Functional CD8 T cell memory responding to persistent latent infection is maintained for life

Abstract

Aging is associated with depressed naive T cell responses, but it is less clear whether T cell memory established early in life also becomes impaired with age. This is particularly important for T cells responding to latent persistent infection, which need to remain functional and capable of controlling the infection over the lifetime; however, repeated stimulation over the lifetime may dysregulate their maintenance or function, potentially contributing to impaired immunity in the elderly. Systemic infection with HSV-1, a persistent latent virus, is associated with memory inflation of virus-specific CD8 T cells. We tested how these inflated memory cells are maintained from adulthood into old age. We found no significant differences in the numbers (i.e., blood, spleen), ex vivo Ag-specific IFN-γ production, and in vivo recall response to HSV-1 (i.e., proliferation, IFN-γ production, cytolysis) between adult and old memory T cells. There was a discrete shift from dominantly effector memory phenotype in the adults to a central memory-like phenotype in the old mice, with fewer old cells expressing the killer cell lectin-like receptor G1 (KLRG1). Adult and old KLRG1(+) memory CD8 T cells were functionally identical: both produced IFN-γ but could minimally proliferate in response to viral challenge. Interestingly, regardless of age, KLRG1(+) cells retained the ability to proliferate and survive in response to homeostatic signals, both in vitro (culture with IL-7 and IL-15) and in vivo (expansion following transfer into lymphopenic recipients). This finding demonstrates that functional effector memory T cells, including those expressing KLRG-1, are maintained and are functional for life, despite the presence of persistent viral infection.

Figure 1. Numbers and phenotype of gB-8p memory CD8 T cells in adult and old mice

Mice were infected i.p. with 10⁶ PFU HSV-1 at 2 months of age. Lymphocytes from blood (A–C) and spleen (D–F) were analyzed by flow cytometry when mice were adult (6 months p.i., 8 months of age, n = 8) and old (20 months p.i., 22 months of age, n = 10). The frequency (A, D) and numbers (B, E) of HSV-1 specific memory CD8 T cells in blood (A, B) and spleen (D, E) were determined based on staining with gB-8p:K^b tetramer and anti-CD8. Expression of KLRG1, PD-1, CD62L, CD127, CD122 and CD27 was determined by flow cytometry and is expressed as percentage of CD8+ gB-8p:K^b tetramer+ T cells expressing indicated phenotype in blood (C) and spleen (D).

Figure 2. Interfering with viral reactivation lowers the extent of KLRG1+ expression by gB-8p memory CD8 T cells in adult but not in old mice

Mice were infected i.p. with 10⁶ PFU HSV-1 at 2 months of age. Some mice remained untreated (HSV group), while the others were continuously treated with famciclovir (famvir, HSV + FVR group) starting on day 5 p.i. as described in Materials and Methods, in order to prevent or reduce viral reactivation. Lymphocytes from blood (A) and spleen (B) were analyzed by flow cytometry when mice were adult (4 months p.i., 6 months of age, n = 7 for both HSV and HSV+FVR groups ) and old (20 months p.i., 22 months of age, n = 3 for both HSV and HSV+FVR groups). Cells were stained with anti-CD8, gB-8p:Kb tetramer, and anti-KLRG1. The data is represented as percentage of CD8+ gB-8p:K^b tetramer+ T cells that express KLRG1. The graphs show average +/− SD per group.

Figure 3. Comparison of ex-vivo IFNγ production in response to antigen by adult and old gB-8p memory CD8 T cells

Splenocytes were harvested from adult (A, B) and old (C, D) HSV-1 infected mice (mouse cohorts as in Figure 1), and were stimulated for 6 hours with 10⁻⁶ M gB-8p peptide in presence of brefeldin A, and then stained using anti-CD8 either gB-8p:Kb tetramer or intracellular IFNγ. A, C. The X values show the percent of tetramer⁺ cells, and the Y values show the percent of IFNγ⁺ cells within total CD8⁺ T-cell pool of individual mice. The values for IFNγ show the net IFNγ (background IFNγ staining from unstimulated wells was subtracted; background was < 0.1%). B, D. Splenocytes were co-stained with anti-CD8, anti-KLRG1, anti-CD127 (surface) and IFNγ (intracellular). The FACS plots are gated on CD8+ T cells, and the numbers represent the percentage of CD8 T cells expressing indicated markers in a representative individual mouse from the tested cohort.

Figure 4. In vivo recall response to HSV-1 by adult and old gB-8p memory CD8 T cells

Splenocytes from adult and old HSV-1 infected mice (cohorts as in Figure 1) were enriched for CD8+ T cells using magnetic bead separation, and were then incubated with antibodies and tetramer. CD8+ CD44+ gB-8p:K^b tetramer+ (referred to as Tet+) T cells were isolated using BD FACSAria II sorter. Indicated number (3×10⁴, 1×10⁴ or 1×10³) of sorted Tet+ cells was injected i.v. into naïve congenic Ly5.2+ recipient mice, which were infected with HSV-1 24 hours later. The frequency and numbers of responding donor memory CD8 T cells were determined on day 7 p.i. by gating on CD8+ Ly5.1+ cells. A–B. Frequency and numbers of gB-8p:K^b tetramer+ T cells in blood. C–D. Frequency and numbers of gB-8p:K^b tetramer+ T cells in spleen. E–F. Splenocytes were stimulated for 6 hours with 10⁻⁶ M gB-8p peptide in presence of brefeldin A, and then stained using anti-CD8 and intracellular IFNγ. Frequency and numbers of memory IFNγ+ T cells in spleen are shown. All values are represented as average +/− SD. G. On day 7 p.i. the responding donor CD8 T cells (effector CTLs derived from the donor memory CD8 T cells) were sorted based on Ly5.1 expression, and their cytotoxic function was tested in an in vivo killing assay as described in Materials and Methods. Briefly, sorted effector CTLs were injected i.v. into naïve congenic recipient mice together with CFSE labeled target cells of two types: specific (gB-8p-pulsed, CFSE^hi) and non-specific (no peptide, CFSE^lo). Four hours later the spleens were harvested and the extent lysis of specific lysis was determined by flow cytometry by comparing elimination of peptide-pulsed targets in the presence and absence of CTLs.

Figure 5. KLRG1+ memory CD8 T cells from adult and old mice fail to proliferate in response to antigenic stimulation, but undergo homeostatic expansion in vivo

Splenocytes from adult and old HSV-1 infected mice (cohorts as in Figure 1) were enriched for CD8+ T cells using magnetic bead separation, and were then incubated with antibodies and tetramer and sorted into CD8+ CD44+ gB-8p:K^b tetramer+ KLRG1+ (labeled as KLRG1+) and CD8+ CD44+ gB-8p:K^b tetramer+ KLRG1− (labeled as KLRG1−) on BD FACSAria II sorter. A–B. The sorted KLRG1+ and KLRG1− cells were adoptively transferred into naïve congenic Ly5.2+ recipients, which were infected with HSV-1 24 hours later. On day 7 p.i. the number of donor gB-8p:K^b tetramer+ cells per spleen was determined based on staining with anti-CD8, anti-Ly5.1, and gB-8p:K^b tetramer. C–D. The sorted KLRG1+ and KLRG1− cells were adoptively transferred into Rag1 knockout recipients, together with congenic Ly5.2 naïve feeder cells. The number of donor cells per ml blood was determined on days 3, 7 and 21 post-transfer based on cell counts and staining with anti-CD8, anti-Ly5.1, and gB-8p:K^b tetramer.

Figure 6. Homeostatic maintenance of KLRG1+ memory CD8 T cells: role of proliferation and survival

Splenocytes from HSV-1 infected adult mice (45 days p.i., n = 5) were enriched for CD8+ T cells using magnetic bead separation, and were then incubated with antibodies and tetramer and sorted into CD8+ CD44+ gB-8p:K^b tetramer+ KLRG1+ (labeled as KLRG1+) and CD8+ CD44+ gB-8p:K^b tetramer+ KLRG1− (labeled as KLRG1−) on BD FACSAria II sorter. Cells were labeled with CFSE and 2×10⁴ sorted cells and 2×10⁶ naïve congenic Ly5.2+ feeder splenocytes were plated per well in RPMI media supplemented with IL-7 (10 ng/ml) and IL-15 (100 ng/ml). The cell proliferation, numbers and phenotype were analyzed on day 7 by flow cytometry. A–B. Proliferation of KLRG1+ (A) and KLRG1− (B)gB-8p:K^b tetramer+ memory CD8 T cells. The FACS plots are gated on CD8+ T cells. C. Live cells from each well counted and the number of the sorted KLRG1+ and KLRG1− memory CD8 T cells was calculated based on staining with anti-CD8, anti-Ly5.1 and gB-8p:K^b tetramer. D–E. The extent of cell division was determined by calculating the average percentage of each population of sorted memory CD8 T cells (KLRG1+ and KLRG1−) that divided at least once (D), or more than two times (E). F–G. Sorted cells retain their original KLRG1 expression. The FACS plots are gated on CD8+ T cells and show co-staining with gB-8p:K^b tetramer+ and KLRG1 on day 7 of culture (F). The average percent of KLRG1+ cells among gB-8p:K^b tetramer+ T cells is shown in G. H–I. The FACS plots are gated on CD8+ T cells and show co-staining with gB-8p:K^b tetramer+ and Bcl-2 on day 7 of culture (H). The average percent of Bcl-2+ cells among gB-8p:K^b tetramer+ T cells is shown in I.

Figure 7. In vivo homeostatic proliferation of KLRG1+ and KLRG1− memory CD8 T cells

Splenocytes from HSV-1 infected adult mice (45 days p.i., n = 5) were sorted as in Figure 6. The sorted cells (KLRG1+ or KLRG1−) were labeled with CFSE, mixed with naïve congenic feeder splenocytes, and adoptively transferred into Rag1 knockout mice (n = 5 per group). Donor memory CD8 T cell numbers (A) and their proliferation (B) were monitored by flow cytometry on day 2, 7 and 21 post-transfer (staining done on lymphocytes isolated from blood). A. Number of donor memory CD8 T cells per ml blood was determined based on staining with anti-CD8, anti-Ly5.2 and gB-8p:K^b tetramer. The average +/− SD is shown. B. Percentage of donor memory CD8 T cells that completed 5 or more divisions. n.d.= not detected. C. KLRG1 expression after transfer into Rag1 knockout mice. KLRG1 expression on the sorted (donor) cells was determined by flow cytometry on day 2 and 21 post-transfer. Data is shown as percent of KLRG1+ cells among the undivided donor cells or donor cells that divided 5 or more times ± SD.