Vaccination with dl5-29 or gD2 is most effective as prophylaxis for acute, latent, and recurrent infection in guinea pigs. Animals received the vaccines on days 21 and 42 before challenge. On day 0, they were challenged vaginally with HSV-2 strain 333. (A) Virus shedding from the vaginal tract. Significant differences between groups are indicated by ★n. ★1, significantly lower titers on day 2 for dl5-29 recipients than for control animals and pgD2 recipients (each P < 0.01); ★2, significantly lower titers on day 4 for dl5-29 recipients than for dl5-29/pgD2 recipients (P = 0.02); ★3, on day 6, significantly lower titers for dl5-29 recipients than for control mice (P = 0.02), pgD2 mice (P = 0.02), and gD2 recipients (P = 0.01); ★4, on day 8, significantly higher titers for control mice than for dl5-29 and pgD2 mice (P = 0.03 each). (B) Group mean lesion scores during acute disease. Recipients of dl5-29 alone, gD2, and dl5-29/pgD2 had milder acute disease than animals given two doses of pgD2 DNA and control animals (each P < 0.01). (C) Mean cumulative numbers of recurrent lesions. The number of recurrences in recipients of dl5-29 alone, gD2, and dl5-29/pgD2 was significantly lower than that in unvaccinated, control guinea pigs (each, P < 0.001), while recurrences in pgD2 recipients remained higher than those in dl5-29 and gD2 recipients (P = 0.02 and 0.03, respectively). (D) Latent viral DNA loads in pooled sacral ganglia. The broken line indicates the limit of detection of this quantitative PCR assay (6 copies of viral DNA/100 ng of DNA). A viral load of <6 copies was assumed to be 3 copies of viral DNA/100 ng of DNA. The horizontal thick and vertical thin bars show the geometric mean numbers of viral genome copies and SEs, respectively. The viral load in control guinea pigs was significantly higher than that in recipients of dl5-29 alone (P = 0.001), dl5-29/pgD2 (P = 0.01), and gD2 (P < 0.001), while pgD2 recipients had higher viral loads than dl5-29 and gD2 recipients (P = 0.03 and 0.002, respectively). Differences between results for gD2 and dl5-29 recipients were not significant (P = 0.21).

When given intravaginally, the dl5-29 vaccine is partially protective but impaired for latency. All animals were vaccinated on days 21 and 42 before challenge. Animals were given PBS mixed with CFA or IFA, dl5-29 vaginally, pgD2, or gD2. Mock-challenged animals were given two doses of dl5-29 vaginally and then challenged with uninfected Vero cell lysate. On day 0, animals other than the mock-challenged group were challenged vaginally with HSV-2 strain 333. (A) Group mean cumulative numbers of recurrent lesions. Control animals experienced more recurrences than recipients of dl5-29 vaginally, pgD2, and gD2 (P = 0.01, 0.04, and <0.001, respectively). gD2 vaccination was superior to vaccination with pgD2, but not significantly (P < 0.07), and was significantly superior to vaginal administration of dl5-29 (P < 0.01). Mock-challenged animals had no recurrences. (B) Latent viral DNA load in pooled sacral ganglia. Thick and thin bars indicate the geometric mean and SE, respectively, for each group. The viral load of gD2 recipients was significantly lower than that of PBS(CFA/IFA), vaginal dl5-29, and pgD2 recipients (P < 0.001, P = 0.001, and P = 0.02, respectively). Animals administered dl5-29 vaginally but mock challenged had no detectable latent viral DNA. Their geometric mean DNA load, assumed to be ≤3 copies of viral DNA/100 ng of DNA, was significantly lower than that for all other groups (P < 0.001).

Immunotherapeutic vaccination of infected guinea pigs. All animals were infected vaginally with HSV-2 strain 333 on day 0. Thereafter, animals were given PBS(CFA/IFA), gD2(CFA/IFA), or dl5-29 on days 7 and 15. (A) Group mean lesion scores indicate comparable levels of acute disease for all groups. (B) Cumulative numbers of recurrent lesions. Recurrences were significantly less frequent for dl5-29 recipients than for control animals (P = 0.04), while there was a trend toward a lower rate of recurrences for gD2 recipients than for controls (P = 0.15). The differences between gD2 and dl5-29 recipients was not significant (P = 0.54).

The latent viral load correlates with acute disease severity and recurrence rates. (A) The mean acute lesion severity scores for every guinea pig studied in this report are plotted against the latent viral load in each of their pooled sacral ganglia. (B) The latent viral load in pooled ganglia for each of the animals is plotted against their cumulative numbers of recurrences during 90 days. Results for animals having no recurrences are represented as log₁₀ 1. The solid line indicates the best linear fit among the data points, while the broken line indicates the limit of detection of the real-time PCR assay.

Heat-killed dl5-29 is not protective in mice. Mice were vaccinated with two doses each of dl5-29 (10⁶ PFU), pgD2 DNA (25 μg), and heat-killed HSV-2 strain 333 (10⁶ PFU) 3 weeks apart or were left unvaccinated. Animals were challenged with wild-type HSV-2 strain 333 (10⁶ PFU), and survival was scored from days 0 to 14 after infection.

HSV-2-specific IFN-γ⁺ CD8⁺ T-cell responses are greater after vaccination with dl5-29 than with gD2. Mice were given two doses each of PBS(CFA/IFA), gD2(CFA/IFA), or dl5-29 10 days apart. (A) Spleen cells were harvested at the indicated days after the second vaccination. Bars show the mean (±SE) number of IFN-γ⁺ CD8⁺ T cells per 10⁶ splenic CD8⁺ T cells responding to HSV-2-infected P815 cells at the indicated time points. An asterisk means that the procedure was not done. (B) All mice were challenged with HSV-2 strain 333 in the right eye 2 weeks after the second vaccination. At the indicated time points, ipsilateral TG were removed, pooled, dispersed into single cells with collagenase, and stimulated with HSV-2-infected P815 cells. Dot plots of flow cytometry, along with the numbers of infiltrating IFN-γ-producing (lower right) and total CD8⁺ T cells (lower left), are shown. Symbols: diamonds, ganglia from control animals; circles, gD2 recipients; squares, dl5-29 recipients.