Hamsters were infected by the intraperitoneal (IP) or subcutaneous (SC) routes, depending on the virulence of the infecting virus for the animals. WNV and YFV were injected IP; JEV and SLEV were administered SC. Infecting doses of the viruses were as follows: WNV 10^4.0 TCID₅₀, YFV 10^6.0 TCID₅₀, JEV 10^6.5 TCID_50, and SLEV 10^6.0 TCID₅₀.

A mouse immune ascitic fluid to WNV was prepared in adult mice. The immunogen was a crude homogenate of brain (10% W/V in PBS) from newborn mice injected intracerebrally (IC) with the B956 prototype strain of WNV (32). The adult immunization schedule consisted of four IP injections of the immunogen mixed with Freund’s adjuvant, given at weekly intervals. Sarcoma 180 cells were given after the final injection to induce ascites formation.

Serum antibodies to WNV and the other three flaviviruses were measured by hemagglutination-inhibition (HI) test and to WNV by immunoglobulin (Ig) M antibody capture enzyme immunoassay (MAC-ELISA) (33). Antigens for both serologic tests were prepared from brains of newborn mice injected IC with each of the flaviviruses; the infected brains were treated by the sucrose-acetone extraction method (33). Hamster sera were tested by HI at serial twofold dilutions from 1:20 to 1:5120 at pH 6.6 (WNV, JEV, and SLEV) or 6.4 (YFV) with 4 units of antigen and a 1:200 dilution of goose erythrocytes, following established protocols (33).

For the MAC-ELISA, microtiter plates were coated with a commercial goat anti-rat IgM (capture) antibody (Kirkegaard & Perry Laboratories, Inc., Gaithersburg, MD), diluted 1:500 in carbonate buffer, pH 9.6. All hamster sera were screened at a 1:40 dilution. The WNV antigen was also used at a 1:40 dilution. The secondary (detecter) antibody was a mouse, anti-Flavivirus, peroxidase-conjugated monoclonal antibody (6B6C-1) at a dilution of 1:6000. Results were read with a SPECTRA shell reader (SLT Labininstruments, Salzburg, Austria). Specimens wells were recorded as positive when the absorbance values at optical density₄₀₅ nm of the specimen wells exceeded 0.20 after subtraction of average background absorbance of control wells (33).

Several groups of Flavivirus-naive (control) hamsters were inoculated IP with 10⁴ TCID₅₀ of WNV to determine the subsequent level and duration of viremia, immune response, and death rate. Table 1 and the Figure show the results of an experiment with a group of 10 hamsters that were bled daily for 6 consecutive days after infection with WNV. Moderate levels of virus were detected in the animals’ blood within 24 hours and persisted for 5 or 6 days. The highest blood virus titers were detected on days 2 and 3 after infection (means 10^5.2 and 10^5.1, respectively). HI antibodies were detected in all the animals by day 5, and the titers had increased substantially by day 6. In general, WNV-specific IgM, as detected by MAC-ELISA, appeared at approximately the same time as the HI antibodies (data not shown).

Table 2 shows the results of a second experiment in which 13 hamsters were infected with WNV. All the animals were bled 6 days after injection, and a subset was bled again at 31, 60, and 90 days. Six days after infection, all the animals had specific HI antibodies to WNV antigen and were negative to the other three flaviviral antigens tested (YFV, SLEV, and JEV). At this time, the animals also had a strongly positive IgM antibody response by MAC-ELISA. Thirty-one days after infection, the HI antibody response had become broadly cross-reactive with the four Flavivirus antigens, although the highest titer was still to WNV, and the IgM antibody had begun to decrease. A similar HI antibody pattern was observed at 60 and 90 days after infection, although by 90 days the HI titers were decreasing. Six of the nine WNV-infected hamsters gave a negative reaction in the WNV MAC-ELISA when tested 60 and 90 days after infection.

Five of the 13 hamsters infected in this second experiment died of WNV encephalitis 7 to 14 days after infection (Table 2). Overall, 14 (47%) of 30 adult hamsters injected IP with 10⁴ TCID₅₀ of WNV died of encephalitis (Table 3). The pathologic reaction of the WNV hamster model has been described (22).

The Figure and Table 4 show the results from another experiment in which 30 adult hamsters were given a single SC injection of approximately 10^6.4 TCID₅₀ of the live attenuated JEV SA14-2-8 vaccine strain. Thirty-eight days later, the animals were injected (challenged) IP with 10⁴ TCID₅₀ of WNV; 10 of the hamsters in this group were bled daily for 6 consecutive days. These blood samples were subsequently titrated to determine the level of WNV viremia. The resulting viremia in the JEV-immune animals was markedly lower than in the naïve hamsters (Figure). Furthermore, the JEV-immune hamsters responded to challenge with WNV by developing a secondary (sequential) type of Flavivirus antibody response. Table 4 shows the HI antibody titers to JEV and WNV antigens in sera of 10 of the SA14-2-8 vaccinated hamsters, 30 days after their JEV immunization. At this time the HI antibody titers to JEV and WNV antigens were characteristic of a primary Flavivirus infection (13,14). On day 38, the animals were challenged with WNV; 6 days later, their sera were tested for HI and WNV-specific IgM antibodies. The boost in HI antibody titers that was observed 6 days after challenge with WNV was typical of a secondary antibody response to Flavivirus infection (13,14). In contrast, IgM antibody response to the second Flavivirus (WNV) infection was minimal (Table 4).

All the JEV-immune hamsters (n = 30) survived challenge with WNV (Table 3). Their infection with WNV was confirmed by the presence of low-level viremia (Figure) and the secondary Flavivirus antibody response following challenge (Table 4). None of these hamsters appeared clinically ill after infection with WNV, in contrast to the naïve animals. Many of the nonimmune hamsters had clinical signs of acute central nervous system injury (somnolence, muscle weakness, paralysis, tremors, and loss of balance) beginning around day 6 after infection, and approximately half died (22). Thus, prior immunization with JEV vaccine reduced the severity of subsequent WNV infection and prevented death.

The Figure and Table 5 summarize the results of another experiment in which 32 adult hamsters were given a single SC injection of approximately 10⁶ TCID₅₀ of SLEV strain BeAr 23379. This wild-type SLEV strain was selected for immunization, since it is not lethal to hamsters. Thirty-two days after injection with SLEV, the animals were inoculated IP with 10⁴ TCID₅₀ of WNV. After this WNV challenge, the hamsters were bled daily for 6 consecutive days, as before. Antibody determinations were also done on blood samples taken 6 days after challenge with WNV.

Titration of daily blood samples from the SLEV-immune hamsters gave results similar to those in the JEV-immune animals. After challenge with WNV, 7 of the 10 SLEV-immune hamsters had brief, low-level viremia (Figure). However, three hamsters had no detectable viremia.

Serologic studies on blood samples taken 30 days after SLEV infection indicated that all the tested animals had been infected (Table 5). The HI response at 30 days was characteristic of primary Flavivirus infection. Six days after WNV infection, HI antibody titers had increased, indicating a secondary flavivirus antibody response. As with the JEV-immune hamsters, the IgM response of the SLEV-immune animals was minimal following the second flavivirus (WNV) infection (Tables 4,5).

Consistent with the low levels of WNV viremia (Figure), all the SLEV-immune hamsters (n = 32) survived subsequent challenge with WNV (Table 3). These animals did not appear clinically ill. These results indicate that prior immunity to SLEV also protected the hamsters from WNV encephalitis and death.

Based on the results obtained with JEV- and SLEV-immune hamsters, we tested the effect of prior immunization with a non-JEV serocomplex Flavivirus on subsequent WNV infection. Accordingly, a group of 30 hamsters was inoculated IP with 10^6.0 TCID₅₀ of the live attenuated 17D YFV strain. Thirty days after immunization, nine of the animals were bled and tested for HI antibodies to YFV and WNV (Table 6). Six days later (36 days after 17D vaccination), the hamsters were inoculated IP with 10⁴ TCID₅₀ of WNV. Ten of these animals were bled daily for 6 consecutive days to determine the level of viremia and subsequent antibody response (Figure) (Table 6).

Following challenge with WNV, YFV-immune hamsters had an intermediate level of viremia (Figure). The mean WNV titers in the YFV-immune hamsters were higher than in the JEV- and SLEV-immune groups, but the titers were lower than in the Flavivirus-naïve (control) hamsters. The death rate in the YFV-immune hamsters was also lower; 4 (13%) of 30 YFV-immune hamsters died after challenge with WNV, compared with 47% in the control group (Table 3).

The HI antibody response after vaccination with YFV 17-D virus (Table 6) was less intense than the primary antibody responses to the other three flaviviruses (Tables 1,2,4,5). Monath (26,34) also observed that immunization with 17-D virus induces a weaker HI and complement-fixing antibody response than infection with a wild-type YFV strain. Nonetheless, 6 days after challenge with WNV, the animals previously immunized with 17-D virus demonstrated a strong secondary-type Flavivirus antibody response. Interestingly, the 17-D immune animals also had a stronger IgM response to WNV infection. These data indicate that 17-D vaccine gives only partial protection against challenge with WNV.