Transcription of hepatitis delta virus RNA by RNA polymerase II

Abstract

Previous studies have indicated that the replication of the RNA genome of hepatitis delta virus (HDV) involves redirection of RNA polymerase II (Pol II), a host enzyme that normally uses DNA as a template. However, there has been some controversy about whether in one part of this HDV RNA transcription, a polymerase other than Pol II is involved. The present study applied a recently described cell system (293-HDV) of tetracycline-inducible HDV RNA replication to provide new data regarding the involvement of host polymerases in HDV transcription. The data generated with a nuclear run-on assay demonstrated that synthesis not only of genomic RNA but also of its complement, the antigenome, could be inhibited by low concentrations of amanitin specific for Pol II transcription. Subsequent studies used immunoprecipitation and rate-zonal sedimentation of nuclear extracts together with double immunostaining of 293-HDV cells, in order to examine the associations between Pol II and HDV RNAs, as well as the small delta antigen, an HDV-encoded protein known to be essential for replication. Findings include evidence that HDV replication is somehow able to direct the available delta antigen to sites in the nucleoplasm, almost exclusively colocalized with Pol II in what others have described as transcription factories.

FIG. 1.

Outline of nuclear run-on assay. HDV replication was initiated in 293-HDV cells by induction with TET for 16 h. Nuclei were then extracted followed by in vitro nuclear run-on transcription in the presence of [³²P]UTP. To inhibit Pol II transcription, 1 μg of amanitin/ml was added to the in vitro reaction. Radioactively labeled run-on RNAs were extracted and hybridized to a membrane on which were immobilized nuclei acids complementary to 18SrRNA, actin mRNA, GAPDH mRNA, U6 RNA, and HDV genomic and antigenomic RNA. Bound ³²P was quantitated by using a Bioimager.

FIG. 2.

Summary of nuclear run-on assays. A representative slot blot hybridization is shown at the left, without or with 1 μg of amanitin/ml added during in vitro transcription, as indicated. The transcripts on the membrane are listed in the first column of the table. The second panel shows the percent resistance to amanitin. This refers to the residual signals in the presence of amanitin relative to untreated controls. These are mean values, with standard deviations, and the number of independent experiments is indicated in parentheses. The third column indicates the host RNA polymerase used in transcription.

FIG. 3.

Association between Pol II and HDV RNAs detected by using IP assay. Cultures of 293-HDV cells after 16 h of TET induction, either with or without formaldehyde cross-linking, as indicated, were used to prepare nuclear extracts for immunoselection. Monoclonal antibodies 4H8 and 8WG16 were used to select for Pol II. A rabbit polyclonal was used to select for δAg. RNAs from fractions that were bound (B) or unbound (u) by protein A beads were (treated to reverse cross-linking, if applicable before being) extracted, and subjected to Northern assay to detect antigenomic and genomic RNAs. The amount of unbound RNA analyzed was 1.25% relative to the bound RNA. After quantitation, we obtained the percentages bound, as indicated.

FIG. 4.

Association between δAg and Pol II transcription complex. Nuclear IPs were performed with rabbit antibody to δAg or, as a negative control, normal rabbit serum. Similarly, we used 4H8, a mouse monoclonal specific for Pol II or, as a negative control, no added antibody (4H8). Immune complexes were selected by using protein A-agarose and then released for gel electrophoresis and immunoblotting. The latter, as indicated at the left side, were developed using antibody to δAg and four antibodies that recognize different forms of Pol II. For each sample, four dilutions of the total unbound material were examined in parallel in order to help determine the indicated percentages of the detected protein that was selected by IP.

FIG. 5.

Rate-zonal sedimentation of nuclear extracts of HDV-293 and 293-δAg cells. Sedimentation was performed on gradients of 10 to 30% sucrose in STE. Fractions were collected and assayed by immunoblot for proteins and qPCR for nucleic acids. The results are expressed in arbitrary units. (A and B) Profiles for SVP, HDV, HBV, and HDV ribonucleoprotein; (C to G) profiles for the indicated nuclear extracts, prepared largely as described for Fig. 4, with Pol II represented by circles and δAg represented by shaded squares. While all extracts underwent a prior DNase treatment, those in panels D and G were also treated with RNase.

FIG. 6.

Immunostaining of HDV-293 and 293-δAg cells. After fixation cells were subjected to double immunostaining as indicated above panels A and B. Confocal images are shown along with the indicated merges, and the separate staining was done with DAPI. In all cases, the detected staining was nuclear, with partitions between the nucleoplasm and nucleoli.