Functional Heterogeneity in the CD4⁺ T Cell Response to Murine γ-Herpesvirus 68

Abstract

CD4⁺ T cells are critical for the control of virus infections, T cell memory and immune surveillance. Here we studied the differentiation and function of murine γ-herpesvirus 68 (MHV-68)-specific CD4⁺ T cells using gp150-specific TCR transgenic mice. This allowed a more detailed study of the characteristics of the CD4⁺ T cell response than previously available approaches for this virus. Most gp150-specific CD4⁺ T cells expressed T-bet and produced IFN-γ, indicating MHV-68 infection triggered differentiation of CD4⁺ T cells largely into the Th1 subset, whereas some became T_FH and Foxp3⁺ regulatory T cells. These CD4⁺ T cells were protective against MHV-68 infection, in the absence of CD8⁺ T cells and B cells, and protection depended on IFN-γ secretion. Marked heterogeneity was observed in the CD4⁺ T cells, based on Ly6C expression. Ly6C expression positively correlated with IFN-γ, TNF-α and granzyme B production, T-bet and KLRG1 expression, proliferation and CD4⁺ T cell-mediated cytotoxicity. Ly6C expression inversely correlated with survival, CCR7 expression and secondary expansion potential. Ly6C⁺ and Ly6C⁻ gp150-specific CD4⁺ T cells were able to interconvert in a bidirectional manner upon secondary antigen exposure in vivo. These results indicate that Ly6C expression is closely associated with antiviral activity in effector CD4⁺ T cells, but inversely correlated with memory potential. Interconversion between Ly6C⁺ and Ly6C⁻ cells may maintain a balance between the two antigen-specific CD4⁺ T cell populations during MHV-68 infection. These findings have significant implications for Ly6C as a surface marker to distinguish functionally distinct CD4⁺ T cells during persistent virus infection.

Introduction

Adaptive immunity to viral infections relies on neutralizing antibodies (Abs), antiviral activity of CD8⁺ T cells and CD4⁺ T cell help. Epstein-Barr virus (EBV) (1) and Kaposi's sarcoma-associated herpesvirus (KSHV) (2) are two γ-herpesviruses that infect humans and are closely associated with the development of malignancies (3). Malignancies associated with EBV and KSHV are commonly found in HIV-infected patients owing to disruption of T cell surveillance (4). Murine γ-herpesvirus 68 (MHV-68) is a naturally occurring rodent pathogen (5), providing an important model to explore γ-herpesvirus infections and immunity (6-10). Mice lacking CD4⁺ T cells lose long-term control of MHV-68 infection (11-13), and CD4⁺ T cells are also thought to contribute to immunity to MHV-68 by more direct mechanisms (14, 15).

CD4⁺ T cells differentiate into various effector cell types depending on the identity of the pathogen, antigen (Ag) characteristics and inflammatory cytokines. The well-known subsets of CD4⁺ T cells include Th1, Th2, Th17, follicular helper T cell (T_FH) and regulatory T cells (Treg) (16). CD4⁺ T helper cells are important for the induction and maintenance of effective humoral immunity (17) and CD8⁺ T cell responses (18). CD4⁺ T cells also contribute to the antiviral response by production of cytokines, such as IL-2 and IFN-γ (14, 19). In addition to being helpers and regulators in antiviral immunity, effector CD4⁺ T cells can directly kill infected cells; these cells are termed cytolytic CD4⁺ T cells or CD4⁺ CTLs (20). Master transcription factors regulate distinct fates of Ag-specific CD4⁺ T cells during viral infection, and T-bet, GATA3, RORγt, Bcl6, eomesodermin (eomes) and Foxp3 can drive CD4⁺ T cell lineage differentiation into Th1, Th2, Th17, T_FH, CTL and Treg, respectively (16).

Upon first Ag encounter, naïve CD8⁺ T cells become activated, expand and develop into short-lived effector cells (SLECs) or memory precursor effector cells (MPECs) (21). SLECs are more terminally differentiated effector cells, conferring immediate protection and decline following Ag clearance. In contrast, MPECs have the ability to respond to survival signals and develop into memory cells. Memory cells are composed of at least two functionally distinct subsets: effector memory (T_EM) and central memory (T_CM) (22). T_EM cells can migrate to inflamed tissues and display immediate effector function, but proliferate poorly in response to Ag. In contrast, T_CM cells mainly home to lymphoid organs and vigorously re-expand upon Ag re-encounter, but lack immediate effector function. Unlike CD8⁺ T cells, however, CD4⁺ T cell differentiation is less well characterized. Lymphocyte antigen 6C (Ly6C) and P-selectin glycoprotein ligand-1 (PSGL1) are considered surface markers to distinguish subsets of CD4⁺ T cells in acute lymphocytic choriomeningitis virus (LCMV) infection (23). Ly6C^hiPSGL1^hi cells have a more terminally differentiated Th1 phenotype; Ly6C^loPSGL1^hi cells are Th1 that have more potential to become memory cells; and Ly6C^loPSGL1^lo identifies T_FH. However the identity of the virus infection can have a marked impact on many aspects of T cell differentiation. These range from altered distribution among phenotypic subsets, to altered differentiation kinetics to T cell exhaustion. Therefore it is important to determine if this model holds true for diverse virus infections. This is particularly true for persistent, reactivating infections such as those of the herpesvirus family, where Ag exposure is chronic yet sporadic.

MHV-68 initially replicates in the lung after intranasal infection and then establishes lifelong latency, primarily in B cells (24), but also in dendritic cells (DCs), macrophages (25) and lung epithelial cells (26). CD4⁺ T cells control MHV-68 infection, not only by providing help for CD8⁺ T cell responses (11), but also through IFN-γ production (14) and direct cytotoxicity (15). It has been shown that ovalbumin (OVA)-specific CD4⁺ T cells have B cell- and CD8⁺ T cell-independent antiviral functions in the control of MHV-68 infection (27). However this was in the context of a virus expressing the strong exogenous antigen OVA, and it is not clear if the same is true for CD4⁺ T cell responses directed toward endogenous viral epitopes. Prior studies have shown persistent MHV-68 infection can induce two effector populations: IFN-γ producers and CD107⁺ cytolytic effectors (28). Recently multiple CD4⁺ T cell epitopes were identified in MHV-68 lytic and latent proteins (29). However detailed functional characteristics, lineage identity and subset distribution within the antiviral CD4⁺ T cell responses have not yet been elucidated at a cellular level. Recently a TCR transgenic mouse with specificity for the MHV-68 gp150_67-83/I-A^b epitope was generated by Blackman and colleagues (30). The gp150_67–83 peptide is an MHC class II (MHC-II)-restricted epitope and activation of gp150-specific CD4⁺ T cells has been confirmed during acute and latent infections. The gp150-specific TCR transgenic mouse provides a powerful model for dissecting the development and antiviral function of MHV-68-specific CD4⁺ T cells.

There are 20~200 naïve CD4⁺ T cells for a specific antigenic epitope in the mouse T cell pool (31). The low precursor frequency, functional heterogeneity and a paucity of utilizable MHC-II tetramers (32) are great barriers for CD4⁺ T cell study, especially for characterizing the CD4⁺ T cell response to a specific viral Ag. In this study, using the gp150-specific TCR transgenic mouse model, we characterized in detail the differentiation, phenotype and antiviral function of virus-specific CD4⁺ T cells. Our study demonstrates that the CD4⁺ T cell response directed toward the endogenous gp150 epitope can protect against virus infection independent of CD8⁺ T cells and B cells. The response in vivo was markedly heterogeneous, divisible by differential Ly6C expression. Antiviral Th1 effector functions were mostly present in the Ly6C⁺ population, whereas the Ly6C⁻ population had the hallmarks of memory cells. Bidirectional interconversion between these populations was observed, suggesting the division of labor within the antiviral CD4⁺ T cell response is dynamically regulated during infection.

Materials and Methods

Mice, generation of CD45.1⁺ gp150-specific TCR transgenic mice and adoptively transferred mice

C57BL/6 (B6) and B6Ly5.2 mice were purchased from the National Cancer Institute (Bethesda, MD). CD45.2⁺ MHV-68 gp150_67–83I-A^b–specific TCR transgenic mice on the recombination-activating gene (RAG)^−/− background were obtained from Dr. Marcia Blackman (Trudeau Institute, Saranac Lake, NY) (30). Mice were maintained under specific pathogen-free conditions in the Dartmouth Center for Comparative Medicine and Research. The Animal Care and Use Committee of Dartmouth College approved all animal experiments. We obtained CD45.1⁺ MHV-68 gp150_67–83I-A^b–specific TCR transgenic (gp150-Tg) mice by breeding the CD45.2⁺ transgenic mouse to B6Ly5.2 (CD45.1⁺) and selecting mice with all CD4⁺ T cells expressing Vα11 and Vβ12. To generate gp150-specific CD4⁺ T cell adoptively transferred mice, gp150-specific CD4⁺ T cells (CD45.1⁺) were prepared from splenocytes of naïve gp150-Tg mice by using EasySep mouse CD4⁺ T cell enrichment kit (StemCell Technologies, Vancouver, Canada). In this study, 10⁴ gp150-specific CD4⁺ T cells were adoptively transferred intravenously (i.v.) into naïve B6 (CD45.2⁺) mice unless otherwise noted, which is consistent with the recommended numbers from previous studies (33).

Virus, infection and plaque assay

MHV-68 clone G2.4 was obtained from Dr. A.A. Nash (University of Edinburgh, Edinburgh, UK). MHV-68 FS73R and FS73 were obtained from Dr. Stacey Efstathiou (University of Cambridge, Cambridge, UK). The viral strains were propagated and titered by using NIH 3T3 cells as previously described (34). ORF73 of MHV-68 encodes mLANA, an homologous protein to KSHV LANA (latency-associated nuclear Ag) (35) that is involved in establishment and maintenance of latency (36). MHV-68 FS73 is a mutant virus with a frameshift in the ORF73 gene, and is unable to establish latency although it replicates normally during the acute phase of the infection. MHV-68 FS73R is a revertant virus with the frameshift repaired, which have similar replication kinetics in vitro and in vivo as WT strains (36). Mice were infected intranasally (i.n.) with 4000 PFU MHV-68 under anesthesia with isoflurane. MHV-68 G2.4 was used unless otherwise noted. In gp150-specific CD4⁺ T cell adoptively transferred mice, infection was performed one day after cell transfer. Virus titers in the lungs were measured by plaque-forming assay as previously described (37). Mice were euthanized when body weight decreased to 75% of their original weight.

Depletion of CD4⁺ T cells and blockade of IFN-γ signaling

To deplete CD4⁺ T cells, mice were administered 500μg anti-CD4 Ab (GK1.5) intraperitoneally (i.p.) at days -1 and 0 post-infection, followed by 250μg twice weekly thereafter until the mice were sacrificed. To block IFN-γ signaling, mice were given i.p. 1mg IFN-γ-blocking Ab (R4-6A2, BioXcell, West Lebanon, NH) at days -1 and 0 post-challenge, and then 500μg every alternate day after infection until mice were sacrificed. Control mice were either untreated or given RatIgG (Jackson ImmunoResearch Laboratories, West Groove, PA).

Tissue and cell preparations

Splenocytes were prepared by passing spleens through cell strainers and red blood cells were lysed using Gey's solution. Lungs were digested with collagenase (2.33 mg/ml) (Sigma-Aldrich, Milwaukee, WI,) and DNase (0.2 mg/ml) (Roche Diagnostics, Indianapolis, Indiana,) for 30 min.

Abs and flow cytometry

Abs for flow cytometric analysis were purchased from eBioscience or BioLegend (San Diego, CA) unless otherwise noted: CD4 APCeFlour780 (GK1.5), CD44 FITC (IM7), CD45.1 allophycocyanine (A20), Ly6C PEcy7 (HK1.4), CXCR5 purified (2G8), KLRG1 PE (2F1 KLRG1), CD127 FITC (SB/199), CD122 PE (TM-b1), PD-1 FITC and allophycocyanine (RMP1-30), CCR7 AlexaFluor488 (4B12), CD62L FITC (MEL-14), IFN-γ FITC and PerCPCy5.5 (XMG1.2), TNF-α FITC (MP6-XT22), IL-2 PE (JES6-5H4), GzmB PE (GB12, Invitrogen, Carlsbad, CA), T-bet Alexa Fluor488 (4B10, Santa Cruz Biotechnology, Santa Cruz, CA), GATA3 PE (TWAJ), RORγt allophycocyanine (AFKJS-9), Bcl6 Alexa Fluor647 (K112-91, BD Biosciences, San Jose, CA), eomes PE (Dan11mag), Foxp3 FITC (FJK-16s), biotinconjugated Affinipure goat anti-rat IgG (H+L) (Jackson ImmunoResearch, West Grove, PA) and streptavidin PE. Samples were analyzed using MacsQuant in the Dartlab core facility or Accuri flow cytometers. Data were analyzed using FlowJo software (Tree Star, Ashland, OR) or Accuri software (BD Biosciences).

Staining of surface markers, intracellular and intranuclear molecules

Surface markers on cells were stained with Abs in PBS with 2% bovine growth serum (FACS buffer) at 4°C for 20 min. For T_FH staining, cells were stained with purified CXCR5 for 1 h in FACS buffer + 0.5% BSA + 2% normal mouse serum (NMS), followed by biotin-goat anti-rat IgG for 30 min in FACS buffer + 2% NMS, and then streptavidin and other surface-staining Abs for 30 min in FACS buffer + 2% NMS at 4°C. For intracellular cytokine/molecule detection, splenocytes were restimulated ex vivo with 2.5μg/ml gp150_67–83 peptide (LSNNNPTTIMRPPVAQN) or 50ng/ml PMA (Sigma-Aldrich, Milwaukee, WI) and 1μg/ml ionomycin calcium salt from Streptomyces conglobatus (Sigma-Aldrich) in complete medium with 10U/ml rIL-2 and 10μg/ml brefeldin A (Sigma-Aldrich) at 37°C for 5h. Subsequently, cells were stained with Abs against surface markers for 20 min at 4°C, followed by fixation with 1% formaldehyde at 4°C for 20 min, and then stained with Abs against IFN-γ, TNF-α, IL-2 or GzmB in 0.5% saponin solution at 4°C for 30 min. For transcription factor detection, Foxp3/Transcription Factor Staining Buffer Set (eBioscience) was used.

Proliferation assay by EdU incorporation

Cell proliferation in vivo was measured by 5-ethynyl-2'-deoxyuridine (EdU) incorporation into DNA. Gp150-specific CD4⁺ T cell adoptively transferred mice that were infected with MHV-68 for 14 days were injected with 1mg EdU i.p. Splenocytes were prepared 16h later and cells were stained with anti-CD4, -CD45.1, -Ly6C Abs and EdU was detected using Click-iT Plus EdU Flow Cytometry Assay Kit (Molecular Probes, Carlsbad, CA) following the manufacturer's protocol.

Secondary expansion of Ly6C⁺ and Ly6C⁻ gp150-specific effector CD4⁺ T cells

1×10⁴ gp150-specific naïve CD4⁺ T cells were adoptively transferred i.v. into B6 mice 1 day prior to MHV-68 infection. 12 days later, splenocytes were prepared and stained with anti-CD4, -CD45.1, -Ly6C Abs, and Ly6C⁺ and Ly6C⁻ gp150-specific CD4⁺ (CD45.1⁺) T cells were sorted using a FACSAria cell sorter. Subsequently, 3×10⁴ Ly6C⁺ or Ly6C⁻ gp150-specific effector cells were adoptively transferred into naïve B6 mice. One day later, the mice were infected with MHV-68. At d8 pi, splenocytes were prepared and stained with anti-CD4, -CD45.1, -Ly6C Abs and analyzed by flow cytometry.

Apoptosis assay by Annexin V staining

Splenocytes were prepared from gp150-specific CD4⁺ T cell adoptively transferred mice that had been infected with MHV-68 for 15 days. Cells were stained with anti-CD4, -CD45.1, -Ly6C Abs and Annexin V using an Apoptosis Detection Kit (BD Biosciences). The apoptosis of Ly6C⁺ and Ly6C⁻ CD45.1⁺CD4⁺ T cells was analyzed by flow cytometry.

Cytotoxicity assays in vivo and ex vivo

For in vivo cytotoxicity assay, splenocytes were prepared from B6 mice and were incubated with 2.5μg/ml gp150 peptide in complete medium at 37°C for 1h. The gp150-pulsed and -unpulsed splenocytes were labeled with 3μM or 0.3μM CFSE, respectively, in HBSS at room temperature for 10 min. The gp150-pulsed (CFSE^hi) and -unpulsed (CFSE^lo) cells were mixed at a 1:1 ratio and used as target cells. 5×10⁶ target cells were injected i.v. into MHV-68 infected gp150-Tg mice (at d13 pi) or naïve gp150-Tg mice, and mice were sacrificed 20 or 40h later. Splenocytes were prepared and stained with 10μM 7-Aminoactinomycin D (7-AAD, BD Biosciences) at room temperature for 15 min. CFSE-positive target cells were analyzed by flow cytometry. Specific lysis was quantitated with percent specific killing, calculated according to the formulas: % specific lysis = (1 − [ratio of infected recipients / ratio of naïve recipients]) × 100%, where ratio = number of CFSE^hi / number of CFSE^lo.

For ex vivo cytotoxicity assays, 2×10⁴ gp150-specific CD4⁺ T cells were adoptively transferred into B6 mice and mice were infected with MHV-68 for 13 days. Splenocytes were stained with anti-CD4, -CD45.1, -Ly6C Abs and sorted by FACS for Ly6C⁺ and Ly6C⁻ CD45.1⁺CD4⁺ T cells. Ly6C⁺ or Ly6C⁻ CD4⁺ T cells were cultured with target cells at an effector to target ratio of 6:1 in complete media in 96 well plates at 37°C for 40h. Cells were stained with 10μM 7-AAD at room temperature for 15 min and CFSE-positive target cells were analyzed by flow cytometry. Specific lysis was calculated based on the formulas: % specific lysis = [1- (ratio of Ly6C⁺ or Ly6C⁻ CD4⁺ group / ratio of target alone group)] × 100%, where ratio = number of CFSE^hi / number of CFSE^lo.

Statistical analysis

Statistical differences were determined by Student's t tests (two-tailed, unpaired) except for the survival curve (log-rank analysis), using GraphPad Prism 5 software (GraphPad, La Jolla, CA).

Results

Gp150-specific CD4⁺ T cells respond similarly whether or not MHV-68 establishes a persistent infection

We first determined the kinetics of the gp150-specific CD4⁺ T cell response in mice that received adoptively transferred gp150-specific CD4⁺ T cells. Mice were infected with one of two MHV-68 strains: revertant FS73R or mutant FS73 virus that can or cannot establish latency, respectively (36, 38). Comparison of FS73R and FS73 infection allowed us to elucidate the effect of persistent infection on the CD4⁺ T cell response. Upon FS73R or FS73 infection, the gp150-specific CD4⁺ T cells expanded, and peaked at day 10 post-infection (pi) (Fig. 1A), with average frequencies of 0.8% - 1.9% of CD4⁺ T cells, and total numbers more than 10⁵ cells per spleen (Fig. 1B). Similar frequencies of gp150-specific CD4⁺ T cells were observed in the lung for FS73R and FS73 infection at d10 pi (Suppl. Fig. 1A). We also measured polyclonal CD44⁺ CD4⁺ T cell populations. At d10 pi, CD44⁺ T cells represented approximately 40% of CD4⁺ T cells in the spleen and 60% in lung in both FS73R and FS73 infections (Suppl. Fig. 2A). Over time, the gp150-specific CD4⁺ T population contracted in both FS73R and FS73 infection (Fig. 1B). It was difficult to reliably detect CD4⁺ T cell memory populations after d50 pi. There was a trend toward higher frequencies and larger total numbers of gp150-specific CD4⁺ T cells in FS73R-infected mice compared to FS73-infected mice, but in most cases it did not reach statistical significance. These data show that gp150-specific CD4⁺ T cells expand and are maintained similarly after adoptive transfer during infection with both the persistent and non-persistent strains of MHV-68. The response peaks during the acute stage of infection, then declines during latent infection.

FIGURE 1.

Kinetics of gp150-specific CD4⁺ T cell responses during MHV-68 FS73R and FS73 infection. Gp150-specific CD4⁺ T cells were prepared from splenocytes of MHV-68- gp150-specific TCR transgenic (gp150-Tg) mice (CD45.1⁺) and 1×10⁴ cells were transferred (i.v.) into B6 (CD45.2⁺) mice, one day prior FS73R or FS73 infection (i.n.). Spleens were harvested from infected gp150-specific CD4⁺ T cell adoptively transferred mice, and splenocytes were stained with anti-CD4 and anti-CD45.1 Abs, and analyzed by flow cytometry.(A) Representative FACS plots of gp150-specific CD4⁺ T cell populations at indicated time post-infection (pi). Plots were gated on CD4⁺ T cells. The percentage shown is the frequency of CD45.1⁺ among CD4⁺ T cells. (B) Frequency of gp150-specific CD4⁺ T cells among CD4⁺ T cells and total number of gp150-specific CD4⁺ T cells in spleen. Data are representative of at least two independent experiments for each time point with four mice per group. Error bars on the graphs represent SD. *p < 0.05.

Gp150-specific CD4⁺ T cells differentiate into two populations based on Ly6C expression, and Ly6C⁺ and Ly6C⁻ cells can interconvert upon secondary exposure to Ag

To test whether there was heterogeneity within the gp150-specific CD4⁺ T cell population based on surface marker expression, we stained for Ly6C and PSGL1 as it has been shown that Th1 cells are Ly6C^hiPSGL1^hi while T_FH cells are Ly6C^loPSGL1^lo in LCMV infection (23). Gp150-specific CD4⁺ T cell adoptively transferred mice were infected with MHV-68 FS73R or FS73, and cells from spleen and lung were prepared and stained with anti-CD4, -CD45.1, -Ly6C and -PSGL1 Abs at different time pi. We did not observe a distinct PSGL1 negative population in gp150-specific CD4⁺ T cells (Suppl. Fig. 3A). Based on Ly6C expression, the gp150-specific CD4⁺ T cells from splenocytes could be divided clearly into two populations. Ly6C⁺ and Ly6C⁻ cell populations existed at a ratio of 2:3 at d10 and d14 pi in both FS73R and FS73 infections (Fig. 2A). The gp150-specific CD4⁺ T cells from the lungs could also be divided clearly into the two populations (Suppl. Fig. 1B). Similarly, Ag-experienced CD44⁺CD4⁺ T cells in spleen and lung could also be divided into Ly6C⁺ and Ly6C⁻ populations after FS73R and FS73 infection (Suppl. Fig. 2B). Over time, the Ly6C⁻ cells in the gp150-specific CD4⁺ cell population became more dominant in FS73R-infected mice, changing to a ratio of 1:4 (Ly6C⁺:Ly6C⁻) at d45 pi, a difference that was statistically significant (p<0.05) when d45 and d10 values were compared. However the ratio remained fairly consistent at 2:3 in FS73-infected mice (Fig. 2A). These findings indicate that both FS73R and FS73 infections direct gp150-specific CD4⁺ T cells into two populations: Ly6C⁺ and Ly6C⁻, and the Ly6C⁻ cells are the larger population in the gp150-specific effector CD4⁺ T cell pool. The altered ratio of Ly6C⁺:Ly6C⁻ cells at later times post-infection could be due to persistent MHV-68 preferentially expanding the Ly6C⁻ population.

FIGURE 2.

Two populations of gp150-specific CD4⁺ T cells based on Ly6C expression. (A) Splenocytes were prepared from infected gp150-specific CD4⁺ T cell adoptively transferred mice and stained with anti-CD4, -CD45.1 and -Ly6C Abs. Representative FACS plots (left panel) were gated on CD45.1⁺CD4⁺ T cells (gp150-specific cells). Bar graph shows relative percentage of Ly6C⁺ and Ly6C⁻ cells within CD45.1⁺CD4⁺ T cells (right panel). Data are representative of at least two independent experiments for each time point with four mice per group. (B) Interconversion between Ly6C⁺ and Ly6C⁻ populations. Gp150-specific CD4⁺ T cells were transferred into B6 mice one day prior to infection. At d12 pi, splenocytes were prepared from 8 spleens, and Ly6C⁺ and Ly6C⁻ CD45.1⁺CD4⁺ T cells were sorted, separately transferred into naïve B6 mice (3×10⁴ cells/mouse) and infected with MHV-68. At d8 pi, splenocytes were stained with anti-CD4, -CD45.1 and -Ly6C Abs. Representative FACS plots (left panel) were gated on CD45.1⁺CD4⁺ T cells. Bar graph shows the Ly6C expression conversion of donor cells (right panel). Data are representative of three independent experiments with three to ten mice per group. Error bars on the graphs represent SD.

Next, we assessed stability of Ly6C expression on gp150-specific effector CD4⁺ T cells. Gp150-specific effector CD4⁺ T cells from infected gp150-specific CD4⁺ T cell adoptively transferred mice were sorted into Ly6C⁺ and Ly6C⁻ cells at d12 pi, then transferred separately into naïve B6 mice. The purity of sorting was 98.5% for Ly6C⁺ cells and 99.0% for Ly6C⁻ cells (data not shown). One day post transfer, mice were infected with MHV-68 and spleens were harvested at d8 pi. The transferred effector cells continued to expand and differentiate, and each gave rise to both Ly6C⁺ and Ly6C⁻ populations, but the majority of cells retained their original phenotype: Ly6C⁺ cells differentiated into Ly6C⁺ and Ly6C⁻, at a ratio of 7:3; Ly6C⁻ cells into Ly6C⁺ and Ly6C⁻, at a ratio of 3:7, respectively (Fig. 2B). The data indicate that both Ly6C⁺ and Ly6C⁻ cells maintain plasticity to differentiate into the alternate phenotype.

Gp150-specific CD4⁺ T cells differentiate into Th1, Treg or T_FH, and Ly6C expression is positively associated with the expression of T-bet

To explore the differentiation of Ag-specific CD4⁺ T cells during MHV-68 infection, we measured expression of transcriptional master regulators, T-bet, GATA3, RORγt, Bcl6, eomes and Foxp3, in gp150-specific CD4⁺ T cells. Both Ly6C⁺ and Ly6C⁻ gp150-specific effector CD4⁺ T cells from gp150-specific CD4⁺ T cell adoptively transferred mice expressed more T-bet than naïve CD4⁺ T cells, but no difference was observed in the expression of GATA3, RORγt, Bcl6 or eomes between effector and naïve CD4⁺ T cells (Fig. 3A). The positive controls of GATA3, RORγt, Bcl6 or eomes staining compared to naïve CD4⁺ T cells are shown (Suppl. Fig. 3B). The gp150-specific CD4⁺ T cells from FS73R-infected mice expressed more T-bet than those from FS73-infected mice, and Ly6C⁺ cells expressed significantly more T-bet than Ly6C⁻ cells (Fig. 3B). In the lungs, the gp150-specific Ly6C⁺CD4⁺ T cells in FS73R-infected mice also expressed more T-bet (Suppl. Fig. 1C). Similarly, Ly6C⁺CD44⁺CD4⁺ T cells expressed more T-bet than Ly6C⁻CD44⁺CD4⁺ T cells in FS73R and FS73 infections in both spleen and lung (Suppl. Fig. 2C). A small population of gp150-specific CD4⁺ T cells expressed Foxp3 and the percentage was higher in the Ly6C⁺CD4⁺ T cell population, compared to their Ly6C⁻ counterparts at d14 pi (Fig. 3C).

FIGURE 3.

Differentiation of gp150-specific CD4⁺ T cells into Th1, Treg and T_FH. Splenocytes from infected gp150-specific CD4⁺ T cell adoptively transferred mice were stained with anti-CD4, -CD45.1, -Ly6C, -transcription factors, -CXCR5 and/or -PD-1 Abs. (A) Representative FACS histograms for transcription factor expression in Ly6C⁺ (bold line), Ly6C⁻ (dashed line) CD45.1⁺CD4⁺ T cells at d14 pi and naïve CD4⁺ T cells from B6 mice (filled histogram). (B) T-bet geometric mean fluorescence intensity (GMFI) in Ly6C⁺ or Ly6C⁻ CD45.1⁺ CD4⁺ T cells at the indicated days pi and in naïve CD4⁺ T cells. (C) Representative FACS plots of Foxp3 expression. Plots were gated on Ly6C⁺ or Ly6C⁻ CD45.1⁺CD4⁺ T cells and graph shows frequency of Foxp3⁺ cells in Ly6C⁺ or Ly6C⁻ CD45.1⁺CD4⁺ T cells at d14 pi. (D) Representative FACS plots and frequencies of CXCR5^hiPD-1^hi cells within CD45.1⁺CD4⁺ T cells at d15 pi. Data are representative of at least two independent experiments with three to four mice per group. ns, no significance; *p < 0.05, **p < 0.01, ***p < 0.001.

It has been shown in infection with multiple pathogens, including Listeria monocytogenes (39) and MHV-68 (40), that CXCR5^hiPD-1^hi identifies T_FH cells, and we were able to detect T_FH with these antibodies. A low frequency of T_FH was detected among gp150-specific cells, with higher frequencies of CXCR5^hiPD-1^hi cells in FS73R infection, compared to FS73 infection (8.3% vs. 3.7%) (Fig. 3D). Similarly, the frequency of CXCR5^hiPD-1^hi cells within CD44⁺CD4⁺ T cells was higher in FS73R infection than that in FS73 infection (13% vs. 6.8%) (Suppl. Fig. 2E). These data suggest that most of gp150-specific CD4⁺ T cells differentiate into Th1 cells, but some into T_FH and Tregs during MHV-68 infection.

Ly6C expression is positively associated with the production of effector molecules, IFN-γ, TNF-α and granzyme B

To investigate the function of Ag-specific effector CD4⁺ T cells, we measured effector molecules, IFN-γ, TNF-α, IL-2 and granzyme B (GzmB) produced by Ly6C⁺ and Ly6C⁻ gp150-specific effector CD4⁺ T cells. Splenocytes from MHV-68-infected gp150-specific CD4⁺ T cell adoptively transferred mice were restimulated with gp150 peptide (Ag specific stimulation) or PMA/ionomycin (non-specific stimulation), then stained intracellularly with Abs against effector molecules. Frequencies of gp150-specific CD4⁺ T cells producing IFN-γ or TNF-α in the Ly6C⁺ population were significantly higher than those in their Ly6C⁻ counterparts, regardless of the nature of stimulation (Fig. 4A). Cytokine production was observed from a relatively low frequency of gp150-specific CD4⁺ T cells following peptide stimulation, and this was increased markedly following PMA stimulation. With PMA stimulation, 50% and 45% of Ly6C⁺ gp150-specific effector CD4⁺ T cells produced IFN-γ and TNF-α, respectively, whereas the frequencies were 23% and 14% in Ly6C⁻ cells at d14 pi (Fig. 4A, right panel). While Ly6C⁺ cells produced more IL-2 than Ly6C⁻ cells after peptide stimulation, production was from a low percentage of cells, and no difference was observed following PMA stimulation. Ly6C⁺ gp150-specific CD4⁺ T cells also produced more GzmB on a per cell basis, and geometric mean fluorescence intensity (GMFI) was approximately 2-fold higher than Ly6C⁻ cells (Fig. 4A). In the CD44⁺CD4⁺ T cell population, higher frequencies of Ly6C⁺ cells produced IFN-γ compared to Ly6C⁻ cells (Suppl. Fig. 2D), consistent with data obtained from antigen-specific CD4⁺ T cells. Further, at the memory phase, Ly6C expression was still associated positively with IFN-γ production (Fig. 4B).

FIGURE 4.

Positive correlation between Ly6C expression and IFN-γ, TNF-α and GzmB production. Splenocytes from infected gp150-specific CD4⁺ T cell adoptively transferred mice were restimulated with gp150 peptide or PMA/ionomycin, and then stained with anti-CD4, -CD45.1, -Ly6C, -IFN-γ, -TNF-α, -IL-2 and/or -GzmB Abs. FACS plots were gated on Ly6C⁺ or Ly6C⁻ CD45.1⁺CD4⁺ T cells. Production of IFN-γ, TNF-α and IL-2 is presented as percent positive cells, and production of GzmB is presented as GMFI within Ly6C⁺ or Ly6C⁻ CD45.1⁺CD4⁺ T cells. (A) Production of effector molecules at d14 pi. by gp150-specific CD4⁺ T cells following gp150 peptide (left panel) or PMA (right panel) restimulation. (B) IFN-γ production with PMA restimulation at d30 pi. (C) Co-production of effector cytokines by Ly6C⁺ or Ly6C⁻ CD45.1⁺CD4⁺ T cells with PMA restimulation at d14 pi Pie charts and bar graphs show the proportions of the total response that are positive for three, two, one or no cytokine within Ly6C⁺ or Ly6C⁻ CD45.1⁺CD4⁺ T cells. All possible combinations of responses are shown on the X-axis of the bar graph. Error bars on the graphs represent SD. Data are representative of at least two independent experiments with three to five mice per experiment. *p < 0.05, **p < 0.01, ***p < 0.001.

To determine whether the gp150-spefic CD4⁺ T cells were polyfunctional, we measured co-production of cytokines by staining cells with the combinations of anti-IFN-γ, -TNF-α and -IL-2 Abs at d14 pi. 13% of Ly6C⁺ gp150-specific effector CD4⁺ T cells produced IFN-γ, TNF-α and IL-2, and 27% produced IFN-γ and TNF-α. By contrast, 6% of Ly6C⁻ cells produced IFN-γ, TNF-α and IL-2, and 5% produced IFN-γ and TNF-α (Fig. 4C). Notably, in the Ly6C⁺ population, 43% of the gp150-specific CD4⁺ T cells did not produce any of the three cytokines, whereas this frequency was 75% in the Ly6C⁻ population (Fig. 4C). The data show that Ly6C⁺ cells are more polyfunctional than Ly6C⁻ cells during MHV-68 infection.

Ly6C expression is positively associated with expression of KLRG1 and CD122 but negatively associated with CCR7 expression

Killer cell lectin-like receptor G1 (KLRG1) is described as a marker for replicative senescence of murine CD8⁺ T cells (41). CD127 (IL-7Rα) is a marker to identify memory cell precursors (42). In CD8⁺ T cells, KLRG1^hiCD127^lo and KLRG1^loCD127^hi cells are considered SLECs and MPECs, respectively (43). In CD4⁺ T cells, more terminally differentiated cells also express more KLRG1 in Mycobacterium tuberculosis infection (44) and higher KLRG1 mRNA in LCMV infection (23). However, unlike CD8⁺ T cells, CD127 expression is not associated with CD4⁺ T cell memory potential (23). In this study, we measured KLRG1 and CD127 expression on Ly6C⁺ and Ly6C⁻ gp150-specific CD4⁺ T cells in MHV-68 infection. KLRG1 expression was upregulated in the Ly6C⁺ gp150-specific effector CD4⁺ T cell population at d15 pi (Fig. 5A). By contrast, the proportion of KLRG1⁺ cells in the Ly6C⁻ population was similar to that in the naïve CD4⁺ T cell population. CD127 expression on Ly6C⁺ gp150-specific effector CD4⁺ T cells was higher than Ly6C⁻ cells. However even naïve Ly6C⁺CD4⁺ T cells had high expression relative to Ly6C⁻ cells.

FIGURE 5.

Expression of surface markers on Ly6C⁺ and Ly6C⁻ gp150-specific CD4⁺ T cells. Splenocytes from infected gp150-specific CD4⁺ T cell adoptively transferred mice were stained with anti-CD4, -CD45.1, -Ly6C and -KLRG1, -CD127, -CD122, -CD44, -PD-1, -CD62L or -CCR7 Abs. Naïve CD4⁺ T cells were prepared from B6 mice. Graphs show percentage or GMFI of cells expressing the molecule indicated within Ly6C⁺ or Ly6C⁻ CD45.1⁺CD4⁺ T cells. (A) Expression of KLRG1 and CD127 at d15 pi. (B) Expression of CD122, CD44 and PD-1 at d15 pi. (C) Expression of CD62L at d15 and d55 pi and expression of CCR7 at d55 pi. Data are representative of two independent experiments with three to seven mice per experiment. *p < 0.05, **p < 0.01, ***p < 0.001.

We also measured the expression of CD122, CD44 and PD-1 on gp150-specific effector CD4⁺ T cells. Compared with Ly6C⁻ cells, Ly6C⁺ cells had higher CD122 expression, but no differences were observed in CD44 and PD-1 expression (Fig. 5B). CD122, the β chain of the IL-2/IL-15 receptor, is critical for IL-2 and IL-15 signaling (45) and can be upregulated by T-bet expression (46). Higher expression of CD122 on Ly6C⁺ gp150-specific effector cells may increase their responsiveness to IL-2 and IL-15.

CD62L and CCR7 are important for T cell circulation through lymph nodes, and are differentially expressed on central and effector memory T cells (47-49). In this study, CD62L expression was lower on cells from infected animals compared with naïve cells, and frequencies of CD62L-expressing gp150-specific CD4⁺ T cells were higher in the Ly6C⁺ population in naïve cells and effector/memory cells at d15 and d55 pi, when compared with Ly6C⁻ cells (Fig. 5C). Therefore differential expression of CD62L is not a post-activation feature of cell populations defined by Ly6C expression, but exists even in the naïve CD4 population. Unlike CD62L, CCR7 expression on naïve CD4⁺ T cells was similar between Ly6C⁺ and Ly6C⁻ populations (data not shown). During the memory phase, CCR7 expression was higher on the Ly6C⁻ gp150-specific CD4⁺ T cells than their Ly6C⁺ counterparts at d55 pi (Fig. 5C), indicating that Ly6C expression was negatively associated with CCR7 at the memory stage.

Ly6C expression is associated with proliferation and survival potential of gp150-specific CD4⁺ T cells

To assess whether Ly6C expression correlated with proliferation capacity of Ag-specific CD4⁺ T cells, we first measured the proliferation of Ly6C⁺ and Ly6C⁻ gp150-specific effector CD4⁺ T cells during the primary response by EdU incorporation. MHV-68-infected gp150-specific CD4⁺ T cell adoptively transferred mice were injected at d14 pi with EdU, and 16h later mice were sacrificed. The frequencies of EdU-positive cells were significantly higher in the Ly6C⁺gp150-specific CD4⁺ T cell population than those in their Ly6C⁻ counterparts (Ly6C⁺ vs. Ly6C⁻: 33% vs. 12%) (Fig. 6A), indicating the Ly6C⁺ population proliferated more quickly. Since Ly6C⁺ and Ly6C⁻ cells can interconvert, we cannot rule out the possibility that the Ly6C⁺ population contains cells that converted from the Ly6C⁻ population during proliferation and vice versa. However, our previous data (Fig. 2B) indicates that approximately 30% of the cells interconverted during 8 days, whereas the current experiment was conducted over a period of only 16 hours.

FIGURE 6.

Correlation between Ly6C expression and potential for survival and proliferation. (A) Proliferation during the primary response. Infected gp150-specific CD4⁺ T cell adoptively transferred mice were injected with EdU i.p. at d14 pi. Spleens were harvested 16h later and stained with anti-CD4, -CD45.1 and -Ly6C and EdU incorporation was detected. Representative FACS plots (left) were gated on Ly6C⁺ or Ly6C⁻ CD45.1⁺CD4⁺ T cells. Graph (right) show percent of EdU positive cells within Ly6C⁺ or Ly6C⁻ CD45.1⁺CD4⁺ T cells. Data are representative of two independent experiments with six mice per experiment. (B) Secondary expansion. Gp150-specific effector CD4⁺ T cells from infected gp150-specific CD4⁺ T cell adoptively transferred mice were sorted into Ly6C⁺ and Ly6C⁻ cells at d12 pi and separately transferred into naïve B6 mice (3×10⁴ cells/mouse). Mice were infected with MHV-68 and splenocytes were stained at d8 pi. Graph shows total number of donor gp150-specific effector CD4⁺ T cells after re-encounter with MHV-68. Data are representative of two independent experiments with eight to ten mice per group. (C) Apoptosis assay. Splenocytes were prepared from infected gp150-specific CD4⁺ T cell adoptively transferred mice at d15 pi, and stained with anti-CD4, -CD45.1, -Ly6C Abs and Annexin V. Representative FACS histogram were gated on Ly6C⁺ or Ly6C⁻ CD45.1⁺CD4⁺ T cells. Graph shows frequency of Annexin V⁺ cells in Ly6C⁺ or Ly6C⁻ CD45.1⁺CD4⁺ T cells. Data are representative of two independent experiments with three mice per experiment. *p < 0.05, **p < 0.01, ***p < 0.001.

To assess correlation between Ly6C expression and the proliferative capacity of Ag-specific CD4⁺ T cells upon secondary exposure to Ag, we compared expansion by re-exposing the Ly6C⁺ and Ly6C⁻ populations to MHV-68. Ly6C⁺ and Ly6C⁻ gp150-specific effector cells were sorted and Ly6C⁺ or Ly6C⁻ cells were separately adoptively transferred into naïve B6 mice. One day post transfer, recipient mice were infected with MHV-68 and sacrificed at d8 pi. Upon secondary exposure to Ag, the gp150-specific effector CD4⁺ T cells proliferated rapidly, and the pool of Ly6C⁻ cells was much larger, approximately 6-fold, than that of Ly6C⁺ cells (Ly6C⁻ vs. Ly6C⁺: 6×10⁴ cells/spleen vs. 1×10⁴ cells/spleen) (Fig. 6B). This showed that the Ly6C⁻ gp150-specific effector CD4⁺ T cell population had higher proliferative potential compared with the Ly6C⁺ population. These data show that Ly6C expression is positively associated with cell proliferation during the primary response, but inversely associated with proliferative potential upon secondary exposure to antigen.

To assess whether Ly6C expression correlated with the survival capacity of Ag-specific CD4⁺ T cells, we measured apoptosis in Ly6C⁺ and Ly6C⁻ gp150-specific effector CD4⁺ T cells, using Annexin V staining. At d15 pi, over 40% of Ly6C⁺ gp150-specific effector CD4⁺ T cells from infected gp150-specific CD4⁺ T cell adoptively transferred mice were positive for Annexin V, compared with approximately 20% for their Ly6C⁻ counterparts (Fig. 6C). This indicated that the Ly6C⁺ gp150-specific CD4⁺ T cell population contained more cells undergoing apoptosis compared with the Ly6C⁻ population.

Gp150-specific CD4⁺ T cells contribute to antiviral responses independent of CD8⁺ T cells and B cells, and this activity depends on IFN-γ

Gp150-specific TCR transgenic (gp150-Tg) mice were on the RAG^−/− background, thus we could exploit this animal model to study contribution of Ag-specific CD4⁺ T cells to antiviral immunity in the absence of CD8⁺ T cells and B cells. Intact or CD4-depleted gp150-Tg mice were infected with MHV-68. Titers of virus in the lungs and body weight were measured. At d15 pi, viral titers in the lungs of CD4-depleted mice were 1600-fold higher than those of intact mice (Fig. 7A). The body weight of CD4-depleted mice rapidly decreased after d10 pi (Fig. 7B, middle panel) and these mice exhibited poorer survival compared with intact mice (Fig. 7C). Despite this clear protection, all mice eventually succumbed to infection, presumably due to the lack of antiviral CD8⁺ T cells and the absence of an antiviral Ab response. We concluded transgenic gp150-specific CD4⁺ T cells could provide relative protection from MHV-68 infection in the absence of indirect effects mediated by help to CD8⁺ T cells or B cells.

FIGURE 7.

Antiviral activity mediated by IFN-γ. Intact (RatIgG), CD4⁺ T cell-depleted (αCD4) and IFN-γ-blocked (αIFN-γ) gp150-Tg mice were infected with MHV-68. (A) Viral titers in lungs of infected mice at d15 pi. (B) Body weight change, calculated relative to original weight (before infection). (C) Endpoint for survival (75% of original weight). Data are representative of two independent experiments with four mice per group. *p < 0.05, **p < 0.01.

To explore mechanisms of CD4⁺ T cell-mediated protection from MHV-68 infection, we tested the role of IFN-γ. IFN-γ is considered a hallmark Th1 cytokine (16), and plays a pivotal role in the outcome of many virus infections as inducer, regulator and effector for both innate and adaptive antiviral responses (50, 51). To inhibit IFN-γ signaling, MHV-68-infected gp150-Tg mice were given IFN-γ-blocking Ab. Similar to CD4-depletion, IFN-γ-blockade greatly increased the viral load in the lungs, and the titer was approximately 1400-fold higher than that of intact mice at d15 pi (Fig. 7A). Moreover, IFN-γ-blockade caused more rapid weight loss following infection (Fig. 7B, right panel) and poorer survival (Fig. 7C). These data demonstrate the critical role of IFN-γ for gp150-specific CD4⁺ T cells to control viral replication and thereby prolong the survival of infected mice.

Gp150-specific CD4⁺ T cells display direct cytolytic activity in vivo, and Ly6C⁺ effector cells have stronger cytolytic activity than Ly6C⁻ cells ex vivo

To explore whether Ag-specific CD4⁺ T cells can differentiate into cells with cytotoxic potential, we assessed the cytolytic activity of gp150-specific CD4⁺ T cells by performing an in vivo cytotoxicity assay. Gp150 peptide-pulsed (CFSE^hi) and -unpulsed (CFSE^lo) splenocytes from B6 mice were mixed 1:1 as target cells, and were injected into naïve or infected gp150-Tg mice (d13 pi). At the time indicated, spleens were harvested, CFSE-labeled target cells were detected by flow cytometry, and the percent specific lysis was calculated. Compared to unpulsed target cells, the proportion of gp150-pulsed target cells decreased and the specific lysis was approximately 44% (Fig. 8A), indicating that gp150-specific CD4⁺ T cells had the ability to elaborate cytotoxic functions.

FIGURE 8.

Positive correlation between Ly6C expression and cytotoxicity. Naïve B6 splenocytes were incubated with or without gp150 peptide and labeled with 3μM or 0.3μM CFSE. The gp150-pulsed (CFSE^hi) and non-pulsed (CFSE^lo) cells were mixed at a 1:1 ratio as target cells. (A) Cytotoxic activity in vivo. Target cells were transferred to naïve or infected gp150-Tg mice (at d13 pi). 20 or 40h later, spleens were harvested and CFSE-labeled target cells were analyzed by flow cytometry. Representative FACS histograms of live target cells at 40h (left panel) and percent specific lysis (right panel). Data are representative of two independent experiments with three mice per group. (B) Cytotoxic activity ex vivo. Ly6C⁺ and Ly6C⁻ effector gp150-specific CD4⁺ T cells were sorted from splenocytes of infected gp150-specific CD4⁺ T cell adoptively transferred mice at d13 pi. The Ly6C⁺ or Ly6C⁻ effector cells were cultured with target cells at a ratio of 6:1 for 40h and CFSE-labeled target cells were analyzed. Representative FACS histograms of live target cells and percent specific lysis are shown. Error bars on the graphs represent SD. *p < 0.05.

To determine whether Ly6C expression was associated with CD4⁺ CTL activity, we performed an ex vivo cytotoxicity assay. Target cells were prepared in the same way as those in the in vivo cytotoxicity assay. Ly6C⁺ and Ly6C⁻ effector CD4⁺ T cells were FACS sorted from the splenocytes of infected gp150-specific CD4⁺ T cell adoptively transferred mice at d13 pi, and co-cultured with target cells for 40h. Ly6C⁺ gp150-specific effector CD4⁺ T cells displayed stronger cytotoxicity, with approximately 25% specific lysis, whereas their Ly6C⁻ counterparts displayed only 7% specific lysis (Fig. 8B), indicating that Ly6C⁺CD4⁺ T cells had stronger CTL activity.

Discussion

Gp150 was the first characterized membrane glycoprotein of MHV-68 (52) and displays homology with EBV gp350/220 and KSHV K8.1 (35). Its expression is associated with late lytic infection (53), and gp150-deficient MHV-68 shows defective virion release but establishes normal latency (54). Although gp150 is associated with lytic infection, gp150-specfic CD4⁺ T cells can be activated during latency, indicating the gp150 epitope is recognized by CD4⁺ T cells during periodic virus reactivation (30). Gp150 is the dominant epitope of MHV-68 for recognition by CD4⁺ T cells and vaccination against this protein can result in a reduction of viral titers in the lungs (55). In this study, we characterized the differentiation and function of Ag-specific CD4⁺ T cells as well as their association with Ly6C expression in viral infection by using MHV-68 gp150-transgenic mice.

Ly6C is expressed on monocytes, DCs, neutrophils and lymphocytes, and the function of Ly6C may be cell type-specific (56). Ly6C on monocytes is considered a differentiation Ag: Ly6C⁺ monocytes showing an inflammatory or immature phenotype and Ly6C⁻ monocytes showing a resident or mature phenotype (56, 57). Ly6C is also expressed on the majority of peripheral CD8⁺ T cells, primarily on Ag-experienced cells (58), and supports preferential homing of CD8⁺ T_CM cells into lymph nodes (59). In this study, MHV-68 infection triggered gp150-specific CD4⁺ T cells to proliferate and differentiate into Ly6C⁺ and Ly6C⁻ cells in adoptively transferred mice (Fig. 2A). The gp150-specific effector CD4⁺ T cells expressed more T-bet than naïve cells, and Ly6C⁺ cells expressed more T-bet than Ly6C⁻ cells (Fig. 3A, 3B), suggesting that Ly6C expression is directly correlated with T-bet, which is in agreement with a report from the LCMV infection model (23). Additionally, a small population of gp150-specific CD4⁺ T cells expressed Foxp3 (Fig. 3C). These data indicate that the overwhelming majority of gp150-specific CD4⁺ T cells differentiate into the Th1 subset, and Ly6C expression is positively correlated with T-bet expression.

In CD8⁺ T cells, high expression of T-bet induces SLECs, whereas low expression of T-bet promotes MPECs (43). In CD4⁺ T cells, T-bet has been shown to associate with terminally differentiated Th1 cells (23). In this study, Ly6C expression was positively correlated with T-bet expression (Fig. 3A, B) and KLRG1 expression (Fig. 5A). Moreover, compared with their Ly6C⁻ counterparts, Ly6C⁺ cells expressed more CD122 (Fig. 5B), produced more IFN-γ, TNF-α and GzmB (Fig. 4), proliferated more quickly (Fig. 6A) and lysed target cells more strongly (Fig. 8B), which benefits control of virus during acute infection. By contrast, Ly6C⁻ cells displayed stronger survival potential (Fig. 6C) and more effective expansion upon secondary exposure to MHV-68 (Fig. 6B). These data demonstrate that the Ly6C⁺ population may contain more terminally differentiated effector cells, while the Ly6C⁻ population may contain the population with an increased potential for forming memory cells. Interestingly, Ly6C⁺ and Ly6C⁻ cells were not fixed in these cell fates, but could interconvert in vivo upon secondary exposure to the virus (Fig. 2B), showing that both populations maintained plasticity.

Our data relating to the different functional abilities of Ly6C⁻ and Ly6C⁺ cells lead to similar conclusions to those reported during acute LCMV infection (23). This is important because it shows very similar processes of CD4⁺ T cell differentiation take place in disparate virus infections. In comparison with data obtained from LCMV infection, the CD4 response was weaker, and reached the limit of detection by approximately 50 days post-infection, whereas LCMV-specific memory CD4⁺ T cells were detectable for more than 150 days. This was not due to the persistent nature of MHV-68 infection, as the same pattern was seen for both the persistent FS73R and non-persistent FS73 strains. Similarly the differentiation profile of the CD4⁺ T cell response did not appear to be affected by virus persistence, except for T_FH cell differentiation. One major difference between the current study and the LCMV study was the absence of a distinct PSGL1 negative population that was identified as T_FH cells. However, in this study, we detected T_FH populations based on CXCR5 and PD-1 expression (Fig. 3D). We noted there were more gp150-specific T_FH after FS73R infection compared to FS73 infection. This is consistent with a previous study in which FS73 mutant MHV-68 induced less T_FH among the total CD4⁺ T cell population compared with WT virus (60). As latent infection with MHV-68 drives germinal center B cell expansion, this likely also enhances the T_FH response, which in turn further supports germinal center expansion. Of note, it has also been shown in the LCMV infection model that there are more T_FH after chronic LCMV clone 13 infection compared to acute Armstrong infection (61). In this study, the authors showed that prolonged TCR signaling was essential for the T_FH differentiation during viral persistence. In addition, it has been shown that high-affinity TCR interactions, leading to stronger TCR signaling, are important components of T_FH differentiation (62). Therefore the persistence of viral antigen in the FS73R infected group may provide prolonged TCR signaling, leading to increased T_FH cells.

Memory T cell subsets can be identified based on cell surface markers, in particular, the expression of the lymph-node homing receptor CCR7. It has been described that CCR7⁺CD62L^hi T cells are T_CM cells, based on their ability to enter lymph nodes through high endothelial venules, while CCR7⁻CD62L^hi/lo CD4⁺ T cells are T_EM cells, as they cannot efficiently enter these sites (63). In this study, Ly6C⁻ gp150-specific CD4⁺ T cells at the memory time point had higher CCR7 expression. On the other hand, Ly6C⁺ CD4⁺ T cells contained more CD62L-expressing cells (Fig. 5C), but even naïve Ly6C⁺ CD4⁺ T cells had high expression levels of CD62L, making it problematic to use CD62L in combination with Ly6C to differentiate memory cell populations. The discordance between CD62L and CCR7 expression on memory CD4⁺ T cells has also been reported in LCMV and L. monocytogenes infection models (64). In the current study, at d55 pi, the frequencies of memory CD4⁺ T cells expressing CD62L were only 38% and 15% in the Ly6C⁺ and Ly6C⁻ populations, respectively. This is consistent with a previous study in which most MHV-68-specific CD4⁺ T cells have a CD62L^lo phenotype up to 14 month pi (65). This is a similar phenotype to the MHV-68-specific CD8⁺ T cell response, which is dominated by CD62L^lo effector memory cells (66).

We also evaluated the contribution of gp150-specific CD4⁺ T cells to virus control independent of CD8⁺ T cells and B cells. Protection could not be transferred to naïve mice with the small numbers of cells obtainable from infected gp150-specific CD4⁺ T cell adoptively transferred mice (data not shown), and therefore we could not evaluate the relative protective potential of Ly6C⁺ and Ly6C⁻ populations. Similar to other viral systems, MHV-68-specific CD4⁺ T cells have a lesser direct antiviral effect than CD8⁺ T cells, likely necessitating the transfer of large numbers of cells, which were not obtainable in our experimental system. However infection of RAG^−/− TCR transgenic mice showed gp150-specific CD4⁺ T cells had direct antiviral activity (Fig. 7). Although direct infection of RAG^/- TCR transgenic mice is not physiological, this provides proof-of-principle that gp150-specific CD4⁺ T cells can contribute to protection. Blockade of IFN-γ signaling completely abrogated the antiviral response, indicating the CD4⁺ T cells function through IFN-γ in vivo (Fig. 7). This finding was surprising in light of the fact that a relatively small population of these cells produced IFN-γ in response to peptide stimulation ex vivo (Fig. 4A, left panel). Additionally, gp150-specific CD4⁺ T cells produced GzmB (Fig. 4A) and displayed direct cytotoxicity against Ag-pulsed cells (Fig. 8A), therefore some gp150-specific CD4⁺ T cells differentiated into CTLs. While we did not assess the relative importance of IFN-γ and cytotoxicity to viral control in vivo, the beneficial effect of CD4⁺ T cells was completely abrogated after IFN-γ blockade (Fig. 7), suggesting CD4-mediated cytotoxity may play only a minor role, at least in mice of the C57BL/6 genetic background.

Currently it is not clear whether CD4⁺ CTLs represent a unique CD4⁺ T cell subset or are a part of the Th1 subset. Previous reports have shown populations of CD4⁺ cells to be directly cytolytic in MHV-68 infection (15, 29). Several studies have also shown that Th1-polarized clones have cytotoxic activity in influenza virus (67), poliovirus (68), West Nile virus (69), and EBV (70) infection models. Eomes is a critical master transcription factor for directing lytic effector differentiation of CD8⁺ T cells (71). Eomes expression renders CD4⁺ T cells cytotoxic by activating both perforin- and FasL-pathways (72). In this system, we did not observe a population of gp150-specific effector CD4⁺ T cells expressing more eomes than naïve CD4⁺ T cells (Fig. 3A). Therefore, further study is needed to clarify the molecular profile involved in the differentiation of CD4⁺ CTL.

In conclusion, phenotypically and functionally distinct subsets of effector and memory CD4⁺ T cells can arise from a clonal population specific for gp150. MHV-68 infection drives the majority of gp150-specific CD4⁺ T cells to differentiate into Th1 cells, and a minority into T_FH cells or Tregs. The CD4⁺ T cells have direct antiviral activity through provision of IFN-γ and direct cytotoxicity. Moreover, Ly6C expression on gp150-specific CD4⁺ T cells is correlated with a more terminally differentiated phenotype and stronger antiviral activity, but inversely associated with some memory cell hallmarks. Both Ly6C⁺ and Ly6C⁻ gp150-specific CD4⁺ T cells are able to interconvert upon secondary Ag exposure in vivo, which may be of benefit to maintain a balance between the two Ag-specific CD4⁺ T cell populations during MHV-68 infection. These findings have significant implications for Ly6C as a surface marker to distinguish functionally distinct CD4⁺ T cells and define CD4⁺ T cell terminal differentiation. CD4⁺ T cells are important for controlling herpesvirus infections. Understanding the phenotype, function and memory properties of CD4⁺ T cells is critical for the development of vaccines and novel antiviral therapies.

Abbreviations used in this article

MHV-68

murine γ-herpesvirus-68

Ly6C

lymphocyte antigen 6C

gp150-Tg mice

gp150-specific TCR transgenic mice

gp150-transferred mice

gp150-specific CD4⁺ T cell-transferred mice