Latent infection with herpes simplex virus is associated with ongoing CD8+ T-cell stimulation by parenchymal cells within sensory ganglia

Abstract

CD8+ T-cell persistence can be seen in ganglia harboring latent herpes simplex virus (HSV) infection. While there is some evidence that these cells suppress virus reactivation, this view remains controversial. Given that maintenance of latency by CD8+ T cells would necessitate ongoing exposure to antigen within this site, we sought evidence for such chronic stimulation. Initial experiments showed infiltration by activated but not naïve CD8+ T cells into ganglia harboring latent HSV infection. While such infiltration was independent of T-cell specificity, once recruited, only virus-specific T cells expressed high levels of preformed granzyme B, a marker of ongoing activation. Moreover, bone marrow replacement chimeras showed that these elevated granzyme levels were totally dependent on presentation by parenchymal cells within the ganglia. Overall, this study argues that activated CD8+ T cells are nonspecifically recruited into latently infected ganglia, and in this site they are exposed to ongoing antigen stimulation, most likely by infected neuronal cells.

FIG. 1.

Persistence of CD8⁺ T cells in the DRGs during latent HSV-1 infection. A cohort of C57BL/6 mice received 5 × 10⁴ resting gBT-I × B6.Ly5.1 CD8⁺ T cells 24 h prior to flank scarification infection. Groups of 4 to 12 mice were then sacrificed at various time points following infection, and the DRGs innervating the site of cutaneous infection were analyzed via flow cytometry for the presence of gB-specific CD8⁺ T cells. The average number of gB-specific CD8⁺ T cells obtained from an individual mouse is shown. Lytic virus has previously been shown to clear from the DRGs by days 7 to 8 postinfection, leaving a persistent latent infection.

FIG. 2.

Infiltration and persistence of OT-I CD8⁺ T cells in the DRGs following flank scarification infection. (A) A cohort of C57BL/6 mice received 5 × 10⁴ resting OT-I CD8⁺ T cells 24 h prior to flank scarification infection. Groups of four mice were then sacrificed on days 4 to 7 p.i., and the DRGs innervating the site of cutaneous infection were analyzed via flow cytometry for the presence of OVA-specific CD8⁺ T cells. Cells were stained with anti-CD8 and K^b-OVA tetramer to identify the double-positive OT-I CD8⁺ T cells. A positive control group of four mice that had received resting gBT-I CD8⁺ T cells prior to flank infection were also analyzed on day 7 postinfection; these samples were stained with anti-CD8 and K^b-gB tetramer. (B) C57BL/6 mice were infected on the flank with HSV-1 and then received either 10⁷ in vitro-activated gBT-I × B6.Ly5.1 or 10⁷ in vitro-activated OT-I × B6.Ly5.1 CD8⁺ T cells on day 4 postinfection. Mice in each group were sacrificed on day 6 and day 20 p.i., lytic and latent time points, respectively. The DRGs innervating the site of cutaneous inoculation were removed from the mice at each time point and treated with collagenase, and the released cells were stained with anti-CD8 and anti-Ly5.1 and analyzed via flow cytometry for the presence of transgenic CD8⁺ T cells. The number of gB-specific (closed circles, n = 14 and 8 for infected and mock-infected mice, respectively) or OVA-specific (open circles, n = 17 and 8 for infected and mock-infected mice) CD8⁺ T cells in the DRGs on day 6 (i) and day 20 (n = 12 and 11 for gB and OVA specific T cells, respectively, into infected mice, and n = 4 for both sets of mock-infected mice) (ii) is shown for each mouse, with averages indicated (−). Mock-infected animals received in vitro-activated gB-specific or OVA-specific CD8⁺ T cells on day 4 p.i.

FIG. 3.

Infiltration of CD8⁺ T cells into the DRGs during HSV-1 latency. A cohort of C57BL/6 mice was infected on the flank with HSV-1. On day 18 p.i., when lytic virus had been cleared and latency established, the mice received either 10⁷ in vitro-activated gBT-I × B6.Ly5.1 (n = 8) or 10⁷ in vitro-activated OT-I × B6.Ly5.1 (n = 8). The mice were then sacrificed on day 20 p.i., and the DRGs innervating the site of cutaneous infection were analyzed for transgenic CD8⁺ T cells via flow cytometry. Cells were stained with anti-CD8 and anti-Ly5.1. The number of gB-specific (closed circles) or OVA-specific (open circles) CD8⁺ T cells is shown for each mouse, with averages indicated (−). The percentage of mice that were positive for transgenic T cells in their DRGs is also shown (open bars). Two negative control groups were included in this experiment. First, mock-infected animals that received 10⁷ in vitro-activated gBT-I × B6.Ly5.1 or OT-I × B6.Ly5.1 CD8⁺ T cells on day 18 after mock infection. The DRGs from these mice were analyzed on day 20 postinfection. Second, HSV-1-infected C57BL/5 mice that received resting gBT-I × B6.Ly5.1 CD8⁺ T cells on day 18 p.i. The DRGs from these mice were analyzed for transgenic T-cell infiltration on day 23 p.i. (5 days posttransfer).

FIG. 4.

CD69 expression on CD8⁺ T cells in HSV-infected DRGs. CD8⁺ T cells from gBT-I × B6.Ly5.1 and OT-I × B6.Ly5.1 mice were stimulated in vitro for 4 days. Following this stimulation, the cells were analyzed for their expression of the early activation marker, CD69, compared to naïve C57BL/6 CD8⁺ T cells (A). The activated T cells were then transferred into separate groups of C57BL/6 mice that had been infected on the flank with HSV-1 4 days earlier (10⁷ cells per mouse). The gB-specific and OVA-specific CD8⁺ T cells that were isolated from the DRGs on day 6 p.i. (B) and day 20 p.i. (C) were analyzed for their expression of CD69.

FIG. 5.

Preformed granzyme B in CD8⁺ T cells in HSV-infected DRGs. CD8⁺ T cells from gBT-I × B6.Ly5.1 and OT-I × B6.Ly5.1 mice were stimulated in vitro for 4 days and then analyzed for their levels of preformed granzyme B via flow cytometry. An example is shown of in vitro-activated gBT-I × B6.Ly5.1 compared to naïve C57BL/6 CD8⁺ T cells (A). The activated T cells were then transferred into separate groups of C57BL/6 mice that had been infected on the flank with HSV-1 4 days earlier (10⁷ cells per mouse). The DRGs were removed from these mice on day 6 (B) and day 20 (C) p.i. The gB-specific (closed circles) and OVA-specific (open circles) CD8⁺ T cells that were isolated from the DRGs at each time point were analyzed for their levels of preformed granzyme B. The results are presented as the geometric mean of fluorescence intensity (MFI); averages are indicated (−). The difference in scale between the two time points is due to a different batch of anti-human granzyme B antibody being used.

FIG. 6.

In vivo activation of gBT-I CD8⁺ T cells following flank scarification infection. B6.Ly5.1→B6, B6.Ly5.1→bm3, and bm3→B6.Ly5.1 bone marrow chimeras were seeded with 5 × 10⁴ resting gBT-I × B6.Ly5.1 CD8⁺ T cells 24 h prior to flank scarification infection. On day 9 p.i., blood was collected from these mice and the gB-specific CD8⁺ T cells in the blood were analyzed for their level of activation, as determined by their level of preformed granzyme B. Blood from naïve gBT-I × B6.Ly5.1 mice was also analyzed as a negative control. The results are presented as the geometric mean of fluorescence intensity (MFI), with averages indicated (−).

FIG. 7.

The role of parenchymal cells in the infiltration and activation of gB-specific CD8⁺ T cells in the peripheral nervous system during HSV-1 latency. B6.Ly5.1→B6, B6.Ly5.1→bm3, and bm3→B6.Ly5.1 bone marrow chimeras were seeded with 5 × 10⁴ resting gBT-I × B6.Ly5.1 CD8⁺ T cells 24 h prior to flank scarification infection. On days 30 to 35 p.i. the mice were sacrificed and the DRGs innervating the site of cutaneous infection were removed. (A) The number of gB-specific CD8⁺ T cells in the DRGs of each mouse was determined using flow cytometry. The results for each individual mouse are shown with averages indicated (−). (B) The gB-specific CD8⁺ T cells within the DRGs were also analyzed for their levels of preformed granzyme B. The results are presented as the geometric mean of fluorescence intensity (MFI), with averages indicated (−). Resting gBT-I × B6.Ly5.1 CD8⁺ T cells were included in this analysis as a negative control. There was no granzyme B analysis for the bm3→B6.Ly5.1 chimeras, as gB-specific CD8⁺ T cells were not detected in the DRGs of these mice.