IL-17-secreting CD8 T cells (Tc17) have been described in several settings, but little is known regarding their functional characteristics. While Tc1 cells produced IFN-γ and efficiently killed targets, Tc17 cells lacked lytic function in vitro. Interestingly, the small numbers of IFN-γ-positive or IL-17/IFN-γ-double-positive cells generated under Tc17 conditions also lacked lytic activity and expressed a similar pattern of cell surface proteins to IL-17-producing cells. As is the case for Th17 (CD4) cells, STAT3 is important for Tc17 polarization, both in vitro and in vivo. Adoptive transfer of highly purified, Ag-specific IL-17-secreting Tc17 cells into Ag-bearing hosts resulted in near complete conversion to an IFN-γ-secreting phenotype and substantial pulmonary pathology, demonstrating functional plasticity. Tc17 also accumulated to a greater extent than did Tc1 cells, suggesting that adoptive transfer of CD8 T cells cultured in Tc17 conditions may have therapeutic potential for diseases in which IFN-γ-producing cells are desired.

A fairly recently described Th subset, CD4 T cells that IL-17 (Th17), has been shown to play an important role in several models of infectious and autoimmune diseases (1, 2). Culture of CD8 T cells in Th17 polarizing conditions results in IL-17 secretion, but the functional and phenotypic characteristics of IL-17-secreting CD8 T cells have not been systematically investigated. One recent study showed that CD8 T cells that lack both T-bet and eomesodermin are shunted into an IL-17-producing phenotype (3), and that these cells are nonfunctional in vivo, with double-knockout animals developing a viral wasting disease when infected with lymphocytic choriomeningitis virus. IL-17-secreting CD8 T cells can also be observed in mice deficient in T-bet alone, where they appear to play a role in allograft rejection (4). While these data suggest that IL-17-secreting CD8 T cells (Tc17) may represent a subset that is distinct from IFN-γ-secreting Tc1 CD8 T cells (5, 6), they are generated under somewhat artificial conditions and may say little about the existence or function of naturally occurring Tc17 cells.

Recent studies by the Dong group showed that CD4 Th subsets display considerable functional plasticity; that is, Treg cells can be converted to Th17 cells, and vice versa, under appropriate polarizing conditions (7). While the Th1 and Th2 subsets were previously considered to be less functionally plastic (8), recent studies show that Th2 cells can be converted into a unique IL-9-secreting subset (9). Supporting the concept of functional plasticity, complete CD4 T cell subset polarization is rarely observed in vivo. As Tc17 cells represent a recently described CD8 population, plasticity of this subset has yet to be investigated.

To address these issues, we polarized CD8 T cells toward IL-17 production (Tc17) in vitro and in vivo. As is the case for CD4 T cells (10, 11), expression of the transcription factor STAT3 appears to be important for Tc17 polarization. Phenotypic analysis shows that Tc17 cells do not express granzyme B and are not able to mediate lysis in vitro. However, adoptive transfer of highly purified, Ag-specific, IL-17-secreting Tc17 cells to Ag-expressing hosts demonstrated functional plasticity in this CD8 T cell subset, with loss of IL-17 expression and acquisition of IFN-γ expression after transfer, as well as increased accumulation and substantial pulmonary pathology as compared with Tc1 CD8 T cells. These studies suggest that adoptive transfer of Tc17 cells may have a role in the therapy of infectious disease and/or cancer.

B10.d2 and BALB/c mice were obtained from The Jackson Laboratory. C3HAhigh mice express hemagglutinin (HA)3 as a self-Ag on a B10.d2 background was previously described (12). CD8+ TCR transgenic mice (Clone 4), which express a TCR recognizing a Kd-restricted HA epitope (518IYSTVASSL526) (13), were either on a BALB/c background or were backcrossed >10 generations onto the B10.d2 background, with Thy1.1 as a congenic marker. Clone 4 mice on the B10.d2 background were backcrossed onto a STAT3flox background and then further crossed to CD4-Cre mice to obtain STAT3−/− Clone 4 animals. All animal studies were performed in accordance with protocols approved by the Animal Care and Use Committee of the Johns Hopkins University School of Medicine.

Spleens and peripheral lymph nodes were harvested from wild-type (WT) (B10.d2), Clone 4, or STAT3−/− Clone 4 donors. Naive CD8 T cells (CD8+CD62Lhigh) were obtained via magnetic bead separation according to the manufacturer’s protocol (Miltenyi Biotec). CD3/CD28 Dynalbeads (Invitrogen) were used for polyclonal activation. For Ag-specific activation, irradiated splenocytes were pulsed with 2 μg/ml HA class I Kd peptide (IYSTVASSL) and cocultured with naive HA-specific CD8 T cells at a ratio of 5:1. Culture medium used was IMDM (Invitrogen) supplemented with 1.0 mM sodium pyruvate (Sigma-Aldrich), 0.1 mM nonessential amino acids (Invitrogen), 100 IU/ml penicillin, 100 μg/ml streptomycin, 0.25 μg amphotericin B (Cellgro), 50 μM 2-ME (Sigma-Aldrich), and 5% heat-inactivated FBS (Quality Biological). Skewing conditions were as follows: Tc0 (media alone), Tc1 (6 ng/ml IL-12, 10 μg/ml anti-IL-4), Tc2 (20 ng/ml IL-4, 10 μg/ml anti-IFN-γ), Tc17 (20 ng/ml IL-6, 5 ng/ml TGF-β, 20 ng/ml IL-1β, 20 ng/ml IL-23, 10 μg/ml anti-IFN-γ, and 10 μg/ml anti-IL-4). Cells were cultured for 2 days and then rested for 3 days before restimulation with PMA (50 ng/ml) and ionomycin (500 ng/ml) before analysis or sorting. All cytokines were from Peprotec except for IL-23 (R&D Systems). Anti-IFN-γ was from eBioscience, and anti-IL-4 was provided by the National Cancer Institute Biological Resources Branch (Frederick, MD).

mAbs were purchased from BD Biosciences or eBioscience, with the exception of anti-IL-18R (R&D Systems). For ICS, cells were restimulated for 5 h as above in the presence of GolgiStop (BD Biosciences). For phospho-STAT3 ICS, 106 cells were stimulated with 20 ng/ml IL-6 for 20 min, followed by fixation in 4% paraformaldehyde. After washing, cells were permeabilized with 90% methanol at 4°C then stained with phospho-STAT3 (BD Biosciences). Flow cytometry was performed using a FACSCalibur (BD Biosciences) and analyzed using FlowJo software (Tree Star). Statistical analyses were performed using Prism 4 (GraphPad Software).

After in vitro polarization and restimulation with PMA (50 ng/ml) and ionomycin (500 ng/ml) for 2 h, IL-17 or IFN-γ-secreting CD8 T cells were stained with bispecific CD45/cytokine mAbs according the manufacturer’s protocol (Miltenyi Biotec) and then sorted to >90% purity using a BD FACSAria instrument. Propidium iodide staining was utilized to exclude dead/dying cells, and CD8 T cells were sorted into four discrete populations: IL-17+IFN-γ, IFN-γ+IL-17, IL-17/IFN-γ double-positive cells and double-negative CD8 T cells.

A20 lymphoma cells labeled with HA peptide (2 μg/ml) were used as targets. Clone 4 CD8 T cells on a BALB/c background polarized into Tc1 or Tc17 effectors were incubated for 4 h with 5 × 103 Cr51-loaded targets at the indicated ratios in triplicate wells. Percentage lysis was calculated as previously described (14).

A recombinant vaccinia virus strain encoding full-length HA (VacHA) from the influenza virus strain A/PR8/34 (H1N1) was a gift from Frank Guarnieri (The Johns Hopkins University) as previously described (15, 16, 17). VacHA was constructed by recombining HA gene into the thymidine kinase gene of WT vaccinia virus. VacHA was expanded on Hu-TK cells in the presence of 25 μg/ml BrdU (Sigma-Aldrich), purified from cell lysates by sucrose banding, titered by plaque assay on BSC-1 cells, and stored at a concentration of 1 × 108 PFU/ml in the −80°C freezer. For a priming dose, VacHA was thawed on ice and diluted in PBS to 1 × 106 PFU per mouse in a volume of 200 μl for intraperitoneal injection. A L. monocytogenes mutant strain engineered to express the influenza HA class I epitope 518IYSTVASSL526 (LMHA) was a gift from Cerus as previously described (18). Briefly, L. monocytogenes carrying the HA epitope (LM-actA INLB uvr PPL2) was derived from wild-type L. monocytogenes strain 10403S and was attenuated via in-frame deletions in the actA, INLB, uvr, and PPL2 genes, respectively. LMHA was grown in Difco brain-heart infusion medium (BD Diagnostic Systems), harvested at mid-log phase of growth, purified by standard methods, formulated in phosphate-buffered sodium chloride solution (PBS)/8% DMSO at a concentration of <1 × 1010 CFU/ml, and stored at −80°C. To prime the mice, LMHA was thawed on ice and diluted in PBS to 1 × 107 CFU per mouse in a volume of 200 μl, corresponding to 0.1 median lethality (0.1 × LD50), for i.v. injection.

In initial adoptive transfer experiments, 2 × 106 CFSE-labeled, Thy1.1+, naive wild-type Clone 4 or STAT3−/− Clone 4 CD8 T cells (B10.d2) were adoptively transferred into Thy1.2+ B10.d2 or C3HAhigh recipients and harvested 4 days posttransfer as previously described (10, 18). In additional experiments, we transferred highly purified, four-way sorted CD8 T cells. For these studies, 1 × 105 Tc1 or Tc17 cells (purified based on cytokine profiles) were adoptively transferred to C3HAhigh mice and harvested 7 days posttransfer. To model the CD8 T cell response to infectious agents, 1 × 106 PFU recombinant vaccinia virus carrying full length HA (VacHA) (15, 16, 17) or 1 × 107 CFU recombinant L. monocytogenes carrying HA Ag (LMHA) (18) were administered to B10.d2 mice 1 h after adoptive transfer (i.p. and i.v, respectively), and cells were harvested 4 days later, as described above.

Total RNA was extracted using the RNeasy Micro Kit (Qiagen), and cDNA was synthesized with the SuperScript III enzyme (Invitrogen). All primers were purchased from Applied Biosystems; reactions were performed in triplicate using an Applied Biosystems 7900 instrument.

Mice were anesthetized using avertin (0.5 mg/g body mass; Sigma-Aldrich), followed by intracardiac perfusion with PBS. The lungs were infused with Bouin’s solution (Sigma-Aldrich) via trachea at a constant pressure for homogeneous expansion of lung parenchyma, and then fixed with 10% formalin. Tissues were paraffin-embedded and sectioned (5 μm), then stained with H&E for analysis. An inflammation score was assigned in a blinded manner (R.A.A.) as described previously (19). Briefly, a cumulative pathological score was calculated according to the following: peribronchiolar thickening (0, normal; 1, abnormal), >50% alveolar septum widening (0, normal; 1, abnormal), hemorrhage (0, normal; 1, abnormal). Negative control tissues were from C3HAhigh mice that did not receive an adoptive CD8 T cell transfer.

Based on recent reports suggesting that IL-17-secreting CD8 T cells (Tc17) may have a nonfunctional phenotype in vivo (3), we optimized cell culture conditions to polarize Ag-specific CD8 cells from TCR transgenic mice (Clone 4) to secrete IL-17 so that we could perform further phenotypic and functional evaluation. As shown in Fig. 1,A, optimized culture conditions were not dramatically different from those used to generate IL-17-secreting CD4 T cells (Th17) (1, 2, 10), with the possible exception that addition of Il-1β did not appear to contribute markedly to Tc17 polarization (20). Notably, blockade of IL-4 and IFN-γ contributed significantly to Tc17 polarization (Fig. 1,A, right panels). As compared with CD8 T cells polarized toward a Tc1 phenotype, Tc17 cells secreted more IL-17 and less IFN-γ, although we were unable to completely abrogate IFN-γ secretion from Tc17 cells under any combination of conditions tested (Fig. 1, A and B). Interestingly, both Tc1 and Tc17 cells secrete significant quantities of TNF-α, and neither subset produces IL-10 (Fig. 1,C). Quantitative real-time PCR analysis (Fig. 1 D) suggested that these differences in IL-17 and IFN-γ secretion were at least partially controlled at the transcriptional level. These data also showed dramatic up-regulation of the transcription factor retinoic acid-related orphan receptor (ROR)γt in Tc17 cells, suggesting a role for this molecule in IL-17 secretion by CD8 T cells, as is the case for CD4 T cells (21). Interestingly, although T-bet message was up-regulated in Tc1 as compared with Tc2 cells (22), its expression was only slightly decreased in Tc17 as compared with Tc1 cells, similar to a recent finding in an Ag-presenting cell-driven Tc17 subset (23), but different from a report on mixed lymphocyte culture (24). As expected, GATA-3 was highly expressed in Tc2 cells, but not Tc1 or Tc17 cells (25). Tc17 CD8 T cells also expressed a high level of IL-23R but not IL-12Rβ2, consistent with the notion that these cells represent a population distinct from Tc1 CD8 T cells.

FIGURE 1.

In vitro polarization of IL-17-secreting CD8 T cells (Tc17). A, HA-specific CD8 T cells were activated in vitro for 2 days with cognate peptide plus irradiated APC in the presence of indicated cytokines. ICS was performed after 3 days of rest followed by 5 h of restimulation with PMA/ionomycin. B, Statistical analysis of IL-17 and IFN-γ production from polarized HA-specific CD8 T cells. Means ± SEM from four independent experiments are shown. ∗, p < 0.05 and ∗∗, p < 0.01 vs IL-6 + TGF-β. C, Comparison of cytokine secretion profiles of in vitro polarized Tc1, Tc2, and Tc17 CD8 cells. D, Quantitative real-time PCR analysis of CD8 T cell subsets; data represent fold change relative to naive (CD8+CD62Lhigh) cells. Means ± SEM are shown, performed in triplicate, repeated four times.

FIGURE 1.

In vitro polarization of IL-17-secreting CD8 T cells (Tc17). A, HA-specific CD8 T cells were activated in vitro for 2 days with cognate peptide plus irradiated APC in the presence of indicated cytokines. ICS was performed after 3 days of rest followed by 5 h of restimulation with PMA/ionomycin. B, Statistical analysis of IL-17 and IFN-γ production from polarized HA-specific CD8 T cells. Means ± SEM from four independent experiments are shown. ∗, p < 0.05 and ∗∗, p < 0.01 vs IL-6 + TGF-β. C, Comparison of cytokine secretion profiles of in vitro polarized Tc1, Tc2, and Tc17 CD8 cells. D, Quantitative real-time PCR analysis of CD8 T cell subsets; data represent fold change relative to naive (CD8+CD62Lhigh) cells. Means ± SEM are shown, performed in triplicate, repeated four times.

Close modal

Since previous work by our group (10) and others (11) demonstrated a requirement for the transcription factor STAT3 in the generation of IL-17-secreting CD4 T cells (Th17), we postulated a similar requirement in Tc17 polarization. These experiments were facilitated by the fact that the CD4 promoter used to generate CD4-Cre mice is expressed at high levels in the thymus in CD4+CD8+ precursors, resulting in Cre-mediated excision in both CD4 and CD8 T cells (Fig. 2,A). In vitro-polarized, Ag-specific (Fig. 2,C) or bulk naive non-TCR transgenic CD8 T cells (Fig. 2,B) from CD8 STAT3−/− mice showed a marked attenuation of IL-17 secretion, although not as complete as was observed in CD4 T cells (10). This discrepancy most likely stems from incomplete deletion of STAT3 in the CD8 compartment (Fig. 2,A). This in vitro finding by skewing CD8 T cells from the conditional STAT3 knockout mice is similar to a recent publication using small interfering RNA to silence STAT3 gene expression (26). Although IL-17-producing CD8 T cells has been found in different mouse models (3, 4, 26, 27) and in humans (28, 29), the in vivo requirement of STAT3 to produce IL-17 in CD8 T cells has not been shown. Therefore, we further explored this STAT3 requirement in vivo by adoptively transferring 2 × 106 HA-specific wild-type or STAT3−/− CD8 T cells into mice expressing HA as a self-Ag (C3HAhigh), or as an infectious Ag (VacHA and LMHA). These data showed nearly complete abrogation of IL-17 secretion by the STAT3−/− CD8 T cells in vivo (Fig. 2 D). Ag-specific STAT3−/− CD8 cells were not markedly different from wild-type cells in terms of proliferation (as assayed by CFSE dilution); >90% of either wild-type or STAT3−/− cells divided in response to self-Ag or infectious Ag by day 4. Survival of these cells was not markedly affected either, as persistence in nontransgenic (B10.d2) hosts was not grossly altered. Thus, as is the case for CD4 T cells (10, 11), STAT3 appears to be required for IL-17 secretion by CD8 T cells both in vitro and in vivo.

FIGURE 2.

STAT3 is required for Tc17 differentiation in vitro and in vivo. A, Phospho-STAT3 expression is abrogated in both CD4 and CD8 T cells from CD4-Cre × STAT3flox mice. B, Polyclonal naive CD8 T cells from wild-type and CD8 STAT3KO mice activated in vitro with CD3/CD28 microbeads under Tc17 skewing condition. ICS was performed post-PMA/ionomycin restimulation. C, HA-specific (Clone 4) CD8 T cells from wild-type and CD8 STAT3KO donors stimulated in vitro with specific peptide and irradiated APC under Tc1, Tc2, and Tc17 skewing conditions. D, WT but not STAT3KO Clone 4 CD8 T cells produce IL-17 after recognition of self-Ag in vivo. CFSE-labeled donor cells (gated on Thy1.1) analyzed 4 days after adoptive transfer to indicated recipients: B10.d2 = WT, VacHA = WT + vaccinia expressing HA, LMHA = WT + Listeria expressing HA, C3HAhigh = transgenic animals widely expressing HA as self-Ag. Doses of infectious agents: VacHA, 1 × 106 PFU i.p, LMHA-1 × 107 CFU i.v. Data are representative of two independent experiments using three to five mice per group.

FIGURE 2.

STAT3 is required for Tc17 differentiation in vitro and in vivo. A, Phospho-STAT3 expression is abrogated in both CD4 and CD8 T cells from CD4-Cre × STAT3flox mice. B, Polyclonal naive CD8 T cells from wild-type and CD8 STAT3KO mice activated in vitro with CD3/CD28 microbeads under Tc17 skewing condition. ICS was performed post-PMA/ionomycin restimulation. C, HA-specific (Clone 4) CD8 T cells from wild-type and CD8 STAT3KO donors stimulated in vitro with specific peptide and irradiated APC under Tc1, Tc2, and Tc17 skewing conditions. D, WT but not STAT3KO Clone 4 CD8 T cells produce IL-17 after recognition of self-Ag in vivo. CFSE-labeled donor cells (gated on Thy1.1) analyzed 4 days after adoptive transfer to indicated recipients: B10.d2 = WT, VacHA = WT + vaccinia expressing HA, LMHA = WT + Listeria expressing HA, C3HAhigh = transgenic animals widely expressing HA as self-Ag. Doses of infectious agents: VacHA, 1 × 106 PFU i.p, LMHA-1 × 107 CFU i.v. Data are representative of two independent experiments using three to five mice per group.

Close modal

We next sought to determine whether Tc17 cells displayed a cell surface phenotype distinct from Tc1 cells. Indeed, cell surface expression of IL-18R, CD45RB, CD38, CD103, and CCR6 were different between Tc17 and Tc1 cells (Fig. 3,A). However, while these markers distinguish Tc1 from Tc17 cells, none was capable of discriminating between the IL-17- and IFN-γ-secreting populations that emerge under Tc17 polarizing conditions. In terms of CTL functional molecules, we found that expression of granzyme B was markedly attenuated in both IL-17- and IFN-γ-producing Tc17 populations, while other functional markers were not grossly different between Tc17 and Tc1 cells (Fig. 3 B). Additionally, the IL-17- and IFN-γ-secreting cells from Tc17 conditions were nearly identical in expression of these functional molecules.

FIGURE 3.

Phenotypic and effector molecule analysis of Tc1 and Tc17 subsets. A, Cell surface phenotype of clonotypic Tc1 cells as compared with IL-17- or IFN-γ-secreting CD8 cells differentiated in vitro under Tc17 conditions. Isotype controls are shown in gray. B, Expression of effector molecules in Tc1 vs IL-17 or IFN-γ-producing Tc17 cells. All experiments were repeated three times.

FIGURE 3.

Phenotypic and effector molecule analysis of Tc1 and Tc17 subsets. A, Cell surface phenotype of clonotypic Tc1 cells as compared with IL-17- or IFN-γ-secreting CD8 cells differentiated in vitro under Tc17 conditions. Isotype controls are shown in gray. B, Expression of effector molecules in Tc1 vs IL-17 or IFN-γ-producing Tc17 cells. All experiments were repeated three times.

Close modal

We next optimized four-way sorting techniques to obtain >98% pure IFN-γ-secreting Tc1 cells, as well as each of the four distinct populations that emerge under Tc17 skewing conditions (Fig. 4,A). Postsort analyses showed that ∼1% of the sorted Tc17 IL-17 producers were IFN-γ-positive, and ∼4% of the sorted Tc17 IFN-γ producers were IL-17-positive. In-depth quantitative real-time PCR analyses of these cells revealed a complex pattern of cytokine, transcription factor, and effector molecule transcription. We found a 300,000-fold relative increase in IL-17 message in the sorted IL-17-producing cells, as well an approximate 15,000-fold increase in IFN-γ message in IFN-γ-producing cells. As was the case for bulk-cultured Tc17 cells (Fig. 1,D), we found a significant up-regulation of RORγt message in all of the populations that secreted IL-17. These results also support our earlier data regarding T-bet message, which was not statistically up-regulated in Tc17 IFN-γ producing cells as compared with Tc1 cells. Perforin and granzyme B message levels were significantly decreased in IL-17-secreting Tc17 cells, with the possibility that these cells might lack lytic function. To assay the functional capacity of these sorted cells, we performed classical in vitro CTL analyses, using HA-pulsed A20 lymphoma cells as targets (Fig. 4 C). Consistent with the observed reduction in effector molecule levels, Tc17 cells were unable to mediate significant lysis in vitro, regardless of their cytokine production patterns. Taken together, these results are consistent with the notion that, regardless of cytokine secretion pattern, Tc17 CD8 T cells represent a cell population without classic CTL function.

FIGURE 4.

Differential gene expression and in vitro CTL function of Tc1 and Tc17 cells. A, Sorting strategy. After in vitro differentiation, clonotypic CD8 T cells were restimulated with PMA/ionomycin for 2 h, followed by staining for cytokine expression and four-way sorting to isolate populations of Tc17 IL-17, IFN-γ, IL-17/IFN-γ double-positive (DP) and double-negative (DN) CD8 T cells as well as Tc1 IFN-γ producers. B, Quantitative real-time PCR analyses of sorted, cytokine-expressing Tc1 cells as compared with Tc17 cells that secrete IL-17, IFN-γ, both, or neither. Expression level relative to naive (CD8+CD62Lhigh) cells calculated using ΔΔCT with 18S as an internal control. Reactions were performed in triplicate and repeated four times (∗, p < 0.05 and ∗∗, p < 0.01 vs Tc1 IFN-γ). C, In vitro lytic function of Tc subsets. In vitro polarized HA-specific CD8 T cells were sorted by cytokine secretion assay, and 4-h 51Cr-release assay was performed using A20 cells pulsed with HA class I peptide as targets, with unlabeled A20 as negative control. Mean ± SEM of triplicate wells, repeated three times.

FIGURE 4.

Differential gene expression and in vitro CTL function of Tc1 and Tc17 cells. A, Sorting strategy. After in vitro differentiation, clonotypic CD8 T cells were restimulated with PMA/ionomycin for 2 h, followed by staining for cytokine expression and four-way sorting to isolate populations of Tc17 IL-17, IFN-γ, IL-17/IFN-γ double-positive (DP) and double-negative (DN) CD8 T cells as well as Tc1 IFN-γ producers. B, Quantitative real-time PCR analyses of sorted, cytokine-expressing Tc1 cells as compared with Tc17 cells that secrete IL-17, IFN-γ, both, or neither. Expression level relative to naive (CD8+CD62Lhigh) cells calculated using ΔΔCT with 18S as an internal control. Reactions were performed in triplicate and repeated four times (∗, p < 0.05 and ∗∗, p < 0.01 vs Tc1 IFN-γ). C, In vitro lytic function of Tc subsets. In vitro polarized HA-specific CD8 T cells were sorted by cytokine secretion assay, and 4-h 51Cr-release assay was performed using A20 cells pulsed with HA class I peptide as targets, with unlabeled A20 as negative control. Mean ± SEM of triplicate wells, repeated three times.

Close modal

We next tested whether Tc1 and Tc17 phenotypes were stable; that is, could they be maintained after adoptive transfer into an Ag-expressing environment? We found that after adoptive transfer into C3HAhigh mice, sorted, cytokine-secreting Tc1 cells were rapidly lost, as previously described in a pmel-1 tumor Ag system (30), but Tc17 cells appeared to persist to a greater extent (Fig. 5,A). Surprisingly, both IL-17- and IFN-γ-secreting Tc17 cells persisted (Fig. 5,B), and significant IFN-γ secretion was detected after transfer of either population (Fig. 5,A). In fact, the most robust IFN-γ secretion came from in vitro purified IL-17-secreting Tc17 cells. Such robust conversion is similar to that recently described for CD4 Th17 cells in an in vivo diabetes model (31). Additionally, negligible IL-17-secreting cells were detected after transfer, either in the lung (Fig. 5,C) or the draining lymph nodes (Fig. 5,D). These data indicate that IL-17-secreting Tc17 cells generated in vitro can convert to IFN-γ secretion in a self-Ag environment and are consistent with the notion that T cell phenotypes may be functionally plastic (7, 9, 31, 32). Although we cannot completely deplete IFN-γ producers from Tc17 IL17 secretors, or completely eliminate IL-17 producers from Tc17 IFN-γ secretors before adoptive transfer (Fig. 4,A), a side-by-side comparison using Tc17 IL-17 and Tc17 IFN-γ generated from the same culture condition enables us to exclude the possibility of a dramatic expansion of a small population of IFN-γ secretors in the purified IL-17-secreting Tc17 population, since greater numbers of lung-infiltrating clonotypic cells were detected after transfer of IL-17-secreting Tc17 than for IFN-γ-secreting Tc17 IFN-γ transfer (mean of absolute number ± SEM: Tc17 IL-17 = 1271 ± 164.8 vs Tc17 IFN-γ = 644.0 ± 148.9; p = 0.0379) (Fig. 5,B). Thus, IL-17-secreting Tc17 cells have a greater in vivo expansion/persistence capacity than IFN-γ-secreting Tc17 cells. Both Tc17 populations expressed CD107 after transfer (Fig. 5 E), consistent with potential restoration of effector function.

FIGURE 5.

In vivo plasticity and persistence of Tc17 cells. A, FACS analysis of lung-infiltrating lymphocytes. Seven days after adoptive transfer of 1 × 105 sorted Tc1 or Tc17 cells, recipient mice were sacrificed and donor cells (Thy1.1) from a CD8 gate were analyzed by ICS after ex vivo restimulation; n = 3–5 mice per group, representative FACS plots shown. Experiments were repeated twice. B, Absolute number and percentage of Tc17 vs Tc1 cells in the lungs of C3HAhigh mice (∗, p < 0.05 and ∗∗, p < 0.01 vs Tc1 IFN-γ). Means ± SEM are shown; n = 3–5 mice per group. C and D, Cytokine production by clonotypic Tc17 vs Tc1 cells in the lungs (C) or hilar draining lymph nodes (D) of C3HAhigh mice (∗, p < 0.05 and ∗∗, p < 0.01 vs Tc1 IFN-γ). Means ± SEM are shown; n = 3–5 mice per group. E, CD107a expression on Tc17 lung-infiltrating lymphocytes. Mean fluorescence intensity (MFI) ± SEM is shown.

FIGURE 5.

In vivo plasticity and persistence of Tc17 cells. A, FACS analysis of lung-infiltrating lymphocytes. Seven days after adoptive transfer of 1 × 105 sorted Tc1 or Tc17 cells, recipient mice were sacrificed and donor cells (Thy1.1) from a CD8 gate were analyzed by ICS after ex vivo restimulation; n = 3–5 mice per group, representative FACS plots shown. Experiments were repeated twice. B, Absolute number and percentage of Tc17 vs Tc1 cells in the lungs of C3HAhigh mice (∗, p < 0.05 and ∗∗, p < 0.01 vs Tc1 IFN-γ). Means ± SEM are shown; n = 3–5 mice per group. C and D, Cytokine production by clonotypic Tc17 vs Tc1 cells in the lungs (C) or hilar draining lymph nodes (D) of C3HAhigh mice (∗, p < 0.05 and ∗∗, p < 0.01 vs Tc1 IFN-γ). Means ± SEM are shown; n = 3–5 mice per group. E, CD107a expression on Tc17 lung-infiltrating lymphocytes. Mean fluorescence intensity (MFI) ± SEM is shown.

Close modal

We next tested whether the observed in vivo conversion of Tc17 cells from IL-17 to IFN-γ secretion (Fig. 5) resulted in pulmonary pathology in our self-Ag model. These studies were complicated by the relatively small numbers of highly purified cytokine-secreting CD8 cells (1 × 105) available for adoptive transfer (Fig. 4,A). Surprisingly, H&E sections from recipients of Tc17 IL-17-secreting cells revealed infiltration, hemorrhage, widened alveolar septae, and peribronchiolar thickening (Fig. 6,A). Blinded scoring of these sections confirmed that inflammation was significantly increased in mice receiving IL-17 secreting Tc17 cells (Fig. 6 B).

FIGURE 6.

Adoptive transfer of IL-17-secreting Tc17 cells results in pulmonary pathology. A, In vitro-polarized HA-specific CD8 T cells were adoptively transferred into C3HAhigh mice. Seven days after adoptive transfer of 1 × 105 sorted Tc1 or Tc17 cells, recipient mice were sacrificed and lung tissues were stained with H&E. Representative sections at ×20 magnification shown. B, Mean pathology score (0–3) is shown; 6–12 mice per group. Mean ± SEM shows ∗, p < 0.05 vs Tc1 IFN-γ.

FIGURE 6.

Adoptive transfer of IL-17-secreting Tc17 cells results in pulmonary pathology. A, In vitro-polarized HA-specific CD8 T cells were adoptively transferred into C3HAhigh mice. Seven days after adoptive transfer of 1 × 105 sorted Tc1 or Tc17 cells, recipient mice were sacrificed and lung tissues were stained with H&E. Representative sections at ×20 magnification shown. B, Mean pathology score (0–3) is shown; 6–12 mice per group. Mean ± SEM shows ∗, p < 0.05 vs Tc1 IFN-γ.

Close modal

A CD4 T cell subset producing IL-17 (Th17) has been shown to play important roles in inflammatory and autoimmune diseases (1, 2). However, few studies have focused on IL-17-secreting CD8 T cells (Tc17), especially with regard to their phenotype and function (3, 4, 26, 27, 28, 29, 33). In the current study we specifically generated IL-17-secreting CD8 cells (Fig. 1) and further purified this subset to characterize its phenotype (Figs. 3 and 4, A and B) and function in vitro (Fig. 4,C) and in vivo (Fig. 5). As is the case for CD4 cells, we showed that CD8 T lymphocytes can be polarized to secrete IL-17 under conditions similar to those used to differentiate Th17 cells. Using a CD4-Cre × STAT3flox system, we showed for the first time that the transcription factor STAT3 appears to be required for CD8 T cells to be polarized toward IL-17 production both in vitro and in vivo (Fig. 2). We also showed that the transcription factor RORγt, a crucial transcription factor for Th17 differentiation, is up-regulated in Tc17. We found that Tc17 differentiation can occur in vivo as well; naive Ag-specific CD8 cells produced IL-17 after adoptive transfer into mice that express their cognate Ag as a self-Ag in the lung (Fig. 2 D).

A previous study using knockout mice suggested that naturally arising or differentiated Tc17 cells might prove to be nonfunctional (3). In mice lacking both T-bet and eomesodermin, CD8 T cells secreted IL-17 but not IFN-γ, and mice developed a chronic viral wasting syndrome when infected with lymphocytic choriomeningitis virus (3). Similarly, when extrinsic IL-12 and type I IFN signals were absent, a population of IFN-γ/IL-17-double-positive CD8 T cells emerged in response to L. monocytogenes infection (33). However, using an allograft rejection model (4), the Bishop group showed that, in T-bet-deficient mice, IL-17-producing CD8 T cells do appear to be functional and play a role in graft rejection. Further data suggesting that IL-17 producing CD8 T cells play a functional role in vivo came from studies of contact dermatitis; here, Ab-mediated neutralization of IL-17 mitigated dermatitis in vivo, and depletion studies suggested that CD8 T cells were a major source of this cytokine (27). In vitro, our data seemed to corroborate data from previous reports (3, 24, 26), as in vitro-polarized Tc17 cells lacked granzyme B expression (Fig. 3,B) and did not mediate lysis in an in vitro CTL assay (Fig. 4,C). However, our in vivo results painted a somewhat different picture. After transfer into animals where HA is widely expressed as a self-Ag, HA-specific IL-17-secreting Tc17 cells expanded significantly, with the emergence of an IFN-γ-secreting phenotype (Fig. 5). This switch in cytokine production was accompanied by a significant increase in accumulation as compared with Tc1 CD8 T cells and substantial pulmonary inflammation, suggesting a functional role for these cells in vivo.

The in vivo conversion of Tc17 IL-17 secretors to IFN-γ-secreting cells is surprising and does not appear to be explained by differential expansion of the small (<2%) population of IFN-γ-secreting precursors in the highly purified adoptive transfer product, since sorted IFN-γ-producing cells (from Tc17 conditions) did not out-proliferate IL-17-secreting Tc17 cells under identical, side-by-side in vivo conditions. Additionally, this conversion occurred rapidly (<1 wk), arguing against a massive expansion of a small subpopulation contaminating the purified cells used for transfer. Instead, our data are consistent with the notion that Tc17 cells demonstrate functional plasticity, as has recently been reported for CD4 T cells by three separate groups (7, 9, 31, 32). In the most relevant example, the Weaver group recently showed that IFN-γ-producing CD4 cells could emerge from Th17 precursors in an IL-12-driven manner (32). Additionally, the Dong group showed functional interconversion between regulatory CD4 T cells and Th17 in vitro (7). Finally, the Stockinger group demonstrated that Th2 cells could be converted to a unique IL-9-secreting phenotype under appropriate conditions (9) and, using a similar methodology to that employed here, showed that highly purified TCR-transgenic CD4 Th17 cells could convert to Th1 cells in a self-Ag diabetes model (31). So, although functional plasticity of CD4 T cells has been reported, our data are the first of which we are aware showing a similar plasticity in CD8 T cells in vivo. Note that during the review process for this paper, the Restifo group obtained similar results in a tumor model system using bulk-cultured (but not highly sorted and purified) Ag-specific CD8 T cells (23). The molecular mechanisms underlying this plasticity have yet to be elucidated, but may involve differential expression of transcription factors or epigenetic modifications within effector cytokine gene promoter and enhancer regions, as has been demonstrated for CD4 T cells (34).

Overall, these studies provide novel data regarding the plasticity and function of IL-17-secreting CD8 T cells (Tc17). Polarization of CD8 T cells in the presence of IL-6 and TGF-β results in populations that secrete IL-17 and IFN-γ (Fig. 1), and these two populations cannot be readily differentiated by cell surface phenotype (Fig. 3 A). As is the case for IL-17-secreting CD4 T cells (Th17) (10, 11), the transcription factor STAT3 appears to be required for Tc17 polarization. After transfer into animals where HA is widely expressed as a self-Ag, HA-specific IL-17-secreting Tc17 demonstrated a novel functional plasticity, with the emergence of an IFN-γ-secreting phenotype and the development of pulmonary pathology. Our present data suggest that, similar to CD4 T cells (35), adoptive transfer of IL-17-secreting CD8 T cells in an infectious disease or tumor setting may have therapeutic benefit.

We are grateful to S. Akira and K. Takeda (Osaka University, Osaka, Japan) for providing the Stat3flox mice and Dr. Linda A. Sherman (The Scripps Research Institute, La Jolla, CA) for Clone 4 mice. We also thank James Flook, Ada J. Tam, and Richard Lee Blosser for assistance with cell sorting.

The authors have no financial conflicts of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

C.G.D. is a Damon Runyon-Lilly Clinical Investigator. This work was also supported by National Institutes of Health Grants R01 CA127153 (to C.G.D.) and K08 CA096948 (to C.G.D.) and the Patrick C. Walsh Fund. D.M.P. is a Januey Scholar, holds the Seraph Chair for Cancer Research, and is supported in part by gifts from William and Betty Toperer, Dorothy Needle, and the Commonwealth Foundation. H.R.Y. is the recipient of a Taiwan Merit Scholarship from the National Science Council and Ministry of Education, as well as Chang Gung Medical Research Grants (CMRPG 350821 and 350822) from Chang Gung Memorial Hospital, Taiwan.

3

Abbreviations used in this paper: HA, hemagglutinin; ICS, intracellular staining; ROR, retinoic acid-related orphan receptor.

1
Dong, C..
2008
. TH17 cells in development: an updated view of their molecular identity and genetic programming.
Nat. Rev. Immunol.
8
:
337
-348.
2
Weaver, C. T., L. E. Harrington, P. R. Mangan, M. Gavrieli, K. M. Murphy.
2006
. Th17: an effector CD4 T cell lineage with regulatory T cell ties.
Immunity
24
:
677
-688.
3
Intlekofer, A. M., A. Banerjee, N. Takemoto, S. M. Gordon, C. S. Dejong, H. Shin, C. A. Hunter, E. J. Wherry, T. Lindsten, S. L. Reiner.
2008
. Anomalous type 17 response to viral infection by CD8+ T cells lacking T-bet and eomesodermin.
Science
321
:
408
-411.
4
Burrell, B. E., K. Csencsits, G. Lu, S. Grabauskiene, D. K. Bishop.
2008
. CD8+ Th17 mediate costimulation blockade-resistant allograft rejection in T-bet-deficient mice.
J. Immunol.
181
:
3906
-3914.
5
Mosmann, T. R., L. Li, S. Sad.
1997
. Functions of CD8 T-cell subsets secreting different cytokine patterns.
Semin. Immunol.
9
:
87
-92.
6
Becker, C., H. Pohla, B. Frankenberger, T. Schuler, M. Assenmacher, D. J. Schendel, T. Blankenstein.
2001
. Adoptive tumor therapy with T lymphocytes enriched through an IFN-γ capture assay.
Nat. Med.
7
:
1159
-1162.
7
Yang, X. O., R. Nurieva, G. J. Martinez, H. S. Kang, Y. Chung, B. P. Pappu, B. Shah, S. H. Chang, K. S. Schluns, S. S. Watowich, et al
2008
. Molecular antagonism and plasticity of regulatory and inflammatory T cell programs.
Immunity
29
:
44
-56.
8
Murphy, E., K. Shibuya, N. Hosken, P. Openshaw, V. Maino, K. Davis, K. Murphy, A. O'Garra.
1996
. Reversibility of T helper 1 and 2 populations is lost after long-term stimulation.
J. Exp. Med.
183
:
901
-913.
9
Veldhoen, M., C. Uyttenhove, J. van Snick, H. Helmby, A. Westendorf, J. Buer, B. Martin, C. Wilhelm, B. Stockinger.
2008
. Transforming growth factor-β“reprograms” the differentiation of T helper 2 cells and promotes an interleukin 9-producing subset.
Nat. Immunol.
9
:
1341
-1346.
10
Harris, T. J., J. F. Grosso, H. R. Yen, H. Xin, M. Kortylewski, E. Albesiano, E. L. Hipkiss, D. Getnet, M. V. Goldberg, C. H. Maris, et al
2007
. Cutting edge: an in vivo requirement for STAT3 signaling in TH17 development and TH17-dependent autoimmunity.
J. Immunol.
179
:
4313
-4317.
11
Yang, X. O., A. D. Panopoulos, R. Nurieva, S. H. Chang, D. Wang, S. S. Watowich, C. Dong.
2007
. STAT3 regulates cytokine-mediated generation of inflammatory helper T cells.
J. Biol. Chem.
282
:
9358
-9363.
12
Adler, A. J., D. W. Marsh, G. S. Yochum, J. L. Guzzo, A. Nigam, W. G. Nelson, D. M. Pardoll.
1998
. CD4+ T cell tolerance to parenchymal self-antigens requires presentation by bone marrow-derived antigen-presenting cells.
J. Exp. Med.
187
:
1555
-1564.
13
Morgan, D. J., R. Liblau, B. Scott, S. Fleck, H. O. McDevitt, N. Sarvetnick, D. Lo, L. A. Sherman.
1996
. CD8+ T cell-mediated spontaneous diabetes in neonatal mice.
J. Immunol.
157
:
978
-983.
14
Kumaraguru, U., S. Suvas, P. S. Biswas, A. K. Azkur, B. T. Rouse.
2004
. Concomitant helper response rescues otherwise low avidity CD8+ memory CTLs to become efficient effectors in vivo.
J. Immunol.
172
:
3719
-3724.
15
Staveley-O'Carroll, K., E. Sotomayor, J. Montgomery, I. Borrello, L. Hwang, S. Fein, D. Pardoll, H. Levitsky.
1998
. Induction of antigen-specific T cell anergy: an early event in the course of tumor progression.
Proc. Natl. Acad. Sci. USA
95
:
1178
-1183.
16
Higgins, A. D., M. A. Mihalyo, P. W. McGary, A. J. Adler.
2002
. CD4 cell priming and tolerization are differentially programmed by APCs upon initial engagement.
J. Immunol.
168
:
5573
-5581.
17
Yang, Y., C. T. Huang, X. Huang, D. M. Pardoll.
2004
. Persistent Toll-like receptor signals are required for reversal of regulatory T cell-mediated CD8 tolerance.
Nat. Immunol.
5
:
508
-515.
18
Goldberg, M. V., C. H. Maris, E. L. Hipkiss, A. S. Flies, L. Zhen, R. M. Tuder, J. F. Grosso, T. J. Harris, D. Getnet, K. A. Whartenby, et al
2007
. Role of PD-1 and its ligand, B7-H1, in early fate decisions of CD8 T cells.
Blood
110
:
186
-192.
19
Tong, J., H. S. Bandulwala, B. S. Clay, R. A. Anders, R. A. Shilling, D. D. Balachandran, B. Chen, J. V. Weinstock, J. Solway, K. J. Hamann, A. I. Sperling.
2006
. Fas-positive T cells regulate the resolution of airway inflammation in a murine model of asthma.
J. Exp. Med.
203
:
1173
-1184.
20
Sutton, C., C. Brereton, B. Keogh, K. H. Mills, E. C. Lavelle.
2006
. A crucial role for interleukin (IL)-1 in the induction of IL-17-producing T cells that mediate autoimmune encephalomyelitis.
J. Exp. Med.
203
:
1685
-1691.
21
Ivanov, I. I., B. S. McKenzie, L. Zhou, C. E. Tadokoro, A. Lepelley, J. J. Lafaille, D. J. Cua, D. R. Littman.
2006
. The orphan nuclear receptor RORγt directs the differentiation program of proinflammatory IL-17+ T helper cells.
Cell
126
:
1121
-1133.
22
Takemoto, N., A. M. Intlekofer, J. T. Northrup, E. J. Wherry, S. L. Reiner.
2006
. Cutting edge: IL-12 inversely regulates T-bet and eomesodermin expression during pathogen-induced CD8+ T cell differentiation.
J. Immunol.
177
:
7515
-7519.
23
Hinrichs, C. S., A. Kaiser, C. M. Paulos, L. Cassard, L. Sanchez-Perez, B. Heemskerk, C. Wrzesinski, Z. A. Borman, P. Muranski, N. P. Restifo.
2009
. Type 17 CD8+ T cells display enhanced antitumor immunity.
Blood
114
:
596
-599.
24
Liu, S. J., J. P. Tsai, C. R. Shen, Y. P. Sher, C. L. Hsieh, Y. C. Yeh, A. H. Chou, S. R. Chang, K. N. Hsiao, F. W. Yu, H. W. Chen.
2007
. Induction of a distinct CD8 Tnc17 subset by transforming growth factor-β and interleukin-6.
J. Leukocyte Biol.
82
:
354
-360.
25
Chen, X. P., D. H. Falkner, P. A. Morel.
2005
. Impaired IL-4 production by CD8+ T cells in NOD mice is related to a defect of c-Maf binding to the IL-4 promoter.
Eur. J. Immunol.
35
:
1408
-1417.
26
Huber, M., S. Heink, H. Grothe, A. Guralnik, K. Reinhard, K. Elflein, T. Hunig, H. W. Mittrucker, A. Brustle, T. Kamradt, M. Lohoff.
2009
. A Th17-like developmental process leads to CD8+ Tc17 cells with reduced cytotoxic activity.
Eur. J. Immunol.
39
:
1716
-1725.
27
He, D., L. Wu, H. K. Kim, H. Li, C. A. Elmets, H. Xu.
2006
. CD8+ IL-17-producing T cells are important in effector functions for the elicitation of contact hypersensitivity responses.
J. Immunol.
177
:
6852
-6858.
28
Ortega, C., A. S. Fernandez, J. M. Carrillo, P. Romero, I. J. Molina, J. C. Moreno, M. Santamaria.
2009
. IL-17-producing CD8+ T lymphocytes from psoriasis skin plaques are cytotoxic effector cells that secrete Th17-related cytokines.
J. Leukocyte Biol.
86
:
435
-443.
29
Kondo, T., H. Takata, F. Matsuki, M. Takiguchi.
2009
. Cutting edge: phenotypic characterization and differentiation of human CD8+ T cells producing IL-17.
J. Immunol.
182
:
1794
-1798.
30
Gattinoni, L., C. A. Klebanoff, D. C. Palmer, C. Wrzesinski, K. Kerstann, Z. Yu, S. E. Finkelstein, M. R. Theoret, S. A. Rosenberg, N. P. Restifo.
2005
. Acquisition of full effector function in vitro paradoxically impairs the in vivo antitumor efficacy of adoptively transferred CD8+ T cells.
J. Clin. Invest.
115
:
1616
-1626.
31
Bending, D., H. De La Pena, M. Veldhoen, J. M. Phillips, C. Uyttenhove, B. Stockinger, and A. Cooke. 2009. Highly purified Th17 cells from BDC2.5NOD mice convert into Th1-like cells in NOD/SCID recipient mice. J. Clin. Invest. [Epub ahead of print].
32
Lee, Y. K., H. Turner, C. L. Maynard, J. R. Oliver, D. Chen, C. O. Elson, C. T. Weaver.
2009
. Late developmental plasticity in the T helper 17 lineage.
Immunity
30
:
92
-107.
33
Curtis, M. M., S. S. Way, C. B. Wilson.
2009
. IL-23 promotes the production of IL-17 by antigen-specific CD8 T cells in the absence of IL-12 and type-I interferons.
J. Immunol.
183
:
381
-387.
34
Wilson, C. B., E. Rowell, M. Sekimata.
2009
. Epigenetic control of T-helper-cell differentiation.
Nat. Rev. Immunol.
9
:
91
-105.
35
Muranski, P., A. Boni, P. A. Antony, L. Cassard, K. R. Irvine, A. Kaiser, C. M. Paulos, D. C. Palmer, C. E. Touloukian, K. Ptak, et al
2008
. Tumor-specific Th17-polarized cells eradicate large established melanoma.
Blood
112
:
362
-373.