As a first step in determining whether there is an age-associated decline in immune control of the persistent γ-herpesviruses, we monitored latent viral load in γHV68-infected mice over time following infection. We used two independent assays- the limiting dilution PCR assay (LDA/PCR) which allows determination of the frequency of latently-infected cells 19 20 and the quantitative PCR assay which allows determination of genome copy number within a standard amount of genomic DNA 21 . Latency assessed by either method was shown to decline after the early peak of latency in the spleen and then stabilize for more than a year after infection (Figure 1). Consistent with the stable latent load in aged mice, we failed to detect recrudescent lytic virus in the lung, spleen or a variety of other anatomical sites (data not shown). These data suggest that immune control over γHV68 latency is maintained with age.

It has been shown that mice latently infected with γHV68 are protected from a second, homologous viral challenge 22 . Therefore, as another way of assessing protective anti-viral immunity in aged γHV68-infected mice, we tested whether these mice could become re-infected with homologous virus. Young mice (3 months post infection) and aged mice (19 months post infection) were intranasally challenged with 400 or 3000 PFU γHV68, and lytic viral titers were monitored in the lung at 3 and 6 dpi (Figure 2). The data show high viral titers in young naïve mice at 3 and 6 dpi, as expected. However, only a low level of viral replication was detected at 3 dpi in the lungs of both young and aged latently infected mice and the level of viral replication did not vary with the dose of the challenge inoculum. By 6 dpi, virus was mostly cleared from the lungs of mice in both age groups. Similarly, there was no significant increase in viral titers in the spleens of either young or aged latently-infected mice after re-infection (data not shown). Together, the absence of viral recrudescence with increasing age and the comparable resistance of young and aged mice to re-infection with homologous virus suggest that there is no overall loss of immune control of γHV68 with age.

As it has been shown that both cellular and humoral immunity contribute to immune control of γHV68 17 19 23 24 25 26 27 28 , we sought to investigate the impacts of ageing on the functionality of each arm of the immune system during viral latency. In order to assess the impact of ageing on cellular immunity, we determined the numbers of γHV68-specific CD8 T cells present in latently-infected mice at different timepoints after infection using MHC class I tetramers specific for two well-characterized γHV68 epitopes p56 (ORF6_487-495) and p79 (ORF61_524-531)(Figure 3). The data show that p56- and p79-specific CD8 T cells are maintained in latently-infected mice over time, up to 21 months post infection. Although occasional individual mice were found to express higher than normal numbers of tetramer-positive cells as early as 6 months post infection, in general we did not observe an age-associated development of virus-specific T cell inflation as has been described for CMV infection 29 30 31 , or virus-specific clonal expansions as we had previously observed in Sendai virus-infected mice 32 . We also assessed the cytolytic activity of the ORF61-specific CD8 T cells present in latently-infected mice at various timepoints following infection using an in vivo CTL assay. As shown in Figure 4, the cytolytic activity against ORF61-pulsed targets measured in individual mice remained stable for up to two years after infection. In addition, in agreement with a previous report 33 , the numbers and CTL function of γHV68-specific CD8 T cells specific for ORF-6-pulsed targets were also maintained with age (data not shown).

We next determined the impact of ageing on humoral immunity. It has previously been shown that following intranasal infection with γHV68, virus-specific IgM antibodies peak on day 10 and subsequently subside, and class-switched antibody levels increase over the first 3 weeks of infection and are then maintained at constant levels for at least 90 days 34 35 . As shown in Figure 5A, we found that the serum titers of virus-specific IgG were similar between recent (1-3 months p.i.) and long-term (18-22 months p.i.) latently infected mice. However, serum neutralization titers were found to be greatly reduced in the majority of mice 18-22 months postinfection (Figure 5B). The reduction in neutralizing activity was an unexpected result, as humoral immunity has been shown to be remarkably long-lived for other viruses, including small pox and influenza virus 36 37 38 . This loss of neutralizing activity with age was specific for γHV68 infection, as serum antibody titers and neutralizing activity were maintained long-term in influenza virus-infected mice (1-3 and 18-22 m.p.i; Figure 5C, D). Interestingly, the loss of neutralizing activity was not restricted to aged mice, rather the decline was gradual over the course of γHV68 infection (Figure 6). Importantly, the decline in neutralizing titer had functional implications, as passively-transferred sera from aged γHV68-infected mice was less efficient in protecting naïve mice from de novo γHV68 challenge than sera transferred from young γHV68-infected mice (Figure 7). Taken together, whereas we found no defect in T cell immunity or loss of overall immune control of latency, there was a clear age-associated change in humoral immunity over the course of latent γHV68 infection.

Despite the decline in neutralizing antibody titers, our data indicate that T cell function is intact and immune control over γHV68 latency is maintained with ageing. This observation is consistent both with our previous finding that T cells can control reactivation in the absence of neutralizing antibodies 19 and the notion that T cell memory generated when young retains function into age 39 40 ). However, de novo infection in the absence of CD4 help has been shown to generate defective CD8 T cell memory 41 42 43 , and the degree of CD4 dependence may vary with the pathogen. Specifically for γHV68, it has been shown that CD4-deficient mice are able to clear lytic virus and viral latency is established normally, however these mice progressively lose immune control over viral latency resulting in recrudescence of lytic virus around 42 dpi 17 . Because there is a well-established age-associated decline in CD4 T cell function 44 45 46 47 48 49 , we examined the ability of aged naïve mice to control de novo infection as another measure of the impact of ageing on functional γHV68-specific immunity. The data show that clearance of lytic virus from the lungs was only slightly delayed (not statistically significant) in aged mice compared to young mice following the initial infection (Figure 8) and that comparable, stable levels of latent virus were detected in the spleens of both groups of mice, as assessed by both infective center and genome copy assays (Figure 9). However, analysis of γHV68-specific CD8 T cell responses elicited in infected mice using MHC class I tetramers revealed generally reduced numbers of virus-specific p56 and p79 CD8 T cells in aged compared with young mice after de novo infection (Figure 10).

Analysis of virus-specific serum antibody titers at 15, 30 and 60 dpi showed that although titers of virus-specific IgM were comparable in young and aged mice early after infection (days 15 and 30), virus-specific IgM titers were increased in aged mice at 60 dpi (Figure 11A), consistent with an age-associated deficiency in the generation of class-switched antibodies at all timepoints examined (Figure 11B). In addition, serum neutralization titers were significantly compromised in aged mice at all timepoints examined (Figure 11C). Taken together, despite statistical differences in cellular and humoral immunity after infection of aged compared with young mice, the aged mice were able to control infection comparably to young mice (Figures 8 and 9).

Young (6-8 weeks) and aged (≥ 18 months) C57BL/6 mice and B6.SJL-Ptprc^aPepc^b/BoyJ (B6.SJL, CD45.1) congenic mice were obtained from the Trudeau Institute animal breeding facility. All mice were housed under specific pathogen free conditions before and after infection. All animal procedures were approved by the Institutional Animal Care and Use Committee at the Trudeau Institute.

Clone WUMS of γHV68 14 was propagated in NIH-3T3 cells. Virus titers were determined by plaque assay on NIH-3T3 cells. Influenza virus A/HK-x31 (×31, H3N2) was grown, stored, and titrated as described previously 87 . Female mice were anesthetized with 2,2,2-tribromoethanol (200 mg/kg) prior to intranasal (i.n.) infection with 400 PFU γHV68 or 300 EID₅₀ of A/HK-x31 influenza virus.

Peripheral blood (100-200 μl), obtained by nicking the tail, was collected in PBS containing 10 U/ml heparin (Sigma-Aldrich, St. Louis, MO). Single cell suspensions from the spleens of individual mice were prepared by mechanical disruption and straining through nylon mesh, and lung cells from individual mice were processed with collagenase D. Samples were depleted of erythrocytes by treatment with buffered ammonium chloride solution. Lymphocytes were enriched from lung cell suspensions by Percoll gradient centrifugation (Hogan, 2002) and from spleen cell suspensions by panning against goat anti-mouse IgG (Jackson ImmunoResearch Laboratories, West Grove, PA). Isolated cells were incubated with Fc block (BD Biosciences, San Jose, CA) for 15 minutes on ice, and then stained with PE- or APC-labeled MHC class I tetramers specific for γHV68 ORF6_487-495/D^b or ORF61_524-531/K^b for one hour at room temperature. All tetramers were generated by the Molecular Biology Core Facility at the Trudeau Institute, as previously described (Altman, 1996). Tetramer-labeled cells were then stained with fluorescently-conjugated antibodies specific for mouse CD8α and mouse CD44 (BD Biosciences) for 30 minutes on ice, washed, and then fixed in 1% paraformaldehyde. Data were collected on a Becton Dickinson FACSCalibur flow cytometer (BD Biosciences) and then analyzed using FlowJo software (Tree Star, Inc., Ashland, OR).

The concentration of lytic virus in lung tissue was determined using a standard plaque assay on NIH-3T3 mouse fibroblasts 17 . Lung tissue obtained at various times post infection was mechanically homogenized and serial-dilutions of the homogenates were prepared. Diluted homogenates were incubated on monolayers of 3T3 cells for one hour at 37°C/10% CO₂, after which the monolayers were overlaid with carboxymethyl cellulose (Sigma-Aldrich) and incubated again at 37°C/10% CO₂. Six days later, monolayers were fixed with methanol, stained with 8% Giemsa stain, and plaques counted.

The frequency of latently infected cells capable of spontaneous in vitro reactivation was assessed using an infective center assay, as previously described 88 . 10-fold serial dilutions (in triplicate) of splenocytes starting at 10⁶ cells/well were plated onto monolayers of NIH-3T3 mouse fibroblast cells. Monolayers were incubated overnight at 37°C and then overlaid with carboxymethyl cellulose. Plaques were quantitated 6 days later after methanol fixation and Giemsa staining. Samples were also assayed following one cycle of freeze/thaw to determine the contribution of lytic virus to the overall viral titers. The number of latently infected cells was then calculated as the difference between the total number of infected cells and the number of lytically infected cells.

The frequency of cells carrying viral genome was determined using a limiting dilution nested PCR assay (LDA/PCR) for the γHV68 open reading frame 50 (ORF50) gene as described 19 89 . Briefly, CD19⁺ B cells were purified from the spleens of infected mice as previously described 90 . B cells were resuspended in isotonic buffer and starting at 10⁴ or 10⁵ cells/well, were diluted with uninfected NIH-3T3 cells, and then transferred into 96 well plates. Twelve replicate reactions were performed for each cell dilution per experiment. Subsequently, the ORF50 gene was amplified from cell lysates with two rounds of nested PCR. The final PCR product was electrophoresed on a 3% agarose gel and stained with ethidium bromide. The reciprocal frequency of cells carrying viral genome was determined using linear regression with a 95% degree of confidence.

Genome copy was estimated as previously described 21 , with the following changes. DNA was extracted from 5 × 10⁶ B cells using a DNeasy kit (Qiagen), according to manufacturer's protocol. ORF50 gene copy number was determined for 200 ng of DNA per sample, using a standard curve quantitation method on an Applied Biosystems 7500 Real-Time PCR System (Applied Biosystems, Foster City, CA). Reactions were run in duplicate, and naïve C57BL/6 splenocyte DNA was used as a negative control. A low ROX AB Taqman Gene Expression master mix (Applied Biosystems, Foster City, CA) was used, providing repeatable detection of copy numbers as low as 3 per 200 ng of DNA. Primers, probes, and reaction cycles used are described 21 .

Virus specific IgG titers in sera were determined by ELISA 19 . Nunc ImmunoMaxisorp plates were coated with purified virions at a concentration of 0.5 μg/well. Following an overnight incubation at 4°C, plates were washed with PBS-Tween (0.05%), and subsequently blocked with PBS/BSA (3%) overnight at 4°C. Dilutions of sera, starting at 1/20, were prepared in PBS/0.05%Tween/0.5%BSA and added to the antigen coated plates. After an overnight incubation at 4°C, virus-specific IgG was detected using alkaline phosphatase-conjugated goat anti-mouse IgG antibody (Sigma-Aldrich) and p-nitrophenyl phosphate (Sigma-Aldrich). Optical densities were read at 405 nm using a VMax^® microplate reader (Molecular Devices, Sunnyvale, CA). Serum antibody titers are expressed as the highest reciprocal dilution of serum giving an OD₄₀₅ reading at least two times greater than the normal mouse serum control.

For the γHV68 neutralization assay, virus (10² PFU) was incubated with serial dilutions of heat-inactivated serum samples in 96-well plates for 1 h at 37°C. Following the incubation, virus/sera mixtures were transferred to 96 well flat-bottom plates containing monolayers of 3T3 cells. After 7 days of incubation at 37°C, 10% CO₂, the monolayers were fixed and plaque formation revealed by staining with 5% Giemsa stain. Neutralization titers are expressed as the highest reciprocal dilution of sera required to cause a 50% reduction in CPE (cytopathic effect).

For the influenza virus neutralization assay, ×31 virus (100 EID₅₀/well) was incubated with serial dilutions of heat-inactivated sera, prepared in Zero-Serum media (Diagnostic Hybrids, Athens, OH) containing 4 μg/ml trypsin, in 96-well round-bottom plates for 1 h at 37°C. Following the incubation, virus/sera mixtures were transferred to 96 well flat-bottom plates containing monolayers of MDCK cells. Plates were then centrifuged at 800 × g for 1.5 hrs. Following centrifugation, the virus/sera mixtures were removed, fresh media added to each well, and the plates incubated at 33°C, 10% C0₂ for 14-16 hours. Monolayers were fixed in 80% acetone and infected cell foci were revealed by incubating with a combination of biotinylated mouse anti-influenza A monoclonal antibodies (Millipore, Billerica, MA), alkaline-phosphatase conjugated goat anti-mouse IgG (Sigma-Aldrich), and Sigma-Fast BCIP/NBT substrate (Sigma-Aldrich). Neutralization titers are expressed as the highest dilution of sera prior to which the number of foci observed is equivalent to the number of foci counted in normal mouse sera/virus-only control wells.

Serum obtained from latent γHV68-infected young (2-4 months post infection) and aged (18-24 months post infection) C57BL/6 mice was administered intravenously to young, naive C57BL/6 mice one day prior to intranasal γHV68 infection (400 PFU). As a control, mice were given normal mouse serum or convalescent serum obtained from C57BL/6 mice previously infected with influenza A/PR8/34 virus (300 EID₅₀). Lungs were harvested from recipient mice on day 5 post infection and viral loads were measured using the standard γHV68 plaque assay.

Target splenocytes harvested from B6.SJL (CD45.1⁺) congenic donor mice were incubated with peptides (10 μg/ml) at 37°C in 10% CO₂ for 5 h with occasional mixing. The peptides included γHV68 ORF61_524-531 and an irrelevant peptide (either influenza NP_366-374 or Sendai NP_324-332). Cells were washed and labeled with 2.5 μM or ≤ 0.5 μM of 5,6-carboxyfluorescein diacetate succinimidyl ester (CFSE; Molecular Probes, Eugene, OR) to obtain CFSE^hi (ORF61_524-553) and CFSE^low (negative control peptide) groups. Following CFSE labeling, cells were combined at equal ratios, washed three times in PBS, and resuspended at a final concentration of 2 × 10⁸ cells/ml. Twenty million cells (100 μl) were injected intravenously into C57BL/6 mice (CD45.2⁺) previously infected with γHV68, or naïve mice as a negative control. At 4 h or 16-17 h post transfer, spleens were harvested from recipient mice and single cell suspensions were prepared for flow cytometric analyses as described above. The individual populations of peptide-pulsed donor CD45.1⁺ cells present in the spleens of recipient mice were identified and enumerated using a combination of APC-conjugated antibodies specific for mouse CD45.1 (eBioscience, San Diego, CA) and CFSE staining intensity. Percent specific lysis was calculated using the following formula: [1-(ratio uninfected/ratio infected)] × 100, where "ratio" refers to the percentage of irrelevant peptide-pulsed cells divided by the percentage of relevant peptide-pulsed cells.

Where indicated, groups were compared statistically using the Student's t-test, the nonparametric Mann-Whitney rank test, or linear regression, or the Chi-square test. P values ≤ 0.05 were considered significant. All statistical analyses were performed using Prism software (GraphPad, La Jolla, CA).