Cutting Edge: Caspase-11 Limits the Response of CD8⁺ T Cells to Low-Abundance and Low-Affinity Antigens

Caspase-11 in mice and caspase-4 and caspase-5 in humans are members of the family of inflammatory caspases that also includes caspase-1 and caspase-12 (1). Unlike other caspases, caspase-11 is expressed at low levels in resting cells and is upregulated upon activation via a type I IFN–dependent process (2). Caspase-11 can bind directly to intracellular LPS, resulting in caspase-11 processing and activation (3) and leading to downstream events that can include caspase-1 activation, cell death, and inflammatory cytokine processing and release (1). Caspase-11 can also act in its proform to regulate actin dynamics; caspase-11 promotes actin depolymerization by facilitating interactions among actin-interacting protein 1, cofilin, and F-actin (4). Accordingly, caspase-11–deficient (Casp11^−/−) cells have altered migration and display reduced fusion of lysosomes to pathogen-containing vacuoles (4, 5).

Surprisingly, activation-dependent expression of caspase-11 is also observed in CD4⁺ and CD8⁺ T cells; both effector and memory T cells have increased expression of caspase-11 compared with naive cells (6–8). Caspase-11 processing was not observed in activated T cells, but T cell migration was affected, with Casp11^−/− cells migrating less efficiently into lymphoid tissues (4). Modulation of actin polymerization by caspase-11 could regulate additional aspects of T cell biology, including TCR signaling (9). The strength of signals received through the TCR can have affects on the phenotype and function of T cells after activation. Activation of CD8⁺ T cells with high-affinity peptides results in increased expansion and effector function compared with stimulation with lower-affinity peptides (10). The strength of TCR stimulation can also affect the phenotype of CD4⁺ T cells, with CD4⁺ T cells receiving higher levels of stimulation preferentially developing into T follicular helper cells (11, 12) and low concentrations of high-affinity peptide favoring Foxp3 expression (13).

We addressed the role of caspase-11 in the activation and function of CD8⁺ T cells and found that Casp11^−/− cells proliferate more readily in response to suboptimal levels of TCR stimulation, leading to a larger effector and memory pool and increased effector function in response to both low-abundance and low-affinity ligands in vivo. However, in the presence of low concentrations of self-Ag, the increased sensitivity of Casp11^−/− cells results in more rapid deletion in tissues in which the Ag is expressed. These data indicate that, in addition to promoting cell death and inflammatory cytokine production, caspase-11 acts as a negative regulator of TCR signaling and limits the expansion and function of T cells in response to low-abundance or low-affinity TCR ligands.

C57BL/6 and Nur77-GFP mice were purchased from The Jackson Laboratory. Nur77-GFP, OT-I, iFABP-tOVA 232-6 (14), and Caspase-11^−/− (15) mice were maintained at the University of Washington and used in accordance with Institutional Animal Care and Use Committee guidelines.

Mice received 1–2 × 10⁴ naive OT-I T cells and were immunized with 2 × 10³ Listeria monocytogenes expressing OVA (LM-OVA), 150 μg OVA, or 1–2 × 10⁶ dendritic cells (DCs) pulsed with LPS and 1 μM N4, Q4, or T4 peptide, as previously described (16). Mice were pulsed with 1 mg BrdU i.p. 2.5 h before sacrifice. OT-I T cells were stimulated ex vivo with peptide for 4 h in the presence of brefeldin A prior to intracellular cytokine staining.

Cells were stained with the indicated Abs (BD Biosciences and eBioscience) or Alexa Fluor 647 phalloidin (Life Technologies), flow cytometry was performed on a FACSCanto (BD), and data were analyzed using FlowJo software (TreeStar). For cell sorting, thymocytes were stained with CD4, CD8α, and TCRβ, and effector OT-I T cells were stained using CD8 and CD45.1. Cells were sorted on a FACSAria.

RNA was isolated using a QIAGEN RNeasy Kit, per the manufacturer’s instructions. Quantitative RT-PCR was performed using the Quantitect SYBR Green RT-PCR Kit (QIAGEN) and the following primers: casp11F: 5′-CCTGAAGAGTTCACAAGGCTT-3′; casp11R: 5′-CCTTTCGTGTACGGCCATTG-3′; actbF: 5′-GGCTGTATTCCCCTCCATCG-3′, and actbR: 5′-CCAGTTGGTAACAATGCCATGT-3′.

Splenocytes were pulsed with 1 nM to 1 μM N4, Q4, or T4 peptide or no peptide at 37°C for 40 min and washed thoroughly with media. A total of 5 × 10⁴ splenocytes was mixed with 1–2 × 10⁴ naive OT-I T cells. T cells were labeled with 2 μM CFSE (Invitrogen) where indicated.

All graphs depict mean ± SD. The two-tailed Student t test was used to determine statistical significance.

Caspase-11 is upregulated in a variety of cell types undergoing activation, including CD8⁺ T cells (6). We examined the levels of caspase-11 mRNA during the development of CD8⁺ T cells by sorting double-negative, double-positive, and CD8 single-positive thymocytes, as well as naive CD8⁺ peripheral T cells. Casp11 was upregulated as thymocytes matured and increased further after cells entered the periphery (Fig. 1A). We also examined whether casp11 expression was upregulated after Ag-dependent activation in the periphery. Naive OT-I T cells, which express a TCR specific for the OVA peptide SIINFEKL, were transferred into mice that were then infected with LM-OVA or immunized with OVA protein. Activated OT-I T cells were sorted, and casp11 expression was analyzed. Casp11 was upregulated an additional 5-fold in activated OT-I T cells in response to both protein immunization and infection (Fig. 1B) compared with naive cells.

To analyze the function of caspase-11 during the activation of CD8⁺ T cells, caspase-11-sufficient (wild-type [WT]) and Casp11^−/− OT-I T cells were transferred into mice, followed by challenge with LM-OVA or OVA protein. There were equal numbers of WT and Casp11^−/− OT-I T cells 7 d after LM-OVA challenge (data not shown). However, after OVA immunization, Casp11^−/− cells were present at higher numbers than WT OT-I T cells as early as 4 d after immunization, and this difference became more pronounced at 6 d (Fig. 1C). The increased number of Casp11^−/− cells compared with WT cells was maintained at memory time points (Fig. 1C). Previous reports indicated that caspase-11 is able to regulate the migration of T cells (4), and we hypothesized that failure of Casp11^−/− T cells to migrate out of the spleen might explain their increase relative to WT cells. However, we saw no increase in the ratio of WT/Casp11^−/− CD8⁺ T cells in the blood or inguinal lymph node, suggesting that WT cells were not exiting the spleen more readily than Casp11^−/− cells (data not shown). These data suggest that caspase-11 limits the accumulation of CD8⁺ T cells in response to protein immunization but has little effect on their migration out of central lymphoid organs.

We hypothesized that the increased numbers of Casp11^−/− OT-I T cells after protein immunization, but not LM-OVA infection, could be due to their enhanced response to ligands of low abundance. This was addressed in vitro by examining the expansion of WT and Casp11^−/− OT-I T cells in response to splenocytes pulsed with various concentrations of a high-affinity ligand (SIINFEKL, N4). In response to high concentrations of N4, WT and Casp11^−/− OT-I T cells diluted CFSE similarly, and WT cells showed a mild defect in accumulation (Fig. 2). In response to low concentrations of N4, we saw an increased percentage of CFSE^lo cells in the Casp11^−/− population compared with WT (Fig. 2A) and increased accumulation of Casp11^−/− cells (Fig. 2B). When a lower-affinity ligand (SIITFEKL, T4) was used, we also observed an increase in proliferation and outgrowth of Casp11^−/− cells compared with WT cells (Fig. 2). These data indicate that there is a cell intrinsic enhancement in the ability of Casp11^−/− T cells to proliferate in response to low levels of antigenic stimulation, because WT and Casp11^−/−cells have equal access to Ag, cytokines, and other factors in this in vitro setting.

Higher levels of TCR signaling can lead to differences in effector function, and we hypothesized that the increased TCR sensitivity of Casp11^−/− cells may result in enhanced effector function. To address this, mice were primed with DCs pulsed with N4 or T4, and the expansion of WT and Casp11^−/− OT-I T cells was examined. T4-DC immunization was not sufficient to activate all transferred cells (data not shown); therefore, we used the intermediate-affinity peptide SIIQFEKL (Q4) to pulse DCs. Consistent with what we observed in vitro, we saw increased accumulation of Casp11^−/− cells compared with WT cells after both N4-DC and Q4-DC immunization, and the difference was enhanced with lower-affinity Ag (Fig. 3A). BrdU incorporation at 3 d after Q4-DC immunization revealed that Casp11^−/− OT-I T cells proliferated more than WT cells (Fig. 3B). No difference was observed in annexin V staining (data not shown), indicating that accumulation of Casp11^−/− cells was due to enhanced proliferation and not decreased sensitivity to cell death.

Cytokine production by Casp11^−/− OT-I T cells relative to controls was examined 4 d after Q4-DC immunization. OT-I T cells were restimulated in vitro with high and low concentrations of N4 peptide, and the production of IFN-γ and IL-2 was measured. A similar percentage of WT and Casp11^−/− cells produced IFN-γ after stimulation with both high and low concentrations of N4 (Fig. 3C, 3D). However, the percentage of IFNγ⁺ cells producing IL-2 was increased in Casp11^−/− OT-I T cells in response to low concentrations of N4 (Fig. 3C, 3D). We also examined the expression of CD62L after Q4-DC immunization and found that Casp11^−/− OT-I T cells had reduced expression compared with WT cells (Fig. 3E). These data indicate that Casp11^−/− T cells expand more readily in response to in vivo peptide-DC stimulation and have enhanced effector function, including an increased ability to produce IL-2.

We hypothesized that caspase-11 modulated the strength of the TCR signal received during priming. This was measured in WT and Casp11^−/− OT-I T cells using a Nur77-GFP reporter (17). Casp11^−/− and WT cells had similar GFP expression in the absence of peptide stimulation (data not shown). WT and Casp11^−/− Nur77-GFP OT-I T cells were stimulated with N4, Q4, or T4 peptides for 12 h, and the fold change in GFP expression in the GFP^hi population over unstimulated WT OT-I T cells was compared. We found that Casp11^−/− T cells had significantly increased GFP expression over a range of peptide affinities (Fig. 4A) and concentrations (data not shown). Additionally, more cells in the Casp11^−/− population responded to low-affinity ligands; a higher percentage of Casp11^−/− cells became Nur77-GFP^hi within 12 h of stimulation (Fig. 4B). By 24 h postactivation, this phenotype was less apparent (data not shown), suggesting a difference in the rate of activation rather than the inability of a subset of WT cells to respond. This data could be explained by increased expression of the TCR and associated molecules in Casp11^−/− T cells; however, naive WT and Casp11^−/− OT-I T cells had similar surface expression of Vα2, Vβ5, CD3, and CD8α (data not shown). These data indicate that TCR signaling in response to low-abundance/affinity ligands is inhibited by caspase-11, resulting in weaker TCR signaling overall and delayed activation.

Caspase-11 was shown to positively regulate actin depolymerization, leading to increased levels of F-actin in Casp11^−/− cells (4, 5), and we speculated that this could lead to enhanced TCR signaling in Casp11^−/− T cells. Using phalloidin staining, we verified increased F-actin levels in Casp11^−/− OT-I T cells compared with WT OT-I T cells after stimulation with low-affinity ligands (Supplemental Fig. 1A, 1B). In addition, the onset of Nur77-GFP expression in WT OT-I T cells after T4 stimulation correlated with higher levels of F-actin compared with Nur77-GFP^lo cells (Supplemental Fig. 1C).

To determine whether self-reactive Casp11^−/− T cells respond more strongly to endogenous ligands, we cotransferred WT and Casp11^−/− OT-I T cells into mice expressing OVA under the control of the fatty acid binding promoter (iFABP-tOVA), which leads to OVA production by mature enterocytes of the small intestine (14). At 6–7 d posttransfer, Casp11^−/− OT-I T cells outnumbered WT cells in the intestinal epithelium (IEL) and mesenteric lymph node (MLN) (Fig. 5). Thirty days after transfer, OT-I cell numbers were significantly lower in the MLN, as reported (14, 18). At this time, 50% of mice had no OT-I T cells of either genotype in the MLN, and the remaining mice had equal numbers of WT and Casp11^−/− cells (Fig. 5). Notably, in the IEL compartment at this late time point, WT OT-I T cells outnumbered Casp11^−/− cells (Fig. 5). These data suggest that the sensitivity of Casp11^−/− OT-I T cells to low-abundance endogenous ligands allowed preferential early expansion and increased numbers in the intestinal tissue relative to WT cells; however, Casp11^−/− OT-I T cells were more efficiently deleted in the intestine over time.

Caspase-11 expression is upregulated in CD8⁺ T cells as they mature, and it functions to negatively regulate TCR signaling following antigenic stimulation, resulting in decreased expansion and effector function, especially in response to low-level Ag stimulation. The stimulus for caspase-11 upregulation in this setting is still unclear. Our data suggest that it is not a type I IFN–dependent process, which is observed in other cell types. Upregulation of caspase-11 is equivalent during priming with Listeria, which results in robust type I IFN production, and protein immunization. Reactive oxygen species also stimulate caspase-11 upregulation (19), and TCR stimulation results in a metabolic shift that generates reactive oxygen species that could account for the upregulation of Casp11 during T cell activation (20).

The stepwise increase in the expression of caspase-11 by CD8⁺ T cells correlates with the reduced sensitivity of the TCR. Naive CD8⁺ T cells that have recently left the thymus are more sensitive to TCR signals than are those that have undergone maturation in the periphery (21), and Ag sensitivity leading to cell division is further reduced in CD8⁺ memory cells (22). Reduced TCR signaling in maturing T cells is more readily apparent when low-abundance and low-affinity Ags are used (21, 22). Lower expression of the TCR, CD3, and other surface molecules associated with robust TCR activation were implicated in reduced signaling (21–23). In addition, protein tyrosine phosphatases that negatively regulate TCR signaling are upregulated in CD8⁺ memory T cells (22). Our data suggest that caspase-11 represents an additional regulatory mechanism for tuning the sensitivity of maturing T cells to both foreign and endogenous Ags.

Overall, these data indicate that caspase-11 is upregulated as T cells become more Ag experienced and acts as a negative regulator of TCR signaling, particularly in response to low-abundance and low-affinity TCR ligands. Caspase-11 can regulate actin polymerization, which is known to play a critical role in regulating TCR signaling (9). These data suggest that modulating the function of caspase-11 in T cells could be used in the context of immunization to regulate their sensitivity to low-affinity Ags, allowing increased expansion and enhanced effector function without the risk of developing autoimmunity.

We thank Vishva Dixit and Timothy Hla for providing Caspase-11^−/− mice.

The authors have no financial conflicts of interest.