Guey-Chuen Perng

The role of natural killer cells in protection of mice against death and corneal scarring following ocular HSV-1 infection

C57BL/6 mice depleted of NK (natural killer) cells with anti-asialo-GM1 antibody were more susceptible to lethal HSV-1 ocular challenge (12% survival) than control C57BL/6 mice (100% survival), CD4+ depleted mice (100% survival), CD8+ depleted mice (80% survival), or macrophage depleted mice (85% survival). NK depletion also resulted in significantly higher levels of HSV-1 induced corneal scarring than was seen with any of the other groups. C57BL/6 mice depleted of NK cells with PK136 (anti-NK1.1 antibody which is more specific for NK cells than is anti-asialo-GM1 antibody) were also more susceptible to HSV-1 ocular challenge than T cell or macrophage depleted mice. Vaccination completely protected NK depleted mice against death and corneal scarring. In contrast to C57BL/6 mice, in BALB/c mice, NK depletion had no effect on survival or corneal scarring following ocular HSV-1 challenge. Experiments with IFN-gamma knockout mice (IFN-gamma(o/o) mice) suggested that IFN-gamma played a minor role in protection of naïve mice against death following HSV-1 challenge. However, IFN-gamma did not appear to be an important factor in protection against HSV-1 induced eye disease. Thus, protection against HSV-1 induced corneal scarring in naive mice appeared to be due to a non-INF-gamma NK function. Our results therefore suggest that NK cells were very important in protecting naive C57BL/6 mice but not vaccinated C57BL/6 mice against corneal scarring and death following ocular HSV-1 challenge.

The effect of latency-associated transcript on the herpes simplex virus type 1 latency-reactivation phenotype is mouse strain-dependent.

Herpes simplex virus type 1 (HSV-1) latency-associated transcript (LAT) null mutants reactivate poorly in the rabbit ocular model. The situation in mice is less clear. Reports concluding that LAT null mutants reactivate poorly in the mouse explant-induced reactivation (EIR) model are contradicted by a similar number of reports of normal EIR of LAT(-) mutants in mice. To determine if the EIR phenotype might be mouse strain-dependent we infected BALB/c and Swiss Webster mice with LAT(-) or LAT(+) virus and assessed EIR in individual trigeminal ganglia. Compared to LAT(+) virus, LAT(-) virus reactivated poorly in Swiss Webster mice (P<0.05). In contrast, the EIR phenotype of these viruses was similar in BALB/c mice (P>0.1). Thus, LAT appeared to have a much greater impact on the EIR phenotype in Swiss Webster mice than in BALB/c mice. The mouse strain therefore appeared consequential in the HSV-1 EIR phenotype in mice.

The Role of Interleukin (IL)-2 and IL-4 in Herpes Simplex Virus Type 1 Ocular Replication and Eye Disease

To assess the relative effect of interleukin (IL)-2- and IL-4-dependent immune responses on herpes simplex virus (HSV)-1 infection, naive, vaccinated, and mock-vaccinated IL-20/0 and IL-40/0 knockout mice were challenged ocularly with HSV-1. Naive IL-20/0 mice were significantly more susceptible to lethal infection than IL-40/0 or parental BALB/c mice. Vaccinated, IL-20/0, IL-40/0, and BALB/c mice induced similar neutralizing antibody titers and were completely protected against HSV-1-induced death and corneal scarring. Vaccinated and mock-vaccinated IL-20/0 mice had significantly higher HSV-1 titers in their eyes than BALB/c mice, while vaccinated and mock-vaccinated IL-40/0 mice had significantly lower HSV-1 titers in their eyes than BALB/c mice. Recombinant (r) IL-2 treatment of the IL-20/0 mice significantly reduced ocular HSV-1 replications, but rIL-4 treatment of IL-40/0 mice significantly increased ocular HSV-1 replications. Th1 (IL-2) cytokine responses may help protect mice against ocular HSV-1 challenge and reduce ocular HSV-1 replication.

Antibody-dependent enhancement of HSV-1 infection by anti-gK sera.

We previously reported that vaccination of BALB/c mice with the baculovirus expressed HSV-1 glycoprotein K (gK) or passive transfer of gK purified IgG to naive BALB/c mice causes severe exacerbation of HSV-1 induced corneal scarring following ocular challenge. In addition, a productive chronic infection, rather than a latent infection, is found in most trigeminal ganglia. These phenomena are accompanied by a very high T(H)1+T(H)2 response in the eye (Ghiasi, H., Cai, S., Nesburn, A.B., Wechsler, S.L., 1996. Vaccination with herpes simplex virus type 1 glycoprotein K impairs clearance of virus from the trigeminal ganglia resulting in chronic infection. Virology 224, 330-333; Ghiasi, H., Cai, S., Slanina, S., Nesburn, A. B., Wechsler, S.L., 1997. Nonneutralizing antibody against the glycoprotein K of herpes simplex virus type-1 exacerbates herpes simplex virus type-1-induced corneal scarring in various virus-mouse strain combinations. Invest. Ophthalmol. Vis. Sci. 38, 1213-1221; Ghiasi, H., Hofman, F.M., Cai, S., Perng, G.C., Nesburn, A.B., Wechsler, S.L., 1999. Vaccination with different HSV-1 glycoproteins induces different patterns of ocular cytokine responses following HSV-1 challenge of vaccinated mice. Vaccine 17, 2576-2582). In the studies reported here, we investigated the hypothesis that anti-gK serum produces antibody-dependent enhancement (ADE) of ocular HSV-1 infection. We found that gK vaccinated mice had significantly higher HSV-1 titers in their eyes than gD or mock-vaccinated mice and that anti-gK sera enhanced HSV-1 infection in the macrophage cell line U937. In addition, passive transfer of anti-gK sera to naive mice 24 h prior to ocular HSV-1 challenge also increased viral replication. These results were consistent with ADE of HSV-1 by sera to gK. This suggests that the severely exacerbated corneal disease seen following HSV-1 ocular challenge of gK vaccinated mice is a result of ADE. The ability of gK sera to cause harmful ADE may impact HSV-1 vaccine development.

Corneal macrophage infiltrates following ocular herpes simplex virus type 1 challenge vary in BALB/c mice vaccinated with different vaccines

Macrophage cell infiltrates in the cornea were examined following ocular herpes simplex virus type 1 (HSV-1) challenge of vaccinated BALB/c mice. Mice were vaccinated with individual HSV-1 glycoproteins, cocktails of different HSV-1 glycoproteins, or live avirulent HSV-1 (strain KOS). Cryostat sections of cornea were taken at different times after challenge and reacted with M1/70, F4/80, BM8, or MOMA-1 monoclonal antibodies. The pattern of macrophage responses in the cornea differed depending on the vaccine that was given prior to HSV-1 ocular challenge. No macrophage response was detected in mice vaccinated with the highly protective 5gPs consisting of the five glycoproteins gB, gC, gD, gE, and gI. In contrast, mock vaccinated mice and mice vaccinated with gK, which is known to exacerbate HSV-1 induced eye disease, had high sustained macrophage responses. Mice vaccinated with 7gPs (5gPs+gG and gH) had moderate levels of macrophages. It appeared that (1) the most effective vaccines induced no detectable infiltrating macrophages in the eyes, while the least efficacious vaccines had very high levels of infiltrating macrophages; (2) presence of CD11b(+) cells in the cornea appeared to correlate with enhanced blepharitis, but did not appear to affect corneal scarring; and (3) presence of F4/80(+) cells in the cornea tended to correlate with increased corneal scarring.

The UL3 open reading frame of herpes simplex virus type 1 codes for a phosphoprotein

Based on sequence analysis, the protein encoded by the UL3 open reading frame (ORF) of herpes simplex virus type 1 (HSV-1) was predicted to contain an N glycosylation site and to be a glycoprotein. To determine if this prediction was correct, we cloned and expressed the DNA encoding the complete sequence of the UL3 ORF in a baculovirus expression system. Western blotting was done using polyclonal antibody raised against synthetic UL3 peptides. Two major baculovirus-UL3 expressed protein bands with apparent molecular weights of 30 kDa and 31 kDa, and two minor protein bands with apparent molecular weights of 29 kDa and 33 kDa were detected. None of the expressed UL3 protein species were susceptible to tunicamycin treatment, suggesting that they were not N-linked glycosylated. Cell fractionation studies indicated that the UL3 protein was localized in the cytoplasmic and nuclear portion of the cells, rather than the cell membrane, again suggesting a lack of glycosylation. In contrast, the baculovirus expressed UL3 protein was phosphorylated as judged by 32Pi-labeling. Immunoprecipitation followed by SDS-PAGE demonstrated a single 32Pi-labeled UL3 related band with an apparent molecular weight of 33 kDa, indicating that the UL3 protein was a phosphoprotein. Antibodies produced in mice vaccinated with baculovirus-UL3 protein reacted with two UL3 related HSV-1 bands on Western blots. These protein bands had apparent molecular weights of 27 and 33 kDa and presumably represent the unphosphorylated and phosphorylated forms of UL3.

The persistent elevated cytokine mRNA levels in trigeminal ganglia of mice latently infected with HSV-1 are not due to the presence of latency associated transcript (LAT) RNAs

Trigeminal ganglia (TG) from mice latently infected with wild type HSV-1 contain detectable levels of cytokine transcripts that are not present in TG from uninfected mice. This suggests that during HSV-1 neuronal latency, the immune system is stimulated by the production of one or more viral proteins. Since the LAT (latency associated transcript) gene is essential for wild type levels of spontaneous reactivation and is the only highly active viral gene during latency, the stimulation of cytokines may indicate the presence of a LAT encoded latency protein. We therefore compared the cytokine transcript profiles in the TG of mice latently infected with wild type and LAT negative viruses. Mice were latently infected with either: (1) the LAT null mutant dLAT2903; (2) its marker rescued virus dLAT2903R; or (3) the parental wild type HSV-1 strain McKrae. As expected, reactivation following explant cultivation of TG from latently infected mice was significantly decreased with dLAT2903 (P < 0.05)(40 +/- 8%, n = 24) compared with dLAT2903R (85 +/- 7.6%, n = 36) or the parental virus (70 +/- 10.0%, n = 36). The relative levels of various cytokines was determined by RT-PCR analysis of TG extracts. None of the cytokine transcripts detected in mice latently infected with the wild type or marker rescued viruses were missing or decreased in mice latently infected with the LAT null mutant 30 or 60 days post infection. There were also no differences in the HSV-1 antibody titers induced by the LAT negative virus compared to the LAT positive viruses. Thus, although LAT facilitated reactivation of HSV-1 from explanted mouse TG, expression of LAT during latency did not appear to be involved in persistent cytokine expression in TG. This suggests that during latency, HSV-1 does not produce a highly antigenic abundant LAT encoded protein.