Comparative study of regulation of RTA-responsive genes in Kaposi's sarcoma-associated herpesvirus/human herpesvirus 8

Abstract

Replication and transcription activator (RTA) (also referred to as ORF50), an immediate-early gene product of Kaposi's sarcoma-associated herpesvirus (KSHV)/(human herpesvirus 8), plays a critical role in balancing the viral life cycle between latency and lytic replication. RTA has been shown to act as a strong transcription activator for several downstream genes of KSHV. Direct binding of RTA to DNA is thought to be one of the important mechanisms for transactivation of target genes, while indirect mechanisms are also implicated in RTA transactivation of certain selected genes. This study demonstrated direct binding of the DNA-binding domain of RTA (Rdbd) to a Kaposin (Kpsn) promoter sequence, which is highly homologous to the RTA-responsive element (RRE) of the PAN promoter. We undertook a comparative study of the RREs of PAN RNA, ORF57, vIL-6, and Kpsn to understand how RTA regulates gene expression during lytic replication. Comparing RNA abundance and transcription initiation rates of these RTA target genes in virus-infected cells suggested that the transcription initiation rate of the promoters is a major determinant of viral gene expression, rather than stability of the transcripts. RTA-mediated transactivation of reporters containing each RRE showed that their promoter strengths in a transient-transfection system were comparable to their transcription rates during reactivation. Moreover, our electrophoretic mobility shift assays of each RRE demonstrated that the highly purified Rdbd protein directly bound to the RREs. Based on these results, we conclude that direct binding of RTA to these target sequences contributes to their gene expression to various extents during the lytic life cycle of KSHV.

FIG. 1.

RTA DNA-binding domain protein (Rdbd) binds a Kpsn promoter sequence which is highly homologous to the RRE of the PAN promoter. (A) Comparison of the pPAN RRE and the homologous Kpsn promoter sequence. The pPAN RRE (PAN) shares significant homology to the Kpsn promoter sequence, and matched sequences are aligned. MJmulti is a mutant version of pPAN RRE with 5 nt mismatched (44), as indicated with arrows. For Kpsn/TG, a mutation of Kpsn (CC→TG) was introduced to further liken Kpsn to pPAN RRE. (B) EMSA of PAN*, MJmulti*, Kpsn*, and Kpsn/TG*. End-labeled probes were incubated with 0, 20, or 50 ng of Rdbd, as indicated. Rdbd was expressed in bacteria with a FLAG peptide at the N terminus and 6× histidine residues at the C terminus and purified using a Ni⁺-nitrilotriacetic acid column. (C) Quantitative analysis of Rdbd binding. Rdbd binding affinities to Kpsn* and Kpsn/TG* were calculated relative to those for PAN*. The values represent averages of relative binding affinities from three independent EMSAs, with the standard deviation shown. A diamond symbol (♦) indicates statistically significant difference in Rdbd binding between Kpsn* and Kpsn/TG* (P < 0.05, t test).

FIG. 2.

Transcript abundance and transcription rates of RTA-responsive genes in KS-1 cells during lytic replication. (A) Abundance of RTA-responsive genes during lytic replication. Total RNA was extracted from KS-1 cells untreated (−) or treated (+) with 3 mM sodium butyrate at 18 h postinduction and denatured with formaldehyde. In addition to 5 μg of RNA from KS-1 cells, different amounts of denatured DNA corresponding to the gene of interest were dotted onto nylon membranes and served as the standard (STD). The membranes were subjected to hybridization with gene-specific probes, including PAN RNA, Kpsn, ORF57, and vIL-6. GAPDH served as a control. (B) Nuclear run-on assays of KS-1 cells untreated (−) or treated (+) with 3 mM sodium butyrate. Nuclei were isolated from uninduced and induced KS-1 cells at 18 h postinduction and incubated with NTPs and [α-³²P]UTP. Nylon membranes containing 5 μg of indicated plasmids, PAN, Kpsn, ORF57, vIL-6, and GAPDH (see Materials and Methods), were hybridized with nuclear run-on products. Equal amounts of run-on products (10⁷ cpm) were added to each membrane. Representative images of the hybridized membranes are shown. (C) Quantitative analysis of nuclear run-on assays in uninduced (left panel) and induced (right panel) KS-1 cells. The intensity of a signal represents the amount of labeled nascent transcripts, which reflects the transcription initiation rate of the gene. The signal for GAPDH in uninduced KS-1 cells was set as 1, and the signals for other genes were analyzed relative to the value of GAPDH.

FIG. 3.

Expression of RTA-responsive genes in KS-1 cells transfected with RTA. Total RNA was extracted from KS-1 cells transfected with pcDNA3 vector alone (Vec) or with pcDNA3/RTA (RTA) at 40 h posttransfection and denatured with formaldehyde. In addition to 5 μg of RNA from transfected KS-1 cells, different amounts of denatured DNA corresponding to the gene of interest were dotted onto nylon membranes and served as the standard (STD). The membranes were subjected to hybridization with gene-specific probes, as described in the legend to FIG. 2.

FIG. 4.

Promoter strengths of reporters containing each RRE under a heterologous promoter, the E4 TATA box. (A) Schematic representation of a reporter construct. One copy of each RRE was cloned into a reporter construct containing the adenovirus E4 minimal TATA box. Reporter constructs were tested in 293T (B), BJAB (C), and KS-1 (D) cells, using a transient-transfection system. At 24 h posttransfection, transfected cells were harvested and subjected to dual luciferase assays. Promoter activities were calculated from levels of luciferase activity of reporters transfected with pcDNA3 or pcDNA3/RTA, in addition to a control vector, pRLSV40, levels of which constitutively expresses Renilla luciferase. Relative to that of pE4T/PAN, the promoter activities of the other pE4T/RREs were calculated. The values represent averages of at least three independent transfections, with the standard deviation shown.

FIG. 5.

Promoter strengths of reporters containing each RRE under another heterologous promoter, the SV40 promoter. (A) Schematic representation of a reporter construct. One copy of each RRE was cloned into a reporter construct containing the SV40 promoter. (B) Reporter constructs were transiently transfected into 293 cells, and levels of luciferase activity were assayed as described in the legend to FIG. 4.

FIG. 6.

Rdbd binds with various affinities to the RREs of PAN, Kpsn, ORF57, and vIL-6 promoters. (A) Sequences of the RREs from the promoters of PAN, Kpsn, ORF57, and vIL-6 of KSHV. Numbers indicate the locations of each RRE in the KSHV genome. (B) Dose-dependent binding of Rdbd to the RREs. Each probe contained an RRE with common flanking sequences at both ends. Increasing amounts of Rdbd (0, 50, 150, and 500 ng) were incubated with end-labeled probes, PAN* (lanes 1 to 4), Kpsn* (lanes 5 to 8), ORF57* (lanes 9 to 12), and vIL-6* (lanes 13 to 16), respectively. Arrows indicate Rdbd-specific binding.

FIG. 7.

Cross-competition assays confirm relative affinities of Rdbd binding to the RREs. (A) Cross-competition assays between PAN and ORF57 RREs. Rdbd (10 ng) was incubated without (lane 2) or with excess unlabeled oligonucleotide, PAN (5- and 50-fold; lanes 3 and 4), or ORF57 (50-, 500-, and 1,000-fold; lanes 5 to 7), for competition assays prior to addition of PAN*. Indicated amounts of unlabeled oligonucleotide, PAN (0.1-, 0.5-, and 5-fold; lanes 9 to 11) or ORF57 (fivefold; lane 12), were mixed with 1 μg of Rdbd, followed by addition of ORF57*. Note that the amounts of the unlabeled oligonucleotide PAN used to compete for Rdbd binding were far less than those of the unlabeled oligonucleotide ORF57. (B) Cross-competition assays of PAN* (10 ng of Rdbd) were performed with a set of unlabeled oligonucleotides (PAN, Kpsn, ORF57, and vIL-6). As indicated, different amounts of cold competitors were used to produce similar levels of competition. Unlabeled oligonucleotides were preincubated with Rdbd, and an end-labeled probe, PAN* or Kpsn*, was added (lanes 2 to 9). A monoclonal antibody against the FLAG peptide was used to supershift protein-DNA complexes containing Rdbd (lanes 10). (C) Cross-competition assay of Kpsn* (50 ng of Rdbd) was performed as described for panel B. (D) Specificity of Rdbd binding to the RRE. The labeled PAN* probe were incubated in excess of unlabeled oligonucleotides NS1 and NS2, both of which contain nonspecific sequences, as negative controls. Coincubation of PAN* with unlabeled oligonucleotides containing its own sequence (PAN) resulted in efficient competition of the complex formation (lanes 2 and 3), while NS1 and NS2 did not show any significant competition (lanes 4 to 7). Supershift assays were performed with anti-FLAG antibody and polyclonal rabbit sera against RTA (lanes 9 and 10). Normal rabbit sera and polyclonal rabbit sera against irrelevant protein (lanes 11 and 12) were also included as negative controls.