CD4 T Cell-Mediated Protection from Lethal Influenza: Perforin and Antibody-Mediated Mechanisms Give a One-Two Punch

Mice

TCR Tg mice in which CD4 cells recognize a peptide 126-138 of HA protein from PR8 influenza were a gift from Dr. D. Lo (The Scripps Research Institute, La Jolla, CA) and have been described in detail (18). IFN-γ^−/− mice on the BALB/c background have been described previously (19) and were bred to TCR Tg mice at the Trudeau Institute Animal Breeding Facility. Pfn^−/− (Pfn-deficient) mice on the BALB/c background were a gift from Dr. J. Harty (University of Iowa, Iowa City, IA) and were bred to HA-specific CD4 TCR Tg mice at the Trudeau Institute Animal Breeding Facility. In some experiments, CD8 TCR Tg mice (a gift from Dr. L. Sherman, The Scripps Research Institute) specific for peptide 518–526 of PR8 HA and OVA 323–339-specific (DO11.10) TCR Tg mice were used as controls. WT TCR Tg, IFN-γ^−/− TCR Tg, BALB/c/Thy1.1, and B cell-deficient mice homozygous for a disruption at the J_H locus (J_HD) (20), backcrossed to BALB/c mice were all maintained at Trudeau Institute. BALB/c By and BALB/c nu/nu mice were purchased from The Jackson Laboratory. All experimental animal procedures were conducted in accordance with the Trudeau Institute Animal Care and Use Committee guidelines.

Medium and peptides

All cells were grown in RPMI 1640 (Invitrogen Life Technologies) containing 2 mM l-glutamine, 100 IU penicillin, 100 μg/ml streptomycin (all obtained from Invitrogen Life Technologies) 10 mM HEPES (Research Organics), 50 μM 2-ME (Sigma-Aldrich), and 8% FBS (HyClone). HA peptide 126–138 (HNTNGVTAACSHE), HA peptide 518–526 (IYSTVAASL), and OVA peptide 323–339 (ISQAVHAAHAEINEAGR) were synthesized by New England Peptide.

Isolation of CD4 T cells and in vitro activation

Naive CD4 cells were isolated from TCR Tg mice using a positive selection protocol with anti-CD4 beads (clone GK1.5) according to the manufacturer’s protocol (Miltenyi Biotec). B cell blasts were generated as previously described (21), pulsed with 5 μM HA 126–138 peptide and combined with naive CD4 cells at a 1:1 ratio to give a final concentration of 3 × 10⁵ cells/ml. IL-2 at 11 ng/ml, IL-12 at 2 ng/ml, and anti-IL-4 (clone 11B11) at 10 μg/ml were added to polarize CD4 cells to the Th1 phenotype characteristic of high IFN-γ secretion (17). Additional IFN-γ at 100 ng/ml was added to IFN-γ^−/− cultures to inhibit IL-4 and IL-5 secretion (Ref. 22 and see Fig. 3⇓A). After 2 days, an equal volume of complete medium containing IL-2 at 11 ng/ml was added, and cultures were incubated for an additional 2 days. To assay for activation and polarization, WT, IFN-γ^−/− or Pfn^−/− CD4 effectors at 1 × 10⁶ cells/ml were restimulated with immobilized anti-CD3 for 48 h, and supernatants were analyzed for cytokine secretion using Beadlyte cytokine beadmates (Upstate Signaling) according to the manufacturer’s instructions. Cytokines were detected using a Luminex 100 (Luminex). Standard curves were analyzed and unknowns calculated using Prism software (GraphPad).

CD4 effector transfer and infection of mice

A total of 5 × 10⁶ CD4 effectors in PBS was injected via tail vein into BALB/c mice, unless otherwise noted. Eighteen to 24 h after CD4 effector injection, mice were infected intranasally with 5000 egg infectious units (EIU) PR8 virus in 50 μl of PBS, unless otherwise indicated. Mice were monitored for weight loss and survival over the course of the experiment. Mice were euthanized when they became moribund according to the Trudeau Institute Animal Care and Use Committee guidelines.

FACS analysis of donor and host populations

Lethally infected BALB/c. Thy1.1 mice were sacrificed at various times postinfection. Lungs were perfused with PBS, dissociated, and single-cell suspensions were incubated with fluorescently labeled anti-CD4 (clone RMA-4.4) and anti-Thy1.2 for 30 min in FACS buffer (PBS with 1% BSA, 0.1% Na azide) at 4°C to mark donor cells. Lung cells were also incubated with FITC-labeled anti-CD8α (clone 53-6.7) to identify host CD8 cell populations. Cells were fixed in 1% paraformaldehyde and analyzed using a BD Biosciences FACSCalibur, and data were processed using FlowJo software (Tree Star).

Quantitation of viral RNA by TaqMan PCR

Viral RNA was detected in a manner similar to previously published protocols (23, 24). RNA was isolated from lung homogenates using TRIzol (Sigma-Aldrich) and 2.5 μg of RNA was reverse-transcribed into cDNA using random hexamers and Superscript II reverse transcriptase (Invitrogen Life Technologies). Fifty nanograms of cDNA was then used for amplification by quantitative real-time PCR (ABI Prism 7700; Applied Biosystems). The following primers and probe were used: 5′-CGGTCCAAATTCCTGCTGAT-3′ (forward), 5′-CATTGGGTTCCTTCCATCCA-3′ (reverse), and 5′- 6-FAM-CCAAGTCATGAAGGAGAGGGAATACCGCT-3′ (probe). The acid polymerase (PA)-gene copy number was calculated from a known concentration of PA-containing plasmid (a gift from Dr. R Webster, St. Jude Children’s Research Hospital, Memphis, TN) used as a standard curve in all reactions. PA copies per lung were then calculated based on the total amount of RNA in each lung sample.

ELISA for detection of anti-influenza Abs

ELISA plates were coated overnight at 4°C with live influenza PR8 virus diluted 1/200 (5 × 10⁶ EIU/ml) in PBS. Plates were washed four times with PBS and 200 μl/well of PBS containing 2% FBS, and 10 mM HEPES was added to block nonspecific binding sites. Serum was then added in blocking buffer starting at a 1/20 dilution and serially diluted 2-fold. After 2–3 h at 20°C, alkaline phosphatase-conjugated goat anti-mouse IgG (Southern Biotechnology Associates) was added at a 1/1000 dilution. For isotype-specific ELISAs, alkaline phosphatase-conjugated goat anti-mouse IgG1, IgG2a, and IgG2b (Southern Biotechnology Associates) were used at 1/1000 dilution. Plates were developed using p-nitro phenyl phosphate (Sigma-Aldrich) tablets dissolved in DEA buffer (10% diethanolamine). Absorbance was read at 405 nm, and end point titers were determined by the dilution that gave two times the background OD using normal mouse serum (NMS) as a negative control.

Treatment of mice with immune sera

NMS was collected from the blood of uninfected BALB/c mice and used as a negative control and as a carrier for preparations containing immune sera. Immune sera were collected from mice that received either WT or IFN-γ^−/− effectors and survived a lethal dose of influenza >30 days. Ten or 20 μl/mouse immune serum was injected i.p. into J_HD mice in a total volume of 100 μl of PBS containing 10% NMS 7 days after lethal infection. Anti-influenza IgG Ab titers in these sera ranged from 20,000 to 50,000.

In vitro cytotoxicity assay

CD4 effectors were washed and plated at decreasing concentrations in 96-well round-bottom plates. H-2^d-expressing B lymphoma cells (A20), labeled for 4 h with 4 μCi/ml [³H]thymidine (Amersham Biosciences), were added to effectors at a concentration of 1 × 10⁴ cells/well. HA_126–138 peptide, or irrelevant peptide as a control, was added at a final concentration of 2 μg/ml. After 4 h at 37°C, plates were harvested, and incorporated [³H]thymidine was measured using a beta plate scintillation counter (Wallac). In some experiments, concanamycin A (CMA) (Sigma-Aldrich) was added to inhibit Pfn exocytosis (25). In other experiments, anti-Fas ligand (FasL) at 10 μg/ml was added to inhibit Fas-mediated killing (26). Percentage of specific cytotoxicity was calculated as (spontaneous cpm − experimental cpm)/spontaneous cpm) × 100.

In vivo cytotoxicity

In vivo cytotoxicity assays were performed as described with minor modifications (27). Three days posteffector transfer, naive BALB/c splenocytes were pulsed with the class II-restricted HA_126–138 peptide for 1 h at 37°C and labeled with 5 μM CFSE (Molecular Probes). Naive BALB/c splenocytes were also pulsed with the class I-restricted HA_518–528 peptide for 1 h at 37°C and labeled with 0.5 μM CFSE. Targets were combined at a 1:1 ratio, and a total of 5 × 10⁶ cells/mouse were injected i.v. Eighteen hours after target injection, mice were sacrificed, spleens were removed, and red cells were lysed and resuspended in FACS buffer. Cells were analyzed with a BD Biosciences FACSCalibur, and data were processed using FlowJo software (Tree Star). Percentage of specific cytotoxicity was calculated as follows: 100 − (((percentage of peptide pulsed in transferred/percentage of unpulsed in transferred)/(percentage of peptide pulsed in naive/percentage of unpulsed in naive)) × 100).

Statistical analyses

Statistical analyses were performed by Prism 4.0 software (GraphPad) using Student’s t tests for parametric data. Survival curves were plotted and analyzed by Prism 4.0 software using the Kaplan-Meier method and log-rank test.

Primed CD4 effectors protect against lethal influenza infection in an Ag-specific manner

Previous studies have demonstrated that repeatedly stimulated CD4 T cell clones transferred to intact mice could protect against a lethal dose of influenza A/Japan/57 (14). We first investigated whether CD4 effectors primed in vitro could promote survival against a highly pathogenic influenza strain such as PR8. In these experiments, we use the term “protection” to signify survival for >20 days after infection with a dose of influenza that routinely kills 80% or more of normal, unprimed mice. Naive CD4 cells were isolated from TCR Tg mice, and effectors were generated by priming with influenza HA peptide-pulsed B cell blasts and Th1-polarizing cytokines for 4 days. CD4 effectors stimulated in this way show decreased levels of CD62L and increased expression of CD43, CD11a, and CD49d compared with naive cells (data not shown). The cell surface profile of in vitro-primed CD4 effectors was similar to that seen when CD4 cells are primed in vivo during a sublethal dose of influenza (12). A total of 5 × 10⁶ in vitro-primed CD4 effectors was transferred into adoptive hosts that were subsequently infected with 5000 EIU (≈2 LD₅₀) PR8. Fig. 1⇓A demonstrates that Th1-polarized, HA peptide-specific CD4 cells protected BALB/c mice against an otherwise lethal dose of influenza PR8. In contrast, OVA-specific Th1 CD4 effectors generated from DO11.10 TCR Tg mice did not provide protection. Thus, CD4 T cell-mediated protection is Ag specific and is not due to nonspecific activation of the transferred cells at the site of infection. Equivalent numbers of naive HA-specific cells did not promote survival after a similar challenge with PR8 (Fig. 1⇓B), demonstrating that CD4 cells must first be primed to provide protection. Furthermore, the protective effect (as measured by the fraction of surviving mice) of primed CD4 cells was shown to be dependent on the number of cells transferred (Fig. 1⇓C). We found that transfer of 5 × 10⁶ cells provided optimal protection, whereas a 10-fold lower number of CD4 effectors was not effective. Therefore, primed but not naive, CD4 T cells effectively promote survival to lethal influenza infection in an Ag-specific manner.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Primed CD4 effectors protect against lethal influenza infection in an Ag-specific manner. A total of 5 × 10⁶ HA-specific (▪) or OVA-specific (▵) cells was transferred to normal BALB/c mice (n = 5/group) and 18 h later, infected with 5000 EIU (2 LD₅₀) PR8 virus (A). B, A total of 5 × 10⁶ naive HA-specific cells (▪) or no cells (▵) was injected into BALB/c mice (n = 5/group) followed by infection with 5000 EIU PR8 virus. C, A total of 5 × 10⁶ (▪), 1 × 10⁶ (□), or 5 × 10⁵ (♦) CD4 effector cells was injected into BALB/c mice (n = 5/group) that were then infected with 5000 EIU PR8. Control mice did not receive cells (▵). Mice were monitored for survival and were euthanized when they became moribund in accordance with Trudeau Institute’s Animal Care and Use Committee guidelines.

CD4 effectors localize to the lung, decrease viral titer, and provide protection independent of host T cells

Visualization of the anti-influenza CD4 response in situ suggests that naive cells are primed in the draining lymph nodes (LN), and then highly differentiated effectors migrate to the lung and release cytokines in response to antigenic challenge (12). Earlier studies in our laboratory suggest that in vitro-generated CD4 effectors transferred to unprimed mice enter both secondary lymphoid and nonlymphoid sites within 24 h, even in the absence of inflammatory stimuli (28). Because CD4-mediated protection occurred only when the cells were primed (Fig. 1⇑), it is likely that CD4 effectors migrate to the lung early in the response and act directly to promote recovery. To determine the localization and kinetics of CD4 effector migration to nonlymphoid sites, Thy1 disparate CD4 effectors were transferred into BALB/c mice that were then infected with PR8 virus. The absolute number of donor CD4 effectors in the lung was determined by FACS analysis at various times postinfection and is shown in Fig. 2⇓A. CD4 effectors were recovered from the lung as early as 1 day after infection, and the absolute numbers of CD4 effectors peaked at day 2 postinfection (Fig. 2⇓A). By day 4 postinfection, the number of CD4 effectors had decreased ∼4-fold from the day 2 peak, and few donor T cells were detectable 1 wk after infection. This pattern was also observed in the draining LN, but with ∼4-fold lower donor cell numbers than in the lung (data not shown). This result suggests that primed CD4 T cells exert their protective effects within the first 4 days after infection.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

CD4 effectors localize to the lung, decrease viral titer, and promote survival independent of host T cells. Mice were injected with 5 × 10⁶ WT effectors (▪) or PBS (▵) and subsequently infected with 5000 EIU (2 LD₅₀) PR8. At 1, 2, 4, and 7 day postinfection, mice were sacrificed, lungs were removed, and lung cells were incubated with Abs to Thy1.2 and CD4 to track donor cells (A) and host CD4 cells (D) or with Ab to CD8 to enumerate the host cell population (C). Shown is the average of the absolute number of donor cells and host CD4 or CD8 cells ±SD from three mice per group. B, At day 4 postinfection, mice were sacrificed, lungs were removed, RNA was isolated, and quantitative PCR was performed as described in Materials and Methods. PA copy number per lung is shown for six mice per group. ∗, Denotes statistical significance; p < 0.05. E, A total of 5 × 10⁶ WT (▪) effectors was injected into athymic nu/nu mice (n = 10 mice/group) and infected 18 h later with 5000 EIU (2 LD₅₀) PR8. Control mice did not receive effectors (▵). Survival was monitored, and mice were sacrificed when they became moribund. These experiments were repeated with similar results.

The rapid influx of CD4 effectors to the site of infection suggested that these cells might function early to reduce viral titers. It has been demonstrated that cytolytic CD8 T cells attack virally infected epithelium and are primarily responsible for decreasing viral titers (29). However, it has also been shown that CD4 cells can promote viral clearance in the absence of CD8 cells (30) and that memory CD4 cells residing in the respiratory tract also act to decrease viral titer (31). To determine whether transfer of CD4 effectors had an impact on viral replication in the lung, a quantitative PCR-based assay that detects the influenza A PA gene copy number was used (23). Fig. 2⇑B shows PA gene copy number per lung at day 4 postinfection and demonstrates that mice given CD4 effectors had significantly (p < 0.05) less viral RNA than PBS-treated controls, suggesting that CD4 effectors did act in some way to reduce viral titers.

Results shown thus far indicated that CD4 effectors localized to the lung and decreased viral titers; however, it was not clear whether the transferred cells functioned directly on virally infected cells, or indirectly via actions on host T cells. To determine whether CD4 effectors accelerated the migration of host T cells into the lung, or increased the magnitude of the host response after lethal infection, host CD8 and CD4 cell populations were enumerated using FACS analysis. Fig. 2⇑, C and D, show that the number of host CD8 or CD4 cells do not increase substantially until 7 days postinfection, regardless of the presence of CD4 effectors. Although addition of CD4 effectors had no impact on the number of host T cells in the lung, it was still unclear whether the host T cell response was contributing to the decrease in viral burden. To determine whether host T cells were required for CD4-mediated protection, CD4 effectors were transferred to BALB/c athymic nu/nu recipients, lacking both CD4 and CD8 cells, followed by lethal PR8 infection. Fig. 2⇑E demonstrates that T cell-deficient mice given CD4 effectors survived a 5000 EIU dose of PR8, whereas T cell-deficient mice given PBS succumbed to lethal influenza infection. Taken together, these data demonstrate that protection mediated by CD4 effectors did not require host T cells, suggesting that CD4 effectors may be acting in a direct mechanism to decrease viral titers at the site of infection.

Protection mediated by CD4 effectors does not require IFN-γ in the priming or effector phase

Using an adoptive transfer model, our laboratory has previously demonstrated that naive TCR Tg CD4 T cells specific for HA of PR8 influenza respond vigorously to sublethal (500 EIU) infection (12). These studies demonstrated that naive CD4 cells proliferate and differentiate in the draining LN and spleen and acquire a Th1 effector profile. The most differentiated cohort of effectors migrate to the lung and produce IFN-γ, but little to no IL-2 at the peak of the response (12). Based on these data, we hypothesized that CD4-derived IFN-γ might be important for CD4-mediated protection to lethal influenza infection. To test this theory, we generated effectors from TCR Tg mice that were deficient in IFN-γ. Because WT CD4 effectors were normally exposed to IFN-γ during effector generation in vitro, we generated IFN-γ^−/− CD4 effectors that were also primed in the presence of exogenous IFN-γ during the 4-day culture period and compared these cells to IFN-γ^−/− effectors primed in the absence of exogenous IFN-γ. FACS analysis of the cell surface markers CD62L, CD25, CD43, and CD27 did not reveal differences in the activation state between WT and IFN-γ^−/− effectors generated with or without IFN-γ (data not shown). As shown in Fig. 3⇓A, both WT and IFN-γ^−/− CD4 effectors produced similar levels of TNF-α, RANTES, and IL-10. However, IFN-γ^−/− effectors generated under Th1-polarizing conditions, but without exogenous IFN-γ during the priming phase, produced some IL-4, IL-5, and IL-6 (≈1 ng/ml) as has been described previously (22). IFN-γ^−/− cells exposed to IFN-γ during the culture period lost the ability to produce IL-4 but still secreted IL-5, albeit at lower levels. Interestingly, IFN-γ^−/− effectors primed in the presence or absence of IFN-γ produced more of the inflammatory cytokine IL-17 than WT cells.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Protection mediated by CD4 effectors does not require IFN-γ in the priming or effector phase. A, Cytokine release, after 48-h restimulation with immobilized anti-CD3, by WT effectors (▪), IFN-γ^−/− effectors cultured in the presence (▧) or absence (dotted bar) of exogenous IFN-γ after the initial 4-day culture. B, A total of 5 × 10⁶ WT cells (▪), IFN-γ^−/− cells cultured in the presence (•), or IFN-γ^−/− cells in the absence (⋄) of exogenous IFN-γ was injected i.v. into BALB/c mice (n = 5/group) that were then infected with 5,000 (2 LD₅₀) or 10,000 (≈5 LD₅₀) EIU PR8 18 h later. C, A total of 5 × 10⁶ WT cells (▪) or IFN-γ^−/− cells cultured in the presence of exogenous IFN-γ (•) were injected into BALB/c mice and infected with 20,000 EIU (≈10 LD₅₀) PR8 18 h later. Control mice did not receive CD4 effectors (▵). Mice were monitored for weight loss and morbidity over time.

To determine whether CD4-derived IFN-γ was required for protection to lethal influenza infection, WT and IFN-γ^−/− effectors grown in the absence or presence of exogenous IFN-γ were transferred to BALB/c mice that were then infected with a lethal dose of influenza. Fig. 3⇑B shows weight loss (top panels) and survival (bottom panels) of mice infected with 5,000 or 10,000 EIU (≈5 LD₅₀) PR8 over a 28-day period. All mice lost weight at the same rate in the first 6–7 days postinfection; however, mice that received CD4 effectors began to recover at ∼7-day postinfection, whereas untreated mice continued to lose weight and eventually succumbed at 5,000 EIU PR8. Increasing the viral dose to 10,000 EIU induced a greater weight loss in all mice, but again, mice that received CD4 effectors began to recover at approximately day 10 postinfection. Importantly, there was no difference in the ability of WT or IFN-γ^−/− effectors to alter weight loss or promote survival, and furthermore, IFN-γ^−/− effectors primed with or without IFN-γ could protect mice against lethal infection at both doses of PR8.

Because it was possible that IFN-γ could be more important for protecting mice against higher doses of pathogenic virus, mice were given CD4 effectors and subsequently infected with 20,000 EIU PR8. At this high dose (∼10 LD₅₀), all mice lost ∼35% of body weight, and mice given CD4 effectors did not begin to recover until day 15 postinfection. Both WT and IFN-γ^−/− effectors were equally effective in abrogating weight loss and promoting survival at this high dose, and no difference in the protective capacity of the effectors was observed (Fig. 3⇑C). Together, these data demonstrate that protection mediated by CD4 effectors does not require IFN-γ in the priming or effector phase, suggesting that CD4 cells use other mechanisms to promote survival against lethal challenge.

CD4 effectors induce protective anti-influenza Abs

A hallmark of CD4 effector function is their ability to promote B cell differentiation and Ab production. Therefore, we hypothesized that increased Ab production, driven by CD4 effectors, may be important for CD4-mediated protection. To evaluate the impact of CD4 effector transfer on Ab production by host B cells, we assayed influenza-specific Ab in serum from lethally infected mice given CD4 effectors or PBS as a control. In BALB/c mice that did not receive effectors, anti-influenza titers were negligible at day 6 postinfection, and total anti-influenza IgG titers were <100 by the time mice became moribund. Transfer of CD4 effectors significantly increased anti-influenza IgG Ab titers in response to lethal PR8 infection compared with mice given PBS (Fig. 4⇓A). Indeed, anti-influenza IgG titers in lethally infected mice given CD4 effectors were 10- to 50-fold higher (p < 0.01) than in mice given PBS and subsequent lethal infection. Cytokine production by CD4 cells is known to influence the Ab isotype profile to T-dependent Ags, and IFN-γ can enhance class switching to IgG2a (32). Consistent with this, Fig. 4⇓B demonstrates that mice given WT CD4 effectors produce significantly higher amounts of anti-influenza IgG2a (p = 0.01) and less IgG1 compared with mice given IFN-γ^−/− effectors. Although the ratio of anti-influenza IgG2a to IgG1 was different in mice receiving WT or IFN-γ^−/− effectors, the ability of CD4 effectors to enhance early Ab production by B cells was not dependent on IFN-γ.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

CD4 effectors induce protective anti-influenza Abs. BALB/c mice were given 5 × 10⁶ WT (▪) or IFN-γ^−/− (•) effectors, or PBS (▵) as a control. Mice were then infected with 5000 EIU (2 LD₅₀) PR8 virus, and serum was collected at 6, 8, 11, 13, and 20 days postinfection. Shown in A is the mean end point anti-influenza IgG titer of three individual mice (±SD), and B shows the mean end point anti-influenza IgG1, IgG2a, and IgG2b titer of four mice ± SD at day 20 postinfection. This experiment was repeated twice with similar results. A total of 5 × 10⁶ WT (▪) or γ^−/− (•) effectors were injected into BALB/c J_HD mice (n = 5/group), and 18 h later, mice were infected with 2500 EIU PR8. Control mice received PBS (▵). C, Shows weight loss and survival of J_HD mice receiving CD4 effectors only. PBS control mice and mice that received γ^−/− CD4 effectors demonstrated a statistically significant difference in survival (p < 0.005) by the Kaplan-Meier log-rank test. This experiment was repeated twice with similar results and statistical differences. J_HD mice receiving CD4 effectors were given naive serum (D), 10 μl of immune sera/mouse (E), or 20 μl of immune sera/mouse (F) 7 days postinfection. Mice receiving IFN-γ^−/− effectors and 10 μl of immune serum showed a statistically higher survival rate than mice receiving 10 μl of immune serum alone (p < 0.05). These experiments were repeated with a different batch of immune sera with similar results.

If CD4-dependent production of anti-influenza Ab is required for CD4-mediated protection, then B cell-deficient mice should not survive a lethal challenge with PR8. Fig. 4⇑C shows weight loss and survival curves of B cell-deficient (J_HD) mice that received PBS, WT, or IFN-γ^−/− effectors before lethal infection. Fig. 4⇑C demonstrates that WT CD4 effectors were unable to protect J_HD mice from lethal infection. Although CD4 effectors did not provide complete protection in B cell-deficient mice, J_HD mice receiving IFN-γ^−/− CD4 effectors were able to significantly (p = 0.005) delay mean time to death compared with J_HD mice given PBS alone.

The observation that CD4 effectors did not protect B cell-deficient mice from lethal infection suggested that either B cells, or the Abs they produce, were required for CD4-mediated protection to lethal influenza. In B cell-deficient mice, we observed a plateau in weight loss at day 5–10 postinfection followed by a sharp drop in weight (Fig. 4⇑C). This coincided with the appearance of anti-influenza Abs (Fig. 4⇑A) and the loss of donor CD4 T cells (Fig. 2⇑A) in intact BALB/c mice. This suggested that CD4 effectors may act via a B cell-independent mechanism during the 5- to 10-day window postinfection and that anti-influenza Ab may act primarily during late stages of lethal infection to ultimately clear virus. To test this hypothesis, J_HD mice were given CD4 effectors, infected with a lethal dose of PR8, and at 7 days postinfection, mice were given convalescent phase sera (i.e., sera from BALB/c mice that had received either WT or IFN-γ^−/− effectors and survived a lethal dose of influenza). Fig. 4⇑D recapitulated our earlier observation and confirmed that J_HD mice given CD4 effectors followed by NMS did not survive a lethal dose of PR8. In contrast, J_HD mice given IFN-γ^−/− CD4 effectors in conjunction with delayed addition of 10 μl of convalescent sera resulted in partial protection (60% survival; Fig. 4⇑E). B cell-deficient mice given a combination of CD4 effectors and 20 μl of immune serum were almost completely protected (80–100% survival) from lethal influenza challenge, whereas only 20% of J_HD mice receiving passive serotherapy at day 7 without CD4 effectors, recovered from lethal PR8 infection (Fig. 4⇑F). Therefore, CD4 effectors, in combination with small amounts of passive sera administered 1 wk postinfection, conferred protection to lethal influenza infection, suggesting that cell-mediated and humoral immunity can synergize to prevent mortality.

CD4 effectors are cytolytic in vitro and in vivo and contribute to Ab-independent protection

The above evidence suggested that CD4-mediated protection did not require IFN-γ or host T cells. However, the plateau in weight loss consistently observed at day 5–10 postinfection, combined with a significant delay in death (Fig. 4⇑C), suggested that CD4 effectors exerted some level of protection in the early stages of infection. There is increasing evidence that CD4 cells can acquire cytolytic activity during infections in mice (33) and humans (34). If CD4 effectors developed cytolytic activity during the 4-day in vitro-priming protocol, they may be able to directly kill virally infected cells. To determine whether primed CD4 effectors acquired cytolytic activity, we examined the ability of in vitro-generated CD4 effectors to kill HA peptide-coated targets in a 4-h in vitro assay (35). Fig. 5⇓A (left panel) demonstrates that both WT and IFN-γ^−/− CD4 effectors were highly cytolytic to peptide-pulsed A20 target cells. This lytic activity was Ag specific because unpulsed A20 cells were not killed. Furthermore, IFN-γ^−/− CD4 effectors had a greater capacity to kill peptide-coated targets compared with WT CD4 effectors (Fig. 5⇓A) and were as potent as CD8 effectors in killing target cells (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

CD4 effectors are cytolytic in vitro and in vivo and contribute to Ab-independent protection. A, WT (▪) or γ^−/− (•) effectors were cultured at various E:T ratios with [³H]thymidine-labeled peptide-pulsed (closed symbols) or unpulsed (open symbols) target cells in a 4-h cytotoxicity assay. Right panel, WT CD4 (▪), γ^−/− CD4 (▧), or CD8 (□) effectors were cultured with peptide-pulsed targets in the presence of 10 μg/ml anti-FasL or increasing concentrations of CMA in a 4-h assay to inhibit Pfn-mediated lysis. Shown is the percentage of inhibition of cytotoxicity at an E:T ratio of 10:1 in the presence of anti-FasL or CMA. B, An in vivo cytotoxicity assay was performed as described in Materials and Methods. A total of 5 × 10⁶ WT CD4, γ^−/− CD4, or CD8 effectors was transferred to BALB/c mice (n = 3/group). Three days posttransfer, CFSE^low, HA_518–526-pulsed targets (I), and CFSE^high HA_126–138-pulsed targets (II) were combined and injected into mice that previously received effectors. Naive mice served as controls. Eighteen hours later, mice were sacrificed, and spleens were removed and monitored for CFSE fluorescence by flow cytometry. Histograms from a representative mouse in each group show the average percentage of cytotoxicity ±SD for each group in a bar graph. This experiment was repeated with similar results. C, A total of 5 × 10⁶ WT (▪) or γ^−/− (•) effectors was injected into J_HD B cell-deficient mice (n = 5/group) and infected 18 h later with 1500 EIU PR8 virus. Mice were monitored for weight loss and morbidity.

Peptide-specific cytolytic activity can occur via two main mechanisms that result in DNA fragmentation and apoptosis of the target cell (36). One mechanism involves the degranulation of intracellular vacuoles and secretion of Pfn and granzymes, whereas the other involves receptor-mediated binding of FasL on the effector cell to Fas on the target cell (37). These two mechanisms of cellular lysis can be distinguished in vitro by the action of the drug CMA, which blocks exocytosis of intracellular vacuoles, thus inhibiting the Pfn-dependent pathway (25), or by addition of a blocking Ab to FasL (38). Because Fas:FasL is known to be a major mechanism of lysis used by CD4 cells in vitro (39), anti-FasL (38) was added during the cytotoxicity assay to block Fas:FasL-mediated killing. Fig. 5⇑A, right panel, shows that blocking of Fas:FasL interactions using anti-FasL Ab did not inhibit CD4-mediated cytotoxicity of peptide-coated A20 cells, even in the 18-h assay, indicating that Fas:FasL was not a major mechanism of cytolytic activity used by these cells.

To determine whether HA-specific CD4 effectors generated in vitro could kill targets using a Pfn-dependent mechanism, we used CD8 cells from HA-specific TCR Tg mice, also primed in vitro under type 1 conditions, as a positive control for Pfn-mediated cytolytic function (40). CMA was added during the 4-h cytotoxicity assay to specifically block Pfn-dependent killing (25). Fig. 5⇑A, right panel, demonstrates that peptid-specific killing by CD4 effectors was inhibited by as little as 100 nM CMA, whereas 500 nM CMA inhibited 60–70% of lysis at an E:T ratio of 10:1, suggesting that CD4 effectors could use Pfn as a mechanism of cytolytic activity.

It was important to confirm that CD4 effectors retained cytolytic activity upon adoptive transfer in vivo. Therefore, we coinjected CFSE^high targets pulsed with HA_126–138 peptide (recognized by CD4 effectors) and CFSE^low targets pulsed with HA_518–528 peptide (recognized by CD8 effectors) to evaluate lytic activity in vivo (27). In this assay, the disappearance of peptide-pulsed CFSE-labeled targets in mice given effectors compared with naive mice indicates peptide-specific cytotoxicity. Fig. 5⇑B shows that the HA_126–138 peptide-loaded targets were preferentially deleted in mice given CD4 effectors (compare 0.43 to 0.97%), whereas the HA_518–528 peptide-pulsed targets disappeared in mice given CD8 effectors (compare 0.19 to 1.04%). IFN-γ^−/− CD4 effectors also lysed HA_126–138 peptide-pulsed targets (compare 0.55 to 0.97%) upon adoptive transfer. Percentages of the resulting CFSE^high and CFSE^low targets were converted into percentage of cytotoxicity, as described in Materials and Methods, and are shown as a bar graph in Fig. 5⇑B. This experiment was repeated using an irrelevant class II binding peptide (OVA_323–339) as a negative control with similar results (data not shown). Therefore, deletion of HA-specific, class II peptide-pulsed targets indicated that CD4 effectors retain peptide-specific cytolytic activity in vivo.

The ability of CD4 effectors to acquire Pfn-mediated cytolytic activity led us to postulate that these cells may promote survival independent of Ab production by B cells. Because B cell-deficient mice succumb to lower doses of influenza PR8 (41), we evaluated CD4-mediated protection after J_HD mice were infected with 1500 EIU PR8 virus, a 2-fold lower dose than that used in the previous experiments (Fig. 4⇑). Fig. 5⇑C demonstrates that 1500 EIU of PR8 virus induced a rapid, progressive weight loss and was fatal to all untreated B cell-deficient mice by day 15 postinfection. In contrast, J_HD mice that received CD4 effectors again showed a plateau in weight loss from days 5–10 but continued to regain weight, and, ultimately, 80% survived an otherwise lethal dose of virus. Therefore, CD4 effectors alone provided protection, without the addition of passive sera, suggesting that cytolytic activity of CD4 effectors may contribute to Ab-independent mechanisms of protection to lethal influenza.

CD4 effectors demonstrate Pfn-mediated cytolytic activity that is required for Ab-independent protection to lethal infection

To directly test the requirement for Pfn in CD4-mediated protection, CD4 effectors were generated from TCR Tg mice deficient in Pfn (Pfn^−/−) (42). Pfn^−/− effectors displayed an activated phenotype (data not shown) and secreted high levels of IFN-γ, IL-10, TNF-α, and RANTES, similar to WT effectors (Fig. 6⇓A). However, Pfn^−/− effectors did not lyse peptide-pulsed A20 target cells in a 4-h cytotoxicity assay, whereas WT effectors demonstrated high, Ag-specific cytolytic activity (Fig. 6⇓B, left panel). This result confirmed the data shown in Fig. 5⇑A and verified the ability of CD4 effectors to lyse peptide-specific targets via Pfn-mediated cytotoxicity. Interestingly, Pfn^−/− effectors acquired FasL-mediated cytolytic activity in an 18 h assay, which was blocked by an Ab to FasL (Fig. 6⇓B, right panel). Again, lytic activity by WT effectors was not blocked by anti-FasL, indicating that the major mechanism of target cell lysis used by CD4 cells involves Pfn.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

CD4 effectors demonstrate Pfn-mediated cytolytic activity that is required for Ab-independent protection to lethal infection. A, Cytokine release of 1 × 10⁶ WT effectors (▪) or Pfn^−/− effectors (stippled bars) after restimulation with immobilized anti-CD3 for 48 h. B, WT (▪) or Pfn^−/− (▾) effectors were cultured at various E:T ratios with [³H]thymidine-labeled peptide-pulsed (closed symbols) or unpulsed (open symbols) A20 target cells in a 4-h cytotoxicity assay. Right panel, WT (▪) or Pfn^−/− (□) effectors were cultured with peptide-pulsed targets in the presence of 10 μg/ml anti-FasL for 4 or 18 h to inhibit Fas-mediated cytotoxicity. A total of 5 × 10⁶ WT (▪) or Pfn^−/− (▾) effectors were injected into BALB/c mice (n = 5/group) (C) or J_HD mice (n = 10/group) (D). Eighteen hours later, mice were infected with 10,000 (≈5 LD₅₀) or 20,000 (≈10 LD₅₀) EIU PR8 virus (C) or 1,500 EIU PR8 virus (D) and monitored for morbidity. This experiment was repeated with similar results.

Because CD4 effectors demonstrated Pfn-mediated lysis in vitro, and exhibited lytic activity in vivo, we tested whether Pfn^−/− CD4 effectors could mediate protection to lethal influenza. Intact BALB/c mice were injected with WT or Pfn^−/− effectors and subsequently infected with very high doses of PR8 virus. Fig. 6⇑C demonstrates that Pfn^−/− effectors were less effective than WT cells in promoting survival at 10,000 and 20,000 EIU PR8; however, 60% of mice given Pfn^−/− effectors still survived a very high (≈10 LD₅₀) dose of PR8. Although this difference did not reach statistical significance, the interpretation of this result was complicated by the Ab response in intact BALB/c mice. Therefore, experiments were repeated in B cell-deficient J_HD mice at a dose of virus where CD4-mediated protection is Ab independent. Fig. 6⇑D shows the weight loss and survival of J_HD mice injected with WT or Pfn^−/− effectors and subsequently infected with 1,500 EIU PR8 virus. As seen in Fig. 5⇑C, J_HD mice given WT CD4 effectors lost weight until about day 5, at which point, weight loss was reversed, and mice regained their body weight by day 15 postinfection. In contrast, mice given Pfn^−/− effectors rapidly lost body weight, did not demonstrate a plateau in weight loss, and eventually succumbed to lethal infection. These data indicate that Ab-independent, CD4-mediated protection to lethal influenza infection requires Pfn, and suggest that CD4 effectors may mediate direct cytotoxicity at the site of lethal infection.