert with the early EICP22 and EICP27 proteins to turn on the expression of the γ1 and some γ2 L genes (27, 38, 48, 57). Taken together, these observations provide the basis for our hypothesis that the EICP0 protein is a regulatory factor that contributes to the switch from E to L gene expression.

Our recent studies with the EICP0 null mutant virus showed that the switch from early to late viral gene expression for the representative genes examined does not require the EICP0 protein (55). However, mRNA expression of IE, E, and L genes as well as the synthesis of these viral proteins were greatly reduced in cells infected with EICP0-deleted virus compared to those infected with wild-type virus (55). The reduction in late gene expression is pronounced in the absence of the EICP0 protein. As mentioned above, a proposed function of the EICP0 protein is promoting the switch from E to L gene expression. The EICP0 protein may specifically activate late promoters for virus genes involved in cell-to-cell spread and pathogenesis. In transient transfection assays, the EICP0 protein is able to release the γ2 L gK promoter from repression mediated by the IE protein by binding to the transcription initiation site of the gK promoter (33).

The IE protein inhibits the trans-activation mediated by the EICP0 protein (35). The antagonism between the IE and EICP0 proteins (35) may help govern the temporal regulation of gene expression. The IE protein may prevent the EICP0 protein from trans-activating L genes inappropriately at E times. As shown in Fig. 1, the IE and EICP27 proteins trans-activated strongly the EICP0 promoter; however, the EICP0 protein is not able to trans-activate its own promoter. As mentioned before, at an early stage of infection, the IE protein alone or in combination with the EICP27 protein turns on the expression of EICP0 gene. Our studies in this paper showed that the NRE of the EICP0 promoter specifically abolished the EICP0 protein's trans-activation of its own promoter. These results indicate that the expression of the EICP0 gene is negatively regulated by the NRE present in its own promoter. Overall, the expression level of the EICP0 protein would be maintained by the EICP27 and IE regulatory proteins.

To ascertain the molecular phenotype of an NRE knockout virus in infected cells, an NRE-null virus is being generated by bacterial artificial chromosome technology. In addition, yeast one-hybrid assays (Clontech) will be performed to isolate the gene encoding the cellular protein (NREBP) that binds to the NRE of the EICP0 promoter. In vitro and in vivo binding assays with the NREBP and the NRE of the EICP0 promoter may give insight into the mechanism by which the NRE governs the EICP0 promoter and also how the IE protein overcomes the negative regulatory function of the NREBP.