C57BL/6 (B6, H-2^b) and TCRβKO male mice were purchased from The Jackson Laboratory (Bar Harbor, ME, USA) and Taconic Farms (Germantown, NY, USA), respectively. Mice were used at 2–8 months of age. All mice were maintained under specific pathogen-free conditions in the University of Massachusetts Medical School, Department of Animal Medicine. All experiments were done in compliance with institutional guidelines as approved by the Institutional Animal Care and Use Committee at the University of Massachusetts Medical School.

Lymphocytic choriomeningitis virus (clone 13) is a RNA virus in the Old World Arenavirus family, and was propagated in BHK21 baby hamster kidney cells (8).

Mice were inoculated i.v. with low dose, 2 × 10⁴ pfu; medium dose, 2 × 10⁵ pfu; and high dose, 2 × 10⁶ pfu of LCMV clone 13. Control naïve mice were either left uninoculated or were inoculated with HBSS. The control mice were always age matched to the experimental group and housed exactly the same in pathogen-free conditions.

Lymphocytic choriomeningitis virus titers in 10% tissue homogenate from spleens and serum were determined by plaque assays on ATCC vero cells, as described elsewhere (9).

Lymphocytic choriomeningitis virus epitopes NP₃₉₆, D^b (FQPQNGQFI), GP₃₃, D^b (KAVYNFATC), NP₂₀₅ epitope (YTVKYPNL). Synthetic peptides were provided by Biosource International (Camarillo, CA, USA) or 21st Century Biochemicals (Marlboro, MA, USA) and used at a 90% level of purity.

Single cell suspensions were prepared from splenocytes or peripheral blood. Erythrocytes were lysed with 0.84% NH₄Cl solution. FACS staining was done as previously described in 96-well plates with fluorochrome-labeled mAbs, anti-CD8 (clone 53-6.7, BD Pharmingen), anti-CD44 (clone IM7), and PD-1 (clone J43). Tetramer staining was done as previously described using phycoerythrin (PE) and/or allophycocyanin (APC) labeled tetramers (10). Samples were analyzed with a Becton Dickinson FACSCalibur flow cytometer (San Jose, CA, USA) or a LSRII (San Jose, CA, USA) and FlowJo software (Tree Star, Inc., Ashland, OR, USA). All surface mAbs were purchased from BD Pharmingen, San Diego, CA, USA or eBioscience, San Diego, CA, USA. MHC class I peptide tetramers specific for LCMV-NP₂₀₅/K^b, LCMV-NP₃₉₆/D^b, and LCMV-GP₃₃/D^b were prepared as previously described (10).

Cells (10⁶) were stimulated either with medium, or 1–5 μM synthetic peptide as previously described (10). Intracellular cytokine-producing cells were detected with anti-IFNγ (clone XMG1.2), anti-TNFα (clone MP6-XT22), and anti-IL-2 (clone JES6-5H4) mAbs. IgG-isotype mAbs were used in the same assay. The mAbs were purchased from BD Pharmingen, San Diego, CA, USA or eBioscience, San Diego, CA, USA. The samples were analyzed as described above.

Mice were injected i.p. on day-1 and -7 of LCMV clone 13 infection with 100 μg of CD8 (clone 2.43) antibody.

At day 14 after LCMV challenge, lungs and livers from LCMV clone 13-infected mice were collected, fixed in 10% neutral buffered formaldehyde and paraffin-embedded. Tissue sections (5 μm) were stained with hematoxylin and eosin and analyzed microscopically by a pathologist. Scoring of lung pathology was graded based on a scale outlined below and based on previously published studies (11–13). Scoring of lung pathology was graded by one pathologist, who was blinded in regards to treatments the mice had received.

The scoring on each individual mouse was done on four to five different sections of the lung representing four to five different lobes of the whole lung assessing both qualitative and quantitative changes in histology. The histology photographs showing high power views of a small portion of one lobe of the lung demonstrate examples of the types of pathology observed in different treatment groups, but evaluations of the whole lung were used for scoring.

Serum alanine aminotransferase (ALT) levels were determined using an ALT UV-kinetic method reagent set (D-Tek, Bensalem, PA, USA). Fifteen microliters of serum was mixed with 150 μL of substrate and absorbance at was read at 340 nm. ALT concentration was calculated as the average change in absorbance/min multiplied by 1768. ALT (IU/L) is equal to the change in absorbance/min multiplied by total reaction volume (0.165 mL) *1000 to convert IU/mL to IU/L. This is divided by the millimolar absorptivity of NADH (6.22) multiplied by the sample volume (0.015 mL) multiplied by the light path (1.0 cm). This equation simplifies into the change in absorbance/min multiplied by 1768.

The one-way ANOVA test with a Bonferroni post-test was used when there were more than two groups. The Student’s t-test was used as indicated in the figures when only comparing two groups. The Mantel–Cox Mortality test was used for mortality studies.

We infected C57BL/6 mice intravenously with low dose (2 × 10⁴ pfu), medium dose (2 × 10⁵ pfu), and high dose (2 × 10⁶ pfu) LCMV clone 13 i.v. There was significantly higher mortality in the medium dose group, with only 25% of mice surviving at Day 13 (Figure 1A). Both the medium and high dose groups experienced a rapid and significant >20% weight loss beginning at day 6 after infection, with the high dose group starting to recover after day 10 while the medium dose group did not recover even by day 13 (Figures 1B,C). All of the mice that survived the medium dose infection were persistently infected with LCMV, similar to the high dose group. The dramatic immunopathology was mediated by the T cell response, as demonstrated by the fact that TCRβ KO mice infected with the medium dose did not have any significant weight loss or death, in contrast to WT controls (Figures 1A,C). The TCRβ KO mice had a slightly higher viral load than the WT controls at day 13 post infection and yet did not have pathology (log10 pfu/spleen: B6 5.7 ± 0.13 SEM; TCRβ KO 6.3 ± 0.13; n = 5/group; p < 0.02). The CD8 T cells were the major T cell population involved in mediating this effect, as demonstrated by depletion of CD8 T cells with mAb. B6 mice depleted of CD8 T cells prior to infection with the medium dose of LCMV also had significantly less weight loss and decreased mortality than the WT controls (Figure 1C) without a significant difference in viral load in the spleen at day 13 post infection (log10 pfu/spleen: B6 4.9 ± 0.5 SEM; anti-CD8 tx 5.3 ± 0.07 n = 5–9/group).

Persistent infection leads to a disruption of the normal immunodominance hierarchy and function of CD8 T cell responses during high dose LCMV clone 13 infection referred to as clonal exhaustion (6, 14). Clonal exhaustion occurs by a stepwise loss of function due to high antigen load and loss of CD4-help. Cells first lose cytotoxic abilities, then IL-2 production followed by the ability to produce TNFα and finally IFNγ. The final stage of exhaustion is the deletion of antigen-specific cells by apoptosis making exhaustion both a qualitative and quantitative. CD8 T cell functional impairment occurs in a hierarchical fashion in chronically infected mice. Production of IL-2 and the ability to lyse target cells in vitro are the first functions compromised, followed by the ability to make TNFα, while IFNγ production is most resistant to functional exhaustion. Antigen appears to be the driving force for this loss of function, since a strong correlation exists between the viral load and the level of exhaustion (15). Epitopes presented at higher levels in vivo result in physical deletion, such as NP396, while those presented at lower levels, such as GP33 and NP205, induce functional exhaustion. These published data would suggest that antigen levels drive the hierarchical loss of different CD8 T cell effector functions during chronic infection, leading to distinct stages of functional impairment and eventually to physical deletion of virus-specific T cells. Thus, we determined the functionality of the epitope-specific CD8 T cells in the three groups of mice by intracellular cytokine staining (ICS) for the presence of IFNγ- and/or TNFα-producing cells and calculating the ratio of IFNγ+TNFα+/IFNγ+ cells at days 7 and 13 post infection. The high dose clone 13-infected mice had significant exhaustion of NP396-, GP33-, and NP205-specific responses, as defined by loss of the double positive IFNγ/TNFα producing cells by day 13 (Figures 2A–E). The medium dose mice had evidence of only partial exhaustion of their CD8 T cell response in this same time frame. They had significantly more of the double positive IFNγ/TNFα producing T cells by all three epitope-specific CD8 T cell responses on day 7 and day 13 than did high dose-infected mice, although significantly less than the low dose group. Expression of the prototypic cell surface marker of exhaustion, PD-1 (14, 16) on epitope-specific CD8 T cells at day 10 post infection was intermediate in the medium dose as compared to the high and low dose groups, consistent with partial exhaustion (Figure 2F). There was also a significantly greater portion of the functional LCMV-GP33 and -NP205-specific CD8 T cells in the medium dose mice producing three cytokines including IL-2 (Figure 2G). These results would suggest that at the medium dose of virus the viral load is high enough to result in a partial exhaustion phenotype (6) and these sub-optimally functioning CD8 T cells are able to mediate the induction of severe immunopathology leading to death.

The lung immunopathology in the surviving medium dose mice as assessed by histology demonstrated severe pulmonary edema and interstitial infiltrates with consolidation, enhanced bronchus associated tissue (BALT), and necrotizing bronchiolitis (Figure 3A). Using our established method of scoring lung pathology (11–13) the medium dose mice had significantly more pathology than either the high and low dose mice or the TCRβ KO and the CD8 T cell-depleted mice, all of which had minimal interstitial infiltrates, consistent with this being CD8 T cell-mediated pathology (Figure 4A).

The livers of the medium dose mice also had significantly increased liver pathology (Figures 3B and 4B) as demonstrated by increased liver enzyme, alanine aminotransferase (ALT), in their serum as compared to low dose, TCRβ KO, and CD8 T cell-depleted mice. Histological sections of the liver showed serious bridging necrosis in the medium dose mice (Figure 3B), while the high and low dose mice as well as the TCRβ KO and CD8 depleted medium dose mice had only periportal and sinusoidal inflammatory infiltrates. We also examined the lymph nodes of these mice and found that with all three doses of LCMV clone 13 they essentially maintained their normal architecture meaning that there was recruitment of lymphocytes in T cell zones and increased numbers of germinal centers with infection but there was no overall destruction or necrosis of the lymphoid structures as has been reported for high dose LCMV-WE infection (17, 18) (data not shown).