Long-term gene expression in memory cells is not due to viral latency. B6 mice were infected with 2 × 10⁵ PFU of LCMV Armstrong. (A) Intracellular granzyme B expression was analyzed on days 8 and 60 postinfection for the indicated epitope-restricted populations (black line). The isotype control staining of the same cell populations also is shown (dashed line). The profile shown is representative of results obtained from three individual mice. (B) An in vivo cytotoxicity assay also was performed on days 8 and 60 postinfection, and specific cytotoxicity is depicted; the results shown are the mean of data from three individual mice, and the error bars indicate standard errors of the means.

Impact of high-dose naïve Tg transfers on the endogenous response. B6.Ly5.1 mice left untreated or were injected with 5 × 10⁵ P14 Tg cells (Ly5.2⁺) and were infected simultaneously with 2 × 10⁵ PFU of LCMV Armstrong and studied at days 8 and 60 after infection. (A) CD62L and IL-7R expression in cells of with different peptide specificities at day 8 after infection. Histograms compare CD62L and IL-7R expression levels of CD8 cells with the indicated peptide specificities in 1 P14 injected (inj.) (open graphs) and 1 noninjected B6 mouse (gray) of 12 mice studied in two independent experiments. On the far left, P14 cells (open histogram) are compared to GP33-specific noninjected mice. (B) Variation of CD62L expression in GP33-specific cells 2 months after infection. Graphs compare Tg cells (upper) to endogenous cells present in the same mouse (middle). The lower graphs show endogenous cells in the mice that were not injected with P14 cells. (C) Absolute numbers of cells of different peptide specificity at day 8 (left) and 2 months (right) after infection. Results show individual mice from one experiment out of two with equivalent results. (D) IFN-γ expression after in vitro stimulation with NP396 and GP276 peptides at day 8 (left) and 2 months (right) after infection.

Rapid progress of CD8⁺ T-cell differentiation during early infection can be monitored using a combination of TCR Tg cells, CFSE division profiles, and CD69 expression. B6.Ly5.1 mice were injected with 5 × 10⁵ P14 Tg cells (Ly5.2⁺) and infected with 2 × 10⁵ PFU of LCMV Armstrong. (A) At day 3 postinfection, splenocytes were stained with CD69, and division profiles were analyzed by the evaluation of CFSE expression. Cells that had divided four or more times were sorted into two subpopulations (CD69⁺ and CD69⁻) based on the indicated gates. (B) Gene expression in individually sorted CD69⁺ and CD69⁻ P14 cells was analyzed on day 3 postinfection; the results are presented in the same format as that used for Fig. 1B. (C) The progression of gene expression in individually sorted P14 cells was analyzed as the linear maturation of cells occurred (i.e., in the following sequence: day 3 CD69⁺ cells [white]→ day 3 CD69⁻ cells [grey]→ day 4 [black]). Statistically significant differences between progressive differentiation stages are marked (** and ***, P < 0.01 and P < 0.001, respectively.

Differentiation patterns of CD8 cells recognizing dominant and subdominant LCMV epitopes show marked similarity. B6 mice were infected with 2 × 10⁵ PFU of LCMV Armstrong. (A) The number of NP396-tet⁺, GP33-tet⁺, and GP276-tet⁺ cells in the spleen was analyzed at different time points after infection. The results shown are the means of values from three to six mice tested in two separate experiments, and the error bars indicate one standard error above the means (SEM). (B) Individual cells of each epitope specificity were recovered at the indicated times (days) postinfection from six individual mice in two independent experiments and were tested directly ex vivo for the coexpression of the indicated effector mRNAs. Forty-five to 90 cells of each specificity were evaluated per time point. Only wells that were positive for CD3ɛ (indicating that they contained a cell) were included in the analysis. Since we did not find significant variation between mice and between experiments, the data were pooled. Each horizontal row represents the pattern of gene expression in the same single cell; representative results from 40 cells are shown. Gene expression is indicated in black, and negative results are shown in white. Cells are ordered by the number of cytotoxic effector genes they expressed. The percentages at the bottom of each column represent the frequency at which the indicated gene was expressed in the whole population analyzed. (C) On day 8 postinfection, NP396-, GP33-, and GP276-reactive cells were identified by tetramer staining, and granzyme B expression in each population was analyzed directly ex vivo by intracellular staining. The filled histograms represent granzyme B staining, and the white histograms show the staining of the same populations with an isotype-matched control antibody. (D) The number of mRNAs for cytotoxic effector genes (Prf1, Gzma, Gzmb, and FasL) coexpressed by each cell was calculated (0 to 4). The results are expressed cumulatively as the percentage of cells specific for a given epitope that coexpressed mRNAs for ≥1, ≥2, ≥3, or 4 of these genes at the indicated times postinfection (a cell expressing two genes would be included in both the ≥2 and ≥1 categories. Statistically significant differences (as determined using Fisher's exact test) are marked (*, P < 0.05). (E) Phenotypes of LCMV- specific T cells at day 8 after infection. Graphs are from one individual mouse out of six mice studied in two independent experiments showing overlapping results. (F) Phenotypes of LCMV-specific memory cells. Results are from one mouse out of seven studied in three independent experiments. We found considerable variation in the expression of CD62L.

Differentiation patterns of adoptively transferred TCR Tg cells and endogenous cells recognizing the same epitope in the same mice. (A) B6.Ly5.1 mice were injected with 5 × 10⁵ or 5 × 10³ P14 Tg cells (Ly5.2⁺) and infected with 2 × 10⁵ PFU of LCMV Armstrong. At day 90 postinfection, splenocytes were removed and restimulated in vitro with the GP33 peptide, and the proportion of P14 cells producing IFN-γ and TNF-α was analyzed by intracellular cytokine staining. The dot plots show representative results from one animal in each group and are gated on P14⁺ cells. The percentage of P14⁺ cells secreting each cytokine is indicated within the dot plots. (B to E) B6.Ly5.1 mice were injected with 5 × 10⁵ P14 Tg cells (Ly5.2⁺) and infected with 2 × 10⁵ PFU of LCMV Armstrong. (B) The proportion of P14 and endogenous GP33-specific cells in the spleen was analyzed over time in the same mice. The mean results from three individual mice are shown, and the error bars indicate standard errors of the means. The inset shows a magnification of the endogenous cell graph. (C) Gene expression in individually sorted P14 and endogenous GP33-specific cells was analyzed; the results are presented as described for Fig. 1B. (D) The pattern of the coexpression of cytotoxic effector mRNAs also was analyzed and is presented as described for Fig. 1D. (E) The percentage of cells on day 60 expressing each of the indicated genes (cytotoxic effector genes as shown in Fig. 3B) was analyzed and is shown for 90 individually sorted P14 and 70 endogenous GP33-specific cells. The cells were obtained from three individual mice.

MHC-I tetramer labeling during the expansion phase. B6.Ly5.1 mice left untreated or receiving Ly5.2⁺ P14 Tg cells were infected with 2 × 10⁵ PFU of LCMV Armstrong and studied at different time points after infection. (A and B) Mice were injected with 5 × 10⁵ Tg cells. (A) Results compare GP33 tetramer (tet) binding in Ly5.2⁺ P14 naïve cells and in P14 cells at on day 3 (top) and day 5 (bottom) after infection. Staining is from individual mice representative of four experiments with two to three mice per time point. (B) On day three after infection, P14 cells were arbitrarily subdivided into tet^neg (gray) and tet^hi (white) subsets, and each population was tested for the expression of the indicated cell surface molecules. Gates for tet^neg cells were established in noninfected B6 mice and for tet^hi in naïve Tg cells. (C) B6 mice were injected with either 5 × 10⁵ or 10⁴ Ly5.2⁺ P14 Tg cells and studied at days 4 and 5 after infection. Results are for GP33 tetramer binding in P14 Tg cells. Naïve mice (upper left), cells from mice injected with 5 × 10⁵ naïve cells (upper right), and cells from three individual mice injected with 10⁴ naïve cells (lower graphs) are shown. At day 4, very few Tg cells were detected in the latter mice. (D) Granzyme B expression in cells expressing different tet binding intensity in normal mice. CD8 cells were recovered 5 days after infection, labeled with GP276 tetramers, and subdivided into tet^int and tet^hi populations. Results show the gates used for such subdivision and intracellular granzyme B staining for tet^int (gray) and tet^hi (white) populations; the dashed lines represent the staining of the same cells with an isotype control antibody.