Transient TCR stimulation induces multiple rounds of CD8 T cell division without further requirement for Ag. The mechanism driving Ag-independent proliferation, however, remains unclear. In this study, we show that the initial duration of TCR stimulation positively correlates with the number of divisions that CD8 T cells subsequently undergo. We find that increased periods of Ag stimulation result in enhanced CD25 up-regulation and greater IL-2 production by CD8 T cells. Depletion of IL-2 from T cell cultures with specific Abs dramatically impairs programmed proliferation. Consistent with this result, IL-2-deficient T cells undergo markedly attenuated Ag-independent proliferation in vitro. Although IL-2 production by stimulated CD8 T cells appears to be essential for in vitro proliferation, upon transfer into recipient mice, IL-2-deficient CD8 T cells undergo extensive proliferation in vivo after transient stimulation. Furthermore, the extent of in vivo proliferation correlates with the duration of in vitro Ag stimulation. These results indicate that the requirements for autocrine IL-2 production by CD8 T cells differs between in vitro and in vivo conditions and suggests that factors in addition to IL-2 can support Ag-independent CD8 T cell proliferation.
The CD8 CTL is critical in defense against a broad range of viral and intracellular bacterial infections (1). Upon TCR recognition of foreign pathogen-derived peptides bound to class I MHC molecules on the surface of APC and with appropriate costimulation, CTLs become activated, divide, and acquire effector functions required to eliminate infected cells. A number of groups have characterized the CD8 T cell response to invasive bacterial and viral pathogens. Infection with these pathogens results in massive, coordinate expansion of Ag-specific CTLs over the course of a week followed by a contraction phase involving substantial apoptosis of activated T cells (2, 3, 4, 5). Recent evidence indicates that the expansion and contraction of CD8 T cells is programmed during initial TCR stimulation, and occurs thereafter independently of Ag (6, 7, 8, 9, 10). For example, in one study, early administration of antibiotics rapidly terminated murine infection with the Gram-positive intracellular bacterium, Listeria monocytogenes, yet Ag-specific CD8 T cells continued to expand in a manner similar to that observed during a normal, unabated infection (6). Moreover, it was demonstrated that brief exposure to Ag was sufficient to induce proliferation and differentiation of CD8 T cells into effector and memory cells (7, 8, 9). The implication of these data is that short contact with Ag programs CD8 T cells to divide for several generations in the absence of further antigenic stimulation.
What factors drive the autonomous expansion of CD8 T cells following Ag encounter? IL-2, produced by activated T cells, is a prime candidate because numerous in vitro studies have established its role as a potent T cell growth factor (11). Recent evidence argues that autocrine IL-2 signals are involved in Ag-independent proliferation (7, 12), but the role of IL-2 in CD8 T cell expansion in vivo is complicated by a recent study of antiviral CD8 T cell responses that suggests that IL-2 production by CD8 T cells is not needed for their expansion in secondary lymphoid organs but may instead be more critical for proliferation of cells within nonlymphoid tissues (13). Other cytokines may be able to induce CD8 T cell proliferation as well. Type I IFNs elaborated during microbial infections stimulate the production of IL-15, which can drive bystander proliferation of activated/memory CD8 T cells (14, 15). It was recently shown that IL-15Rα signaling through CD8 T cells themselves is not required for bystander proliferation (16), suggesting that factors other than IL-15 can stimulate CD8 T cell proliferation in this process. IL-7 is involved in homeostatic proliferation of CD8 and CD4 T cells (17, 18), but its role in CD8 T cell expansion during Ag-induced proliferation appears to be minor (17).
In this report, we have investigated the role of autocrine IL-2 production by CD8 T cells in Ag-independent T cell expansion. We find that the proliferative program acquired by CD8 T cells is dramatically influenced by the duration of the initial antigenic stimulus. With increased periods of TCR stimulation, CD8 T cells increasingly up-regulate the high affinity IL-2 receptor and produce greater amounts of endogenous IL-2, both of which contribute to in vitro expansion of T cells. Autocrine IL-2 production, however, appears to be unnecessary for Ag-independent CD8 T cell expansion in vivo, suggesting that other IL-2-independent mechanisms also drive programmed T cell proliferation.
Materials and Methods
Mice and bacteria
Wild-type and IL-2−/− BALB/cJ mice were obtained from The Jackson Laboratory (Bar Harbor, ME). BALB/c Thy1.1-congenic mice were provided by C. Surh (The Scripps Research Foundation, La Jolla, CA). Recombination-activating gene (RAG)3-deficient (RAG°) WP11.12 and L9.6 transgenic mice expressing the L. monocytogenes p60449–457/Kd-specific TCR and p60217–225/Kd-specific TCR, respectively, were generated as previously described (6, 19, 20). All cell cultures were grown at 37°C, 5% CO2, in RPMI 1640 medium (Life Technologies, Grand Island, NY) supplemented with l-glutamine, HEPES, 2-ME, antibiotics (penicillin/streptomycin), and 10% FCS. L. monocytogenes strain 10403S was provided by D. Portnoy (University of California, Berkeley, CA) and grown in brain-heart infusion broth (Becton Dickinson, Sparks, MD). Sublethal infections of mice were conducted by injecting 2000 bacteria per animal in the tail vein. In some experiments, ampicillin (Sigma-Aldrich, St. Louis, MO) was given to mice at 2 mg/ml in their drinking water 24 h postinoculation of bacteria to terminate the infection.
Cell sorting and CFSE labeling
CD8 T cells were enriched to ∼90% purity from the spleen by magnetically activated cell sorting using anti-CD8a-conjugated microbeads (Miltenyi-Biotec, Auburn, CA). Cells were washed with PBS and resuspended at 5 × 107/ml in PBS containing 2 μM (for in vitro stimulation experiments) or 10 μM (for adoptive transfer experiments) CFSE (Molecular Probes, Eugene, OR). After incubating at 37°C for 10 min in the dark, cells were immediately washed with cold RPMI 1640/10% FCS before plating. For transfer experiments, cells were resuspended in PBS for injecting i.v. into mice.
Abs and flow cytometry
The following mAbs were purchased from BD PharMingen (San Diego, CA): anti-CD8a-PerCP (53-6.7), anti-CD25-PE (PC61), anti-Thy1.2-PE (53-2.1), anti-Thy1.1-PerCP, anti-IL-2 PE (JES6-5H4), anti-CD3ε (145-2C11), and anti-CD28 (37.51). Tetrameric H-2Kd/peptide complexes were generated as described (2). For 4-color analyses, ∼5 × 105–1 × 107 cells per sample were incubated on ice for 60 min with saturating concentrations of mAbs and tetramers in FACS staining buffer (PBS, 1% FCS, 0.05% sodium azide). Labeled cells were washed with FACS buffer and fixed in PBS containing 2% paraformaldehyde before analysis. Analyses were conducted on a BD-LSR flow cytometer (BD Biosciences, Mountain View, CA) using CellQuest and FlowJo (Tree Star, San Carlos, CA) software. Flow sorting was conducted on a MoFlo cytometer (DAKOCytomation, Fort Collins, CO).
In vitro T cell proliferation and IL-2 production
In Ab stimulation experiments, purified, CFSE-labeled CD8 T cells were plated at 1–5 × 105/well in 96-well plates coated at 30 μg/ml with anti-CD3 and anti-CD28 mAbs. In some experiments, recombinant murine IL-2 (R&D Systems, Minneapolis, MN) was added at 20 ng/ml or neutralizing anti-IL-2 Ab (R&D Systems) was added at 10 μg/ml. Ag-independent expansion following TCR stimulation was then assessed by transferring T cells into new wells that were not coated with Abs. T cell proliferation was analyzed by flow cytometric analysis of CFSE dilution after 3–6 days of culture. IL-2 production was measured by harvesting culture supernatants from stimulation assays and directly adding the supernatant to cultures of the IL-2-dependent CTLL line. Proliferation of CTLL cells in response to IL-2 in culture was measured 24 h later by pulsing cells with 1 μCi [3H]thymidine/well for 20 h before harvesting cells onto glass-fiber filters and determining incorporated radioactivity using a TopCount liquid scintillation counter (Packard Instrument, Meriden, CT). IL-2 amounts were determined from cpm values by establishing a standard curve using known amounts of IL-2. For intracellular IL-2 analysis, purified CD8 T cells were stimulated by 30 μg/ml plate-bound anti-CD3/CD28 for 0–19 h and transferred to noncoated wells. At 4 days after the initiation of cultures, the cells were restimulated with PMA (50 ng/ml) and ionomycin (500 ng/ml) in the presence of brefeldin A for 4–5 h. Cells were then washed, incubated with Abs against cell surface markers for 30 min, fixed, permeabilized, and incubated with anti-IL-2 Ab for 30 min before washing and analysis on the flow cytometer.
Purified, CFSE-labeled WP11.12 CD8 T cells stimulated 0–43 h in vitro with plate-bound anti-CD3/CD28 Abs were washed extensively and injected i.v. into naive Thy1.1+ BALB/c mice at ∼6 × 105 cells per recipient. Analysis of donor cell proliferation in vivo was assessed 6 days after the initiation of T cell stimulation by harvesting recipient spleens and analyzing CFSE dilution among CD8+Thy1.2+ gated donor cells by flow cytometry. In analyses of Ag-independent expansion of wild-type vs IL-2-deficient WP11.12 CD8 T cells, 8 × 105 purified, CFSE-labeled IL-2+/+ or IL-2−/− T cells were injected i.v. into naive Thy1.1 BALB/c mice following in vitro stimulation for 0 or 24 h with plate-immobilized anti-CD3/CD28 Abs. Donor cell proliferation was analyzed 4 days posttransfer. In other experiments, 1 × 106 CFSE-labeled splenocytes from IL-2+/+ or IL-2−/− WP11.12 mice were injected i.v. into naive Thy1.1 BALB/c mice, which were infected 24 h later with 2000 L. monocytogenes. A subset of recipients was given drinking water containing 2 mg/ml ampicillin 24 h later to terminate the infection. Analysis of donor cell proliferation was conducted 4 days after infection.
Quantitative real-time PCR
Total RNA was extracted from purified CD8 T cells using the RNeasy kit (Qiagen, Valencia, CA) according to the manufacturer’s protocol, and reverse transcribed with oligo(dT) primers using the SuperScript First Strand Synthesis System (Invitrogen, Carlsbad, CA). Real-time PCR was then performed with the cDNA using the following primer pairs: IL-2 (forward), 5′-GCAGGATGGAGAATTACAGG-3′, (reverse), 5′-CAGAAAGTCCACCACAGTTG-3′; HPRT (forward), 5′-CTGGTGAAAAGGACCTCTCG-3′, (reverse), 5′-TGAAGTACTCATTATAGTCAAGGGCA-3′; GAPDH (forward), 5′-ACCACAGTCCATGCCATCAC-3′, (reverse), 5′-TCCACCACCCTGTTGCTGTA-3′. Serial dilutions of known amounts of cDNA from anti-CD3-stimulated BALB/c CD8 T cells were used to generate a standard curve to determine the relative quantity of IL-2 mRNA in each sample at a given fluorescence emitted during product amplification. Values for IL-2 mRNA were normalized to the amount of HPRT or GAPDH message, which yielded similar results. Amplification and analyses were conducted with the LightCycler kit and LightCycler Instrument (Roche, Indianapolis, IN).
Duration of initial TCR stimulation determines extent of programmed proliferation of CD8 T cells
We and others previously showed that a brief period of TCR stimulation can induce either polyclonal or TCR transgenic CD8 T cells to undergo extensive proliferation over the course of several days in the absence of Ag. To determine the role of cytokines in programmed T cell proliferation using a uniform population of cells of defined specificity, we labeled purified L. monocytogenes-specific WP11.12 TCR transgenic CD8 T cells with CFSE and stimulated them with plate-bound anti-CD3 and anti-CD28 Abs in the presence or absence of exogenous IL-2 for various lengths of time before transferring cells to noncoated wells, in which their proliferation was monitored for the next 4 days. T cells stimulated in vitro for as little as 1 h divided extensively (>7 rounds) over the course of 4 days when cultured with IL-2 in the absence of Ag (Fig. 1). An increased duration of TCR stimulation led to recruitment of a greater number of T cells into the response, as reflected by the progressive decrease in the undivided, CFSEhigh population. Cells recruited into the response underwent similar degrees of proliferation regardless of the duration of stimulation, which is consistent with the finding that escalating doses of Ag induce greater numbers of T cells to proliferate, rather than driving similar numbers of T cells to undergo greater proliferation (7). In the absence of exogenous IL-2, however, cell recoveries were generally decreased and we observed very little proliferation unless CD8 T cells were stimulated for at least 6 h. Interestingly, in the absence of exogenous IL-2, the extent of CD8 T cell division correlated with the duration of TCR stimulation (Fig. 1).
Ag-independent CD8 T cell proliferation in vitro requires CD8 T cell production of IL-2
Because Ag-independent CD8 T cell proliferation was enhanced with prolonged TCR stimulation when IL-2 was not provided, we hypothesized that extended periods of TCR stimulation induced CD8 T cells to produce endogenous IL-2, which would facilitate their continued division. To investigate this further, we stimulated CFSE-labeled CD8 T cells with plate-immobilized anti-CD3/CD28 for various lengths of time and immediately measured the amount of IL-2 secreted into the culture supernatants. We found that the amount of IL-2 produced was proportional to the duration of TCR stimulation (Fig. 2,A). A closer examination revealed that a longer duration of TCR stimulation yielded a greater frequency of IL-2-producing CD8 T cells as assessed by intracellular cytokine staining (Fig. 2,B). In addition, when we analyzed expression of CD25, the α-chain of the high affinity IL-2 receptor, we found that prolonged TCR stimulation enhanced CD8 T cell expression of this receptor (Fig. 3 B). Thus, the duration of TCR stimulation significantly influenced IL-2 expression by CD8 T cells as well as their ability to respond to IL-2.
To determine whether the IL-2 produced as a result of extended TCR stimulation of CD8 T cells is critical for supporting Ag-independent proliferation, CFSE-labeled TCR transgenic CD8 T cells were stimulated between 0 and 19 h with plate-bound anti-CD3/CD28 Abs in the presence or absence of neutralizing anti-IL-2 Ab before removal from Ag. When cells were analyzed 4 days later, we found that the addition of anti-IL-2 to the cultures significantly inhibited Ag-independent division of as well as CD25 up-regulation on CD8 T cells (Fig. 3). A small amount of proliferation could be observed when cells were stimulated for >14 h before removal from Ag, which may be the result of incomplete IL-2 neutralization or IL-2-independent proliferation (21).
Ag-independent CD8 T cell proliferation in vivo occurs in the absence of IL-2
We next determined whether IL-2 was required for Ag-independent CD8 T cell proliferation to occur in vivo. Purified, CFSE-labeled WP11.12 CD8 T cells were stimulated for various lengths of time in vitro with plate-bound anti-CD3/CD28, washed free of any exogenous IL-2, and then injected into naive mice. Surprisingly, when we analyzed division of transferred cells 6 days later, we found that, similar to our in vitro observations, CD8 T cells had divided in proportion to the duration of initial TCR stimulation before adoptive transfer (Fig. 4). Furthermore, the daily injection of 100 μg neutralizing anti-IL-2 Ab into recipient mice did not block the autonomous in vivo expansion of transferred CD8 T cells following in vitro activation (data not shown). This was not a phenomenon exclusive to our TCR transgenic T cells because we could also observe in vivo proliferation following brief in vitro activation of polyclonal BALB/c CD8 T cells (data not shown).
It is possible that, after transient TCR activation, CD8 T cells continue to synthesize endogenous IL-2 that drives their in vivo proliferation and which might be incompletely neutralized by anti-IL-2 Abs. To address this, we analyzed real-time IL-2 mRNA expression in purified, in vitro activated, polyclonal CD8 T cells either immediately after removal from Ag or after a 24–48 h culture in the absence of Ag. As predicted, cells stimulated for 2, 6, or 18 h expressed readily detectable levels of IL-2 mRNA; however, when cells were assessed for IL-2 message levels 24–48 h after removal from TCR stimulation, no significant IL-2 expression was found (Fig. 5,A). To show that this was also true in vivo, we transferred WP11.12 TCR transgenic CD8 T cells into naive Thy1.1 congenic BALB/c mice and infected recipients with a sublethal dose of L. monocytogenes. We terminated the infection by administering antibiotics to recipients 24 h postinoculation and then assessed the expression of IL-2 within donor CD8 T cells at various points postinfection to determine whether IL-2 continued to be synthesized as these cells underwent Ag-independent expansion in vivo. Although a small amount of IL-2 mRNA could be detected in donor CD8 T cells 24 h after antibiotic administration, no IL-2 could be detected on subsequent days (Fig. 5,B), despite their continued expansion (Fig. 5 C).
Endogenous IL-2 production by CD8 T cells is not required for Ag-independent expansion in vivo
To assess more rigorously whether IL-2 production by CD8 T cells is required to drive their Ag-independent expansion, we crossed WP11.12 mice to the IL-2−/− background and assessed the ability of IL-2-deficient T cells to proliferate after transient antigenic stimulation. After activating CD8 T cells for 8 h in vitro on wells coated with anti-CD3/CD28 and removing cells from stimulus, we found that the recovery of IL-2-deficient transgenic CD8 T cells was relatively low and these cells exhibited markedly reduced proliferative capacity as compared with wild-type CD8 T cells when CFSE dilution was assessed 4 days later (Fig. 6,A). Consistent with results obtained using anti-IL-2 Ab (Fig. 3 A), a small degree of IL-2-independent proliferation could be observed if cells were stimulated for a longer period of time, but in general the expansion of CD8 T cells in vitro following brief Ag stimulation is impaired in the absence of endogenous CD8 T cell production of IL-2.
To examine whether IL-2−/− CD8 T cells could undergo Ag-independent proliferation in vivo, we stimulated CFSE-labeled IL-2+/+ and IL-2−/− WP11.12 CD8 T cells 24 h with plate-bound anti-CD3/CD28 in vitro, washed, and transferred cells into naive Thy1.1 congenic BALB/c recipients and assessed their proliferation 4 days later. Stimulated T cells underwent Ag-independent expansion in vivo regardless of whether they expressed endogenous IL-2, but the recovery of IL-2−/− cells was ∼25% of the IL-2+ cells, suggesting that endogenous IL-2 production may contribute to cell survival rather than proliferation (Fig. 6,B). Finally, we transferred IL-2+/+ RAG−/− WP11.12 cells or IL-2−/− RAG−/− WP11.12 cells into naive Thy1.1 congenic BALB/c mice and infected recipients 24 h later with L. monocytogenes. Half of the recipients were given ampicillin in their drinking water 24 h later to terminate the infection. When donor CD8 T cell expansion was assessed 4 days postinfection, we found that both IL-2+/+ and IL-2−/− WP11.12 CD8 T cells were recovered in similar numbers and had divided similarly whether or not mice had been treated with antibiotics (Fig. 6 C). Moreover, both IL-2+/+ and IL-2−/− WP11.12 CD8 T cells had acquired the ability to produce IFN-γ in response to cognate peptide stimulation (data not shown). Thus, IL-2 production by CD8 T cells is not required for in vivo Ag-independent division and differentiation.
The expansion of CD8 T cells following Ag encounter is programmed during the period of initial TCR stimulation and occurs without any further requirement for Ag. It has been unclear, however, what factors drive this autonomous expansion in vivo. The role of IL-2 in Ag-independent T cell expansion has been investigated in recent studies but the exact requirement for IL-2 in vivo is controversial. In this study, we examined the role of autocrine IL-2 in supporting Ag-independent CD8 T cell proliferation in vitro and in vivo. We showed that, in the absence of exogenous IL-2, the duration of initial TCR signaling instructs CD8 T cells as to the extent of division they must subsequently undergo. This result is consistent with recent studies showing that the length of Ag stimulation determines the magnitude of T cell expansion (12, 22, 23, 24). We also demonstrated that longer periods of T cell activation induced greater production of endogenous IL-2 by CD8 T cells, which was necessary for their continued expansion in vitro, but not required for their division in vivo.
The importance of endogenous CD8 T cell production of IL-2 in programmed expansion was implicated in two recent studies, but these experiments primarily examined the importance of IL-2 in vitro (7, 12). In contrast, another report suggested that autocrine IL-2 production by CD8 T cells did not play a major role in expansion within secondary lymphoid organs following viral infection (13). Our results are consistent with both of these findings and indicate that other factors must exist in vivo that can support the continued expansion of CD8 T cells. It is possible that CD4 T cell production of IL-2 or IL-4 (25) may promote CD8 T cell expansion during the immune response; however, we have observed that the CD8 T cell response to L. monocytogenes infection is intact in mice lacking CD4 T cells (19), suggesting that Th cell provision of these cytokines is not necessary. It has also been suggested that activated dendritic cells can transiently produce IL-2 (26), which may act on activated CD8 T cells. This may occur during the early stages of the immune response but does not likely persist, because dendritic cell production of IL-2 depends on continued microbial stimuli that wane as organisms are cleared (27). Moreover, we have previously shown that the swift acquisition of effector functions by CD8 T cells after priming results in rapid elimination of professional APC within 3–4 days of infection (28), whereas CD8 T cell expansion continues until 7–9 days of infection.
It is likely that factors other than IL-2 must provide support for Ag-independent CD8 T cell expansion in vivo. We speculate that such factors may exist in a tonic state in naive animals, because we were able to transfer preactivated CD8 T cells into noninfected mice and observe their continued expansion over several days (Fig. 4). Interestingly, it was recently shown that CD4+CD25+ regulatory T cells lack the capacity to produce IL-2 and expand in vitro, yet could be induced to expand in vivo (29). In another study, IL-2-deficient TCR transgenic CD4 T cells, which proliferated poorly in vitro, were found to proliferate in vivo upon immunization with specific Ag in the presence of LPS as an adjuvant (30). These data support the notion that additional T cell growth factors may be extant in an intact animal that may not be available in traditional cell cultures. In preliminary experiments, we tested the role of IL-15 using a neutralizing reagent in vivo, as well as examined the role of CD40-CD40 ligand interactions using CD40 ligand−/− mice, and did not find any impact of their absence on Ag-independent CD8 T cell expansion. Based on recent reports, both IL-15 and IL-7 may be unnecessary for primary Ag-driven CD8 T cell expansion (16, 17). Indeed, Ag-driven proliferation of CD4 T cells has been shown to proceed independently of the common cytokine receptor γ chain used by IL-2, IL-4, IL-7, IL-9, and IL-15 (31, 32). Although it is not known whether CD8 T cells follow similar rules, it appears that T cells can use other soluble or membrane-bound interactions during their proliferative response in vivo. Additional experiments are underway to identify the factors that are involved in this process.
We thank Ingrid Leiner, Rielle Giannino, An Tran, Jessica Vega, and Ewa Menet for excellent technical support.
This work was supported by National Institutes of Health Grants AI-39031 and AI-42135. P.W. is supported by an Arthritis Foundation postdoctoral fellowship.
Abbreviation used in this paper: RAG, recombination-activating gene.