A Pioneer Transcription Factor and Type I Nuclear Hormone Receptors Synergistically Activate the Bovine Herpesvirus 1 Infected Cell Protein 0 (ICP0) Early Promoter

ABSTRACT

Following bovine herpesvirus 1 (BoHV-1) acute infection of ocular, oral, or nasal cavities, sensory neurons within trigeminal ganglia are an important site for latency. Stress, as mimicked by the synthetic corticosteroid dexamethasone, consistently induces reactivation from latency. Expression of two key viral transcriptional regulatory proteins, BoHV-1 infected cell protein 0 (bICP0) and bICP4, are regulated by sequences within the immediate early promoter (IEtu1). A separate early promoter also drives bICP0 expression, presumably to ensure sufficient levels of this important transcriptional regulatory protein. Productive infection and bICP0 early promoter activity are cooperatively transactivated by Krüppel-like factor 4 (KLF4) and a type I nuclear hormone receptor (NHR), androgen receptor, glucocorticoid receptor, or progesterone receptor. The bICP0 early promoter contains three separate transcriptional enhancers that mediate cooperative transactivation. In contrast to the IEtu1 promoter, the bICP0 early promoter lacks consensus type I NHR binding sites. Consequently, we hypothesized that KLF4 and Sp1 binding sites are essential for type I NHR and KLF4 to transactivate the bICP0 promoter. Mutating KLF4 and Sp1 binding sites in each enhancer domain significantly reduced transactivation by KLF4 and a type I NHR. Chromatin immunoprecipitation (ChIP) studies demonstrated that occupancy of bICP0 early promoter sequences by KLF4 and type I NHR is significantly reduced when KLF4 and/or Sp1 binding sites are mutated. These studies suggest that cooperative transactivation of the bICP0 E promoter by type I NHRs and a stress-induced pioneer transcription factor (KLF4) promote viral replication and spread in neurons or nonneural cells in reproductive tissue.

IMPORTANCE Understanding how stressful stimuli and changes in the cellular milieu mediate viral replication and gene expression in the natural host is important for developing therapeutic strategies that impair virus transmission and disease. For example, bovine herpesvirus 1 (BoHV-1) reactivation from latency is consistently induced by the synthetic corticosteroid dexamethasone, which mimics the effects of stress. Furthermore, BoHV-1 infection increases the incidence of abortion in pregnant cows, suggesting that sex hormones stimulate viral growth in certain tissues. Previous studies revealed that type I nuclear hormone receptors (NHRs) (androgen, glucocorticoid, or progesterone) and a pioneer transcription factor, Krüppel-like factor 4 (KLF4), cooperatively transactivate the BoHV-1 infected cell protein 0 (bICP0) early promoter. Transactivation was mediated by Sp1 and/or KLF4 consensus binding sites within the three transcriptional enhancers. These studies underscore the complexity by which BoHV-1 exploits type I NHR fluctuations to enhance viral gene expression, replication, and transmission in the natural host.

INTRODUCTION

Bovine herpesvirus 1 (BoHV-1) acute infection results in high levels of virus production and causes conjunctivitis as well as upper respiratory tract disease (reviewed in reference 1). Infection also suppresses host immune responses (2), which can lead to life-threatening bacterial pneumonia (3). A BoHV-1 entry protein is a bovine respiratory disease complex (BRDC) susceptibility gene for Holstein calves (4). These observations confirm that BoHV-1 is a cofactor of the polymicrobial disease BRDC (1, 3). BRDC is the most important disease in cattle and costs the beef industry in the United States more than $500 million in direct costs and $5 billion in indirect costs each year (5). BoHV-1, including modified live vaccines, is also an important causative agent of abortions (reviewed in reference 6).

BoHV-1 genes, like other Alphaherpesvirinae subfamily members, are expressed in a temporal cascade following infection of permissive cells (reviewed in reference 1). Three immediate early (IE) genes express mRNAs that are translated into BoHV-1 infected cell protein 0 (bICP0), bICP4, and bICP22 (7–10). Promoter sequences within the immediate early transcription unit 1 (IEtu1) control IE expression of IE/2.9 and IE/4.2 mRNAs produced from a single alternatively spliced transcript (Fig. 1A). IE/2.9 mRNA is translated into the bICP0 protein, and IE/4.2 mRNA is translated into the bICP4 protein (7–9). A separate E promoter drives expression of E/2.6, an early transcript translated into the bICP0 protein (7–9, 11). The bICP0 E promoter is predicted to sustain bICP0 protein levels during productive infection. BoHV-1 mutants containing point mutations in the bICP0 RING finger or a deletion of the bICP0 amino terminus grow poorly relative to the rescued mutant or wild-type (wt) BoHV-1 (12, 13). Deleting the amino-terminal region of the bICP0 gene resulted in a virus that established a “persistent-like” infection in bovine kidney cells.

FIG 1.

Schematic of bICP0 E deletion constructs used in this study. (A) Schematic of BoHV-1 genome. L, unique long sequences; S, unique short sequences; open rectangles, repeats. Locations of IE transcripts and promoters that drive their expression during productive infection are also shown (7–9). IE/4.2 encodes the bICP4 protein, and IE/2.9 encodes the bICP0 protein. The IEtu1 promoter activates IE/4.2 and IE/2.9 expression. Expression of a second bICP0 transcript (E/2.6) is driven by the bICP0 early promoter (bICP0 E). The IEtu2 promoter (IEtu2) drives expression of the IE/1.7 transcript that is translated into bICP22. Solid lines in the transcript position map denote exons (e1, e2, or e3), and dashed lines denote introns. ori, origin of replication in the repeats. (B) bICP0 E full-length promoter construct (EP-943) was constructed as previously described (69) and cloned upstream of the luciferase vector (pGL3-Basic Vector; Promega) as a SacI-HindIII fragment. The position of the TATA box is shown, an arrow denotes the position of the transcriptional start site, and the 5′ end of fragment extends to the nucleotide upstream of the initiating ATG. The positions of putative transcription factor binding sites are indicated. (C and D) Putative models of how type I NHRs interact with KLF4 and/or Sp1 transcription factors to recruit transcriptional coactivators and cooperatively transactivate the bICP0 E promoter. Chromatin is represented by DNA (gray lines) wrapped around histone core molecules (light purple).

The ability of BoHV-1 to establish a lifelong latent infection in neurons within trigeminal ganglia (TG) and other neurons is essential for virus transmission and complicates the development of more effective vaccines. The only known viral genes expressed in latently infected neurons are the latency-related (LR) gene (14) and ORF-E, which is upstream of the LR gene (15). Although expression of LR gene products is critical for stress-induced reactivation from latency in calves (16), it is unlikely that they directly facilitate reactivation, because LR gene products are dramatically reduced during dexamethasone-induced reactivation (17, 18). Consequently, we believe that LR gene products support establishment and maintenance of latency, which sustains a pool of latently infected neurons that can reactivate from latency.

The synthetic corticosteroid dexamethasone (DEX) consistently induces reactivation in calves and rabbits latently infected with BoHV-1 (14, 19). Stress correlates with an increased incidence of reactivation episodes in cattle (reviewed in references 1 and 14). Certain stress-induced cellular transcription factors identified in TG during reactivation (20) cooperate with the glucocorticoid receptor (GR) and DEX to stimulate key viral promoters and productive infection (21–23). We predict that these cellular transcription factors mediate early stages of reactivation from latency by stimulating lytic cycle viral gene expression. During DEX-induced reactivation from latency, bICP0 and VP16 proteins are readily detected within 1 h after DEX treatment; conversely, other late proteins (gC and gE) are seen only in a few TG neurons at 6 h after DEX treatment (24, 25). bICP0 and VP16 also appear to be expressed in TG neurons prior to bICP4 and bICP22 (26), suggesting that the bICP0 E promoter and VP16 promoter are active prior to the IEtu1 promoter during DEX-induced reactivation.

During latency, the BoHV-1 genome is likely organized as heterochromatin, suggesting that many transcription factors are unable to bind and transactivate viral promoters. During DEX-induced reactivation from latency, expression of four KLF family members (KLF4, KLF6, KLF15, and PLZF) are stimulated within 3 h after DEX-induced reactivation from latency (20). KLF family members resemble the Sp1 transcription factor family, and both families of transcription factors interact with GC-rich motifs, including certain consensus Sp1 binding sites (GGGCGG) (reviewed in references 27 and 28). The Sp1 transcription factor binds herpes simplex virus 1 (HSV-1) promoters and stimulates viral transcription (29), suggesting that Sp1 and certain KLF family members can activate BoHV-1 promoters. Interestingly, KLF4 is a pioneer transcription factor that can specifically bind silent chromatin (heterochromatin) and subsequently activate transcription (30–33). Conversely, most transcription factors bind “relaxed” chromatin or DNA lacking nucleosomes. In summary, stress-induced transcription factors like KLF4 are predicted to activate the BoHV-1 lytic cycle program of gene expression following stressful stimuli that trigger reactivation from latency in neurons and promote viral replication in nonneuronal cells that do not support high levels of viral replication.

The androgen receptor (AR), GR, and progesterone receptor (PR) belong to the type I family of nuclear hormone receptors (NHRs). AR and PR specifically bind many GR response elements (GREs) and can transactivate promoters containing GREs (34–36). Type I NHR-hormone complexes translocate to the nucleus, interact with a GRE or similar hormone response elements (HREs), and stimulate transcription. GR and PR are also considered to be pioneer transcription factors (31–33). The BoHV-1 IEtu1 promoter contains two functional GREs (22, 23). Interestingly, GR and PR cooperate with Krüppel-like transcription factor 4 (KLF4) or KLF15, two stress-induced transcription factors in TG, to transactivate the IEtu1 promoter and stimulate productive infection (22, 37–40). The IEtu1 promoter is transactivated only by GR or PR and KLF4 or KLF15 when both GREs are present (23, 38). The bICP0 E promoter is cooperatively transactivated by KLF4 plus AR, GR, or PR more efficiently than other stress-induced transcription factors (37, 39, 40). In contrast to the IEtu1 promoter, the bICP0 E promoter does not contain a consensus GRE and the half GREs are not important for transactivation by type I NHRs. Independent studies have demonstrated that type I NHRs cooperate with certain KLF family members to cooperatively activate specific cellular promoters (41–44). The objective of this study was to identify bICP0 E promoter sequences necessary for cooperative transactivation by type I NHRs and KLF4.

RESULTS

Identification of bICP0 E promoter sequences that mediate KLF4 and type 1 NHR cooperative transactivation.

The bICP0 E promoter (EP-943), which spans the 943 nucleotides directly 5' of the translation initiating AUG, contains three separate transcriptional enhancers, nucleotides 172 to 328, 328 to 638, and 638 to 943, that mediate cooperative transactivation by KLF4 and GR (37), PR (39), or AR (40) (Fig. 1B). The half GREs located between nucleotides 638 and 943 do not play a role in transactivation mediated by AR, GR, or PR and KLF4 (37, 39, 40). When the individual enhancer fragments are cloned upstream of a heterologous minimal promoter, cooperative transactivation by type I NHRs and KLF4 is not recapitulated, in part because basal enhancer activity is high (data not shown) (37). Although there are numerous reasons why this approach was not useful, we predict that bICP0 E promoter-proximal sequences are important for mediating cooperative transactivation. For these studies, we used cells from a mouse neuroblastoma cell line (Neuro-2A) because they have neuron-like properties and can be differentiated into dopamine-like neurons (45). Approximately 50% of Neuro-2A cells are transfected, making them an excellent model to examine the effects of stress-induced transcription factors on activation of key BoHV-1 regulatory promoters. While primary neurons are more physiologically relevant, they are poorly transfected, making them unsuitable for identifying promoter/enhancer elements responsive to type I NHR and stress-induced transcription factors. Since BoHV-1 replicates approximately 10,000-fold less efficiently than bovine cells (46), we were unable to conclusively demonstrate that GR and KLF4 occupied bICP0 E promoter sequences following infection with BoHV-1 (data not shown). However, we previously demonstrated that GR, AR, PR, and KLF4 occupy bICP0 E promoter sequences during productive infection of permissive bovine cells (37, 39, 40).

Based on observations discussed above, we hypothesized that type I NHRs are tethered to KLF4 or a complex containing KLF4 plus the type I NHR; consequently, this complex binds KLF4/Sp1 binding sites and stimulates promoter activity (Fig. 1C). Interactions between KLF4 and unknown transcriptional coactivators may also promote interactions with type I NHRs. This protein complex then specifically binds KLF4/Sp1 binding sites and stimulates promoter activity (Fig. 1D). Regardless of these two scenarios, bICP0 E promoter activity is cooperatively stimulated by a type I NHR and KLF4. To test this hypothesis, mutant constructs for each enhancer domain were constructed and examined for their ability to be transactivated by a type I NHR and KLF4.

Initial studies analyzed the effects of mutating a KLF4 consensus binding site (nucleosome depleted) in EP-328 (Fig. 2A) that overlaps a Sp1 binding site (Fig. 2A and B, blue circles and black triangles, respectively) and a KLF4-like binding site that has one base mismatch to a KLF4 nucleosome-enriched site (gray circles). KLF4 can bind distinct sequences via a nucleosome-enriched or -depleted mechanism (34). Since previous studies revealed that a type I NHR or KLF4 had little effect on bICP0 E promoter activity relative to the cooperative effects of these transcription factors (37, 39, 40), these studies focused on the synergistic effects of KLF4 and a type I NHR. Wild-type EP-328 was transactivated by AR, GR, or PR and KLF4 between 8- and 10-fold (Fig. 2C to E), which is consistent with previous results (37). Dihydrotestosterone (DHT), an AR-specific agonist, reduced AR- and KLF4-mediated transactivation of EP-328. Furthermore, DEX and P4 (GR and PR agonists, respectively) also inhibited transactivation, suggesting that EP-328 mediated transactivation by KLF4 and that type I NHR-mediated transactivation occurred by a ligand-independent mechanism (47, 48), which confirmed previous studies (37, 39, 40).

FIG 2.

Localization of EP-328 sequences important for type 1 NHR- and KLF4-mediated transactivation. (A) Schematic of wt EP-328 and mutants used for this study. The locations and sizes of the PCR products using the EP-328 primer sequences are shown. The EP-328 primers used are described in Materials and Methods. Black triangles indicate Sp1 binding sites. Lowercase red letters denote position of mutations in the Sp1 or KLF4 binding sites. (B) Neuro-2A cells were transfected with the designated EP-328 promoter constructs (0.5 μg DNA), type I NHR construct (1 μg DNA for AR and GR and 2 μg for PR; 1 μg of PR-A and PR-B), and KLF4 (0.5 μg DNA for GR or PR, 1 μg DNA for AR). (C to E) Designated cultures were treated with 2% stripped FBS, and DHT (10 nM) (C), DEX (10 μM) (D), or P4 (100 nM) (E) was added to the cultures. At 48 h after transfection, cells were harvested and protein lysate was subjected to a dual-luciferase assay. Promoter activity levels in the sample containing wt EP-328 cotransfected with only empty vector and treated only with DMSO (no hormone) were normalized to a value of 1, and the fold activation for other samples is presented. The results are the average of three independent experiments, and error bars indicate the standard errors. Student's t test was used for analyzing the results. An asterisk indicates significant differences (P < 0.05) in transactivation of wt EP-328 cotransfected with a type 1 NHR and KLF4 relative to Neuro-2A cells transfected with mutant EP-328 constructs.

Mutating the KLF4-like binding site (EP-328ΔKLF4) significantly reduced the ability of AR (Fig. 2C), GR (Fig. 2D), or PR (Fig. 2E) to cooperate with KLF4 relative to the wt EP-328 construct. When the overlapping KLF4/Sp1 binding site in EP-328 was mutated (EP-328ΔSp1), a significant reduction in the ability of AR, GR, or PR to cooperate with KLF4 to transactivate EP-328ΔSp1 relative to wt EP-328 was also observed in transfected Neuro-2A cells. Mutating both sites (EP-328ΔKLF4ΔSp1) did not further reduce transactivation, suggesting that both sites were necessary for cooperative transactivation.

Mutating EP-328 KLF4/Sp1 binding sites impairs NHR and KLF4 interactions.

Occupancy of EP-328 sequences by AR (Fig. 3A), GR (Fig. 3B), and PR (Fig. 3C) was significantly higher than that in EP-328ΔKLF4ΔSp1 following transfection of the promoter construct alone or cotransfection with KLF4 and a type I NHR into Neuro-2A cells. Significantly higher levels of GR occupied EP-328 when transfected with GR and KLF4 than with the empty control as well or when transfected with GR and KLF4 and treated with DEX. In contrast, this was not the case for AR. With respect to PR, significantly higher PR levels occupied EP-328 sequences when cotransfected with PR and KLF4 regardless of DEX treatment relative to the empty control. KLF4 occupancy of EP-328, but not EP-328ΔKLF4ΔSp1, increased when cells were cotransfected with AR and KLF4 regardless of DHT treatment (Fig. 3D). Furthermore, significantly higher levels of KLF4 occupied EP-328 promoter/enhancer sequences when cells were cotransfected with AR and KLF4 regardless of DHT treatment relative to the empty control. KLF4 occupancy of EP-328 was significantly higher than EP-328ΔKLF4ΔSp1 occupancy when Neuro-2A cells were cotransfected with KLF4 and GR regardless of DEX treatment (Fig. 3E). Surprisingly, higher levels of KLF4 occupied EP-328 promoter/enhancer sequences when cells were cotransfected with GR and KLF4 and treated with DEX than when cultures were incubated with DEX and the empty control. While it was clear that significantly more KLF4 occupied EP-328 than the EP-328ΔKLF4ΔSp1 construct, cotransfection with PR and KLF4 had no effect on KLF4 occupancy regardless of P4 treatment (Fig. 3F). Collectively, these studies revealed that KLF4 occupancy of EP-328ΔKLF4ΔSp1 was significantly less than that of wt EP-328 regardless of transfection with AR, GR, or PR and treatment with the hormone agonist.

FIG 3.

Occupancy of EP-328 by a type I NHR and KLF4. Neuro-2A cells were cotransfected with EP-328 (4 μg DNA), AR (3.0 μg DNA) (A and D), GR (3.0 μg DNA) (B and E), or PR (1.5 μg DNA each of PR-A and PR-B expression plasmid) (C and F) and, where indicated, KLF4 (3 μg DNA). Empty vector was added to maintain the same concentration of DNA in each transfection assay. Mock samples are Neuro-2A cells that were not transfected with DNA. After transfection, 2% stripped FBS was added to the medium, and where indicated, the respective hormone was added. Transfected cells were processed for ChIP as described in Materials and Methods, and immunoprecipitation (IP) was conducted using the AR (A), GR (B), PR antibody (C), or KLF 4 antibody (D to F). As a control, isotype antibody was also used for each sample. Detection of immunoprecipitated DNA was performed by PCR using the EP-328 primer set (see Materials and Methods and Fig. 2A). The results are the average of three independent experiments. An asterisk indicates statistically significant differences between the wt and mutant constructs. A dagger indicates a significant difference between the indicated wt samples. Significant differences (P < 0.05) were calculated using Student's t test.

Identification of bICP0 EP-638 promoter sequences important for transactivation by type 1 NHRs and KLF4.

The bICP0 EP-638 promoter (Fig. 4A) contain five distinct Sp1 and/or four putative KLF4 binding sites within nucleotides 328-638, including three overlapping KLF4-like/Sp1 repeats (GGGCGGG) (Fig. 4B, motif C). The KLF4 and Sp1 sites between nucleotides 172 and 328 are designated motif E. Mutants of motifs A to E (EP-638ΔAll) were examined for cooperative transactivation by KLF4, a type1 NHR, and the respective hormone in Neuro-2A cells. In comparison to wt EP-638, the promoter activity of the EP-638ΔC mutant was significantly reduced when cotransfected with KLF4 and AR (Fig. 4C), GR (Fig. 4D), or PR (Fig. 4E). Consistent with previous studies (37, 39, 40), EP-638 exhibited reduced type I NHR- and KLF4-mediated transactivation compared to EP-943. Furthermore, EP-638ΔC and wt EP-328 promoter had similar promoter activities, suggesting that the intact C motif was essential for cooperative transactivation of EP-638 by KLF4 and a type 1 NHR (Fig. 4D and E). When sites A to E were all mutated (EP-638ΔAll), transactivation by KLF4 and a type I NHR was the same as that of the EP-328 promoter and EP-638ΔC construct. Interestingly, the EP-638ΔB construct exhibited significantly less transactivation when cotransfected with KLF4 and PR (Fig. 4E), but not AR or GR and KLF4, than EP-638, suggesting that this site was more important for PR- and KLF4-mediated transactivation. Furthermore, EP-638ΔD was transactivated by GR and KLF4 with reduced efficiency compared to EP-638 (Fig. 4D). As previously reported, KLF4 and type I NHR transactivation occurred in a ligand-independent fashion because addition of the respective hormone significantly reduced promoter activity.

FIG 4.

Identification of sequences in EP-638 responsive to type 1 NHRs and KLF4. (A) Schematic representation of the EP-638 promoter fragment and positions of relevant transcription factor binding sites. Letters beneath transcription factor binding sites denote the names of each mutant. All five individual mutants and the ΔAll mutant, which contains all five mutations, were cloned into pGL3-basic plasmid that drives a firefly luciferase gene. (B) The GC-rich Sp1 and KLF4 binding sites were mutated into EcoRI sites (GAATTC) to disrupt core binding sequences. Lowercase red letters denote mutations in the respective binding sites. For each binding site motif, wt and mutant sequences are shown. Motifs A, D, and E each contain a single Sp1 site, and motifs B and E contain KLF4 sites. Motif C contains three adjacent Sp1 sites. Due to the similarity between Sp1 and KLF4 binding sites, motif C contains three sequential KLF4-like binding sites. All three Sp1 sites were mutated, which also disrupts the KLF4-like sites. (C to E) The denoted bICP0 E promoter construct (0.5 μg plasmid) was transfected into Neuro-2A cells cultured in MEM containing 2% stripped FBS, alone or in combination with KLF4 (0.5 μg DNA for GR or PR, 1 μg for AR) and AR (1 μg DNA) (C), GR (1 μg) (D), or PR (1 μg each PR-A and PR-B) (E). Empty vector plasmid was added where needed to maintain a consistent quantity of DNA in each reaction mixture. Cells were also cotransfected with a plasmid containing a minimal TK promoter driving Renilla luciferase as a transfection control. At 24 h after transfection, certain samples were treated with the corresponding nuclear hormone, DHT (10 nM), DEX (10 μM), or P4 (100 nM). Untreated cultures contained an equal volume of methanol (DHT) or DMSO (P4) to maintain consistency across reactions, as DHT is dissolved in methanol and P4 in DMSO. At 48 h after transfection, cells were harvested and a dual-luciferase assay was performed. Promoter activity was calculated as a ratio of firefly luciferase activity to that of the Renilla luciferase control. Samples with only the promoter construct were normalized to a value of 1, and fold activation over this baseline by KLF4-and type I NHR-transfected cells is presented. The results are the average of three independent experiments, and error bars indicate the standard errors. An asterisk indicates significant differences (P < 0.05) in cells transfected with the EP-638 mutant construct cotransfected with a type 1 NHR and KLF4 relative to the wt promoter using Student's t test.

KLF4/Sp1 binding sites in EP-638 are crucial for NHR and KLF4 binding.

DNA sequences between EP-328 and EP-638 contain 310 bases, and only 29 of these bases are A or T; hence, this fragment is 91% GC rich. Consequently, we failed to produce reliable PCR primers in this region that are suitable for ChIP studies. Although the GC content in EP-328 is 84%, the primers used to ChIP the EP-328 promoter consistently yielded clear-cut results and were used to ChIP wt EP-638 and EP-638ΔAll.

All three type I NHRs studied (AR, GR, and PR [Fig. 5A to C, respectively]) and KLF4 (Fig. 5D to F) occupied EP-638 DNA when the promoter alone was transfected into Neuro-2A cells, indicating that these transcription factors are expressed in Neuro-2A cells. Promoter occupancy of the mutant EP-638ΔAll construct was significantly reduced for all type I NHRs and KLF4 (Fig. 5, white bars). These studies also revealed that cotransfection with AR or GR and KLF4 significantly increased AR and GR occupancy. Cotransfection of EP-638 with KLF4 and AR (Fig. 5D) or GR (Fig. 5E), but not PR (Fig. 5F), significantly increased KLF4 occupancy of EP-638 promoter/enhancer sequences. Treatment of cultures with the indicated hormone significantly reduced AR, GR, and PR occupancy of EP-638 promoter/enhancer sequences. Collectively, these data demonstrate that the Sp1 and KLF4 binding sites within the EP-638 promoter were essential for KLF4 and type I NHR binding. For the most part, there was a correlation between transactivation by KLF4 and a type I NHR (Fig. 4) and increased promoter occupancy of wt EP-638 by these transcription factors relative to EP-638ΔAll.

FIG 5.

Occupancy of EP-638 sequences by a type I NHR and KLF4. Neuro-2A cells were transfected with either the bICP0 wt EP-638 or the EP-638ΔAll mutant (4 μg DNA) promoter, alone or with KLF4 (3 μg DNA) and a type 1 NHR expression plasmid: AR (3.0 μg DNA) (A and D), GR (3.0 μg DNA) (B and E), or PR (1.5 μg DNA each, PR-A and PR-B) (C and F). Empty vector plasmid was added to maintain an equal concentration of DNA in each transfection reaction. Mock samples were untransfected Neuro-2A cells. After transfection, 2% stripped FBS was added to the medium, and where noted, cells were treated with 10 nM DHT, 10 μg DEX, or 100 nM P4. For untreated cells, the appropriate vehicle was added: methanol for DHT or DMSO for P4. Transfected cells were processed for ChIP as described in Materials and Methods and immunoprecipitated (IP) using the AR (A), GR (B), PR (C), or KLF 4 (D to F) specific antibody. Nonspecific rabbit IgG was used as an isotype control antibody. Detection of immunoprecipitated DNA was performed using the EP-328 primer set as described in Fig. 2A and Materials and Methods. The results are the average of three independent experiments. An asterisk indicates statistically significant differences between the wt and mutant constructs. A dagger indicates a significant difference between the indicated wt samples. Significant differences (P < 0.05) were calculated using Student's t test.

KLF4/Sp1 sites in EP-943 are important for type 1 NHR and KLF4 transactivation.

Deleting nucleotides 638 to 943 consistently reduced type I NHR- and KLF4-mediated transactivation relative to that of EP-943 (37, 39, 40): however, the two half GREs play no role in this reduction. A perfect match for a nucleosome-enriched (n.e.) KLF4 binding site (33) is adjacent to a consensus Sp1 binding site in EP-943 (Fig. 6A and B). A nucleosome-depleted (n.d.) KLF4-like binding site with one nucleotide mismatch (C versus G) is also embedded within the Sp1 binding site. Mutations in these sites were examined in Neuro-2A cells. EP-943ΔKLF4, EP-943ΔSp1, and EP-943ΔKLF4ΔSp1 were transactivated significantly less by KLF4 and AR (Fig. 6C), GR (Fig. 6D), or PR (Fig. 6E) in Neuro-2A cells than wt EP-943. Interestingly, mutating one or both sites yielded the same effect as deleting sequences between nucleotides 638 and 943 (EP-638), indicating that the KLF4 and adjacent Sp1 binding sites were essential for cooperative transactivation by KLF4 and type I NHRs. As previously demonstrated (37, 39, 40), the addition of DHT, DEX, or P4 to cultures reduced wt EP-943, EP-638, and the respective mutants. These results are consistent with previous studies demonstrating that type I NHRs cooperate with KLF4 via an unliganded mechanism because addition of the respective hormones was not required for transactivation.

FIG 6.

Identification of EP-638 to EP-943 sequences important for transactivation by KLF4 and type 1 NHRs. (A) Schematic of wt EP-943 mutants used for this study and location of transcription factor binding sites. Primers used to amplify sequences between nucleotides 639 and 943 (EP-943 primers) are described in Materials and Methods, and the location of the 126-bp PCR product is shown. (B) Nucleotide sequences of Sp1 and KLF4 binding sites between nucleotides 638 and 943. Blue shaded nucleotides denote the KLF4/Sp1 binding site and yellow shaded nucleotides denote the KLF4 binding site. Mutants prepared are shown, and red or blue nucleotides denote positions of mutations. n.e., nucleosome enriched; n.d., nucleosome depleted. Neuro-2A cells were transfected with EP-943, EP-638, or the EP-943 mutant promoter constructs (0.5 μg of designated reporter construct) and KLF4 (0.5 μg with AR or GR, 1 μg with PR) plus the denoted type I NHR (1.0 μg DNA AR or GR, 1 μg of PR-A and PR-B). At 24 h after transfection, cultures were treated with 2% stripped FBS, and the indicated hormone, DHT (10 nM) (C), DEX (10 μM) (D), or P4 (100 nM) (E). Untreated samples had vehicle added (DMSO for P4 or methanol for DHT) to maintain consistency across samples. At 48 h after transfection, cells were harvested, and protein lysate was subjected to the dual-luciferase assay. Promoter activity of the respective bICP0 E promoter constructs cotransfected with only an empty vector and not treated with the respective hormone was set at a value of 1, and fold activation above this baseline for other samples is presented. The results are the average of three independent experiments, and error bars indicate the standard errors. Student's t test was used for analyzing the results. An asterisk indicates significant differences (P < 0.05) in cells transfected with wt EP-943 cotransfected with a type 1 NHR and KLF4 relative to the respective mutant promoter constructs.

Occupancy of EP-943 sequences by type I NHRs and KLF4.

Occupancy of EP-943 sequences located between nucleotides 638–943 by AR (Fig. 7A), GR (Fig. 7B), and PR (Fig. 7C) was significantly higher than that of EP-943ΔKLF4ΔSp1 following transfection of these promoter constructs into Neuro-2A cells. Overexpression of AR and KLF4, but not GR and KLF4, significantly increased occupancy of wt EP-943 promoter/enhancer sequences in comparison to the empty control and overexpressing KLF4, AR, and DHT treatment. Surprisingly, significantly higher PR levels occupied EP-943 promoter/enhancer sequences in the empty control relative to cells transfected with PR and KLF4 regardless of PR treatment. KLF4 occupancy of EP-943 was significantly higher than that of EP-943ΔKLF4ΔSp1 when transfected with KLF4 and AR (Fig. 7D), GR (Fig. 7E), or PR (Fig. 7F). KLF4 occupancy was significantly higher when cultures were cotransfected with AR and KLF4 if cultures were treated with DHT relative to the other combinations. Relative to the empty control, significantly higher KLF4 levels occupied EP-943 when cotransfected with GR and KLF4 regardless of DEX treatment. Furthermore, addition of DEX significantly increased KLF4 occupancy following transfection with GR and KLF4 in comparison to cultures not treated with DEX. Finally, PR and KLF4 cotransfection plus P4 treatment significantly increased KLF4 occupancy of EP-943 promoter/enhancer sequences relative to the empty control sample. In general, these studies demonstrated that mutating the KLF4 and Sp1 binding site near the 5′ terminus of EP-943 influenced type I NHR and KLF4 occupancy, which correlated with promoter activation (Fig. 6).

FIG 7.

Occupancy of EP-638 to EP-943 sequences by type I NHRs and KLF4. Neuro-2A cells were cotransfected with the denoted EP-943 luciferase construct (4 μg DNA), AR expression plasmid (3.0 μg DNA) (A and D), GR (3.0 μg DNA) (B and E), or PR (1.5 μg PR-A and 1.5 μg PR-B DNA) (C and F) and, where indicated, the KLF4 expression construct (3 μg DNA). Empty vector was added to maintain the same concentration of DNA in each transfection assay. Mock samples were Neuro-2A cells not transfected with DNA. After transfection, 2% stripped FBS was added to MEM, and cells were treated with the indicated hormone or vehicle control. Transfected cells were processed for ChIP as described in Materials and Methods, and IP was conducted using AR (A), GR (B), PR (C), or KLF4 (D to F) or isotype control antibody as indicated. Detection of immunoprecipitated DNA was performed using the EP-943 primer set, which is described in Materials and Methods, and the location and size of the 126-bp PCR product (Fig. 6A). The results are the average of three independent experiments. An asterisk indicates statistically significant differences (P < 0.05) from ChIP samples that were immunoprecipitated with isotype control IgG using Student's t test. A dagger indicates a significant difference between the indicated wt samples. Significant differences (P < 0.05) were calculated using Student's t test.

DISCUSSION

Increased type I NHR levels generally stimulate KLF4 expression. For example, GR and KLF4 coregulate anti-inflammatory genes in keratinocytes (43). Furthermore, P4 stimulates KLF4 expression (49), which promotes transition of granulosa cells to luteal cells and shifts estrogen production to predominantly P4 (50). In addition, activated AR binds KLF4 promoter sequences and enhances KLF4 expression; consequently, KLF4 reciprocally binds AR promoter sequences to sustain KLF4 expression (44). Thus, normal physiological changes in hormone levels may trigger bICP0 E promoter activity prior to the IEtu1 promoter in certain cell types. Consistent with this prediction, the bICP0 protein (26) and mRNA (51) are detected in TG prior to bICP4 during DEX-induced reactivation from latency. PR (52, 53) and AR are also detected in neurons (54), and P4 sporadically triggers reactivation from latency in rabbits (55) and calves latently infected with BoHV-1 (unpublished studies). Since BoHV-1 is the most frequently diagnosed cause of viral abortion in North American cattle (6), increased P4 levels during pregnancy may promote virus spread to female reproductive tissues, which culminates in reduced live birth rates. Finally, viral DNA has been detected in semen (56–58), suggesting androgen levels enhance virus that is spread to testicular tissue and transmitted to cows via sexual contact. Collectively, these observations suggest that normal physiological changes in hormone levels and their effects on KLF4 expression can stimulate virus replication and shedding in cattle because the BoHV-1 genome contains more than 100 consensus type I NHR binding sites (23).

Our studies generally support the model proposed in Fig. 1C, because KLF4 (nucleosome enriched or depleted) and/or Sp1 binding sites were essential for cooperative transactivation. Hence, we suggest that KLF4, Sp1, and/or unknown transcriptional coactivators, but not type I NHRs, directly bind bICP0 E promoter sequences to mediate cooperative transactivation. This model is also supported by studies demonstrating that GR, and presumably AR and PR, regulate transcription by tethering directly to certain transcription factors or via a mediator-assisted tethering mechanism (reviewed in reference 59). GR does not stably interact with KLF4 (unpublished data); however, KLF4 and PR (39) or AR (40) stably interact with each other. Consequently, GR and KLF4 are predicted to cooperatively transactivate the bICP0 E promoter by the interaction of GR with a transcriptional coactivator that is part of a complex containing KLF4 and/or Sp1, as shown in Fig. 1D.

While overexpression of AR, GR, or PR and KLF4 clearly stimulated wt EP-943, EP-638, or EP-328 promoter activity, there was not always a correlation between AR, GR, or PR occupancy of bICP0 E promoter sequences in ChIP studies when a type I NHR was overexpressed relative to negative controls. In general, this was the case for the EP-328 and EP-943 constructs. Conversely, transfection of a type I NHR, KLF4, and wt EP-638 leads to increased AR, GR, or PR occupancy relative to that of cultures transfected with EP-638 and empty vector. There are low basal levels of type I NHRs in Neuro-2A cells, as well as in most cultured cells. Furthermore, type I NHRs in certain established cell lines may contain mutations that reduce their activity, because these hormone receptors can induce cell death, arrest growth, and/or induce differentiation. Notably, the endogenous GR in Neuro-2A cells is smaller than the GR-alpha protein. Alternative splicing of the GR pre-mRNA can lead to eight GR protein isoforms (60). The GR construct we use expresses only the GR-alpha isoform because it transactivates GRE-containing promoters better than the other seven isoforms. Endogenous GR in Neuro-2A cells does not efficiently transactivate viral promoters or stimulate productive infection: however, it clearly binds to bICP0 E promoter sequences. We have also transfected the two PR isoforms (PR-A and PR-B) in equal concentrations, because certain promoters are transactivated more efficiently by one or the other isoform (61). Based on these observations, we predict there may not be a significant increase in binding of the respective type I NHRs, because the endogenous protein is replaced by the wt type I NHR transfected into Neuro-2A cells. As discussed above, we suggest that GR is tethered to KLF4 and/or other transcriptional coactivators because the bICP0 E promoter does not contain a consensus GRE. Consequently, stringent washing conditions used for ChIP studies may remove proteins from a protein complex (including type I NHRs) that are not directly and tightly bound to bICP0 E promoter sequences. Overexpression of AR or GR and KLF4 consistently led to an increase in KLF4 occupancy of EP-328 and EP-638: however, this effect was not as dramatic as with EP-943. With respect to PR and KLF4 cotransfection studies, KLF4 occupancy of EP-328, EP-638, and EP-943 was not dramatically different from that of the empty control. While interpretation of the results of the ChIP studies is not straightforward, significant differences were consistently detected between the respective wt and mutant constructs using ChIP and promoter strength studies.

Regardless of which type I NHR was examined, cooperative transactivation with KLF4 occurred via an unliganded fashion because hormone addition reduced bICP0 E promoter activity. Although it is unclear how unliganded activation of steroid receptors occurs (62), ligand-independent GR phosphorylation facilitates specific cellular stress signaling pathways (63). For example, ligand-independent GR phosphorylation correlates with increased UV exposure (63, 64); consequently, mRNA that encodes specific enzymes regulated by GR increases (65). Furthermore, β₂-adrenergic receptor agonists, and in part cAMP, induce ligand-independent GR activation (47). Notably, epinephrine (adrenaline), a neurotransmitter released in response to stress (66), also activates GR via a β₂-adrenergic receptor-mediated mechanism (67). We suggest that binding of a type I NHR to its specific ligand impairs an interaction between a type I NHR with KLF4 or other transcriptional coactivators necessary for transactivation of the bICP0 E promoter. Consequently, certain stressful stimuli may preferentially trigger bICP0 E promoter activity and induce reactivation from latency in the absence of high levels of corticosteroids. Since the IEtu1 promoter is cooperatively transactivated more efficiently by KLF15 and GR or PR than by KLF4, increased corticosteroid levels are predicted to preferentially trigger IEtu1 promoter activity and reactivation from latency.

MATERIALS AND METHODS

Cells and virus.

Mouse neuroblastoma cells (Neuro-2A) cells were obtained from the ATCC. These cells were grown in minimal essential medium (MEM) supplemented with 10% fetal bovine serum (FBS), penicillin (10 U/ml), and streptomycin (100 μg/ml).

Plasmids.

A mouse GR expression vector was obtained from Joseph Cidlowski, NIH. An AR expression vector was obtained from Addgene. The human PR-A and PR-B isoforms in the pSG5 expression vector were obtained from Pierre Chambon (University of Strasbourg, Strasbourg, France). The KLF4 expression vector was obtained from Jonathan Katz (University of Pennsylvania).

Construction of the BoHV-1 bICP0 E promoter and deletion constructs (EP-943, EP-638, EP-328) used in the present study were described previously (21, 22, 37, 68). Numbers in the plasmid name refer to the length of the bICP0 E promoter fragment (in nucleotides) cloned into pGL3-Basic Vector (Promega; see Fig. 2, 4, and 6 for a schematic of these constructs). All plasmids were prepared from bacterial cultures by alkaline lysis and two rounds of cesium chloride ultracentrifugation.

Transfection and dual-luciferase reporter assay.

Neuro-2A cells (8 × 10⁵) were seeded into 60-mm dishes containing MEM with 10% FBS at 24 h prior to transfection. Two hours before transfection, the medium was replaced with fresh growth medium lacking any antibiotics. Cells were cotransfected with the designated bICP0 E promoter construct cloned upstream of the firefly luciferase gene (0.5 μg plasmid DNA) in the pGL3-Basic Vector and a plasmid encoding Renilla luciferase under the control of a minimal herpesvirus thymidine kinase (TK) promoter (50 ng DNA) as a transfection control. Where indicated, expression plasmids expressing PR-A and PR-B plasmid (1 μg of DNA of each PR construct) or AR (1 μg plasmid DNA) or GR (1 μg plasmid DNA) and KLF4 (0.5 μg plasmid DNA for PR and GR or 1 μg plasmid DNA for AR) were included in the plasmid mixture. To maintain equal plasmid concentrations in the transfection mixtures, empty expression vector was added as needed. EP-328 and EP-943 (Fig. 2 and 6, respectively) were transfected with Lipofectamine 3000 (Thermo Fisher catalog no. L3000015), while EP-638 samples (Fig. 4 and 5) were transfected with TransIT-X2 (Mirus catalog no. MIR6000) in accordance with the manufacturers’ guidelines. Due to the higher transfection efficiency with TransIT-X2, promoter activity in EP-638 (Fig. 4) samples was normalized to wild-type promoter activation levels in EP-943 (Fig. 6) and EP-328 (Fig. 2) by calculating the average difference between the wild-type promoter activation levels in the two sets of samples and applying this correction factor to the EP-638 luciferase samples in Fig. 4. Neuro-2A cells were incubated in 2% charcoal-stripped FBS after transfection. At 24 h after transfection, Neuro-2A cultures were treated with water-soluble DEX (10 μM; Sigma, D2915), 5α-dihydrotestosterone (DHT) (10 nM; Cerilliant, D-073), or P4 (100 nM; Tocris Bioscience, 2835). Forty-eight hours after transfection, cells were harvested, and protein extracts were subjected to a dual-luciferase assay using a commercially available kit (E1910; Promega). Luminescence was measured with a GloMax 20/20 luminometer (E5331; Promega).

ChIP assay.

Chromatin immunoprecipitation (ChIP) studies were performed as previously described (21, 22, 37, 68). Neuro-2A cells were grown in 100-mm dishes to approximately 80% confluence (1 dish/sample, approximately 2 × 10⁶ cells) and were cotransfected with the designated bICP0 E promoter constructs (4 μg DNA) and a plasmid that expresses PR-A and PR-B plasmid (1.5 μg of DNA of each PR construct) or plasmids that express GR or AR and KLF4 (3 μg DNA of each). For these studies, cells were transfected with the indicated plasmids using Lipofectamine 3000 (Invitrogen) according to the manufacturer’s instructions. Each plate was treated as a separate sample. Twenty-four hours after transfection, Neuro-2A cells were cultured in MEM containing 2% charcoal-stripped FBS. Cultures were treated with vehicle (dimethyl sulfoxide [DMSO] or methanol), DEX (10 uM; Sigma), DHT (10 nM; Cerilliant), or P4 (100 nM; Tocris Bioscience, 2835) for 4 h. Formaldehyde cross-linked cells were lysed in buffer A (50 mM HEPES [pH 7.5], 140 mM NaCl, 1 mM EDTA [pH 8.0], 1% Triton X-100, 0.1% sodium deoxycholate, 0.1% SDS) containing protease inhibitors. Following sonication to generate DNA fragments of approximately 500 bp, cell lysate containing sheared DNA was precleared using salmon sperm DNA-agarose (Millipore) to remove DNA that nonspecifically sticks to agarose. Input samples were collected (10 μl) from the precleared sonicated DNA protein complexes (500-μl sample).

Cleared lysate (one-third of the total lysate) was divided into three tubes and incubated with 5 μl of PR antibody (alpha PR-22; Thermo Fisher Scientific; MA1-412) or 4 μg of anti-GR (3660S; Cell Signaling) or anti-AR (441; Santa Cruz), the anti-KLF4 antibody (ab72543; Abcam) (4 μg), or nonspecific isotype control rabbit IgG (18140; Sigma) in buffer B (50 mM Tris HCl [pH 8.0], 150 mM NaCl, 2 mM EDTA [pH 8.0], 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS) for 12 h at 4°C. This process specifically immunoprecipitated the designated transcription factor bound to sheared DNA. Immunoprecipitates containing sheared DNA fixed to the designated transcription factor were collected using Dynabeads protein A beads (Life Technologies) and washed extensively with buffer C (20 mM Tris-HCl, pH 8.0, 500 mM NaCl, 2 mM EDTA, 1% Triton X-100, 0.1% SDS). Samples were extracted twice with phenol-chloroform-isoamyl alcohol to remove proteins associated with sheared DNA bound to a specific transcription factor. The EP-328 primers used for this study are 5′-GCCCCCCCCCAAAAACAC-3′ and 5′-CAAGGCGAAACCCCCCAC-3′. The locations of these primers are shown in Fig. 2A, and the amplified product is 130 bp. The EP-943 primers used were previously described (21, 22, 37, 68) (5′-TCCGCCCCCCCCCAAAAAC-3′ and 5′-GAAACCCCCCACGCAAGGC-3′). These primers amplify the bICP0 E promoter region spanning sequences between nucleotides 638 and 943 and yield a 127-bp fragment (see Fig. 6A for the region of EP-943 that is amplified). DNA bands were quantified using Image Lab software and are presented as percent input. The input samples represented approximately 2% of the cell lysate.