JOURNAL OF VIROLOGY, July 2006, p. 6469–6477 0022-538X/06/$08.00⫹0 doi:10.1128/JVI.00245-06 Copyright © 2006, American Society for Microbiology. All Rights Reserved.
Vol. 80, No. 13
Genetic Changes in Hepatitis Delta Virus from Acutely and Chronically Infected Woodchucks John L. Casey,1,2* Bud C. Tennant,3 and John L. Gerin2 Department of Microbiology and Immunology, Georgetown University, Washington, D.C. 200571; Division of Molecular Virology and Immunology, Georgetown University Medical Center, Rockville, Maryland 208502; and Department of Clinical Sciences, New York State College of Veterinary Medicine, Cornell University, Ithaca, New York3 Received 2 February 2006/Accepted 12 April 2006
A woodchuck-derived hepatitis delta virus (HDV) inoculum was created by transfection of a genotype I HDV cDNA clone directly into the liver of a woodchuck that was chronically infected with woodchuck hepatitis virus. All woodchucks receiving this inoculum became positive for HDV RNA in serum, and 67% became chronically infected, similar to the rate of chronic HDV infection in humans. Analysis of HDV sequences obtained at 73 weeks postinfection indicated that changes had occurred at a rate of 0.5% per year; many of these modifications were consistent with editing by host RNA adenosine deaminase. The appearance of sequence changes, which were not evenly distributed on the genome, was correlated with the course of HDV infection. A limited number of modifications occurred in the consensus sequence of the viral genome that altered the sequence of the hepatitis delta antigen (HDAg). All chronically infected animals examined exhibited these changes 73 weeks following infection, but at earlier times, only one of the HDV carriers exhibited consensus sequence substitutions. On the other hand, sequence modifications in animals that eventually recovered from HDV infection were apparent after 27 weeks. The data are consistent with a model in which HDV sequence changes are selected by host immune responses. Chronic HDV infection in woodchucks may result from a delayed and weak immune response that is limited to a small number of epitopes on HDAg. studies did not address whether genetic changes in the virus could be important for the establishment of chronic infection. In this study, we analyze genetic changes occurring in the HDV genome following the infection of woodchucks with an inoculum derived from a molecular clone of HDV genotype I. We found that a limited number of sequence changes occur during chronic infection. Some of these modifications alter the consensus sequence of hepatitis delta antigen (HDAg) and could indicate a mechanism for avoidance of host immune responses.
Hepatitis delta virus (HDV) is an obligate subviral satellite of hepatitis B virus (HBV) that causes both acute and chronic liver disease in humans (13). The HDV virion is composed of a ribonucleoprotein core and an envelope formed by the surface antigen protein (HBsAg) of HBV (3–5). The envelope is the sole helper function provided by HBV. HDV is able to replicate within cells in the absence of HBV (17) but requires HBsAg for the packaging and release of HDV virions (29, 33) as well as for infectivity (32). In addition to the envelope of its natural helper HBV, HDV can be packaged by the surface antigen of related hepadnaviruses, including woodchuck hepatitis virus (WHV) (27, 29) and woolly monkey hepatitis B virus (18), and can produce both acute and chronic infection in woodchucks that are chronically infected with WHV (20, 27, 30). There are two modes of HDV infection: superinfection of an individual with chronic HBV infection and coinfection with both viruses of an individual who has not been exposed previously to HBV. While the latter type of exposure typically leads to recovery from both viral infections, the former results in chronic HDV infection with a frequency of 70% to 80% (13). The determinants of the outcome of HDV superinfection are not known. Previous studies of viral genetic changes during infection have analyzed HDV RNA obtained either from the sera of chronically infected patients during the course of HDV infection (7, 14, 19) or from the liver of a woodchuck infected with virus that had been passaged several times (21). These
MATERIALS AND METHODS Plasmid construction. The plasmid construct pGDLx1.2 was created by inserting a 1.26-fold-overlength HDV cDNA obtained from a patient infected with HDV genotype I (GenBank accession no. L22066) between the PstI and SphI sites of pGEM3Zf(⫹). Sequences from positions 654 to 1092, which contain both genomic and antigenomic ribozymes, are duplicated in this construct. Transfection of a woodchuck with HDV cDNA. All animal experiments for the current study were conducted under protocols reviewed and approved by the Cornell University Institutional Animal Care and Use Committee. Woodchuck 4928, a chronic WHV carrier, was transfected with 50 g of the HDV construct pGDLx1.2 in a total volume of 0.5 ml of phosphate-buffered saline by direct injection at several separate sites in the surgically exposed liver. Sera were obtained weekly and analyzed by reverse transcription (RT) and semiquantitative PCR for HDV RNA, as described previously (22, 31). Highly positive sera from weeks 7 to 15 were combined to create the woodchuck hepatitis delta virus type 1 (WHDV-1) pool. This serum pool contained 1 ⫻ 109 genome equivalents of HDV RNA per ml, as determined by blot hybridization. Infection of WHV woodchucks with woodchuck-derived HDV. WHV carrier woodchucks were produced experimentally by administering the standardized cWHV8P1 inoculum (8) to neonatal woodchucks born to WHV-negative females. The WHV carrier status of woodchucks was verified by detection of persistent WHV viremia and antigenemia (8). Chronic WHV carrier woodchucks were infected at 18 months of age with woodchuck-derived HDV by intravenous administration of a 0.5-ml volume of a 1:5 dilution of the WHDV-1 pool. Serum samples were obtained weekly and analyzed for HDV RNA by reverse transcription and semiquantitative PCR (22, 31).
* Corresponding author. Mailing address: Department of Microbiology and Immunology, Georgetown University Medical Center, 3900 Reservoir Rd., NW, Washington, DC 20007. Phone: (202) 687-1052. Fax: (202) 687-1800. E-mail:
[email protected]. 6469
6470
CASEY ET AL.
J. VIROL.
TABLE 1. Oligonucleotide primers used for RT-PCR amplification Primer
5414 6657 3415A 6465
Sequence
5⬘-GAGATGCCATGCCGACCCGAAGAG-3⬘ 5⬘-CAGCAGTCTCCTCTTTACAGA-3⬘ 5⬘-TCTTGTTCTCGAGGGCCTTCCTTC-3⬘ 5⬘-GTTCGCTGAAGGGGTCCTCTGGAG-3⬘
Location (positions)
886–909 1660–1640 1267–1290 331–308
Sequence analysis of HDV RNA. Serum samples were subjected to reverse transcription, followed by PCR with one of two sets of primers (Table 1). Primer pair 5414 and 6657 amplifies a 775-bp product from positions 886 to 1660 (numbering according to GenBank accession no. L22066) that encompasses the entire HDAg coding region; primer pair 3415A and 6465 amplifies a 746-bp product that includes the amino-terminal part of the HDAg coding region and the noncoding portion of the genome that is base paired with it in the unbranched rod structure (Fig. 1). PCR amplification was performed using Pfu polymerase (Stratagene, La Jolla, CA) for 35 cycles. PCR products were either sequenced directly (MWG Biotech, High Point, NC) or cloned into the plasmid pZero (Invitrogen, Carlsbad, CA). When the products were cloned, two separate RT-PCRs were performed to minimize the potential of analyzing sibling clones. Changes at the amber/W site were not included in the sequence comparisons because this site is modified as part of the normal HDV replication cycle (25).
RESULTS Transfection of a WHV-infected woodchuck with a replicating HDV cDNA. In order to obtain a woodchuck-derived, infectious pool of HDV for study of the woodchuck model, the plasmid construct pGDLx1.2 was injected into the liver of woodchuck 4928, which was chronically infected with WHV. This construct contains a 1.26-fold-overlength HDV cDNA insert derived from a genotype I HDV isolate (GenBank accession no. L22066). Six weeks following injection, HDV RNA became detectable in the serum. Viremia peaked between weeks 7 and 15, then declined, and became undetectable at 21 weeks posttransfection. Serum samples obtained during the peak of viremia (weeks 7 to 15) were combined to create the WHDV-1 pool. The concentration of HDV in this pool was determined by Northern blot analysis to be 109 HDV genome equivalents/ml. Analysis of HDV genetic stability in the transfected woodchuck. To analyze the genetic diversity of the WHDV-1 pool, 11 clones of the HDAg coding region were obtained by RT-
FIG. 1. Schematic of the HDV RNA genome showing PCR primers and products analyzed. The flattened circular line represents the HDV genome, the open rectangle the HDAg gene, the filled triangle the location of the genomic ribozyme, and the open triangle the location of the antigenomic ribozyme. Arrows indicate the locations and orientations of primers, and the dashed lines show the PCR products.
TABLE 2. Number and types of sequence changes observed in cDNA clones No. (%) of sequence changes in clones a Nucleotide change
U3C A3G C3U G3A U3G G3C G3U A3U U3A Insertion Deletion All
HDAg coding region Inoculum
Wk 8c
Wk 73d
Noncoding region (wk 73)e
5 (0.23)b 12 (1.2) 0 0 1 (0.05) 0 0 0 0 0 0 18 (0.24)
5 (0.06) 5 (0.06) 0 2 (0.02) 0 0 0 0 0 0 0 12 (0.14)
77 (1.0) 52 (1.4) 29 (0.3) 7 (0.1) 8 (0.1) 8 (0.1) 6 (0.1) 2 (0.1) 2 (0.03) 0 0 191 (0.69)
0 4 (0.6) 3 (0.5) 0 0 0 1 (0.1) 1 (0.1) 0 10 (0.4) 3 (0.1) 22 (1.0)
a Inoculum, number of sequence changes compared with L22066; weeks 8 and 73, number of sequence changes compared with inoculum consensus. b Percentage of specific nucleotides changed. c Animals 4543 and 4569, six clones each. d Animals 4543, 4547, 4553, 4559, and 4569, eight clones each. e Animal 4569, six clones.
PCR, using primers 6657 and 5415 (Fig. 1), and then sequenced. Clones were obtained from two independent RTPCR amplifications to minimize the chances of obtaining sibling clones. Compared with the sequence of the transfected cDNA, a total of 18 sequence changes occurred in the 695nucleotide (nt) region sequenced, or 0.24% (Table 2). Remarkably, 11 of the 18 changes occurred at position 1008 in the genome, which was modified from A to G in all 11 clones. The change at this position alters the predicted HDAg sequence at amino acid 198 from isoleucine to threonine. This amino acid change has been described previously after multiple passages of HDV in woodchucks (9). The remaining seven nucleic acid sequence changes were observed at different positions in different clones. In similar analyses of cDNA clones obtained by RT-PCR, we observed that the error rate due to Pfu polymerase was less than 0.1% (26). Thus, while the modification at position 1008 is clearly not likely to be due to misincorporation during RT-PCR amplification, it is impossible to determine to what extent the other seven changes occurred during the course of infection or during reverse transcription and PCR amplification. Infection of woodchucks with woodchuck-derived HDV. Nine WHV carrier woodchucks were inoculated intravenously with 1 ⫻ 108 genome equivalents of woodchuck-derived HDV by using the WHDV-1 inoculum. By 3 weeks postinoculation, all nine animals became positive for HDV viremia, which reached maximum levels by 11 weeks postinoculation in all but one animal, woodchuck 4562 (Fig. 2). Peak viremia varied from approximately 108 to 1010 genome equivalents per ml. All animals seroconverted to anti-HDV and remained WHV positive throughout the course of the study (data not shown). Three animals (animals 4548, 4552, and 4566) lost detectable serum HDV by 37 weeks postinoculation and remained negative for the remaining 36 weeks. Five animals (animals 4543, 4547, 4553, 4559, and 4569) exhibited detectable viremia for more than 80% of bleed dates through 73 weeks postinoculation. We consider these five animals to be chronically infected
VOL. 80, 2006
HDV GENETIC CHANGES DURING ACUTE OR CHRONIC INFECTION
FIG. 2. Time course of HDV RNA detected in the sera of animals inoculated with WHDV-1. Sera were obtained weekly and analyzed for HDV RNA by reverse transcription and semiquantitative PCR (22). Quantitation was based on a comparison of PCR product yields with standard dilutions of an HDV source of known concentration.
with HDV and WHV. Of these five animals, all but animal 4559 were positive for HDV viremia for the entire post-acutephase period. Animal 4559 was negative for viremia at six times but was positive for six consecutive analyses from weeks 55 through 73 (Fig. 2). Animal 4562 exhibited a relatively late peak of viremia (at approximately 30 weeks postinoculation), and levels of viremia dropped and remained low or undetectable from weeks 35 to 73. Although viremia was sporadic and low in this animal, infection was likely to be chronic. Thus, six of nine (67%) animals became chronically infected
6471
with HDV, a rate of chronic infection similar to that observed in humans (13). Analysis of HDV genetic stability in chronically infected woodchucks. The HDAg coding region was successfully amplified from the sera drawn at 73 weeks postinoculation from five woodchucks chronically infected with HDV. Although weak viremia was detectable in animal 4562 by using a different set of PCR primers (22, 31), we were unable to obtain sufficient product from this animal by using primers that flank the entire HDAg coding region. The sequences of eight clones from each of the other five chronically infected animals were obtained and compared with each other and with the consensus sequence of the inoculum. The extent of sequence diversity in these 40 clones was considerably greater than that in the 11 clones derived from the WHDV-1 inoculum (Table 2). There were a total of 191 (0.69%) sequence differences between the 40 clones and the consensus sequence of the inoculum; these changes occurred at 83 different positions in the 695-nt region analyzed. Most positions were found to be modified in just one clone, but many were modified in numerous clones (data not shown). On average, there were 4.8 sequence changes per clone (range, 1 to 11). All 40 clones obtained from samples from week 73 contained the consensus A-to-G change observed at position 1008 in the inoculum. Of the other seven positions found to be changed in the 11 inoculum-derived clones, just two, nt 1358 and nt 1183, were also found to be modified in clones from week 73; position 1358 was changed in 10 clones and position 1138 in 2. In order to examine how rapidly sequence divergence occurred, six HDV cDNA clones were obtained at 8 weeks postinoculation from two of the animals (animals 4543 and 4569) that eventually became chronically infected. Analysis of the sequences of these 12 clones indicated a much lower degree of nucleotide substitution than at 73 weeks postinoculation (Table 2). Subtracting the frequency of sequence changes observed at 8 weeks from that at 73 weeks gives an overall rate of 4 ⫻ 10⫺3 substitutions per site per year in the HDAg coding region, similar to results previously reported for infected patients (7, 14, 19). Types of sequence changes. In the clones of HDV RNA from week 73, adenosine-to-guanosine changes occurred with the highest frequency (1.4% of adenosines modified), followed by uridine to cytosine (1% of uridines modified) (Table 2). Both of these modifications are consistent with editing by host RNA adenosine deaminase (ADAR) activity. ADAR deaminates adenosines in RNA to inosines, which are subsequently transcribed as guanosines; adenosines that are edited on the genome will appear as A-to-G changes, while those on the antigenome will appear as U-to-C changes on the genome. A-to-G and U-to-C transitions accounted for 67% (129/191) of the sequence substitutions observed in the clones from week 73. These were also the predominant sequence changes observed in the clones obtained during the peak of viremia (Table 2). The high relative frequency of these changes may indicate that ADAR is responsible for much of the genetic divergence that arises during the course of acute and chronic HDV infection. Previous analyses of HDV sequence divergence among different isolates have indicated that the noncoding region of the genome is, in general, more variable than the HDAg coding
6472
CASEY ET AL.
J. VIROL.
Notably, sequence modifications in the noncoding region included insertions and deletions (mostly single nucleotide), and these accounted for the difference in the rates of sequence divergence between the coding and noncoding regions analyzed. No insertions or deletions were found in any of the clones from the HDAg coding region (Table 2). Changes in the HDAg consensus sequence. Comparison of the HDAg sequence of the inoculum with those of the clones obtained at 73 weeks postinoculation from the five chronically infected animals indicated that the distribution of the modifications was not random (Fig. 3). In all five animals, at least one change occurred in the consensus HDAg coding sequence (defined here as the sequence in six or more of the eight clones) (Fig. 3 and Table 3). Most of these consensus changes were in the N-terminal region of the protein. In animal 4543, the consensus was altered at two positions; in the other four animals, there was a single modification in the consensus sequence (Table 3). The consensus changes observed in animals 4543 and 4547 were unique to those two animals. At least six of eight clones from woodchucks 4553, 4559, and 4569 exhibited a change in the consensus at nucleotide position 1480 that changed amino acid 41 from isoleucine to valine. In animals 4543 and 4569, the sequence changes were observed in all eight clones; however, in the other three woodchucks, only six of eight clones contained the identical sequence change. Remarkably, for these three animals, the two clones that were not altered at the position where the consensus changed each contained at least one additional change within 45 bp of the consensus that also changed the HDAg amino acid sequence (Fig. 4). Thus, in the clones from animal 4547 that did not exhibit the Q19-to-E change, serine 4 was changed to proline; in one clone, isoleucine 16 was also changed to valine. The two clones from woodchuck 4553 that retained I41 were modified at position 54 (I3V); one clone was also modified at amino acid 51 (L3M). Finally, the two clones from animal 4559 that were not modified at amino acid 41 exhibited I3T changes at position 37. As can be seen in Fig. 4, these modifications were unique to particular animals. Thus, for each of the chronically infected animals, the HDAg sequence was modified in all clones, either at a single position or at two or more positions within a 15-amino-acid segment (Fig. 4). Furthermore, while there were similarities in the sequence changes observed in some of the woodchucks, the overall changes in each animal were unique. Sequence changes that did not affect the amino acid sequence of HDAg. Ninety (47%) of the 191 sequence modifications in the clones from week 73 did not affect the HDAg
FIG. 3. Histogram of HDAg amino acid sequence changes in HDV cDNA clones obtained from chronically infected animals at 73 weeks postinfection. Sequence changes from the inoculum are indicated by vertical bars; bar heights indicate the number of clones changed at different positions. The dashed horizontal line indicates one clone.
region (22, 28). Moreover, studies of the evolution rate of HDV during the course of chronic infection have shown that sequence changes accumulate more rapidly in the noncoding region (7, 19). To compare the extent of sequence changes in the coding and noncoding regions, the region 1267 to 1681/1 to 308 was amplified with primers 3415A and 6465 (Fig. 1), using serum obtained at 73 weeks postinoculation from animal 4569. Analysis of six cDNA clones of this region indicated a rate of sequence divergence of 1 ⫻ 10⫺2 changes per nucleotide per year (Table 2). This rate was considerably higher than that observed in the coding region over the same period of time.
TABLE 3. Number of specific sequence changes observed in eight cDNA clones obtained at 73 weeks No. of indicated sequence changes in clones Animal
4543 4547 4553 4559 4569 a
1128U3C (159T3A)a
1480U3C (41I3V)
7 1 1
Nucleotide change (HDAg amino acid change).
1546G3C (19Q3E)
1556U3G (15E3D)
8 6 6 6 8
1358U3C
1617A3G
2 3 1
7 5 6 6 3
4
VOL. 80, 2006
HDV GENETIC CHANGES DURING ACUTE OR CHRONIC INFECTION
6473
to the positions that affected the HDAg sequence, changes at 1358 and 1617 appeared to be randomly distributed among clones from different animals (Table 3). Both sequence changes are consistent with activity of the host RNA adenosine deaminase ADAR1 or ADAR2; the U-to-C change at position 1358 is consistent with ADAR editing on the antigenome, and the A-to-G change at position 1617 is consistent with ADAR editing on the genome. Kinetics of sequence changes in animals that either recovered from HDV infection or became chronically infected. To examine the dynamics of sequence changes, RNA was isolated from sera obtained at weeks 8, 17, and 27 postinoculation. Products of RT-PCR amplifications were sequenced directly to determine the consensus sequences at these times. In the animals that eventually became chronically infected, no changes in the majority sequence of HDAg were observed at 8 weeks postinoculation and there was no evidence of minor populations (ca. 10 to 20%) with sequence differences at positions that changed at later times. By 27 weeks postinoculation, the majority sequence had changed in just one of the animals that eventually became chronically infected (animal 4543) (Table 4). In two others (animals 4553 and 4569), there was evidence of some sequence heterogeneity, but the abundance of the new sequence was 20% or less (Table 4). The evolution of sequence changes in woodchuck 4543 was particularly interesting. In all of the clones obtained from this animal at 73 weeks postinoculation, position 1556 was changed from U to G, which changes the HDAg sequence from glutamate to aspartate at amino acid 15. Direct sequencing of HDV RNA from serum obtained at 17 weeks postinoculation indicated that position 1556 was identical to that in the sequence of the original inoculum; however, position 1554 had changed such that the original A was replaced by G in a majority of the population (Fig. 5). This A-to-G modification substitutes threonine for isoleucine at amino acid 14. By week 27, position 1554 had reverted to the original sequence, but the U at position 1556 had been completely replaced by either A or G, which were present at roughly equal levels. Both of these substitutions change glutamate 15 to aspartate. By week 73, guanosine replaced the adenosine that was present at position 1556 in about half of the HDV population at week 27. At week 27 postinoculation, somewhat greater sequence
TABLE 4. Changes at week 27 Type of infection
Animal
Position
% Change
Resolved
4548
1458 1555 1556 1063 1555 1458 1460 1555 1556
⬎90 ⬎90 ⬎90 ⬃40 ⬎90 ⬃20 ⬃20 ⬃20 ⬃10
1556
⬃50
1480
⬃10
1555
⬃20
FIG. 4. Comparison of HDAg sequences in cDNA clones obtained from chronically infected animals at 73 weeks postinfection. The consensus sequence of the inoculum is shown at the top of each panel. Sequence changes are shown by single-letter amino acid codes; dots indicate no change.
coding sequence. Forty percent of these occurred at two positions, 1617 and 1358, which were among the sites most frequently modified. Position 1617, modified in 27 of 40 clones, is outside the HDAg coding region, and the U-to-C modification at position 1358, found in 10 of 40 clones, is silent. In contrast
4552 4556
Chronic
4543 4547 4553 4559 4569
6474
CASEY ET AL.
J. VIROL.
sequencing to what extent these changes were clustered on the same genome or distributed among different genomes. The HDAg amino acid sequence changes found at week 27 in the woodchucks that recovered from infection differed somewhat from the changes that occurred by week 73 in the chronically infected animals (Fig. 6). Only six different modifications were observed that changed the HDAg consensus sequence, and three of these (E15D, I16V, and I41V) were found in two or more animals. An I16V substitution occurred in two of the animals that recovered from infection, and I41V was common among three of the five chronically infected animals. On the other hand, the E15D modification was observed in both animal 4548, which recovered from HDV infection, and animal 4543, which became chronically infected. The E15D change was also detected at the 50% level in animal 4543 at 27 weeks postinfection (Table 4). Thus, it is not clear to what extent the apparent segregation of sequence changes between animals that became chronically infected and those that recovered from infection is due to the outcome of infection, to the timing of the sequence change, or to other factors. DISCUSSION
FIG. 5. Electropherogram of HDV RT-PCR products obtained at weeks 8, 17, and 27 postinfection from animal 4543. Black, G; green, A; blue, C; red, U. Positions 1554 and 1556 are shown by dashed lines. Note that because of the primer used for sequencing, the sequence shown is the reverse complement of the genome.
changes were evident in the animals that recovered from infection than in the animals that became chronically infected. The increased sequence alterations occurred either in the degree of sequence change at particular positions, in the number of positions modified, or both (Table 4). HDV RNA in animal 4548 exhibited three nearly complete (⬎90%) sequence changes at positions 1458, 1555, and 1556 (Table 4). In animal 4552, the sequence was nearly completely modified at position 1555, which also was altered in animal 4548; in addition, position 1063 was 40% modified in this woodchuck. While no majority sequence changes were observed in HDV RNA from animal 4566, partial modifications were detected at four sites, all of which were at or near positions found to be altered in other animals. All of these changes affect the amino acid sequence of HDAg. It is not possible to determine from direct
The use of the molecularly derived, woodchuck-adapted HDV inoculum described in this report afforded analysis of HDV genetic changes during the course of acute and chronic infection. U-to-C and A-to-G transitions dominated the changes that were observed, accounting for 67% of all sequence modifications. These substitutions are consistent with RNA editing by ADAR1, which specifically edits the antigenomic RNA at the amber/W site in an integral part of the HDV replication cycle. The occurrence of A-to-G transitions on the genome indicates that genomic RNA may also be susceptible to editing by ADAR1. In an analysis of sequence diversity occurring during the course of serial acute-phase passage of HDV in woodchucks, Netter et al. (21) observed higher rates of U-to-C and A-to-G sequence changes. Chang et al. (6) recently reported that transitions consistent with ADAR activity were the most frequently observed changes in a modified HDV genome that replicated in cultured cells for 1 year. Editing on the HDV RNA by ADAR1 is highly specific for the amber/W site. During a 2-week course of replication in cultured cells, non-amber/W sites were edited at very low levels (26). However, editing at non-amber/W sites is readily detectable in vitro, with some sites modified more efficiently than others (26). In the clones of the week 73 PCR products, most of the sites at which substitutions consistent with ADAR editing occurred were found to be modified in only a single clone (data not shown), but some, including those at which four of the six HDAg consensus sequence changes occurred, were found to be modified in many clones. The abundance of changes consistent with ADAR editing in the viral population during the course of infection is likely to have been determined by both the susceptibility of different sites to ADAR editing and any long-term selective growth advantage conferred by the sequence changes. Remarkably, a limited number of modifications in the consensus sequence of HDAg occurred in seven of eight animals infected with the molecularly derived inoculum (Fig. 4 and Table 4). Moreover, a consensus change was observed in virus
VOL. 80, 2006
HDV GENETIC CHANGES DURING ACUTE OR CHRONIC INFECTION
6475
FIG. 6. Consensus HDAg sequence changes in animals that either recovered from HDV infection or became chronically infected. Sequences for animals whose infection resolved were obtained by direct sequencing of RT-PCR products obtained from HDV RNA isolated at 27 weeks postinfection (Table 4); two consensus changes were observed for animal 4548 and none for animal 4556. Sequences for animals that became chronically infected were obtained by sequencing of eight clones of RT-PCR products obtained from HDV RNA isolated at 73 weeks postinfection (Fig. 3). Locations of sequence modifications are indicated in their approximate positions along the HDAg coding sequence.
isolated from the woodchuck transfected with the HDV cDNA clone, even though the overall rate of sequence divergence was low in clones isolated from this animal. A previous study of the genetic stability of HDV in woodchucks (21) reported no consensus sequence changes during serial passages over a total length of 253 days. However, only the C-terminal half of the HDAg coding region was analyzed in that study and some changes could have been missed. Indeed, most of the consensus changes that we observed were in the N-terminal half of the protein (Fig. 3). It is difficult to distinguish between the effects of positive and negative selective pressures on changes in the virus sequence. Most likely, both negative selective pressure due to host immune responses and positive selective advantages due to increased replication efficiency play a role. Indeed, there is evidence of both processes in our results. The dominant A-to-G change at position 1008 (amino acid 198) may be the result of adaptation of the virus to the woodchuck from the native human host. This change, which was observed in another passage of HDV to woodchucks (9), was observed in all 11 clones obtained during the acute phase of viremia in animal 4928 following transfection and was found in all sequences obtained from woodchucks that were subsequently infected. It thus appears that this change confers a significant replication advantage in woodchucks. The A1008G sequence modification replaces I198 with T in large HDAg (HDAg-L); I198 is among the additional amino acids present in HDAg-L that are required for particle formation with the hepadnavirus surface antigen. O’Malley and Lazinski found that mutation of I198 to T was fully tolerated for packaging of the mutant HDAg-L by HBV envelope proteins (24). Perhaps the I198-to-T change produces a more favorable interaction between HDAg-L and the WHV surface antigen. That this sequence change was not observed in a serial passage of HDV in WHV-infected woodchucks (21) might be explained by the use of different WHV strains. The results obtained for the woodchucks infected with the molecularly derived inoculum are consistent with a strong influence of host immune responses on the selection of viral genetic changes. The woodchucks used in this study were not
inbred; thus, somewhat diverse immune responses and subsequent evasive genetic changes in the viral sequence could be expected. Particular changes that affected the amino acid sequence were clearly segregated among different animals; however, two other frequent changes, at positions 1358 and 1617, that did not affect HDAg sequence appeared to be distributed randomly among clones from different animals (Table 3). While consensus changes at some HDAg amino acid positions in woodchucks 4543 and 4569 were due to modifications in all clones, the amino acid at the position changed in the consensus sequence was not modified in all clones in woodchucks 4547, 4553, and 4559 (Fig. 4). However, for these three woodchucks, all clones not containing modifications at the position changed in the consensus sequence were modified at one or more sites within 15 amino acids of this position (Fig. 4). Such changes, whether restricted to a single amino acid or clustered within a 15-amino-acid region, could readily affect the immune response by affecting the association of specific HDAg-derived peptide fragments with major histocompatibility complex class I or class II molecules (23). The evolution of sequence changes in animal 4543 is consistent with this interpretation. Initial changes in the HDAg consensus sequence at position 1554 in animal 4543 at 17 weeks postinoculation may indicate a virus-evasive reaction to immune pressure that either did not fully avoid the host immune response or imposed a replication disadvantage. Thus, the subsequent reversion at this position combined with the change at position 1556 would be consistent with either a more successful avoidance of host immune responses (in the former case) or a better combination of immune avoidance and viral replicative capacity (in the latter case). HDV sequence changes observed at 73 weeks postinoculation in the chronically infected animals predominated at a limited number of sites, most of which are in the N-terminal segment of the HDAg coding region (Fig. 3 and 4; Tables 3 and 4). Previous analyses of humoral antibody responses to HDV infection identified an immunodominant region between amino acids 52 to 93 (2); relatively few sequence changes were observed in this region (Fig. 3). However, high percentages of HDV-infected woodchucks also exhibited humoral antibody
6476
CASEY ET AL.
responses to peptides from the amino-terminal region of HDAg (amino acids 1 to 18 and 36 to 52) (2), and sequences in this region were observed to be modified frequently in our study (Fig. 3). The significance of this association is not clear because humoral antibody to HDAg is not likely to be protective; however, the correlation suggests the possibility of immunological cross talk between the development of humoral responses and other immune responses that are involved in the clearance of virus infection, i.e., cytotoxic T-lymphocyte (CTL) responses. Dissection of the immune response to HDV infection in future studies may help to directly determine the extent to which HDAg sequence changes are the result of immunologic pressure. The determinants of the outcome of HDV superinfection are not known. Analyses of the viral genetics of hepatitis C virus have indicated that progression to chronic infection is correlated with sequence changes and increased sequence diversity in the hypervariable region of the envelope protein (11). Previous studies of genetic changes that occur during the course of HDV infection have analyzed sequence modifications that occur over time in HDV RNA isolated either from the sera of patients with established chronic infection (7, 14, 19) or from the liver of a woodchuck infected with HDV that had been passaged in woodchucks several times previously (21). However, none of these studies specifically addressed the role of viral genetic changes in the establishment of chronic infection. In our study, the timing and degree of sequence changes appeared to be correlated with the outcome of infection. Five of five chronically infected woodchucks exhibited consensus sequence changes at 73 weeks postinfection, but only one of the five exhibited sequence substitutions at earlier times (Fig. 4 and Table 4). In contrast, direct sequencing of RT-PCR products indicated that majority sequence changes were apparent by 27 weeks in two of three woodchucks that recovered from HDV infection, and in the third, there was evidence of multiple sequence changes that had not yet dominated the population (Table. 4). Moreover, the number of positions changed was generally greater in the animals that recovered from HDV infection than in those that developed chronic infection (Fig. 4 and Table 4). Although the correlation between sequence changes and the outcome of infection is based on a small number of animals, we note that this correlation is also supported by analysis of HDV infection in woodchuck 4928, which was transfected with the cDNA clone and eventually recovered from HDV infection; like the three animals that recovered from infection initiated by intravenous inoculation, this animal also produced virus in which a consensus change was detected early, between 8 and 15 weeks posttransfection. One explanation of the differences observed between woodchucks that recovered and those that became chronically infected is that the virus sequence changed earlier in the woodchucks that recovered because of a more vigorous host immune response. In the animals that recovered, either the changes in HDAg sequence were insufficient to avoid immune pressure or the immune system was able to recognize additional epitopes that facilitated viral clearance. In the animals that became chronically infected, the delayed appearance and limited nature of the changes in consensus sequence may indicate that the immune response was weak or delayed and may have been limited to only
J. VIROL.
one or two epitopes. Thus, chronic HDV infection in woodchucks may result from a delayed and weak immune response that is limited to a small number of HDAg epitopes. Indeed, analyses of proliferative peripheral blood mononuclear cell responses to HDAg in infected humans and woodchucks have indicated weak responses that are typically limited in most individuals to just one to three epitopes (10, 12, 23). If the explanation is correct that the observed genetic changes are determined by the immune response, then stimulation of a potent immune response, which is inferred by the sequence changes observed at 27 weeks postinfection in the animals that recovered, could be achieved by vaccination and such a response could result in viral clearance. Most likely, the alteration of the course of infection by some vaccines (1, 10, 12, 15, 16) is due to the ability to elicit appropriate cell-mediated immune responses. However, no correlation has been observed between vaccine efficacy and anti-HD T-cell proliferative responses (12). Perhaps, CTL activity is the more important contributor to clearance of HDV-infected cells. Future vaccine studies may benefit from measurement of CTL activity induced by HDV infection and/or vaccines as well as analysis of viral genetics. The ability of the virus sequence to change at a limited number of sites could be a critical factor in the ability to develop a successful vaccine against HDV infection. ACKNOWLEDGMENTS This work was supported by National Institute of Allergy and Infectious Diseases contracts NO1-AI-45179 and NO1-AI-35164 and by grant R01-AI42324. We gratefully acknowledge the expert assistance of Betty Baldwin, Lou Ann Graham, Richard Moore, and Chris Bellezza (Cornell University) and of Frances Wells and James Nupp (Georgetown University). REFERENCES 1. Bergmann, K. F., J. L. Casey, B. C. Tennant, and J. L. Gerin. 1993. Modulation of hepatitis delta virus infection by vaccination with synthetic peptides: a preliminary study in the woodchuck model. Prog. Clin. Biol. Res. 382:181–187. 2. Bergmann, K. F., P. J. Cote, A. Moriarty, and J. L. Gerin. 1989. Hepatitis delta antigen. Antigenic structure and humoral immune response. J. Immunol. 143:3714–3721. 3. Bonino, F., K. H. Heermann, M. Rizzetto, and W. H. Gerlich. 1986. Hepatitis delta virus: protein composition of delta antigen and its hepatitis B virusderived envelope. J. Virol. 58:945–950. 4. Bonino, F., B. Hoyer, E. Ford, J. W. Shih, R. H. Purcell, and J. L. Gerin. 1981. The delta agent: HBsAg particles with delta antigen and RNA in the serum of an HBV carrier. Hepatology 1:127–131. 5. Bonino, F., B. Hoyer, J. W. Shih, M. Rizzetto, R. H. Purcell, and J. L. Gerin. 1984. Delta hepatitis agent: structural and antigenic properties of the deltaassociated particle. Infect. Immun. 43:1000–1005. 6. Chang, J., S. O. Gudima, and J. M. Taylor. 2005. Evolution of hepatitis delta virus RNA genome following long-term replication in cell culture. J. Virol. 79:13310–13316. 7. Chao, Y. C., H. S. Tang, and C. T. Hsu. 1994. Evolution rate of hepatitis delta virus RNA isolated in Taiwan. J. Med. Virol. 43:397–403. 8. Cote, P. J., B. E. Korba, R. H. Miller, J. R. Jacob, B. H. Baldwin, W. E. Hornbuckle, R. H. Purcell, B. C. Tennant, and J. L. Gerin. 2000. Effects of age and viral determinants on chronicity as an outcome of experimental woodchuck hepatitis virus infection. Hepatology 31:190–200. 9. Deny, P., A. L. Zignego, N. Rascalou, A. Ponzetto, P. Tiollais, and C. Brechot. 1991. Nucleotide sequence analysis of three different hepatitis delta viruses isolated from a woodchuck and humans. J. Gen. Virol. 72:735–739. 10. D’Ugo, E., M. Paroli, G. Palmieri, R. Giuseppetti, C. Argentini, E. Tritarelli, R. Bruni, V. Barnaba, M. Houghton, and M. Rapicetta. 2004. Immunization of woodchucks with adjuvanted sHDAg (p24): immune response and outcome following challenge. Vaccine 22:457–466. 11. Farci, P., A. Shimoda, A. Coiana, G. Diaz, G. Peddis, J. C. Melpolder, A. Strazzera, D. Y. Chien, S. J. Munoz, A. Balestrieri, R. H. Purcell, and H. J. Alter. 2000. The outcome of acute hepatitis C predicted by the evolution of the viral quasispecies. Science 288:339–344.
VOL. 80, 2006
HDV GENETIC CHANGES DURING ACUTE OR CHRONIC INFECTION
12. Fiedler, M., M. Lu, F. Siegel, J. Whipple, and M. Roggendorf. 2001. Immunization of woodchucks (Marmota monax) with hepatitis delta virus DNA vaccine. Vaccine 19:4618–4626. 13. Gerin, J. L., J. L. Casey, and R. H. Purcell. 2001. Hepatitis delta virus, p. 3037–3050. In D. M. Knipe and P. M. Howley (ed.), Field’s virology, 4th ed. Lippincott Williams & Wilkins, Philadelphia, Pa. 14. Imazeki, F., M. Omata, and M. Ohto. 1990. Heterogeneity and evolution rates of delta virus RNA sequences. J. Virol. 64:5594–5599. 15. Karayiannis, P., J. Saldanha, A. M. Jackson, S. Luther, R. Goldin, J. Monjardino, and H. C. Thomas. 1993. Partial control of hepatitis delta virus superinfection by immunisation of woodchucks (Marmota monax) with hepatitis delta antigen expressed by a recombinant vaccinia or baculovirus. J. Med. Virol. 41:210–214. 16. Karayiannis, P., J. Saldanha, J. Monjardino, A. Jackson, S. Luther, and H. C. Thomas. 1993. Immunisation of woodchucks with hepatitis delta antigen expressed by recombinant vaccinia and baculovirus controls HDV superinfection. Prog. Clin. Biol. Res. 382:193–199. 17. Kuo, M. Y., M. Chao, and J. Taylor. 1989. Initiation of replication of the human hepatitis delta virus genome from cloned DNA: role of delta antigen. J. Virol. 63:1945–1950. 18. Lanford, R. E., D. Chavez, K. M. Brasky, R. B. Burns III, and R. Rico-Hesse. 1998. Isolation of a hepadnavirus from the woolly monkey, a New World primate. Proc. Natl. Acad. Sci. USA 95:5757–5761. 19. Lee, C. M., F. Y. Bih, Y. C. Chao, S. Govindarajan, and M. M. Lai. 1992. Evolution of hepatitis delta virus RNA during chronic infection. Virology 188:265–273. 20. Negro, F., B. E. Korba, B. Forzani, B. M. Baroudy, T. L. Brown, J. L. Gerin, and A. Ponzetto. 1989. Hepatitis delta virus (HDV) and woodchuck hepatitis virus (WHV) nucleic acids in tissues of HDV-infected chronic WHV carrier woodchucks. J. Virol. 63:1612–1618. 21. Netter, H. J., T. T. Wu, M. Bockol, A. Cywinski, W. S. Ryu, B. C. Tennant, and J. M. Taylor. 1995. Nucleotide sequence stability of the genome of hepatitis delta virus. J. Virol. 69:1687–1692. 22. Niro, G. A., A. Smedile, A. Andriulli, M. Rizzetto, J. L. Gerin, and J. L. Casey. 1997. The predominance of hepatitis delta virus genotype I among chronically infected Italian patients. Hepatology 25:728–734. 23. Nisini, R., M. Paroli, D. Accapezzato, F. Bonino, F. Rosina, T. Santantonio, F. Sallusto, A. Amoroso, M. Houghton, and V. Barnaba. 1997. Human CD4⫹
24.
25.
26.
27.
28.
29. 30.
31.
32.
33.
6477
T-cell response to hepatitis delta virus: identification of multiple epitopes and characterization of T-helper cytokine profiles. J. Virol. 71:2241–2251. O’Malley, B., and D. W. Lazinski. 2005. Roles of carboxyl-terminal and farnesylated residues in the functions of the large hepatitis delta antigen. J. Virol. 79:1142–1153. Polson, A. G., B. L. Bass, and J. L. Casey. 1996. RNA editing of hepatitis delta virus antigenome by dsRNA-adenosine deaminase. Nature 380:454– 456. Polson, A. G., H. L. Ley III, B. L. Bass, and J. L. Casey. 1998. Hepatitis delta virus RNA editing is highly specific for the amber/W site and is suppressed by hepatitis delta antigen. Mol. Cell. Biol. 18:1919–1926. Ponzetto, A., P. J. Cote, H. Popper, B. H. Hoyer, W. T. London, E. C. Ford, F. Bonino, R. H. Purcell, and J. L. Gerin. 1984. Transmission of the hepatitis B virus-associated delta agent to the eastern woodchuck. Proc. Natl. Acad. Sci. USA 81:2208–2212. Radjef, N., E. Gordien, V. Ivaniushina, E. Gault, P. Anais, T. Drugan, J. C. Trinchet, D. Roulot, M. Tamby, M. C. Milinkovitch, and P. Deny. 2004. Molecular phylogenetic analyses indicate a wide and ancient radiation of African hepatitis delta virus, suggesting a deltavirus genus of at least seven major clades. J. Virol. 78:2537–2544. Ryu, W. S., M. Bayer, and J. Taylor. 1992. Assembly of hepatitis delta virus particles. J. Virol. 66:2310–2315. Schlipkoter, U., A. Ponzetto, K. Fuchs, R. Rasshofer, S. S. Choi, S. Roos, M. Rapicetta, and M. Roggendorf. 1990. Different outcomes of chronic hepatitis delta virus infection in woodchucks. Liver 10:291–301. Smedile, A., J. L. Casey, P. J. Cote, M. Durazzo, B. Lavezzo, R. H. Purcell, M. Rizzetto, and J. L. Gerin. 1998. Hepatitis D viremia following orthotopic liver transplantation involves a typical HDV virion with a hepatitis B surface antigen envelope. Hepatology 27:1723–1729. Sureau, C., A. M. Moriarty, G. B. Thornton, and R. E. Lanford. 1992. Production of infectious hepatitis delta virus in vitro and neutralization with antibodies directed against hepatitis B virus pre-S antigens. J. Virol. 66: 1241–1245. Wu, J. C., P. J. Chen, M. Y. Kuo, S. D. Lee, D. S. Chen, and L. P. Ting. 1991. Production of hepatitis delta virus and suppression of helper hepatitis B virus in a human hepatoma cell line. J. Virol. 65:1099–1104.
