The immunoregulatory role of CD244 in chronic hepatitis B infection and its inhibitory potential on virus-specific CD8+ T-cell function

Authors


  • Supported by the Kompetenznetz Hepatitis HepNet (Teilprojekt 10.2.1) (H.D. and M-C.J.), the Research Network “Natural Resistance to Infection” from the German Federal Ministry of Research (BMBF) (N.G., H.D., B.R., A.U.), and the German Research Foundation (DFG) (A.U., H.D.).

  • Potential conflict of interest: Nothing to report.

Abstract

Multiple inhibitory receptors may play a role in the weak or absent CD8+ T-cell response in chronic hepatitis B virus (HBV) infection. Yet few receptors have been characterized in detail and little is known about their complex regulation. In the present study, we investigated the role of the signaling lymphocyte activation molecule (SLAM)-related receptor CD244 and of programmed death 1 (PD-1) in HBV infection in 15 acutely and 66 chronically infected patients as well as 9 resolvers and 21 healthy controls. The expression of CD244, PD-1, and T-cell immunoglobulin domain and mucin domain 3 (TIM-3) was analyzed in virus-specific CD8+ T-cells derived from peripheral blood or liver using major histocompatibility complex class I pentamers targeting immunodominant epitopes of HBV, Epstein-Barr-virus (EBV), or influenza virus (Flu). In chronic HBV infection, virus-specific CD8+ T-cells expressed higher levels of CD244 both in the peripheral blood and liver in comparison to the acute phase of infection or following resolution. CD244 was expressed at similarly high levels in EBV infection, but was low on Flu-specific CD8+ T-cells. In chronic HBV infection, high-level CD244 expression coincided with an increased expression of PD-1. The inhibition of the CD244 signaling pathway by antibodies directed against either CD244 or its ligand CD48 resulted in an increased virus-specific proliferation and cytotoxicity as measured by the expression of CD107a, interferon-γ, and tumor necrosis factor-α in CD8+ T-cells. Conclusion: CD244 and PD-1 are highly coexpressed on virus-specific CD8+ T-cells in chronic HBV infection and blocking CD244 or its ligand CD48 may restore T-cell function independent of the PD-1 pathway. CD244 may thus be another potential target for immunotherapy in chronic viral infections. (HEPATOLOGY 2010)

CD244, also known as 2B4, was recently described as an inhibitory molecule in CD8+ T-cell exhaustion during chronic lymphocytic choriomeningitis virus infection.1 It belongs to the signaling lymphocyte activating molecule (SLAM)-related membrane receptor family and is predominantly expressed on natural killer (NK) cells and CD8+ T-cells. It is known as an activating molecule on CD8+ T-cells interacting with CD48 as a high affinity ligand.2 The immunoregulatory role of CD244 as an activating or inhibitory molecule depends on different factors: (1) the density of surface expression, with costimulatory qualities in the presence of low or moderate but inhibitory qualities with high expression; (2) the coexpression of additional inhibitory molecules; and (3) the presence of the intracellular adaptor-protein SLAM-associated protein (SAP).3-6 The upregulation of inhibitory receptors plays a central role in CD8+ T-cell dysfunction during chronic hepatitis B virus infection (HBV) and the in vitro blockade of programmed death-1 (PD-1) by programmed death ligand-1 (PD-L1) leads to the recovery of T-cell function with the enhancement of proliferation and cytokine release.7 T-cell restoration by blocking inhibitory molecules might have important implications as a novel therapeutic strategy against viral infections. However, the individual susceptibility to in vitro blockade of PD-1 in HBV infection shows a broad variability and remains heterogeneous, which might be explained by the hierarchic coregulation of multiple negative regulatory pathways. In the face of its inhibitory potential, we tested the expression of CD244 on virus-specific CD8+ T-cells in the peripheral blood and in liver tissue of chronically-infected HBV individuals and the results obtained in the peripheral blood were compared to acute and resolved infection. CD244 in chronic HBV was additionally compared to Epstein-Barr virus (EBV) and influenza virus (Flu) infection, two representative candidates for latently persisting and self-limiting infections. The effect of CD244 blockade was investigated with respect to the restoration of T-cell proliferation, cytotoxicity, and T helper 1 cytokine release in dysfunctional HBV-specific CD8+ T-cells.

Abbreviations

ALT, alanine aminotransferase; anti-HBc, antibody to hepatitis B core antigen; APC, allophycocyanin; CFSE, carboxyfluorescein succinimidyl ester; EBV, Epstein-Barr virus; FACS, fluorescence-activated cell sorting; FITC, fluorescein isothiocyanate; HBeAg, hepatitis B e antigen; HBsAg, hepatitis B surface antigen; HBV, hepatitis B virus; HCV, hepatitis C virus; HIV, human immunodeficiency virus; HLA, human leukocyte antigen; ICS, intracellular cytokine staining; IL, interleukin; INF, interferon; LIL, liver-infiltrating lymphocytes; MFI, mean fluorescence intensity; MHC, major histocompatibility complex; NK, natural killer; PBMC, peripheral blood mononuclear cell; PD-1, programmed death-1; PD-L1, programmed death ligand-1; PE, phycoerythrin; rhIL, recombinant human interleukin; RPMI, Roswell Park Memorial Institute medium; SAP, SLAM-associated protein; SLAM, signaling lymphocyte activation molecule; TIM-3, T-cell immunoglobulin domain and mucin domain 3; TNF-α, tumor necrosis factor α.

Materials and Methods

Study Subjects.

Peripheral blood was obtained from study subjects at the University Hospital Munich after institutional review board approval. All patients gave written informed consent. The protocol and the procedures of the study were conducted in conformity with the ethical guidelines of the Declaration of Helsinki. There were 66 patients with chronic HBV infection, 15 patients with acute infection, 9 resolvers, and 21 healthy individuals included in the study. All patients were human leukocyte antigen (HLA)-A*0201-positive and negative for hepatitis C virus (HCV)/human immunodeficiency virus (HIV)-1/2. Patients with chronic infection had been seropositive for hepatitis B surface antigen (HBsAg) and antibody to hepatitis B core antigen (anti-HBc), and had been seronegative for hepatitis B surface (HBs) antibodies. Data from the peripheral blood (n = 44) and liver tissue (n = 4) were collected from 48 chronic patients who had been never treated with nucleos(t)ide analogues or interferon (IFN)-α). These patients were characterized by: (1) HBV DNA: 1.0 × 106 copies/mL; (2) alanine aminotransferase (ALT): 48 U/L; (3) hepatitis B e antigen (HBeAg): positive (n = 7), negative (n = 32), not determined (n = 9); (4) age: 37 years; and (5) gender: female (n = 23), male (n = 25). The Ishak scoring system was used for histopathological grading: Ishak 1/4 (n = 3); Ishak 2/4 (n = 1). Eighteen chronically infected patients, who received nucleo(s)tide therapy, were characterized by: (1) HBV DNA: 3.5 × 103 copies/mL; (2) ALT: 53 U/L; (3) HBeAg: positive (n = 2), negative (n = 12), not determined (n = 4); (4) age: 43 years; and (5) gender: female (n = 3), male (n = 15). Acute infection was diagnosed by the following criteria: acute onset of hepatitis in previously healthy individuals, along with recent onset of jaundice, exclusive of metabolic or toxic causes; ALT at least 10-fold above the limit of normal; HBsAg-positive and anti-HBc immunoglobulin M (IgM)-positive: 80% of acute patients were enrolled within the first 4 weeks, 20% between weeks 4 and 8 after onset of clinical symptoms: (1) HBV DNA: 6.4 × 107 copies/mL; (2) ALT: 1663 U/L; (3) HBeAg: positive (n = 8), negative (n = 4), not determined (n = 3); (4) age: 36 years; and (5) gender: female (n = 1), male (n = 14).

Preparation of Peripheral Blood Mononuclear Cells and Liver Tissue.

Peripheral blood mononuclear cells (PBMCs) were isolated from whole blood as described.8 Liver-infiltrating lymphocytes (LIL) were isolated from liver biopsy, repeatedly washed in Roswell Park Memorial Institute (RPMI) medium and stained with major histocompatibility complex (MHC) class I pentamer for phenotypic analysis.

Synthetic Peptides.

HBV core peptide (c)18-27 and EBV peptide BMLF1 were synthesized by EMC Microcollections (Tübingen, Germany) and ProImmune (Oxford, UK).

Flow Cytometric Analysis.

Fluorescein isothiocyanate (FITC)-conjugated/phycoerythrin (PE)-conjugated CD244, allophycocyanin (APC) anti-PD-1, FITC anti–interleukin-2 (IL-2), PE anti-IFN-γ, PE-Cy7 anti–tumor necrosis factor α (TNF-α), and functional blocking antibodies anti-CD48, anti-PD-L1, anti-PD-L2, isotype control mouse IgG1, anti-CD3, and anti-CD28 were purchased from eBioscience (San Diego, CA). FITC anti-PD-1, FITC/PE-Cy7/APC-H7 anti-CD8, APC anti-CD107a, anti-CD38, anti-CD69, and anti-HLA-DR, peridinin chlorophyll protein complex (PerCP) anti-CD14, anti-CD19, anti-CD3, Via-Probe, and Monensin were purchased from BD Biosciences (San Jose, CA). APC anti–T-cell immunoglobulin domain and mucin domain 3 (TIM-3) was purchased from R&D Systems (Minneapolis, MN). Low endotoxin anti-CD244 (2B4, clone 2B4) was purchased from AbD Serotec (Oxford, UK).9 Micro-Beads for T-cell enrichment were purchased from Miltenyi Biotec (Bergisch-Gladbach, Germany). If the CD244 expression in chronic HBV exceeded 80%, CD244 expression was defined as CD244high.

MHC-Class I Pentamer Staining.

The following PE-labeled/APC-labeled HLA-A*0201-restricted MHC class I pentamers were used: HBV core (c)18-27 (FLPSD FFPSV), HBV envelope (e)183-191 (FLLTRILTI), HBV polymerase (p)573-581 (FLLSLGIHL), EBV BMLF1, and Flu Matrix 1. PBMCs (2 × 106) were incubated for 10 minutes at room temperature in culture medium (RPMI 1640, 2 mM glutamine, 1 mM sodium pyruvate, 5% human AB serum, 100 IU/mL penicillin, 100 μg/mL streptomycin). After wash step surface markers were added for 20 minutes at 4°C. Cells were then washed and incubated with anti-PE/anti-APC Micro-Beads for 15 minutes. After the wash step, 90% of cells were applied to MS columns (Miltenyi Biotec) according to the manufacturer's instructions. The other 10% were reserved for fluorescence-activated cell sorting (FACS) analysis. PE-positive/APC-positive cells were eluted from the column and analyzed by FACS. Cells were gated on the CD8+, CD14−, CD19−, and Via-Probe− population. Frequencies of Pent+ T-cells were calculated as described previously.10

Determination of IFN-γ Producing Cells by Elispot.

The 96-well culture plates were coated with IFN-γ antibody (Mabtech, Stockholm, Sweden). Before use, unbound antibodies were removed and blocked with RPMI containing 10% human AB serum. PBMCs (2.5 × 105) were incubated with HBV core peptide (10 μg/mL) for 48 hours at 37°C in the presence or absence of 10 μg/mL anti-CD244 or 5 μg/mL anti-CD48. Biotin-conjugated anti-IFN-γ was added after a wash step, followed by 2 hours of incubation. The unbound antibodies were washed and cells were incubated in detection solution. The number of spots was scored by an Elispot reader (AID, Straßberg, Germany). If the mean value plus two standard deviations (2SD) in healthy individuals was exceeded, the increase of virus-specific IFN-γ release after CD244 blockade was defined as positive.

CD69 Activation Marker Assay.

PBMCs (1 × 106) were incubated overnight at 37°C in culture medium in the presence or absence of HBV core peptide (10 μg/mL) and 2 μg/mL staphylococcal enterotoxin B (SEB) (Sigma-Aldrich, St. Louis, MO). After incubation, all cells were collected, washed, stained with surface markers, and analyzed by FACS. Patterns of CD69 coexpression were determined by gating on CD3+, CD8+ T-cells.

Carboxyfluorescein Succinimidyl Ester Assay.

Lymphocytes were labeled with 1 μM carboxyfluorescein succinimidyl ester (CFSE) (Molecular Probes, Eugene, OR) as described previously11 and cultured for 7 days in the presence of HBV core peptide (10 μg/mL) and anti-CD244 (10 μg/mL) or anti-PD-L1/2 (5 μg/2.5 μg/mL). Plain medium, isotype control, and phytohemagglutinin (PHA) (Biochrom, Berlin, Germany) (2.4 μg/mL) served as controls. At day 3, 20 IU/mL recombinant human interleukin 2 (rhIL-2) was added. At day 7, cells were collected, washed, stained with surface markers, and analyzed by FACS. The percentage of proliferating CD8+CFSElow T-cells was determined after gating on CD3+ T-cells.

In Vitro Expansion Assay.

In vitro expansion was performed with 2 × 106 PBMCs diluted in 1 mL culture medium. After preincubation with anti-CD244 (10 μg/mL), anti-CD48 (5 μg/mL), anti-CD244/CD48/anti-PD-L1/2 (5 μg/2.5 μg/mL) for 30 minutes at 37°C, cells were stimulated with HBV core or EBV posttranscriptional regulator protein of EBV containing a CD8 epitope (BMFL1) peptide (10 μg/mL). At day 7, 20 IU/mL rhIL-2 was added. Isotype control and healthy donors (n = 9) served as controls. At day 21, cells were collected, washed, stained with pentamers, and analyzed by FACS. If the mean value plus 2SD in healthy individuals was exceeded, the increase of virus-specific CD8+ T-cells was defined as positive.

Intracellular T helper 1 Cytokine Release (Intracellular Cytokine Staining) With CD107a Detection.

After in vitro expansion of 3 × 106 PBMCs with HBV core or EBV peptide in the presence or absence of anti-CD244, cells were re-stimulated with or without antigen (10 μg/mL) in the presence of Monensin (2μM) and anti-CD107a (15 μg/mL). Anti-CD3/CD28 (10 μg/mL/2 μg/mL) was used as positive control. After 6 hours of incubation, cells were collected, washed, stained with surface markers, fixed, and permeabilized. Cells were then stained with anti-IFN-γ, IL-2, and TNF-α for 20 minutes at room temperature and analyzed by FACS after a wash step. Cells were gated on CD3+, CD8+ T-cells. Background-corrected data were shown, with subtraction of individual costimulated control sample.

Statistical Analysis.

Data were shown as mean values. GraphPad Prism was used for analysis of the Mann-Whitney U test, Wilcoxon signed rank test, Fisher's exact test, and Spearman correlation test. P values of less than 0.05 were considered significant.

Results

CD244 Expression in HBV-Specific CD8+ Pent+T-cells of Peripheral Blood and Liver.

We first performed a comparative analysis of CD244 on total and HBV core (c)18-27–specific CD8+ T-cells. The characterization of CD244 distribution was done in the peripheral blood of 27 chronically infected patients, 13 acutely infected patients, 8 resolvers, and 15 healthy individuals. CD244 expression was also investigated in the liver tissue of four chronically infected patients. The mean frequency of CD8+Pentc18-27+ T-cells in chronically infected and untreated patients was 0.02%. Higher frequencies were detected in chronically infected patients under nucleo(s)tide therapy (0.04%), HBV resolvers (0.07%), and acutely infected patients (0.13%). Liver-derived Pent+ T-cell frequencies (n = 3) showed the highest rate (6.5%). No HBV-specific T-cell was detectable in healthy individuals (data not shown). CD244 expression in the peripheral blood was significantly higher on virus-specific CD8+ T-cells of chronically infected untreated patients (78%; mean fluorescence intensity [MFI]: 760) versus acutely infected patients (61%; MFI: 542) (P = 0.01 and P = 0.02, respectively) and patients who spontaneously cleared the virus (51%; MFI: 444) (P = 0.01 and P = 0.005, respectively) (Fig. 1A; Supporting Fig. 1). No difference in virus-specific CD244 expression was detectable in chronic untreated versus treated patients (80%; MFI: 675). CD244 was exclusively higher on peripheral CD8+Pentc18-27+ T-cells of chronically infected patients compared to total CD8+ T-cells but not in acute infection and resolvers (untreated: P = 0.0005; treated: P = 0.01) (Fig. 1A,B). Moreover, virus-specific CD244 expression was significantly higher in the liver (97%; MFI: 1232) compared to the peripheral blood (78%; MFI: 760) (P = 0.03 and P = 0.01, respectively) (Fig. 1A). Liver-derived total CD8+ T-cells (91.7%; MFI: 1117) expressed significantly higher amounts of CD244 compared to the peripheral blood of chronic infection (50%; MFI: 564) (P = 0.005 and 0.002, respectively) (Fig. 1B). No difference in CD244 expression was detected on liver-derived CD8+Pentc18-27+ T-cells (97%; MFI: 1232) in comparison to liver-derived total CD8+ T-cells (91.7%; MFI: 1117) (P = 0.2 and P = 0.6, respectively).

Figure 1.

Distribution of virus-specific (A) and total CD244 expression (B) in PBMCs (filled circles) from cHBV untreated and treated patients (n = 27), aHBV patients (n = 13), HBV resolvers (n = 8), and healthy controls (n = 15). Intrahepatic levels of CD244 expression of four cHBV patients are shown as filled diamonds. CD244 expression was investigated in response to different HBV antigens: HBV core (c)18-27 (left), HBV polymerase (p)573-581 (middle), and HBV envelope (e)183-191 (right). Asterisks mark the significant difference between total and CD8+Pent+ T-cells in cHBV patients. (C) Patterns of virus-specific CD244 in latently persisting EBV infection and self-limiting Flu infection. (D) Representative contour plots of peripheral and liver-derived virus-specific CD8+Pent+CD244+ T-cells in chronic HBV infection, resolved HBV infection, and on Flu-specific CD8+ T-cells. Example graphs show events after gating on CD8+, CD14−, CD19−, and Via-Probe− T-cell population. Black bars mark the mean values from different groups. P-values were calculated by using the Mann-Whitney U test. (aHBV: acutely infected; cHBV: chronically infected.)

CD244 Expression in Relation to Viral Load, HBeAg, and ALT Levels.

We next analyzed the correlation of CD244 to viral load, HBeAg, and ALT in chronically infected and untreated patients. No significant association was found between virus-specific CD244 expression and viral load (P = 0.8), HBeAg (P = 0.4), or ALT (P = 0.1) using linear regression analysis and Fisher's exact test (data not shown).

Patterns of CD244 Expression in CD8+ T-cells With Different Antigen Specificities.

We next investigated the CD244 expression in the peripheral blood of chronically infected (n = 6) and acutely infected patients (n = 6) in response to different HBV antigens. MHC class I pentamers targeting polymerase peptide (p)573-581 and envelope peptide (e)183-191 were used to phenotype virus-specific CD8+ T-cells. Chronic and acute infection were characterized by the following range of Pent+ T-cell frequencies: (1) p573-581; chronic: from 0.0045% to 0.13%; acute: from 0.001% to 0.29% and (2) e183-191; chronic: from 0.001% to 1.5%; acute: from 0.01% to 1%. Although CD244 on CD8+Pentp573-581+ T-cells (87%) of chronically infected patients was comparable to HBV core antigen (78%), lower levels on CD8+Pente183-191+ T-cells (47%) were detectable (P = 0.08) (Fig. 1A). HBV core and polymerase antigens showed significant higher levels of CD244 in chronic infection in comparison to acutely infected patients (P = 0.01, respectively), whereas CD244 on CD8+Pente183-191+ T-cells was similar in both groups (P = 1) (Fig. 1A).

Phenotypic Analysis of CD244 Expression in HBV versus EBV and Flu Infection.

We next analyzed the CD244 expression on EBV-specific and Flu-specific CD8+ T-cells in the same chronically infected HBV patient. Virus-specific CD244 in chronic HBV (78%; MFI: 760) was comparable to latently persisting EBV infection (n = 12) (83%; MFI: 614). Self-limiting Flu infection (n = 5) was characterized by significant lower levels of CD244 (18%; MFI: 225) compared to chronic HBV (percentage: P = 0.001; MFI: P = 0.001) and EBV infection (percentage: 0.001; MFI: 0.0007) (Fig. 1C). Representative FACS contour plots are given in Fig. 1D.

Longitudinal Analysis of CD244 Expression in Chronic and Acute HBV Infection.

To determine the CD244 expression in different phases of HBV infection, we longitudinally investigated acutely infected patients until resolution (n = 3) and chronically infected patients during nucleo(s)tide therapy (n = 3). CD244 expression in acute (Fig. 2A) and chronic infection (Fig. 2B) did not show significant changes in relation to: (1) clinical parameters (HBV DNA, ALT, HBeAg, HBsAg) and (2) immunological features such as CD8+Pentc18-27+ T-cell frequencies. We observed distinct variation in PD-1 and TIM-3 expression during the course of acute infection (Fig. 2A). Both molecules declined in all three acutely infected patients.

Figure 2.

Longitudinal analysis of CD244, PD-1, and TIM-3 expression on virus-specific CD8+ T-cells derived (A) from acutely infected HBV patients (gray background) following resolution (white background) and (B) from chronically infected HBV patients during nucleo(s)tide therapy (gray background). Virus-specific CD8+Pentc18-27+ T-cell frequencies (×) are shown at the bottom of the graphs. CD244 (filled circles), PD-1 (filled squares), and TIM-3 (filled triangles) expression are shown at the top of the graphs. Corresponding time course (day/week); HBV DNA (copies/mL) or ALT (U/L); HBeAg-status and HBsAg-status are listed below the different graphs of acutely and chronically infected HBV patients.

Coexpression of CD244 with CD38, CD69, and HLA-DR.

To determine the correlation of CD244 with the activation status of CD8+Pentc18-27+ T-cells in chronic HBV infection (n = 9), we co-stained CD244 with different activation markers such as CD38, CD69, and HLA-DR. Virus-specific CD244 showed low coexpression with CD38 (17%) and CD69 (12.5%) and modest coexpression with HLA-DR (31%) (Fig. 3A). Subsequently, we determined the induction of CD8+CD69+CD244+ T-cells after stimulation of chronically infected patients (n = 9) with HBV core antigen. CD244+CD8+ T-cells coexpressed lower levels of CD69 after antigenic stimulation (1.7%) compared to CD244-CD8+ T-cells (5.1%) (Fig. 3B). Representative FACS contour plots are shown (Fig. 3C).

Figure 3.

Antigen-specific coexpression of CD244 with different activation markers in chronic HBV infection (n = 9). (A) Direct ex vivo frequencies of CD244+CD38+, CD244+CD69+, and CD244+HLA-DR+ coexpression on CD8+Pentc18-27+ T-cells. (B) In vitro induction of CD244+CD69+ coexpression on CD8+ T-cells after antigenic and SEB stimulation. Gray bars represent the mean increase of CD244 and CD69 double-positive CD8+ T-cells including the SEM. (C) Example graphs show events after gating on CD8+, CD14−, CD19−, Via-Probe−, and Pent+ T-cell populations. Black bars mark the mean values from different groups. (SEB: staphylococcal enterotoxin B; SEM: standard error of the mean.)

Patterns of Virus-Specific Coexpression of CD244 and PD-1.

We next investigated the coexpression of CD244 and PD-1 in the peripheral blood of chronically infected and untreated HBV patients (n = 12), resolvers (n = 6), EBV infection (n = 8), and in the liver tissue of three chronic patients. Peripheral CD244/PD-1 was significantly higher on CD8+Pentc18-27+ T-cells of chronic infection (77%) compared to the total CD8+ T-cells (13.5%) (P = 0.0005) (data not shown). CD244/PD-1 was significantly higher coexpressed on liver-derived virus-specific CD8+ T-cells (96.3%) compared to the peripheral blood (77%) (P = 0.02) (Fig. 4A), whereas intrahepatic total CD8+ T-cells coexpressed lower amounts of CD244/PD-1 (70%) (data not shown). HBV resolution was significantly associated with low coexpression (33%) (P = 0.0009) (Fig. 4A). CD244/PD-1 was significantly lower on EBV-specific CD8+ T-cells (55.5%) compared to chronic HBV infection (P = 0.01), although the virus-specific expression of CD244 was similar in both viral diseases (Fig. 4A). Chronically infected patients highly expressed virus-specific PD-1 in the peripheral blood (85%) and liver (75%), whereas resolvers showed a significant decrease of PD-1 (53%; P = 0.005) (data not shown). Representative FACS contour plots are shown (Fig. 4B).

Figure 4.

Antigen-specific coexpression of CD244 with the inhibitory molecule PD-1. (A) High coexpression of CD244 and PD-1 on CD8+Pentc18-27+ T-cells in the peripheral blood (filled circles) and liver tissue (filled diamonds) of chronic HBV infection (n = 12) in comparison to peripheral EBV-specific CD8+ T-cells (n = 8) and HBV resolvers (n = 6). (B) Representative contour plots of CD244 and PD-1 coexpression on virus-specific CD8+ T-cells. Example graphs show events after gating on CD8+, CD14−, CD19−, Via-Probe−, and Pent+ T-cell populations. Black bars mark the mean values from different groups. P-values were calculated by using the Mann-Whitney U test.

Effect of CD244 Blockade on IFN-γ Release Using Elispot Assay.

Subsequently, we tested the effect of CD244 blockade on IFN-γ production in peripheral blood CD8+ T-cells of chronically infected HBV patients (n = 15) using Elispot assay. The inhibition of CD244 or CD48 significantly enhanced virus-specific IFN-γ production (P = 0.04 and P = 0.01, respectively) (Fig. 5A). Increased IFN-γ production by CD244 or CD48 blockade was restricted to CD244highCD8+ T-cells (P = 0.02) (Fig. 5B). Unspecific IFN-γ secretion was defined by stimulation of PBMCs: (1) from healthy donors with HBV core peptide in the presence of blocking antibodies (n = 11); and (2) from chronically infected patients with isotype control (n = 11). Samples of healthy controls (data not shown) and samples stimulated with isotype control (P = 0.1) did not show unspecific IFN-γ secretion (Fig. 5C).

Figure 5.

Effect of CD244 pathway blockade on HBV-specific IFN-γ secretion using the Elispot assay. (A) PBMCs of chronically infected HBV patients (n = 15) were incubated with the HBV core peptide in the presence or absence of anti-CD48 and anti-CD244. (B) High CD244 ex vivo expression on virus-specific CD8+ T-cells was associated with significant increase of IFN-γ secretion and T-cell expansion using Fisher's exact test. Dashed line marks the cutoff for CD244high and CD244low/intermediate expressing CD8+ T-cells. Black bars mark the mean values from different groups. (C) Gray bars represent the mean increase of HBV-specific IFN-γ production including the SEM after 48 hours of incubation with antigen and blocking antibodies as well as IgG1 isotype control. P-values were calculated by using the Wilcoxon signed rank test. (SEM: standard error of the mean.)

Effect of CD244 Blockade on IFN-γ, TNF-α, IL-2, and CD107a Production using ICS.

We next evaluated the effect of CD244 blockade on the restoration of IFN-γ, TNF-α, and IL-2 production and cytotoxicity in chronic HBV (n = 5), HBV resolvers (n = 3), acutely infected patients (n = 3), and healthy controls (n = 5). The effects on cytokine and CD107a production in chronically infected patients are shown (Fig. 6A). CD244 blockade did especially augment CD107a (3%), which rises 5.2-fold in four of five patients in comparison to antigen stimulation (0.58%) (P = 0.1) (Fig. 6B). CD244 blockade further enhanced virus-specific TNF-α production two-fold in five of five patients from 0.73% to 1.49% (P = 0.06) (Fig. 6C). The effect on IFN-γ release was modest (1.39-fold); however, there was a trend toward higher expression in five of five patients (P = 0.06) (Fig. 6A). CD244 blockade increased virus-specific IL-2 production in three of five patients from 0.01% to 0.055% (P = 0.3) (Fig. 6A). Because such low levels of IL-2 production were close to the detection limit of intracellular cytokine staining (ICS), IL-2 was excluded for further analysis.

Figure 6.

Effect of CD244 blockade on Th-1 cytokine release and cytotoxicity in chronic HBV infection using ICS. (A) Bar graphs show the increase of virus-specific IFN-γ, TNF-α, IL-2, and CD107a expression displayed as percentages of total CD8+ T-cells including the SEM. White bars represent mean values after antigenic stimulation. Gray bars represent mean values after CD244 blockade. Black points below the graph mark the evaluated combination of cytokines and CD107a mobilization. (B,C) The point-to-point increase of CD107a and TNF-α secretion after CD244 blockade (left). (D,E) The point-to-point increase of IFN-γ+TNF-α+ and CD107a+TNF-α+ CD8+ T-cells after CD244 blockade (left). Representative contour plots are shown on the right side. Example graphs show events after gating on CD3+CD8+ T-cells. P-values were calculated by using the Wilcoxon signed rank test. (SEM: standard error of the mean; Th-1: T helper 1.)

CD244 inhibition also induced effects on multifunctionality, as we detected a two-fold enhancement of IFN-γ+TNF-α+CD8+ T-cells from 0.63% to 1.26% in five of five patients (P = 0.06) (Fig. 6D) and a 1.86-fold increase of CD107a+TNF-α+CD8+ T-cells from 0.7% to 1.3% in five of five patients (P = 0.06) (Fig. 6E). There was a modest 1.4-fold increase of CD107a/IFN-γ expressing virus-specific CD8+ T-cells in four of five patients from 1.65% to 2.3% (P = 0.12) (Fig. 6F), and a 1.88-fold increase of CD107a+IFN-γ+TNF-α+ expression from 0.6% to 1.14% in four of five patients (P = 0.12) (Fig. 6F). No increase in cytokine production and CD107a expression was detectable in acute patients and resolvers stimulated with the HBV core peptide as well as healthy individuals stimulated with the EBV peptide, as shown (Supporting Fig. 2). No reaction was measurable in healthy controls stimulated with the HBV core peptide (data not shown).

Effect of CD244 Blockade on T-cell Expansion.

To assess the effect of CD244 on proliferation of dysfunctional CD8+ T-cells, we examined the changes in T-cell expansion of chronically infected (n = 12) and acutely infected HBV patients (n = 3), HBV resolvers (n = 3), and healthy controls (n = 9) using MHC class I pentamers. After antigenic in vitro stimulation of chronically infected patients, we observed a strong increase of CD8+Pentc18-27+ T-cell frequencies from day 0 (0.02%) to day 21 (0.56%). Blocking CD244 with anti-CD48 (Fig. 7A) or anti-CD244 (Fig. 7B) augmented virus-specific CD8+ T-cell frequencies in 5 of 12 (1.62-fold) (P = 0.9) or 5 of 10 (1.65-fold) (P = 0.6) chronically infected patients, respectively. The dual blockade of CD244 and CD48 increased the frequencies in five of six patients, which indicates a susceptibility of 83.3% (2.1-fold) (P = 0.09) (Fig. 7C). In comparison, blocking PD-1 by PD-L1/2 did enhance the CD8+Pentc18-27+ T-cell frequencies in six of eight patients 2.98-fold, which represents the most significant increase (P = 0.01) (Fig. 7D). Single or dual blockade of CD48 and CD244 significantly enhanced T-cell expansion by CD244high expressing CD8+ T-cells (P = 0.01) (Fig. 5B). No increase in T-cell expansion was detectable in acute patients and resolvers stimulated with the HBV core peptide as well as healthy individuals stimulated with the EBV peptide, as shown (Supporting Fig. 3). Unspecific background reaction was determined by: (1) stimulation of healthy donors with HBV core peptide in the presence of blocking antibodies (n = 8), and (2) stimulation of HBV patients with HBV core peptide in the presence of isotype control (n = 9). Samples of healthy controls and samples stimulated with isotype control did not show unspecific CD8+ T-cell proliferation (data not shown).

Figure 7.

Effect of CD244 and PD-L1/2 pathway blockade on T-cell expansion in chronic HBV infection. Increase of HBV-specific CD8+ T-cell frequencies from day 0 (white points) to day 21 (black points) of in vitro culturing displayed as point-to-point graphs after HBV core peptide stimulation alone, and (A) anti-CD48 blockade (n = 12), (B) anti-CD244 blockade (n = 10), (C) anti-CD244/CD48 dual blockade (n = 6), and (D) anti-PD L1/2 blockade (n = 6). All patients were first stimulated with antigen (10 μg/mL), 20 IU rhIL-2 was subsequently added on day 7. Representative contour plots are showing events after gating on CD8+, CD14−, CD19−, and Via-Probe− T-cell population. P-values were calculated by using the Wilcoxon signed rank test.

Effect of CD244 Blockade on T-cell Proliferation Measured by CFSE.

We confirmed the impact of CD244 blockade on the restoration of T-cell proliferation in chronically infected patients (n = 7) using CFSE (Fig. 8A). CD244 blockade led to a four-fold, significantly higher proliferation of virus-specific CD8+ T-cells (6.6%) in comparison to antigen stimulation (1.6%) (P = 0.01) (Fig. 8B). PD-L1/2 blockade augmented T-cell proliferation 2.8-fold from 1.6% to 4.5% (P = 0.03) (Fig. 8C), whereas isotype control (mean: 1.8%) did not induce T-cell proliferation (P = 0.8) (Fig. 8A). Representative FACS contour plots are shown (Fig. 8D).

Figure 8.

Antigen-specific CD8+ T-cell proliferation in chronic HBV infection after CD244 blockade measured by CFSE (n = 7). (A) The effect of CD244 inhibition on the frequencies of CD8+CFSElow T-cell fraction compared to HBV core peptide stimulation alone, IgG1 isotype control, PD-L1/2 blockade, and PHA stimulation is displayed showing bar graphs including the standard error of the mean (SEM). (B,C) Detailed point-to-point change of CD244 and PD-L1/2 inhibition with significant increase of proliferating virus-specific CD8+ T-cells. (D) Example graphs show events after gating on CD3+CD8+ T-cells. P-values were calculated by using the Wilcoxon signed rank test. (PHA: phytohemagglutinin.)

Discussion

CD244 plays a pivotal role in CD8+ T-cell regulation. Early studies demonstrated that CD244 acts as an activating receptor on NK-cells and T-cells in mice and humans. Cross-ligation enhanced lytic activity and IFN-γ secretion.12-15 CD244 is not only known as an activating molecule, as studies using a CD244-knockout mice model highlighted that CD244 can be an inhibitory receptor in NK-cells.4, 16 CD244 was recently described as an inhibitory molecule in chronic lymphocytic choriomeningitis virus infection with high expression on exhausted CD8+ T-cells and high coexpression with PD-1, lymphocyte activation gene-3 (LAG-3), and transcriptional repressor and B lymphocyte differentiation factor (Blimp-1).1, 17 However, its detailed function on CD8+ T-cells during chronic viral infection in humans has remained unknown. Therefore, we addressed CD244 as a potential inhibitory molecule in chronic HBV infection and analyzed its expression and functional influence on impaired virus-specific CD8+ T-cells.

Our results suggest that CD244 can act as an inhibitory receptor on HBV-specific CD8+ T-cells, which is supported by at least four important findings: First, CD244 was higher on virus-specific CD8+ T-cells in chronic HBV compared to total CD8+ T-cells, which could be interpreted as a specific up-regulation. Liver-derived virus-specific and total CD8+ T-cells expressed high amounts of CD244, indicating an up-regulation at the side of viral replication. Second, viral clearance was associated with low expression of CD244. Third, chronically infected HBV patients were characterized by high coexpression of CD244 with PD-1 in the peripheral blood and the liver. Fourth, CD244 blockade recovered T-cell proliferation, cytokine production, and cytotoxicity in chronic infection but not in acute patients, resolvers, and EBV infection. These observations indicate that CD244 contributes to T-cell dysfunction and can act in concert with other inhibitory molecules. In our study, chronically infected HBV patients are characterized by high levels of CD244 in comparison to acutely infected individuals and resolvers. In this context, Peritt et al.18 could show that CD244 expression increases in HIV patients with disease progression, which confirmed our observation of CD244 up-regulation during chronic HBV infection. Notably, CD244 was highly coexpressed with PD-1 in the peripheral blood and the liver of chronically infected patients but not with activation markers. These characteristics underline the inhibitory function of CD244 in HBV persistence and are in line with recently published data in mice.1, 17 A distinct up-regulation of inhibitory molecules such as PD-1 on intrahepatic CD8+ T-cells reflects the specific immunological features in the chronically infected liver as the side of viral replication.19 We therefore investigated CD244 on liver-derived virus-specific CD8+ T-cells and could show that CD244 was up-regulated on intrahepatic CD8+ T-cells with high PD-1 coexpression. Higher CD244 expression in the liver could be due to higher levels of antigen or a different cytokine milieu in the inflamed liver tissue.

The inhibitory function of CD244 was suggested to be mainly dependent on the receptor per cell amount and the presence of SAP.20 Thus, low or intermediate expression was suggested to deliver positive signaling, whereas high expression in the presence of low SAP mediates negative signaling.6 The data collected in our study support these findings and are consistent with an inhibitory function of CD244 on CD244highCD8+ T-cells. High CD244 expression in chronic HBV infection was associated with enhancement of: (1) T-cell proliferation measured by expansion of Pent+ T-cell frequencies in CD8+ T-cell lines and by defining proliferating CD8+CFSElow T-cells; (2) IFN-γ production measured by Elispot; and (3) cytokine release and T-cell degranulation using ICS. Patients with CD244 expression below 80% did not respond to the inhibition, with unaltered or reduced CD8+ T-cell expansion and Elispot IFN-γ secretion. These observations might be explained by the described costimulatory function of CD244low/intermediate-expressing CD8+ T-cells.1 The variability in CD8+ T-cell restoration after blockade of CD244 could be explained by: (1) the complex bidirectional interaction of CD244 and CD48, especially during long-term in vitro conditions as in the case of T-cell lines; (2) the presence of still unknown molecules, which possibly act as ligands of CD244 or vice versa; and (3) the undefined influence of CD244 expressing nonspecific CD8+ T-cells on HBV-specific CD8+ T-cells.

Antigen stimulation in the presence of rhIL-2 enhanced virus-specific CD8+ T-cell frequencies, which highlights the lack of CD4+ T-cell help as a key factor of CD8+ T-cell dysfunction.21 Differences in the response to the blockade of CD244 might be due to the presence of rhIL-2, which may reduce the inhibitory effect of CD244 as described for PD-1.22 Enose-Akahata et al.23 recently reported that rhIL-2 enhances SAP in CD8+ T-cells. High levels of SAP are known to be responsible for mediating costimulatory signaling through CD244, thus the addition of rhIL-2 could diminish CD244 inhibition.

Nevertheless, blockade of CD244 seems to be a promising approach to enhance T-cell proliferation, cytokine release, and cytotoxicity in dysfunctional CD8+ T-cells. Our CD244 blockade experiments suggest that there exist a hierarchical reconstitution. Although the blockade of CD244 and PD-l seemed to have comparable effects on T-cell proliferation, inhibition of CD244 especially augmented “effector” functions. This comparison of different inhibitory molecules was done to classify the role of CD244 in concert of hierarchical coregulation of multiple inhibitory pathways. Our data on CD8+ T-cell expansion after PD-L1/2 blockade are consistent with published data in chronic HCV and HIV.24, 25 However, the detailed coregulation of PD-1 and CD244 remains to be elucidated in further studies. Distinct “downstream” mechanisms could enhance the possibility of additive effects and mark a promising approach to achieve a better recovery of T-cell function than CD244 or PD-1 blockade alone.26-28

In summary, this is the first study that characterizes CD244 as an inhibitory receptor overexpressed on HBV-specific CD8+ T-cells in the peripheral blood and the liver of chronically infected patients. Further studies will be necessary to better define the complex patterns of CD244/CD48 interaction, the detailed contribution of CD244 to CD8+ T-cell dysfunction and the possible therapeutic potential of CD244 for the immunotherapy of chronic viral diseases.

Acknowledgements

We thank the patients, clinicians, and technicians involved in this study.

Ancillary