INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Hepadnaviruses are a group of small, enveloped DNA viruses of which hepatitis B virus (HBV) is the prototype. Although the mature virus contains a circular partially double-stranded 3.2-kb DNA genome, hepadnaviruses replicate solely through reverse transcription from a pregenomic RNA (pgRNA) intermediate within cytoplasmic nucleocapsids (2, 26).

A stem-loop structure that resides near the 5' end of pgRNA is the primary element of the hepadnavirus RNA packaging signal () (4, 7, 11, 13, 21) and serves as the origin for reverse transcription (19, 23, 27, 29). The viral polymerase initiates reverse transcription from the bulge of the stem-loop and synthesizes a 3- to 4-nucleotide (nt) oligomer, which is covalently linked to the polymerase (1, 6, 19). This polymerase-linked oligomer functions as minus-strand DNA primer for the synthesis of minus-strand DNA (5, 22, 28). The polymerase-linked oligomer is then transferred to a complementary UUC motif within direct repeat 1 (DR1*) near the 3' end of the pgRNA, where minus-strand DNA is elongated (19, 22, 23, 27, 29).

Elongation of minus-strand DNA is accompanied by degradation of the pgRNA (17). The terminal 15- to 18-oligonucleotide stretch of the pgRNA is resistant to the RNase H activity of HBV polymerase when the reverse transcription proceeds to the 5' end of the RNA template (14). This short RNA is translocated to a complementary sequence in DR2 near the 5' end of the minus strand, where the plus-strand DNA is initiated. This process leads to the formation of the relaxed circular (RC) DNA genome (25). Part of the plus-strand RNA primer does not transfer to DR2 but initiates the synthesis of plus-strand DNA in situ. This process results in the formation of duplex linear (DL) DNA genome (25).

The mechanism of the translocation of the polymerase-linked oligomer to the primer acceptor site is poorly understood. Previous results showed that complementarity between the polymerase-linked oligomer and the DR1* is required for the transfer to occur (19). Sequence analyses revealed that the adjacent sequence of DR1* contains several copies of the UUC motif as well as DR2. However, the polymerase-linked oligomer does not transfer to such a UUC motif, indicating that the UUC motif alone is not sufficient for polymerase-linked oligomer translocation. In mutants in which the complementarity between the polymerase-linked oligomer and DR1* have been destroyed, the polymerase-linked oligomer still can be transferred to the location of the altered DR1* but not the other UUC motifs (16, 19). Additionally, deletion of DR1* in woodchuck hepatitis virus can lead to the initiation of minus-strand DNA synthesis at an internal site (24). These results strongly suggested that a well-controlled mechanism beyond complementarity may exist to control the specificity of polymerase-linked oligomer transfer. These results also raise the interesting question whether the distance between the priming site (bulge site on ) and the DR1* in the pgRNA is important for polymerase-linked oligomer translocation. To address this issue, a series of mutants with insertions of various lengths of foreign sequence into different sites on the HBV genome were constructed to explore the effects of the altered distance between the priming site and DR1*. Our data show that the polymerase-linked oligomer was transferred to multiple internal acceptor sites in various insertion mutants. Among them, nt 2091 to 2093 (IAS2) is the major one used by the polymerase-linked oligomer. Remarkably, this event resulted in the production of restricted sizes of genome DNA, i.e., approximately 3.2 kb; thus, almost a unit length of HBV genome is maintained. The significance of this finding is discussed.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Plasmids. HBV mutants used in this report were derived from plasmid pMH-9/3091 subtype ayw (10). Plasmid pSHH2.1 and helper plasmids pMTP and pMT1883 were described previously (3, 4, 30). The HBV sequence was numbered according to the system of Pasek et al. (20), beginning with the A residue of the C gene initiation codon. For construction of plasmids X200, X400, X600, X825, and X1021, the SauI-HpaI (nt 236 to 438), AluI (nt 2250 to 2661), HpaI (nt 438 to 1062), EcoRV-AluI (nt 1125 to 1950), or PvuII-EcoRV (nt 104 to 1125) restriction fragments of the lacZ DNA of pCH110 (Pharmacia, Inc., Piscataway, N.J.) were cloned into the Klenow-filled-in XhoI site of pMH9/3091-m8. pMH-9/3091-m8 was a derivative of pMH-9/3091, which contained a created XhoI site at nt 37 on the HBV genome (3). To generate X1215, the 194-bp restriction fragment of HaeIII of the X174 DNA sequence was inserted into the Klenow-filled-in ClaI site located at the lacZ sequence of X1021. The EcoRV-AluI fragment of the lacZ DNA was inserted into three restriction sites on pMH-9/3091 (BspEI [nt 433], AvrII [nt 1461], and SpeI [nt 1962]) to generate B825, A825, and S825 insertion mutants with a positive orientation of the lacZ sequence. To generate plasmid X-GFP, a 745-bp BamHI-NotI (nt 661 to 1406) restriction fragment of plasmid pEGFP-N2 (Clontech, Palo Alto, Calif.) containing the green fluorescent protein gene sequence was inserted into the XhoI-restricted, Klenow-blunted vector pMH-9/3091-m8. The C2093G mutant carried a C-to-G mutation at nt 2093 (see Fig. 4A) and was generated by jumping PCR using plasmid X1021 as the template. The mutagenic primer was mHBV2498 (5'-CGCTGTTACCAATTTTGTTTTG [nt 2077 to 2098]). To construct ^m mutants, PCR fragments were amplified by a primer-carried C-to-G mutation in the priming site (5'-GTCCTACTGTTGAAGCCTCC [nt 3136 to 3155]; mutated base in bold). Then the amplified products were used to replace the corresponding region of plasmid X1021. ^m is similar to plasmid 1 published by Nassal and Rieger (19). Plasmid EV825 was constructed by the following procedure. An EcoRV site was first created by jumping PCR at a position just behind DR1*. The mutagenic primer used in the jumping PCR was mHBV3511 (5'-CTCTGCCTAATGATATCTTGTTC [nt 3111 to 3133]; EcoRV site in bold). The EcoRV-AluI restriction fragment (nt 1125 to 1950) of the lacZ DNA sequence was then inserted into this created EcoRV site on pMH-9/3091. Constructs harboring PCR products were confirmed by DNA sequencing.

Transient transfection. HuH-7 human hepatoma cells (18) were transfected by the calcium phosphate coprecipitation method as described previously (4). For cotransfection, 15 µg of HBV mutant plasmid was cotransfected with 15 µg each of plasmids pMTP and pMT1883 per 15-cm plate.

Isolation of viral core particles. The intracellular core particles and nucleic acid were purified as described previously (4). Core particles were immunoprecipitated with 10 µl of human anti-HBV core protein antiserum which was coated on protein A-agarose beads.

RNA preparation. Total cellular RNA was extracted with RNAzol B (Biotecx, Houston, Tex.) from day 2 posttransfection HuH-7 cells. For detection of encapsidated pgRNA, immunoprecipitated cores from the cytoplasm were treated with micrococcal nuclease for 30 min at 37°C as described previously (11). Then RNA was prepared by digestion with proteinase K (200 µg/ml in 1% sodium dodecyl sulfate) at 37°C for 1 h followed by phenol-chloroform extraction. After ethanol precipitation, nucleic acids were treated with DNase I for 20 min (11).

Detection of HBV nucleic acids. The endogenous polymerase assay (12) was performed as described previously (3) to detect HBV DNA of core particles. Southern and Northern blot analyses were performed as described previously (4). The SalI-SmaI HBV fragment containing the full-length HBV sequence from pMH-9/3091 was labeled by random priming (Promega Corp., Madison, Wis.) to serve as a probe.

Primer extension analysis. Primer extension analysis was carried out to detect the 5' end of the minus-strand DNA as described by Nassal and Rieger (19). The thermocycling parameters were 95°C for 1 min, 56°C for 1 min for primer KN-23 or 52°C for 1 min for primers HBV2414, HBV1771, and HBV3341, and 72°C for 1 min for 15 to 30 cycles. The extension products were mixed with loading buffer and subjected to electrophoresis in a 6% polyacrylamide sequencing gel. After being dried, the gel was autoradiographed at 70°C. Autoradiograms from Southern and Northern analyses were transformed to computer images using Adobe Photoshop version 5.0. Oligonucleotides for primer extension analyses correspond to nt 1933 to 1952 (KN-23), nt 2014 to 2035 (HBV2414), nt 1390 to 1371 (HBV1771), and nt 3041 to 3060 (HBV3341) on the HBV genome. Oligonucleotide sequence are sense strand (plus-strand polarity) except for HBV1771 (minus-strand polarity).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

The HBV genome with foreign DNA inserts affects the formation of RC form DNA. To explore if the insertion of foreign DNA into its genome affects HBV DNA replication, HBV mutants containing lacZ gene sequences, X825 and X1021, were generated. These insertion mutants produced pgRNAs of 4.1 and 4.3 kb, respectively (Fig. 1D, lanes 2 and 3), compared with the wild-type 3.3-kb pgRNA [not including the poly(A) sequence] (lane 1). Southern blot analysis was employed to monitor the viral genome by using cytoplasmic core particles produced from the HBV insertion mutants along with helper HBV genomes, pMT1883 and pMTP, which provided in trans all the viral proteins required for encapsidation. As shown in Fig. 1B, the wild-type HBV genome (pMH-9/3091) produced typical RC and DL DNAs (lane 1). However, mutants X825 and X1021 exhibited only a major band with a molecular size of approximately 3.0 to 3.2 kb (lanes 2 and 3), similar to that of the wild-type DL DNA product. No replication signals were detected when mutant genomes were transfected alone (data not shown). Furthermore, mutants that contain the green fluorescent protein-encoding gene at the same site also gave rise to the same replication pattern as mutants X825 and X1021 (data not shown), suggesting that the size of the insertion sequence, not the context of foreign sequences, contributes to this change.

View larger version (53K): [in this window] [in a new window]   FIG. 1.   Southern and Northern blot analyses of replicate products of HBV insertion mutants. (A) Schematic representation of restriction sites and genome organization of HBV. Four open reading frames of HBV are shown at the top, i.e. genes of core protein (C), surface protein (S), DNA polymerase (P), and X protein (X). The XhoI site was used to insert either 825 or 1,021 bp of the lacZ DNA as indicated. The cis elements required for HBV replication are located on the genome, including direct-repeat elements (DR1, DR1*, and DR2) and the RNA encapsidation signal sequence () located at the 5' end of pgRNA. The polymerase-linked oligomer priming site is located at the bulge region of . (B and C) Southern blot analysis of HBV nucleic acids. Cytoplasmic cores were isolated from HuH-7-transfected cell 5 days posttransfection. HBV DNAs which had been repaired using endogenous polymerase reaction with cold deoxynucleoside triphosphates (12) were isolated from core particles produced by HuH-7 cells transfected with various plasmids as indicated at the top of the figures. Undigested (lanes 1 to 3) or EcoRI-digested (lanes 4 to 6) DNA samples were separated by agarose gel electrophoresis (1.3% agarose), and transferred to a filter, hybridized with 32P-labeled full-length HBV DNA (B) or the whole gene of lacZ DNA (C). Panel C shows the recombinant HBV hybridization with lacZ probe after stripping off the HBV DNA probe. DNA size markers are indicated at the right. (D) Northern blot analysis of HBV transcripts. Total RNAs isolated from HuH-7 cells that were transfected with various plasmids as in panel A were separated through a formaldehyde denaturing gel, transferred to a filter, and hybridized with 32P-labeled full-length HBV DNA. pgRNA and surface (preS1/S2) transcripts are indicated.

Polymerase-linked oligomer transfers to internal novel acceptor sites in insertion mutants. The above-described data suggest that the major replication products of X825 and X1021 were DL DNA of approximately 3.0 to 3.2 kb even though the sizes of the pgRNAs are 4.1 or 4.3 kb. This result could be explained if the polymerase-linked oligomer is transferred not to DR1* but to a new internal sequence on the HBV genome. Analysis of the terminus of minus-strand DNA by primer extension, as shown in Fig. 2, revealed that the most 5' end of the minus-strand DNA of X825 and X1021 was mapped to nt 2093 and minor amounts of extended products ended at nt 2074 and nt 2158 (lanes 1 and 2). A trace amount of extended product that terminated at nt 2093 was also observed in wild-type genomes pMH-9/3091 and pSHH2.1 (lanes 3 and 4). The sites at 2074, 2093, and 2158 were named the internal acceptor sites IAS1, IAS2, and IAS3, respectively. Taken together, the results indicated that the linear DNAs produced by X825 and X1021 were elongated from IASs rather than DR1*. Furthermore, the length from the IAS to the EcoRI site is 0.8 kb, consistent with the EcoRI digestion of mutant genomes.

View larger version (52K): [in this window] [in a new window]   FIG. 2.   Determination of the 5' termini of minus-strand DNA of replicate products by primer extension. Core particle DNAs were prepared as described in the legend to Fig. 1A and hybridized with KN-23 oligonucleotide as indicated in Fig. 1A. A sequencing ladder primed with the same oligonucleotide using cloned HBV DNA as template was loaded in parallel to serve as a DNA marker. Arrows with nucleotide number at the right side indicate the 5' end of minus-strand DNA.

Translocation of the polymerase-linked oligomer to IAS2 in the insertion mutant is not dependent on insertion sites on the HBV genome. To examine whether other sites on the HBV genome of the insertion mutant lead to change in the primer acceptor site from DR1* to IAS, a fragment containing 825 bp of the lacZ gene sequence was inserted into BspEI, AvrII, and SpeI sites on the HBV genome to generate HBV mutants B825, A825, and S825, respectively (Fig. 3A). Northern blotting detected a 4-kb pgRNA in all mutants, as predicted (data not shown). The replication products of these mutants displayed DNA arrays similar to X825 as demonstrated by the endogenous polymerase assay (Fig. 3B). Primer extension analysis of the replicate products revealed that the 5' end of minus-strand DNA mapped primarily to nt 2093, which was similar to that obtained with mutants X825 and X1021 (Fig. 3C). The results indicate that the polymerase-linked oligomer transferring to IAS is not dependent on the insertion site on HBV genome.

View larger version (47K): [in this window] [in a new window]   FIG. 3.   Translocation of the polymerase-linked oligomer to IAS2 is independent of insertion sites on the HBV genome. (A) Schematic representations of the insertion constructs. Symbols are identical to those in Fig. 1, except that X825, B825, A825, S825, and EV825 stand for mutants with the 825-bp insertion at XhoI (X), BstEII (B), AvrII (A), SpeI (S), and EcoRV (EV), respectively. The triple vertical line indicates the position of IASs. (B and D) The performance of the endogenous polymerase assay is described in Materials and Methods. (C) Primer extension was done as described in the legend to Fig. 2, except that oligonucleotide HBV2414 was used as a primer for extension and sequencing. WT, wild type; NC, negative control.

Mutational analysis of the role of IAS2 in DNA replication. To confirm that IASs function as HBV minus-strand primer acceptor sites, a mutant (C2093G) of IAS2 in which C was changed to G at position 2093 in X1021 was constructed. This X1021 mutant resulted in a dramatic decrease (up to 90%) of replicate DNA content, as demonstrated by the endogenous polymerase assay (Fig. 4B, compare lane 2 with lane 1 [parental type]). Analysis of the 5' end of minus-strand DNA by primer extension revealed that the usage of IAS2 by the polymerase-linked oligomer was indeed abolished whereas the usage of IAS1 and IAS3 by the polymerase-linked oligomer was, at most, slightly affected by the IAS2 mutation (Fig. 4C, lane 2). This result indicates that base pairing between the polymerase-linked oligomer and IAS2 plays an important role in polymerase-linked oligomer translocation. Since the C2093G mutant did not demonstrate synthesis of other DNA species (Fig. 4B, lane 2), the results also clearly show that the IAS2 is the major polymerase-linked oligomer acceptor site for X1021. Northern blotting revealed that the quantity and quality of RNAs isolated from core particles produced by each mutant were similar, suggesting that the abolition of DL DNA in the C2093G mutant is not due to the failure of pgRNA encapsidation (Fig. 4D).

View larger version (55K): [in this window] [in a new window]   FIG. 4.   (A) Schematic representation of priming and polymerase-linked oligomer acceptor sites on pgRNA. The shaded oval tailing with the GAA trinucleotide represents the polymerase-linked oligomer; GAA is copied from UUC within the bulge of  as indicated. Nucleotide sequences within the rectangle indicate the sequence around the polymerase-linked oligomer acceptor site, and arrows with nucleotide number indicate the 5' end of minus-strand DNA. The distance from the priming site to the polymerase-linked oligomer acceptor is shown at the top in base pairs. (B and C) Mutation analysis of polymerase-linked oligomer translocation. Replicate DNAs were isolated from intracellular viral core particles produced by HuH-7 cells cotransfected with various constructs as indicated along with pMTP and pMH1883. The endogenous polymerase assay (B) and primer extension (C) were carried out as described in the legends to Fig. 1 and Fig. 2. (D) A portion of the core particles from panel B was prepared for the analysis of encapsidated pgRNA. The HBV sequence between nt 2089 and 2104 is shown at the left side of panel C.

Effects of insertion sizes on polymerase-linked oligomer translocation. To further understand the relevance of the length of the insertion sequence and the acceptor site selection of the polymerase-linked oligomer, various lengths of lacZ gene sequence were inserted into the XhoI site at nt 37 on the HBV genome to generate mutants X200, X400, X600, X825, X1021, and X1215. Several interesting results were obtained from this panel of insertion mutants. (i) Production of RC DNA in each mutant was seriously affected (Fig. 5A, lanes 2 to 7). Analysis of plus-strand DNA by primer extension indicated that the usage of DR2 by the plus-strand primer was very low in mutant X400 and undetectable in mutants with insertions larger than 625 bp (data not shown), a finding consistent with the disappearance of RC DNA in insertion mutants. (ii) The size of the upper DL (largest) DNA produced by insertion mutants increased in parallel with the length of the insertion (Fig. 5A) and the amount was dramatically reduced in mutants containing insertions larger than 825 bp. However, this DL DNA was gradually replaced by novel bands of approximately 2.6 to 3.4 kb as the insertion length increased (Fig. 5A). Primer extension analysis of the 5' end of minus-strand DNA indicated that the usage of DR1* by the polymerase-linked oligomer gradually decreased when the insertion size was increased (Fig. 5B). In contrast, the usage of IAS2 was increased in a parallel manner (Fig. 5B). In control experiments with helper plasmids (pMT1883 plus pMTP) alone, no primer extension products were detected (Fig. 5B, lane 8). The intensities of the primer extension products in Fig. 5B (lanes 1 to 8) were determined by amplification of templates with 15 cycles of one-way PCR; if the reaction was further subjected to another 15 cycles, the intensity (lanes 10 and 12) was twice that of the value obtained after the first 15 cycles (lanes 9 and 11), indicating that the amounts of input primer still can quantify the amounts of template within 15 to 30 cycles under our experimental conditions. (iii) The RC and DL DNA were demonstrated by digesting repaired DNA with EcoRI. As shown in Fig. 5A (lanes 9 to 16), RC DNA migrated down to the position of the upper DL DNA (lanes 9 to 11) and the lower DL DNA produced two fragments smaller than those of undigested DL DNA. One of these fragments, the 0.8-kb DNA fragment, was excised from the EcoRI site to IAS2 (lanes 11 to 15); therefore, all mutants generated this DNA fragment. This result, together with data obtained from primer extension analysis (Fig. 5B), strongly suggested that the upper DL DNAs were elongated from DR1* and the lower DL DNAs were elongated from IAS2. We also noticed that the additional band increased its size in parallel with the insertion length (Fig. 5A). These bands were resistant to EcoRI digestion and therefore might represent a replication intermediate or DNA species that lacks EcoRI site.

View larger version (39K): [in this window] [in a new window]   FIG. 5.   Size effects of polymerase-linked oligomer translocation. (A) The preparation of replicate DNAs of intracellular viral core particles and the performance of Southern blot analysis are the same as in Fig. 1B. *, largest DL DNA produced by insertion mutants; , novel bands produced as the insertion length increased; * (lanes 9 to 11), RC DNA migrating at the DL DNA position; , additional band produced by lower DL DNA. NC, negative control. (B) Primer extension analysis of the 5' end of minus-strand DNA. Viral DNA was digested with EcoRI prior to primer extension. The primers used for primer extension are HBV3341 (for mapping DR1*), HBV2414 (for mapping IAS2), and HBV1771 (for detecting the viral template containing the EcoRI site). A sequencing ladder with cloned HBV DNA as the template primed by the DNA oligomer HBV3341 (left) or HBV2414 (right) was loaded in parallel to serve as the DNA marker. A pilot test was done to determine suitable amounts of DNA template to obtain a similar signal primed by the HBV1771 primer. (C) Frequency of DR1* and IAS2 minus-strand primer acceptor site used by the polymerase-linked oligomer in each individual insertion mutant. The intensities of primer extension products were normalized based on equal amounts of minus-strand DNA derived from the HBV1771 primer. The distances from priming site () to IAS2 and to DR1* are 2.13 and 3.15 kb, respectively, in wild-type pgRNA.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Our results show that the polymerase-linked oligomer was transferred to multiple novel acceptor sites when foreign sequences were inserted into sites between  and DR1* in the intact HBV genome. These identified IASs, which are located approximately 2 kb from the priming site (bulge region of ) on the 5' end of wild-type pgRNA, contain a stretch of poly(U) sequence at the polymerase-linked oligomer binding sequence. They function as the acceptor site for polymerase-linked oligomer similar to that of DR1*. These IASs were also utilized by the wild-type genome although at a very low frequency. However, the utility of these IASs become a major one in our insertion mutants.

The mechanisms of polymerase-linked oligomer translocation to the DR1* acceptor site are largely unknown. It has been proposed that the 5' and 3' ends of the RNA template are juxtaposed within the capsid in a situation that allows efficient strand transfer to proceed (16, 19, 23). Our data presented here may provide a clue to whether this does occur. The results obtained from the experiment in Fig. 5 indicate that the polymerase-linked oligomer could translocate to either DR1* or IAS2 or both depending on the insertion size. In smaller or larger insertion mutants, only one site is preferentially selected by the polymerase-linked oligomer. For example, DR1* is used primarily in X200 and IAS2 is used in insertion mutants with insertions larger than 825 bp, which makes the distance from the priming site to the acceptor site in those mutants approximately 3.1 ± 0.2 kb. Interestingly, in the intermediate-insertion mutants (X400 and X600), both sites are utilized by polymerase-linked oligomer, which makes the distance from IAS2 less than 3.1 kb and that from DR1* greater than 3.1 kb. How could this happen? One explanation is that the priming site may juxtapose with a region on pgRNA within the nucleocapsid and the polymerase-linked oligomer may have the ability to scan the appropriate acceptor site bidirectionally. In other experiments, we showed that the polymerase-linked oligomer produced from the ^m mutant does not match IAS2. Interestingly, our results also show that this mutated polymerase-linked oligomer was translocated to a region near IAS2 (at nt 2098) with a perfect sequence complementary to it (Fig. 4C). Therefore, this result supports the idea that the polymerase-linked oligomer may have the ability to scan appropriate acceptor sites. These results, together with the fact that the IAS2 mutation (C2093G) resulted in a loss of the function of the polymerase-linked oligomer acceptor site, are consistent with the hypothesis that (i) complementarity between the polymerase-linked oligomer and the primer acceptor site is required but not sufficient for primer translocation and (ii) the priming site and IASs of pgRNAs may be juxtaposed with each other within core particles of insertion mutants in order to facilitate primer translocation. The latter hypothesis is further supported by the result that sequence inserted downstream of DR1* did not affect the transfer of the polymerase-linked oligomer to DR1* (Fig. 3D).

Loss of RC DNA in our insertion mutants with insertion sizes greater than 400 bp could be caused by one of two possibilities: (i) minus- or plus-strand primer transfers to an incorrect accepter site or (ii) lack of a terminally redundant sequence on pgRNA. Two types of DL DNA were detected in these mutants. Our results clearly demonstrated that type II DNA were generated by transferring the polymerase-linked oligomer to IASs. This kind of aberrant translocation leads to both a lack of the DR2 sequence in minus-strand DNA and a loss of terminal redundancy, both of which are required for RC DNA formation (15). Our results also showed that type I minus-strand DNA (elongated from DR1*) genomes retain the sequence determinants required for RC DNA formation, i.e., DR2 and terminal redundancy. However, the production of RC DNA in these insertion mutants was hardly detected. At present, we do not know why minus-strand DNA initiated from DR1* could not generate RC DNA in our insertion mutant (Fig. 5A).

On the basis of these discussions, a model is proposed, as shown in Fig. 6, to account for each synthesized DNA product in various insertion mutants as well as in the wild-type genome. Mutants are classified into four classes according to which region or cis element juxtaposes with the priming site within core particles. In class I mutants, the priming site and DR1* are juxtaposed with each other as depicted in Fig. 6a. Minus-strand DNA was elongated primarily from DR1*; thus, the production of RC and DL DNA resembled that in the wild-type genome. EV825 belongs to this phenotype. Class II mutants are those in which the region to be juxtaposed with the priming site is located between DR1* and IAS2, as shown in Fig. 6b. To elongate their minus-strand DNA, the polymerase-linked oligomer may have the ability to scan acceptor sites bidirectionally in these mutants. If the polymerase-linked oligomer is elongated from DR1*, type I DL DNA or/and RC DNA was produced, as in the X200, X400, and X600 mutants. If the polymerase-linked oligomer is elongated from IAS2, type II DL DNA were formed, as in the X400, X600, and X825 mutants. As shown in Fig. 6f, following translocation of the nascent polymerase-linked oligomer from the priming site to the IAS on pgRNA, minus-strand DNA is elongated from the 3' end to the 5' end along pgRNA that started at IAS. Elongation is concomitant with degrading pgRNA by the viral RNase H (17); thus, the 3'-end RNA of IAS is removed from pgRNA [indicated by the RNA fragment with the poly(A) sequence in Fig. 6f]. The polymerase-linked oligomer was not transferred to DR1* of pgRNA, which results in the loss of DR2 on minus-strand DNA. In class III mutants, the priming site is juxtaposed with IAS2, as depicted in Fig. 6c. An example is mutant X1021, in which only type II DL DNA was produced. Class IV mutants resemble class III mutants in terms of production of their DNA phenotype, except that IAS2 is not juxtaposed with priming site. As depicted in Fig. 6d, the polymerase-linked oligomer must back-scan to IAS2 in order for the minus-strand DNA elongation to occur. This model is consistent with our data presented. Therefore, our data also provide indirect evidence that the priming site and DR1* may be juxtaposed with each other within core particles in the wild-type genome. It has been reported that several cellular proteins such as the chaperone complex of heat shock protein 90 and p23 were copackaged and interacted with pgRNA inside core particles (8, 9). These cellular proteins may be involved in maintaining pgRNA in such a particular structure to facilitate polymerase-linked oligomer translocation. Alternatively, pgRNA may organize into a particular structure within core particles, which brings the priming site and DR1* or IAS (in insertion mutants) together to facilitate translocation. These two possibilities may not be mutually exclusive. Finally, our result also shows that IAS2 can be utilized in wild-type pgRNA during polymerase-linked oligomer translocation. Consequently, this process will also contribute to genome heterogeneity or production of defective virus.

View larger version (18K): [in this window] [in a new window]   FIG. 6.   Proposed structure model for minus-strand primer translocation and conversion to replicate DNA products. The bold line in path a to d indicates insertion sequences at different sites or of various sizes. Boxes labeled 1', 2' and IAS' represent DR1, DR2, and IAS on minus-strand DNA. The diagram was grouped into four pathways according to which region or cis elements of pgRNA juxtapose with the priming site. In path a, the priming site is juxtaposed with DR1*; in path b, the priming site is juxtaposed with a region between DR1* and IAS depending on the insertion size; in path c, the priming site is juxtaposed with IAS; in path d, the priming site is juxtaposed with the upstream region of IAS. The elongation of minus-strand DNA initiated at DR1* is further grouped into path e, while DNA elongated at IAS is grouped into path f. The RNA fragment with the poly(A) tail shown in path f represents the 3' end of RNA removed from pgRNA during elongation. For details, see the text.