Quantitating the magnitude of the lymphocytic choriomeningitis virus-specific CD8 T-cell response: it is even bigger than we thought

Abstract

Measuring the magnitudes and specificities of antiviral CD8 T-cell responses is critical for understanding the dynamics and regulation of adaptive immunity. Despite many excellent studies, the accurate measurement of the total CD8 T-cell response directed against a particular infection has been hampered by an incomplete knowledge of all CD8 T-cell epitopes and also by potential contributions of bystander expansion among CD8 T cells of irrelevant specificities. Here, we use several techniques to provide a more complete accounting of the CD8 T-cell response generated upon infection of C57BL/6 mice with lymphocytic choriomeningitis virus (LCMV). Eight days following infection, we found that 85 to 95% of CD8 T cells exhibit an effector phenotype as indicated by granzyme B, 1B11, CD62L, CD11a, and CD127 expression. We demonstrate that CD8 T-cell expansion is due to cells that divide >7 times, whereas heterologous viral infections only elicited <3 divisions among bystander memory CD8 T cells. Furthermore, we found that approximately 80% of CD8 T cells in spleen were specific for ten different LCMV-derived epitopes at the peak of primary infection. These data suggest that following a single LCMV infection, effector CD8 T cells divide > or =15 times and account for at least 80%, and possibly as much as 95%, of the CD8 T-cell pool. Moreover, the response targeted a very broad array of peptide major histocompatibility complexes (MHCs), even though we examined epitopes derived from only two of the four proteins encoded by the LCMV genome and C57BL/6 mice only have two MHC class I alleles. These data illustrate the potential enormity, specificity, and breadth of CD8 T-cell responses to viral infection and demonstrate that bystander activation does not contribute to CD8 T-cell expansion.

FIG. 1.

Phenotype of CD8⁺ splenocytes at peak of LCMV response. C57BL/6 mice were infected with 2 × 10⁵ PFU LCMV Armstrong i.p., and splenocytes were analyzed eight days later. (A) Number of CD44^lo and CD44^hi CD8⁺ splenocytes isolated from naive and infected mice. (B) Coexpression of granzyme B and CD44 on CD8⁺ splenocytes isolated from naive or infected mice. (C) Granzyme B expression levels among tetramer-positive versus tetramer-negative CD44^hi CD8⁺ lymphocytes among infected mice. Splenocytes were stained with granzyme B, CD44, CD8, and either NP396, GP34, GP276, or GP33 H-2^b tetramers, as indicated. Black line: granzyme B expression among tetramer-negative CD8⁺ CD44^hi splenocytes. Shaded histogram: granzyme B expression among tetramer-positive CD8⁺ lymphocytes. Thin gray line: isotype control antibody (Ab) staining of tetramer-positive CD8 T cells. (D) Coexpression of granzyme B and various markers of T-cell activation (gated on CD8⁺ splenocytes). Data in panels A, B, and D are representative of at least three independent experiments (and at least 18 total mice). Data in panel C are representative of two independent experiments (n = 3 per experiment).

FIG. 2.

CD8 T cells of unknown specificity mount a recall response to LCMV infection and divide >7 times. Naive Thy1.2⁺ C57BL/6 mice were infected with LCMV Armstrong (2 × 10⁵ i.p.). Three months later, splenocytes were isolated and 3 × 10⁷ were transferred into naive Thy1.1⁺ C57BL/6 mice. Recipients were infected with LCMV Armstrong (1 × 10⁶ PFU i.v.), and spleen was analyzed seven days later. (A) Summary of experimental design. (B) Percentage of CD44^hi donor and host CD8⁺ T cells prior to and seven days following LCMV challenge. (C) Donor (Thy1.2⁺) and host (Thy1.1⁺) CD8⁺ splenocytes were analyzed for coexpression of granzyme B and CD44 or CD11a. (D) Proportion of donor or host CD8 T cells staining with the indicated tetramers. (E) Donor splenocytes were labeled with CFSE prior to transfer in order to track division following infection. Splenocytes were gated on Thy1.2 and CD8, as shown, and CFSE versus CD44 or tetramer staining was analyzed. Data are representative of two independent experiments (n = 3 per experiment).

FIG. 3.

Memory CD8 T cells do not divide >3 times upon heterologous infections. (A) Thy1.2⁺ VSV immune splenocytes (3 × 10⁷) were labeled with CFSE and transferred into Thy1.1⁺ naive C57BL/6 mice. One day following transfer, recipient mice were either infected with LCMV or left uninfected. (B) Eight days later (peak of LCMV response), the CFSE profile among donor VSV-N tetramer-positive CD8⁺ T cells was analyzed. (C) Splenocytes (3 × 10⁷) from immune P14 chimeras were labeled with CFSE and transferred to naive B6 mice. One day following transfer, recipient mice were either infected with VSV or left uninfected. (D) Six days later (peak of VSV response), the CFSE profile among donor P14 CD8⁺ T cells was analyzed (data representative of two independent experiments; n = 3 per experiment). Plots shown are gated on total CD8⁺ T cells and on Thy1.1⁺ P14 cells, as indicated.

FIG. 4.

Heterologous infection induces transient changes in memory CD8 T-cell phenotype. One hundred fifty days after LCMV infection, P14 chimeric mice were challenged with VSV. Granzyme B and CD69 expression was analyzed on LCMV-specific P14 memory CD8 T splenocytes two and six days after VSV infection and compared to mice that were not infected with VSV. All histograms are gated on Thy1.1⁺ P14 memory CD8 T cells. The experiment was performed three times (n = 2 to 3 mice per group, per experiment) with similar results.

FIG. 5.

Most CD44^hi+ CD8⁺ lymphocytes are specific for LCMV at peak of infection. Eight (A and B) or 90 (C and D) days following LCMV infection, splenocytes were incubated for five hours with the indicated peptides and brefeldin A, surface stained for CD8 and CD44, permeabilized, and stained with anti-interferon gamma. n = 19 (A and B) or 5 (C and D).

FIG. 6.

LCMV-specific CD8 splenocytes undergo approximately 15 divisions but do not displace CD8 T cells of other specificities during the peak of response. The number of LCMV-specific, non-LCMV-specific, and total CD8 T cells within spleen was quantitated prior to LCMV infection, at the peak of the response, and 60 to 90 days postinfection (memory). The number of naive LCMV-specific CD8 T cells is an estimate based on the reported precursor frequency of GP33-specific CD8 T cells (8).