Rapid Production of Neutralizing Antibody Leads to Transient Hepadnavirus Infection

Abstract

Hepatitis B virus (HBV) frequently causes transient infections in adults but chronic infections in infants. The basis of these age-related outcomes is not known. Infection of ducks with duck hepatitis B virus (DHBV) displays a similar dependence of outcome on the age of the host at the time of infection. In this study we compared the infection of ducks at 3 days and 3 weeks of age. We found that the efficiency of infection of hepatocytes by virus in the inoculum was similar between the two age groups but that spread of the infection throughout the liver was severely inhibited in the 3-week-old-old ducks, while a rapid spread of the infection was observed in 3-day-old ducklings. Inhibition of virus spread was accompanied by the appearance in the serum of virus neutralizing activity, as assayed by blocking of infection of primary hepatocyte cultures. Neutralizing activity appeared as early as 1 or 2 days postinfection and increased during the next 2 weeks. Depletion of immunoglobulins from serum eliminated the neutralizing activity. The specific depletion of IgM indicated that IgM appeared as the dominant fraction of neutralizing antibody in the first 2 days postinfection, but declined from day 3 on while IgG antibody rose. We conclude that excess neutralizing antibody arising rapidly in birds inoculated at 3 weeks of age but not in newly hatched ducks prevented secondary cycles of infection, resulting in a limited infection in the liver and contributing to the eventual transient outcome of the infection.

Hepatitis B virus (HBV) causes both transient and persistent infections in humans. It is estimated that more than 350 million people are chronically infected with HBV (10, 16, 32). Transient or persistent HBV infection is largely determined by the age at which the infection is acquired. When the viral infection occurs perinatally, more than 90% of such infections become persistent (1, 20, 26). Persistent HBV infection can lead to chronic hepatitis, liver cirrhosis, and hepatocellular carcinoma (2, 4, 14, 30). In contrast, more than 90% of infections occurring in adults are transient (19, 23). Transiently HBV-infected adults can experience either a clinical acute infection or a subclinical course. The subclinical infection involves a majority of HBV-infected adults, and they are identified only by detection of virus-specific antibodies (5, 21, 25). Virologic and immunologic distinctions between infant and adult groups during the initiation and spread of HBV infection are not defined, and mechanisms responsible for age-related outcomes of HBV infection are not clear.

Duck hepatitis B virus (DHBV) and woodchuck hepatitis virus are members of the hepadnavirus family, and DHBV-infected ducks and woodchuck hepatitis virus-infected woodchucks are naturally infected animal models for studying HBV infection (18, 27-29). Viral infection in both the duck and the woodchuck also shows the same age-related outcomes as does HBV infection (3, 6, 8, 9). To understand the mechanisms for the age-related outcomes by HBV infection, we studied the virologic and immunologic distinctions between young ducklings and adolescent ducks infected with DHBV. Our data indicate that early production of neutralizing antibody interrupts virus spread and leads to a transient infection.

MATERIALS AND METHODS

Preparation of virus inocula.

The D2 cell line, a chicken hepatoma cell line stably transfected with a DHBV-16 wild-type genome-containing plasmid, obtained from W. S. Mason (Fox Chase Cancer Center, Philadelphia, Pa.), was used for production of viruses. Viruses in collected D2 medium were quantitated by phosphorimaging of hybridized Southern blots after selective extraction of viral DNA from enveloped particles (11) and precipitation with 10% polyethylene glycol 8000 to achieve desired viral concentration. Inocula for injection were prepared by appropriate dilution of the concentrated viruses in Dulbecco's modified Eagle's medium-F12 medium (Gibco Invitrogen). Procedures used in the analysis of viral DNA replicative intermediates, agarose gel electrophoresis, and blot hybridization were published previously (12).

Animals, inoculation, and sample collection.

One-day-old Peking ducklings were purchased from Metzer Farms (Redlands, Calif.). Congenitally DHBV-infected ducklings, detected after screening by dot blot (31), were excluded from experiments. Two groups of ducks at different ages were used for DHBV infection: one group was infected at 3 days posthatch (group 1), and the other group was infected at 3 weeks posthatch (group 2). Each 3-day-old and 3-week-old duck was intravenously inoculated with 1 × 108 or 3 × 108 virus particles, respectively, in 0.2-ml volume. In experiment 1, three animals of each group were sacrificed daily to harvest liver tissue for analysis of viral replicative intermediates and histology from day 1 to day 7 postinfection. Two uninfected birds of group 2 were sacrificed at day 7 to serve as controls. Blood samples were drawn daily from each animal until sacrifice. Infection of 3-week-old ducks was repeated in experiment 2. Three transiently infected animals that were sacrificed at day 7 or day 22 postinfection from experiment 2 were included in this study because they provided a large volume of good-quality serum samples for detailed analysis of neutralizing antibody activity.

Antibody staining of viral proteins on liver sections.

Sections from paraffin-embedded liver tissue blocks were cut and used for staining of viral pre-S proteins by specific rabbit polyclonal antibodies (Hans Will, Hamburg, Germany). Procedures for immunochemical staining on liver sections were the same as previously published (13).

Assay for neutralizing antibody by blocking infectivity.

The neutralizing antibody assay was done with primary duck hepatocyte cultures. Preparation of hepatocyte cultures followed the protocol described previously (22). To determine whether a serum contained neutralizing antibody, an inoculum of 107 virus particles was first mixed with 200 μl of serum and incubated at 37°C for 1 h, then added to primary duck hepatocyte cultures 1 day after plating. The medium was changed daily. Cultures were harvested for DNA extraction at 10 or 11 days postinoculation. Blocking of infectivity was judged by comparing the amount of replicative intermediates in primary duck hepatocyte cultures infected with the standard inoculum incubated either with serum from a naïve duckling or with test serum. For evaluation of inhibition of viral replication, 200 μl of serum was added to primary duck hepatocyte cultures at day 4 postinfection and every second day thereafter, and cultures were harvested at day 12 postinfection for analysis of replicative intermediates.

Depletion of serum immunoglobulins.

For immobilization of rabbit anti-duck antibody, 500 μl of protein G-agarose beads (Roche) was mixed with 400 μl of rabbit anti-duck IgG (containing anti-duck IgM activity; Nordic Immunology) or with 120 μl of goat anti-duck IgM (Nordic). The beads were incubated at 4°C overnight, then washed twice with wash buffer (20 mM sodium phosphate, 150 mM sodium chloride, 2 mM EDTA, pH 7.0) to remove unbound antibody. Serum (200 μl) collected at days 1 to 4 postinfection from an animal with a transient infection was added to 250 μl of beads with the immobilized antibody, incubated at room temperature for 1 h, and centrifuged for 1 min in a microcentrifuge. The supernatant was tested for neutralizing activity. Ratios of beads and antibody required for optimum depletion of IgG or IgM were empirically determined.

Assay for precipitation of virus by antibody.

Serum (20 μl) was mixed with free viruses equivalent to 6 × 10⁵ genomes in a 20-μl volume, and 60 μl of 10:1 Tris-EDTA (TE) buffer was added to each tube to bring the volume to 100 μl. The mixture was incubated at 4°C overnight followed by microcentrifuge centrifugation for 10 min, and the supernatants and pellets were carefully separated. The pellets were suspended in 400 μl of 10:1 TE buffer containing 0.2% sodium dodecyl sulfate and 300 μl of the same buffer was added to the supernatants. The pellet and supernatant were subjected to the pronase digestion at a concentration of 500 μg/ml for 1 h at 37°C, phenol extraction, and ethanol precipitation. DNA was dissolved in 10 μl of 10:0.1 TE buffer (10 mM Tris-HCl, 0.1 mM EDTA, pH 7.4) for PCR amplification. The detectable viral genomes in the pellet fraction were calculated as fractions of the total viral DNA detected.

Real-time PCR quantitation of viremia in transient infection.

For real-time PCR analysis of viremia in transiently infected ducks, DNA was extracted from 20 μl of serum at each time point after pronase digestion (500 μg/ml) in the presence of 0.2% sodium dodecyl sulfate. Then 5 μl of extracted DNA equivalent to 10 μl of serum was mixed with 10 μl of 2x SYBR green supermix (Bio-Rad), primers, and water to make a total of 20 μl. The PCR mix contained a final concentration of 50 mM KCl, 20 mM Tris-HCl, pH 8.4, 200 μM each deoxynucleotide triphosphate, 0.5 unit of iTaq DNA polymerase, 3 mM MgCl₂, SYBR Green I, 10 nM fluorescein, and 200 nM each primer.

The sense primer (nucleotide positions 894 to 915: AGG GAC TTT GAC ATG GTC AGG CA) and anti-sense primer (nucleotide positions 1263 to 1238: TTT TAT CAC TGG CAC CGC TGG TTC T) amplified a sequence in the pre-S region of the DHBV 16 genome sequence (17). The PCR amplification cycle consisted of three temperature profiles: denaturation at 95° C for 30 s, annealing at 55° C for 20 s, and elongation at 72° C for 45 s. Distinction of PCR products from primer dimers was made possible by melting curves or, when necessary, by agarose electrophoresis. In order to have accurate standard plasmid controls, the supercoiled plasmid containing the DHBV genome was isolated from an agarose gel and recovered with a commercial gel purification column (Qiagen). The optical density was measured at 260 nm to determine the copy number, and the purified plasmid was linearized by EcoRI cleavage for efficient denaturation.

RESULTS

Rapid spread of infection occurred in livers of 3-day-old but not 3-week-old ducks after initial infection.

Two groups of ducks were infected with DHBV. The ducks in group 1 at 3 days of age were inoculated with 10⁸ virus particles, and the ducks in group 2 were inoculated with 3 × 10⁸ particles at 3 weeks of age. Ducklings were bled daily and tested for viremia. In group 1, viremia was detected by dot blot at 3 days postinfection, and the viremia reached peak levels on days 5 and 6 (Fig. 1, panel A). No viremia was detected by dot blot in group 2 ducks after inoculation, however, by real-time PCR assay, a low viremia was detected beginning at day 2, which peaked around day 6 and decreased thereafter, eventually to undetectable levels at days 14 and 22 (Fig. 1, panel B). Thus, as previously reported (8), newly hatched ducklings are highly susceptible to infection, while older ducks are more resistant and show evidence of only a low-level transient infection when challenged with an equivalent inoculum.

FIG. 1.

Viremia in DHBV-infected ducks. (A) Viremia in group 1 ducks was determined by dot blot hybridization with a radioactive riboprobe specific for detection of the minus strand. The detection limit was about 10⁷ viral genomes per ml. (B) Viremia of group 2 ducks was assayed by real-time PCR. The detection limit was about 10² genomes per ml. At least three ducks were assayed for each time point.

In both groups 1 and 2, small amounts of viral replicative intermediates became detectable on day 2 or 3 postinfection, indicating that the initiation of viral infection occurred in the inocula for both groups (Fig. 2). However, the viral replicative intermediates increased by about 100-fold by day 4 in group 1 (Fig. 2A) and remained high thereafter (not shown). In contrast, no comparable change in the level of the replicative intermediates was observed from days 4 through 7 in the group 2 ducks (Fig. 2B), suggesting that viral infection was not significantly expanded after the first round of infection.

FIG. 2.

Viral replicative intermediates and antigen-staining cells in livers of DHBV-infected ducks. Three ducks each from group 1 (A) and group 2 (B) were sacrificed at the indicated times postinfection, and replicative intermediates were extracted from the liver. The samples loaded contained a total of 20 μg of cellular RNA, and viral DNA was detected by hybridization with a riboprobe specific for detection of the minus strand. The percent antigen-staining hepatocytes in some of the samples is indicated at the top of the lanes.

To investigate the extent of viral infection in the liver in the two groups, viral pre-S proteins were stained on liver sections for viewing the initiation and spread of viral infection. Antigen-staining cells were counted, and these numbers are expressed as the percentage of total hepatocytes in Fig. 2. These numbers roughly reflect the Southern blot results and confirm that the infection in 3-week-old ducks failed to spread significantly. The appearance of antigen-staining cells is shown in Fig. 3. Scattered DHBV-infected cells were detected on days 2 and 3 postinfection in the two groups. However, a massive expansion of infection to every hepatocyte was observed in the group 1 ducks between day 3 and day 4 or 5, consistent with the increase in viremia and replicative intermediates. In contrast, the majority of hepatocytes in group 2 ducks remained uninfected between day 4 and day 7, and small clusters of positive cells, possibly formed by division of infected cells or local spreading to neighboring cells (7), appeared on the sections. The failure of the initial virus to spread throughout the liver suggested a lack of availability of newly released viruses for new rounds of infection despite the fact that the numbers of initially infected hepatocytes were similar between the groups. We therefore investigated whether serum samples from these animals contained virus neutralizing activity.

FIG. 3.

Immunochemical detection of viral pre-S protein in infected liver. Polyclonal rabbit anti-pre-S was used as the primary antibody and horseradish peroxidase-conjugated goat anti-rabbit IgG was used as the second antibody. Color was developed with diaminobenzidine as the substrate. A representative section for each time point from three animals is shown. Magnification, ×400.

Early appearance of virus neutralizing activity in serum samples from animals with transient infections.

We tested for neutralizing activity on a standard inoculum used to infect primary duck hepatocyte cultures. Serial serum samples from individual animals with transient infection were mixed with 10⁷ DNA-containing virus particles before inoculation of hepatocyte cultures. Infected cells were assayed after 10 days for viral replicative intermediates as a measure of the infectivity of the inoculum. As shown in Fig. 4, a reduction in infectivity was observed when inocula were incubated with serum samples, compared with inoculum-only controls or inocula incubated with serum from naïve ducks (upper left panel). The neutralizing activity appeared in the serum as early as day 2 or day 3 postinfection in eight of nine birds and appeared to increase thereafter in some birds over time. Addition of serum containing neutralizing activity to the cell 1 day before or after infection failed to reduce virus replication, indicating that the neutralizing activity acted on the virus rather than on the cells (data not shown, see below).

FIG. 4.

Assay of neutralizing activity present in sera from transiently infected ducks. Serial serum samples from each of nine animals from group 2 were analyzed for their ability to block infection of primary duck hepatocytes with 10⁷ virus particles. After incubation with 200 μl of serum, the inoculum was added to cultures of primary duck hepatocytes. At 10 days postinfection, viral replicative intermediates were extracted and analyzed by Southern blotting, as described in the legend to Fig. 3. A reduction of replicative intermediates compared with inoculum alone (nil) or inoculum incubated with normal duck serum (−) indicated that neutralizing activity was present.

Lack of detectable neutralizing activity in some sera (e.g., bird 98) might be expected from the large size of our standard inoculum (10⁷ viral genomes) relative to the amount of antibody that might be required to prevent the spread of infection. For example, the amounts of antibody sufficient to neutralize only 10⁶ virus particles per ml would not be expected to be detectable in our assay. This panel of serum samples (nine animals) were tested in three independent assays for neutralizing activity, and the results from these tests were consistent with the experiment shown.

We further tested for evidence of neutralizing antibody by determining whether there was available unbound antibody that could aggregate and precipitate free viruses. Virus particles were incubated with serum samples and aggregates were pelleted by microcentrifuge centrifugation. As shown in Fig. 5, serum from a naïve duck failed to precipitate virus particles (7% in the pellet), while serum samples containing virus neutralizing activity as early as day 2 postinfection caused the precipitation of 50 to 90% of added virus.

FIG. 5.

Precipitation of free viruses by unbound antibody. Serial serum samples were collected from a transiently infected 3-week-old duck. Virus particles (5 × 10⁶ genomes) were mixed with serum from each time point, incubated, and subjected to microcentrifugation. DNA in the supernatant and pellet was extracted and assayed by real-time PCR. The fraction of the total virus pelleted is plotted.

Quenching of neutralizing activity by depletion of IgG and IgM.

To determine whether the neutralizing activity was due to immunoglobulins, the neutralizing activity was assayed following depletion of IgG or IgM plus IgG from a set of serial serum samples. More than 90% of heavy and light chains of duck immunoglobulins could be depleted from the serum by our protocol, as examined by Western blots (data not shown). As shown in Fig. 6, the neutralizing activity was completely quenched once the IgG plus IgM fractions were depleted. A loss of neutralizing activity by depletion of IgG plus IgM was observed even in the serum sample from day 1 postinfection, indicating that neutralizing activity as early as 1 day was mediated by immunoglobulins. It appeared that IgM antibody constituted a major fraction of neutralizing antibody at days 1 and 2 postinfection because the majority of quenching of the neutralizing activity was observed following depletion of IgM from sera. However, IgM-mediated activity declined and activity due to IgG became dominant from day 3. The data from this depletion assay are consistent with the conclusion that the neutralizing activity was mediated by antibody present in serum samples at very early time points postinfection.

FIG. 6.

Loss of neutralizing activity on primary duck hepatocyte following depletion of immunoglobulins. Sera collected from a transiently infected 3-week-old duck at the indicated times were depleted of immunoglobulins as described in Materials and Methods. The depleted sera were assayed for neutralizing activity compared with that present in normal duck serum and whole sera. The data represent the normalized amount of replicative intermediates at day 12 present in primary duck hepatocytes infected with a standard inoculum (4 × 10⁷ viral genomes) incubated with the indicated serum samples. 0, normal duck serum; T, not depleted; M, IgM depleted; M/G, IgM and IgG depleted.

Estimate of the amount of unbound neutralizing antibody in the serum.

We determined the titer of neutralizing antibody in serial serum samples from one animal by incubating decreasing amounts of serum with a standard inoculum of 3 × 10⁷ virus particles and determining the inhibition caused. The results are shown in Table 1. Although the amount of antibody present in 0.2 ml of serum at 1 day postinfection was insufficient to neutralize the standard inoculum of 3 × 10⁷ virus particles, by day 14 the neutralizing activity had increased approximately 10-fold (3 × 10⁷ particles neutralized per 0.025 ml). Since the viremia present in these samples was well below the neutralizing capacity of the serum, the neutralizing activity seems to be capable of accounting for the inhibition of spread of the infection.

TABLE 1.

Titration of neutralizing activity of serial serum samples^a

Amt of serum (ml)	Relative infectivityb
	Day 1	Day 3	Day 7	Day 14	Day 22
0.20	0.38	0.00	0.00	0.00	0.00
0.10	0.86	0.03	0.00	0.00	0.00
0.050	0.73	0.22	0.13	0.00	0.00
0.025	0.90	0.31	0.72	0.17	0.00

^aSamples were collected from a 3-week-old duck with a transient infection. Low-level viremia was documented in the sera collected from the transiently infected duck at days 1, 3, 7, 14, and 22 at 3 × 10³, 1 × 10⁴, 8 × 10³, 3 × 10³, and <10 genomes per ml, respectively.

^bValues represent the amount of viral DNA in the cultures at day 10 postinfection divided by the amount in a culture infected with the standard inoculum alone, consisting of 3 × 10⁷ virus particles.

Sera containing neutralizing antibody do not inhibit viral replication.

To further investigate any possible additional antiviral activity exerted by sera from transiently infected animals, we tested whether the sera from group 2 ducks were able to inhibit the accumulation of viral DNA in cells that were already infected. Serial serum samples from two ducks with transient infections were tested for neutralizing antibody as described above. In parallel, primary duck hepatocytes were infected with the standard inoculum in the absence of any serum, and then sera collected from the same animals on days 1, 7, and 14 (duck 47), or days 1, 6, and 7 (duck 48) postinfection were added to the infected cultures at day 4. Cultures were harvested at day 12 and assayed for viral DNA. As expected, the infectivity of the inocula was reduced by added sera from day 6, 7, or 14 of the group 2 birds, indicating the presence of neutralizing antibody in those sera (Fig. 7). However, the sera showing neutralizing activity did not inhibit viral replication when added to primary duck hepatocyte cultures from day 4 onward, when replicative intermediates began to appear in the cultures (not shown). These data indicate that all of the inhibitory activity of those sera could be accounted for by the neutralizing activity on the extracellular viruses, at least in our assay system.

FIG. 7.

Blocking activity in serum is due to virus neutralization alone. The abilities of serum samples to block infection when added to the virus inoculum (filled bars) or to the cells 4 days after infection (open bars) were compared. Serial serum samples (200 μl) from 3-week-old ducks with transient infections were incubated with a standard inoculum (10⁷ genomes) before infection of primary duck hepatocytes. In parallel, the same serum samples were added to cultures at day 4 postinfection with the standard inoculum, and the serum plus medium was renewed every second day for 8 days. Cells were harvested at day 12, and replicative intermediates were assayed by Southern blot hybridization and phosphorimaging. The values plotted are arbitrary units.

DISCUSSION

Virologic and immunologic distinctions between newly hatched (3 days of age) and adolescent (3 weeks of age) ducks infected with DHBV were investigated in this study. We were able to observe infection of liver proceeding in two phases. The first phase was the infection of a small portion of hepatocytes directly by viruses in the inoculum. The second phase was the spread of infection to the rest of the liver by viruses released from initially infected cells. The data we have presented suggest that the massive spread of infection in the livers of 3-week-old ducks inoculated with infectious DHBV can be interrupted by an antibody response detectable as early as 1 or 2 days postinoculation. During the short interval between injection and the appearance of neutralizing antibody, some of the injected virus was able to initiate infection in scattered hepatocytes, and the resulting low level of infected cells persisted during the subsequent 7 days postinfection, spreading only locally in small foci either through second rounds of infection of neighboring cells or division of infected hepatocytes (7).

We demonstrated that the interruption in the spread of infection was accompanied by the appearance of neutralizing activity in the serum and that the neutralizing activity could be removed from the serum by depletion of IgG and IgM. In most cases, the neutralizing activity (sufficient to neutralize 10⁷ viral genomes) was in excess of the concentration of virus in the serum (10³ to 10⁵ viral genomes) of the infected ducks by several orders of magnitude. Virus particles added to serum samples containing neutralizing activity but not to normal duck serum were aggregated and precipitated by low-speed centrifugation, consistent with the presence of specific antibody. Finally, neutralizing activity in the serum acted on the virus rather than the cells, since there was no effect on subsequent virus replication of treated cells with serum samples either before (not shown) or after exposure to the inoculum.

Jilbert et al. previously reported the dose dependence of the outcome of DHBV infection in young and adult ducks: lower doses of virus resulted in a low-level transient infection without detectable viremia, while higher doses produced viremia with or without chronic infection (8). It was not clear from that study how susceptible adult ducks were to the initial inoculum or whether virus spread in the livers of 1- and 2-week-old ducks was aborted. Comparison of our study with that of Jilbert et al. is problematic because of the large difference in the efficiency of infection of ducks from the suppliers that were used, a difference that has been observed repeatedly (A. R. Jilbert, personal communication). Nevertheless, our result is consistent with that study in that older ducks were more resistant to development of viremia after inoculation than newly hatched ducklings.

In an extension of their findings, our study has shown that the difference in outcome of infection between 3-day-old and 3-week-old ducks was due not to a difference in susceptibility to infection by the inoculum but to a difference in the ability of virus to spread throughout the liver and produce significant viremia. It is possible that if a larger number of cells had been infected by a correspondingly larger inoculum, the amount of virus produced initially would have titrated the ability of the immune system to produce early antibody, allowing spread throughout the liver to occur. Whether such a scenario is responsible for the dose dependence of the outcome of infection reported by Jilbert et al. has not been adequately tested.

We do not know whether the protective antibody production in our experiments was due to stimulation by antigen present in the inoculum, by that produced by the initial population of infected cells, or both. The inoculum, which contained 3 × 10⁸ DNA-containing virus particles, could be expected to contain 1,000-fold more empty viral envelopes that may have constituted a significant antigenic dose. However, even the very low doses used by Jilbert et al. (10⁴ genomes) (A. R. Jilbert, personal communication) produced transient infections without viremia in 2-week-old ducks, suggesting that virus spread may be interrupted by antibody produced in response to the spreading virus itself. Alternatively, virus spread may be prevented by other mechanisms when the infection is initiated by a sufficiently small inoculum.

Despite the relatively large size of our inoculum (10⁸ genomes), injection of 3-day-old ducklings did not result in an antibody response sufficient to interrupt the spread of infection. In our experience, 3-day-old ducklings are generally highly susceptible to infection with as few as 10³ virus particles, and the results presented in this study are consistent with the age-related outcome reported by Jilbert et al. (8). The basis for the age-dependent outcome may be related to the delayed appearance of nonmaternal immunoglobulins (15, 24) in the duck until approximately 20 days posthatching. If this delay is associated with an inability of antibody production to be elicited by antigenic stimulation, then the age-related outcome to infection would seem to depend on the age-related ability of the duck to rapidly produce neutralizing antibody.

Although low dose exposures of immunocompetent adults, human or animal, to hepadnaviruses most frequently result in transient infections with no overt antigenemia (5, 22, 25), similar to those studied here, occasionally infection can spread throughout the liver during a course of several weeks or months. In such cases, it seems unlikely that a sufficient neutralizing antibody response would have occurred during this incubation period, since such a response appears capable of arresting the spread of infection, as shown in this study. It may be that under some circumstances, viral envelope antigens released from a small number of infected cells fail to provide any positive stimulation of neutralizing antibody because they are efficiently removed from the circulation by the large number of neighboring uninfected hepatocytes or for other unknown reasons. Alternatively, the spread of infection may be slowed but not prevented by an antibody response that is insufficient in magnitude or specificity, resulting in a prolonged incubation period between exposure and viremia. These possibilities can be tested by examining the rate of spread of infection in a larger number of immunocompetent adult animals.