Individuals with systemic lupus erythematosus show evidence of a significant increase in monocyte apoptosis. This process is mediated, at least in part, by an autoreactive T cell subset that kills autologous monocytes in the absence of nominal Ag. We have investigated the apoptotic pathways involved in this T cell-mediated process. Expression of the apoptotic ligands TRAIL, TNF-like weak inducer of apoptosis (TWEAK), and Fas ligand on lupus T cells was determined, and the role of these molecules in the monocyte apoptotic response was examined. We report that these apoptotic ligands mediate the autologous monocyte death induced by lupus T cells and that this cytotoxicity is associated with increased expression of these molecules on activated T cells, rather than with an increased susceptibility of lupus monocytes to apoptosis induced by these ligands. These results define novel mechanisms that contribute to increased monocyte apoptosis characterizing patients with lupus. We propose that this mechanism could provide a source of potentially antigenic material for the autoimmune response and interfere with normal clearing mechanisms.
Altered expression of cell surface molecules involved in programmed cell death and increased apoptosis have been described in PBMCs from patients with systemic lupus erythematosus (SLE)3 (1, 2, 3, 4). Although most reports have focused on lymphocytes (1, 5, 6), our group (7) and more recently others have (8) reported increased apoptosis of monocytes/macrophages from patients with idiopathic lupus. Our group also identified an MHC-specific autoreactive lupus T cell subset that kills autologous monocytes without added Ag, providing a mechanism for the increased monocyte apoptosis (9). However, the pathways by which the autoreactive cells kill monocytes have not been fully characterized.
We recently reported that murine T cells induce apoptosis in syngeneic stimulating macrophages through the proapoptotic ligands TRAIL, TNF-like weak inducer of apoptosis (TWEAK), and Fas ligand (FasL; Ref. 10). We also found that the same molecules are expressed on lupus-inducing murine T cells made autoreactive with DNA methylation inhibitors (11), suggesting that these pathways could also be involved in the autoimmune response in this model. Whether these proapoptotic molecules are similarly involved in the T cell-mediated monocyte apoptosis that occurs in human lupus is unknown.
In this report, we investigated the pathways involved in autologous monocyte apoptosis induced by lupus T cells. Expression of TWEAK, TRAIL, and FasL on lupus T cells was determined, and the role of these molecules in the monocyte apoptotic response was examined. We report that multiple apoptotic pathways mediate the autologous monocyte death induced by lupus T cells and that this cytotoxicity is associated with increased expression of apoptotic ligands on activated T cells, rather than with increased susceptibility of lupus monocytes to apoptosis induced by these ligands. These results thus define novel mechanisms contributing to the increased monocyte apoptosis characterizing patients with lupus. We propose that this mechanism could provide a source of potentially antigenic material for the autoimmune response. Because cells from the monocytic/macrophage lineage are responsible for clearing apoptotic material, we also propose that this mechanism could contribute to decreased clearance of apoptotic debris in patients with lupus.
Materials and Methods
Abs and reagents
FN14: Fc (murine), a newly characterized receptor for the apoptotic ligand TWEAK (12), was a kind gift from Genentech (San Francisco, CA). Purified PHA was from Murex (Norcross, GA). The following reagents were all obtained from BD PharMingen (San Diego, CA) and, unless specified otherwise, the Abs were all mouse anti-human mAbs: FITC-annexin-V kit, FITC-anti-CD2, PE-anti-CD14, CYC-anti-CD3, purified anti-CD3, APC- and CYC-anti-CD4, purified anti-HLA-ABC, PE-anti-CD29, CYC- and FITC-anti-CD8, anti-human TRAIL/TNFSF10 mAb, biotin-anti-FasL, PE-anti-TRAIL, purified anti-TRAIL, APC-mouse IgG2a, PE mouse IgG1, FITC-goat anti-rabbit Ig-specific polyclonal Ab, FITC-goat anti-mouse Ig-specific polyclonal, streptavidin-CYC conjugate, affinity-purified polyclonal rabbit-human death receptor (DR) 4 (C-terminal) and affinity-purified polyclonal rabbit-human DR3 Abs, purified rat anti-human IL-4, purified mouse anti-human IFN-γ, purified rat anti-human IL-10, and purified mouse anti-human IFN-α. PE mouse anti-human CD19 was obtained from Coulter (Miami, FL). Recombinant human TRAIL R1/Fc chimera, recombinant human TRAIL and TRAIL inhibitor, and rabbit anti-human TRAIL R2/DR5 were obtained from R&D Systems (Minneapolis, MN). Purified mouse anti-human CD28 was from Southern Biotechnology Associates (Birmingham, AL). FITC-conjugated and PE-conjugated affinity-purified F(ab′)2 goat anti-mouse IgG (H + L) were obtained from Immunotech (Marseille, France). FITC-conjugated affinity-purified F(ab′)2 donkey anti-goat IgG (H + L), HRP-conjugated Affinipure F(ab′)2 donkey anti-goat IgG (H + L), purified goat anti-mouse-TWEAK, and recombinant human TWEAK were obtained from RDI (Flanders, NJ). FITC goat F(ab′2) anti-rabbit IgG (H + L) was from Caltag Laboratories (Burlingame, CA). Recombinant human FasL, recombinant human TRAIL kit (including TRAIL enhancer), and ApoSENSOR ADP/ATP Ratio Assay kit were obtained from Alexis Biochemicals (San Diego, CA). Pan T cell isolation kit, CD4 T cell isolation kit, monocyte isolation kit, and CD8 microbeads were obtained from Miltenyi Biotec (Auburn, CA). The pancaspase inhibitor Ac-VAD-CHO was from Biosource International (Camarillo, CA). Chloroquine, hydrocortisone, indomethacin, 6-mercaptopurine (6-MP), PMSF, sodium orthovanadate, leupeptin, and aprotinin were obtained from Sigma-Aldrich (St. Louis, MO). 51Cr was obtained from PerkinElmer (Boston, MA). Hoechst 33342 was from Molecular Probes (Eugene, OR). The CellTiter 96 aqueous one solution cell proliferation assay was from Promega (Madison, WI).
Patients with active SLE, inactive SLE, and rheumatoid arthritis (RA) were recruited from the outpatient Rheumatology clinic and inpatient services at the University of Michigan (Ann Arbor, MI), and from the Michigan Lupus Cohort (Ann Arbor, MI). Healthy controls were obtained by advertising. All SLE and RA patients fulfilled the American College of Rheumatology criteria for these diseases (13, 14). SLE activity was assessed by the SLE disease activity index (SLEDAI; Ref. 15), and those with SLEDAI >5 were considered to have active disease. RA activity was assessed by the modified disease activity score (16). Patient and control T cells were paired and studied in parallel. Information regarding the demographics, disease activity, and use of medications by the subjects is provided in Table I.
|Variable .||Lupus .||Control .||RA .|
|Age (mean± range)||37.6 ± 11.4||28 ± 10.4||45 ± 10.3|
|Females n (%)||47 (90.3)||19 (54.2)||9 (60)|
|Males n (%)||5 (9.7)||16 (45.8)||6 (40)|
|SLEDAI ≤ 5 n (%)||27 (52)|
|SLEDAI ≥ 5 n (%)||25 (48)|
|Prednisone (<0.5 mg/kg/day)||40||0||55|
|Prednisone (0.5–1 mg/kg/day)||21||0||25|
|Prednisone (>1 mg/kg/day)||3.8||0||2|
|Variable .||Lupus .||Control .||RA .|
|Age (mean± range)||37.6 ± 11.4||28 ± 10.4||45 ± 10.3|
|Females n (%)||47 (90.3)||19 (54.2)||9 (60)|
|Males n (%)||5 (9.7)||16 (45.8)||6 (40)|
|SLEDAI ≤ 5 n (%)||27 (52)|
|SLEDAI ≥ 5 n (%)||25 (48)|
|Prednisone (<0.5 mg/kg/day)||40||0||55|
|Prednisone (0.5–1 mg/kg/day)||21||0||25|
|Prednisone (>1 mg/kg/day)||3.8||0||2|
PBMCs and T cell isolation
PBMCs were isolated by Ficoll-Hypaque (Amersham Pharmacia Biotech, Uppsala, Sweden) gradient centrifugation as described (9). T cells were isolated by negative selection using magnetic beads and instructions provided by the manufacturer (pan T cell isolation kit; Miltenyi Biotec), or by rosetting with sheep erythrocytes (Colorado Serum, Denver, CO) as described (9).
Unstimulated and PHA-stimulated human T cells were cultured in 24-well plates in RPMI/5% FBS (100,000 cells/ml) in the presence or absence of graded concentrations (0.1–100 μM) of indomethacin, chloroquine, hydrocortisone, or 6-MP (17, 18). A stock solution of 6-MP was prepared in dimethylformamide at a concentration of 10 mg/ml. The stock solution was diluted in assay diluent (80% culture medium/20% ethanol) to yield a 6-MP working solution of 80 μg/ml or less as indicated. The working solution of 6-MP as well as the other materials prepared in assay diluent were then further diluted 1/40 into the cell cultures for the tests. The final concentrations of ethanol (0.5%) or dimethylformamide (<0.02%) do not yield significant effects on the cell cultures (19). Indomethacin was prepared in a concentration of 500 mM in absolute ethanol then diluted to final concentrations in the cell culture medium immediately before. Control cells were treated with an equal volume of the solvent; the highest concentration of ethanol in the cell culture medium was 0.1%.
PBMCs or purified T cells and monocytes (isolated with magnetic beads) were washed with standard buffer (PBS, 2% horse serum, 0.1% sodium azide) and incubated for 30 min at 4°C with 0.06–0.15 μg/ml fluorochrome-conjugated mAb following the manufacturer’s directions, washed three times again with standard buffer, then fixed in 1% paraformaldehyde and analyzed in a FACScan flow cytometer (BD Biosciences, Mountain View, CA) using previously described protocols (9). For identification of B cells and monocytes, fluorochrome-conjugated anti-CD19 and anti-CD14 were used, respectively.
Where indicated, purified T cells from seven healthy controls were stimulated using 96-well plates coated with 10 μg/ml human anti-CD3 and 1 μg/ml human-anti-CD28 in PBS. Cells were cultured at 37°C overnight, harvested then incubated with human FN14:Fc (anti-TWEAK), anti-TRAIL, anti-FasL-biotin, or control Ig for 30 min, washed and then incubated with goat anti-mouse-FITC, goat-anti-mouse PE, or streptavidin-APC, respectively, for an additional 30 min. Cells were washed and fixed in 1% formaldehyde and analyzed using flow cytometry using a FACScan flow cytometer.
Monocytes were purified by adherence to plastic petri dishes and eluted with cold PBS/EDTA (20). Monocyte yield was usually 10% of the starting number of PBMC. For further purification, negative selection with a magnetic bead monocyte isolation kit (Miltenyi Biotec) was used. Typical purity was ∼93–96% by CD14 staining.
Measurement of apoptosis
Annexin V staining.
Monocytes were isolated using negative selection with magnetic beads. Approximately 50,000 cells were suspended in 100 μl of 1× annexin V binding buffer (0.1 M HEPES/NaOH (pH 7.4), 1.4 M NaCl, 25 mM CaCl2). Cells were then stained with 5 μl of annexin V-FITC and 10 μl propidium iodide (PI; 0.05 μg/ml) for 15 min at room temperature in the dark. Cells were analyzed immediately using a FACScan flow cytometer. In additional experiments, monocyte apoptosis was measured in unfractionated PBMCs by labeling cells with anti-CD14-PE, washing in PBS, and then staining with annexin V/FITC and PI as above. To measure monocyte apoptosis induced by lupus T cells, lupus and control monocytes were isolated using negative selection with magnetic microbeads. Monocytes were cultured with autologous T cells using a ratio of 25:1 T cell:monocyte. Cells were cultured together for 18 h (9) in the presence or absence of increasing concentrations of blocking Abs or fusion proteins to TRAIL, TWEAK, or FasL. Cells were then stained with anti-CD2PE and anti-CD14 APC, washed, resuspended in 1× annexin V binding buffer, and stained with annexin V/FITC as described above. Cells were analyzed immediately using a FACScan flow cytometer.
The ApoSENSOR ADP/ATP Ratio Assay kit was used to measure ADP:ATP ratios. This kit uses bioluminescent detection of the ATP level via luciferase catalyzed reaction for a rapid screening of apoptosis and cell viability in mammalian cells (21). A total of 50,000 monocytes were added to 96-well plates in RPMI/10% FBS in the presence or absence of human recombinant FasL (Alexis Biochemicals), TRAIL soluble (human) recombinant fusion protein or TWEAK soluble (human) recombinant fusion protein (Research Diagnostics, Flanders, NJ), or isotype control (BD PharMingen). Where indicated, an affinity-purified rabbit anti-mouse Ig (Alexis Biochemicals) was added to optimize cross-linking. The cells were cultured for 18 h after which the medium was removed and cells were treated with 100 μl nuclear releasing agent as per manufacturer’s instructions for 5 min at room temperature with gentle shaking. ATP and ADP levels were measured, and ADP:ATP ratios calculated according to the manufacturer’s instructions.
Hoechst 33342 staining.
Monocytes from lupus patients and controls were isolated using negative selection with magnetic beads. A total of 50,000 cells were then cultured in the presence or absence of 1–100 ng/ml TRAIL soluble (human) recombinant fusion protein, human recombinant FasL, TWEAK soluble (human) recombinant fusion protein, or isotype control for 18 h in a total volume of 200 μl RPMI/3% FBS/penicillin/streptomycin. Cells were then washed, resuspended in 1 ml PBS, and incubated with 1 μl Hoechst 33342 stock solution and 10 μl PI. Cells were then incubated on ice for 20–30 min and analyzed immediately using a FACScan flow cytometer.
For direct cytotoxicity assays, 50,000 monocytes/well were incubated with FasL, TRAIL, TWEAK, or a control mAb (mouse IgG1 or IgG2; BD PharMingen) in a total volume of 200 μl/well. Following manufacturer’s instructions, a specific enhancer Ab was added for cross-linking after incubation with TRAIL and FasL recombinant proteins. These enhancers are Abs that interact with TRAIL and FasL recombinant proteins, increasing their activity. The enhancers were used at 2 μg/ml following the manufacturer’s instructions. As additional negative controls, cells were stimulated with enhancer without TRAIL or FasL. The cells were cultured for 18 h, then cell viability was determined by trypan blue exclusion. Quantitative analysis of apoptosis was done with annexin V/PI staining and/or ADP:ATP ratio calculation (see ADP:ATP ratio).
For cell-mediated cytotoxicity, 51Cr release assays were performed as previously described by our group (9). Briefly, 250,000 PBMC were adhered to round bottom microtiter wells in 100 μl RPMI/10% FBS for 1–2 h, after which the nonadherent cells were removed. The adherent cells were labeled with 2 μCi 51Cr/well for 3 h, and then washed. T cells were obtained by rosetting or negative selection using magnetic beads and added to the labeled monocytes at a ratio of 25:1. Cr release was determined 18 h later as described (9).
T cell proliferation assay
T cells from healthy controls or lupus patients were suspended in RPMI/5% FBS/penicillin/streptomycin media and treated with immobilized anti-CD3 (10 μg/ml) for 3 days in 96-well plates in the presence or absence of the following inhibitory Abs or fusion proteins: anti-human TRAIL, human Fn14:Fc (anti-TWEAK), anti-human FasL, and isotype control Ig at concentrations of 1–100 ng/ml. A total of 50,000 T cells/well were used, in a total volume of 200 μl RPMI/3% FBS/penicillin/streptomycin/well. At 72 h, T cell proliferation was measured using a colorimetric method (cell titer 96 Aqueous one solution cell proliferation assay; Promega) following manufacturer’s instructions (22). In brief, 20 μl of the CellTiter 96 solution reagent was pipetted into each well and the plate was incubated for 3 h at 37°C in a humidified, 5% CO2 atmosphere. Absorbance was measured at 490 nm using a 96-well plate reader (Bio-Tek Instruments, Winooski, VT).
The difference between means was analyzed using Student’s t test or ANOVA, using Systat 10 software (Systat, Evanston, IL) and Stata 6.0 (Stata Corporation, College Station, TX). To determine whether current treatment with immunosuppressives was associated with markers of apoptosis, univariate linear regression was performed. Each marker of apoptosis was modeled separately as a dependent variable, with medications modeled as dichotomous independent predictors. Control subjects were excluded from this portion of the analysis because a high degree of colinearity existed between case/control status and treatments due to the lack of immunosuppressive use in controls.
Monocytes from lupus patients undergo accelerated apoptosis
Monocytes from 16 lupus patients with active lupus (SLEDAI >5) and 8 controls were isolated by negative selection with magnetic beads, incubated with annexin V/FITC and PI, then analyzed by flow cytometry. Fig. 1,A shows a representative histogram. Patients with lupus had significantly greater percentages of apoptotic monocytes (annexin V+, PI−) relative to controls (Fig. 1,B). To exclude the possibility that the isolation procedure induced apoptosis in lupus monocytes, these findings were confirmed in unfractionated PBMCs staining with anti-CD14+ and annexin V/PI. Apoptotic monocytes in six controls represented 4 ± 2.5% of CD14+ cells. In six patients with SLE, 45 ± 15.2% of CD14+ cells were apoptotic, p < 0.01. We also confirmed that lupus T cells kill autologous monocytes and that this phenomenon correlates with disease activity. In addition, we found that this phenomenon appears to be disease-specific. Fig. 2 shows that T cells from patients with lupus, but not from RA patients or controls, spontaneously kill autologous monocytes (p < 0.0001, SLE vs RA and controls). In addition, T cells from patients with active SLE kill monocytes better than those from patients with inactive disease (p < 0.001, active SLE vs inactive SLE).
Lupus T cells show increased expression of apoptotic ligands
We then proceeded to investigate the pathways involved in T cell-mediated monocyte death in SLE. Because we had previously reported that the apoptotic ligands TRAIL, TWEAK, and FasL mediate T cell-induced cytotoxicity of Ag-specific macrophages in murine systems (10), we compared expression of these apoptotic ligands on T cells from 35 healthy controls, 52 patients with SLE, and 15 individuals with RA. Using flow cytometry, we found that SLE patients had significantly greater expression of TRAIL, TWEAK, and FasL on CD4+ and CD8+ T cells when compared with controls. Representative histograms for CD4+ T cells are shown in Fig. 3. Fig. 4,A shows that the percentage of CD4+ and CD8+ T cells expressing TRAIL, TWEAK, and FasL is significantly increased. When mean channel fluorescence (MCF) was analyzed, FasL was significantly increased on lupus CD4+ T cells, and TRAIL was significantly increased on lupus CD4+ and CD8+ T cells when compared with controls (Fig. 4 B). RA patients showed increased levels of TRAIL on CD4+ and CD8+ cells when compared with controls (control CD4 81 ± 20 vs RA CD4 475.6 ± 48, and control CD8 89 ± 12 vs RA CD8 362 ± 54 by MCF; mean ± SE; p < 0.01), but not of TWEAK or FasL. However, there were no differences in the percentage of cells expressing these ligands when compared with controls.
The findings were specific for T cells. There was no difference in expression of these apoptotic ligands on B cells from SLE patients when compared with controls (data not shown). Previous studies have found no consistent defect in the expression or function of the Fas receptor in SLE (3). Similarly, we observed no significant differences in the expression of DR3 (also known as TRAMP, LARD, WSL-1, or Apo-3), DR4 (also known as TRAIL-R1 or Apo-2), or DR5 (also known as TRAIL-R2 or Trick2) on monocytes from these same SLE patients when compared with controls (data not shown).
Association of ligand expression with disease activity and T cell activation
When percentage of expression was analyzed according to disease activity, patients with active lupus (SLEDAI >5) had greater expression of TRAIL and FasL on CD4+ and CD8+ T cells compared with individuals with inactive disease (SLEDAI ≤ 5; Fig. 4,C). When expression of apoptotic ligands was analyzed by MCF according to disease activity, lupus patients with more active disease had a higher expression of TRAIL and FasL, but not TWEAK, on CD4+ T cells than patients with inactive lupus (Fig. 4,D). To determine whether these molecules are overexpressed on activated T cells, two-color flow cytometry was performed using CD25 and CD29 as activation markers. CD29 (β1 integrin) is the β-chain associated with the integrin α subunits 1–6 known as very late Ags, and is expressed very late after T cell stimulation (23). The lupus patients had higher numbers of activated T cells than controls (for CD29, 3.8% ± 2.1 vs 32.5 ± 7.9, controls vs lupus, mean ± SEM, p < 0.01; for CD25, 1.2 ± 0.8 vs 10.8 ± 2.4 control vs lupus, mean ± SEM, p < 0.05). TRAIL, TWEAK, and FasL were increased on a subset of activated T cells, as measured by CD29+ T cells and CD25+ T cells, when compared with controls. Fig. 5 shows the average increase of the proapoptotic ligands on the activated cells. Our findings for FasL confirm previous observations by Kovacs et al. (24).
To determine whether T cell activation could contribute to the increased expression of these apoptotic ligands on lupus T cells, T cells from seven healthy control subjects were stimulated with immobilized anti-CD3 and anti-CD28 overnight. The percentage of T cells expressing TRAIL, TWEAK, and FasL expression increased after stimulation (for TRAIL, 6.6% ± 2.8 in unstimulated T cells vs 25.2% ± 13.8 in stimulated T cells; for TWEAK, 5.9% ± 3.4 in unstimulated T cells vs 27.03 ± 12.1 in stimulated T cells; for FasL, 1.4% ± 0.44 in unstimulated T cells vs 16% ± 1 in stimulated T cells; mean ± SEM, p < 0.05 for all variables). Representative histograms are shown in Fig. 6. These results suggested that the increase in expression of apoptotic ligands seen in activated lupus T cells is similar to what is seen in T cells from healthy controls after CD3/CD28 stimulation.
Effect of medications
To exclude the possibility that medications altered the expression of these apoptotic ligands, unstimulated and stimulated human T cells were cultured in the presence or absence of graded concentrations (0.1–100 μM) of medications that are commonly used in lupus treatment: indomethacin (for nonsteroidal anti-inflammatory drugs), chloroquine (for antimalarials), hydrocortisone (for steroids), and 6-MP (for azathioprine), or combinations, as previously reported (17, 18, 19). After culture for 1, 6, 12, and 24 h, TRAIL, TWEAK, and FasL expression was determined by flow cytometry and compared with untreated cells. None of these medications caused significant changes in the expression of these ligands, suggesting that medications do not account for the differences in expression (data not shown). To determine whether current treatment with immunosuppressives was associated with markers of apoptosis, univariate linear regression was performed. Each marker of apoptosis was modeled separately as a dependent variable, with medications modeled as dichotomous independent predictors. Control subjects were excluded from this portion of the analysis because a high degree of collinearity existed between case/control status and treatments due to the lack of immunosuppressive use in controls. There was no correlation of specific medications with proapoptotic molecule expression in these patients. In addition, there was no correlation between medications and SLEDAI, which also correlates with the ability of the T cells to kill monocytes (9).
Lupus monocytes have equal susceptibility to cell death induced by apoptotic ligands when compared with controls
To evaluate whether these apoptotic pathways are functional in monocytes from patients with SLE, and whether they differ from normal controls, we treated monocytes from five lupus patients and five controls with graded concentrations of TRAIL, TWEAK, or FasL recombinant proteins for 18 h. Using 51Cr release assays, we found a significant increase in cell death induced by these ligands (Fig. 7,A, p < 0.05 for the three apoptotic ligands at a concentration of 100 ng/ml, relative to control Ig). There was no statistically significant difference between lupus and control patient monocytes in their susceptibility to apoptosis secondary to recombinant protein stimulation. In addition, TRAIL cytotoxicity was inhibited by adding a specific TRAIL inhibitor (human recombinant TRAIL-R1:Fc), confirming specificity (25 ± 1.2% cytotoxicity for TRAIL agonist vs 1.8 ± 0.9% for TRAIL antagonist + agonist, mean ± SEM). The change in ADP:ATP ratios has been used to differentiate different modes of cell death and viability. Decreased levels of ATP and increased levels of ADP are recognized markers of apoptotic cells (25). Using ADP:ATP ratio assays, we confirmed that monocyte cell death after ligand stimulation demonstrates an apoptotic pattern (Fig. 7 B). This was also confirmed using Hoechst 33342:PI staining (data not shown).
Lupus T cells kill autologous monocytes through apoptotic pathways involving the ligands TRAIL, TWEAK, and FasL
We next asked if lupus T cells use these proapoptotic molecules to kill autologous monocytes. Monocytes from lupus patients and controls were radiolabeled with 51Cr and cultured with autologous T cells at a ratio of 25:1 T cells:monocytes in the presence or absence of graded concentrations of blocking Abs or fusion proteins to TRAIL, TWEAK, FasL, or a control Ig. A pancaspase inhibitor (Ac-VAD-CHO) was used as a positive control. Inhibiting these molecules partially blocked T cell-mediated macrophage killing (Fig. 8 A), with anti-FasL being the most potent (p < 0.005). Anti-TRAIL and anti-TWEAK also gave significant inhibition (p < 0.05 at 100 ng/ml). The combination of anti-FasL, TRAIL, and TWEAK was not more inhibitory than any of the Abs alone.
T cells from lupus patients were then fractionated into CD4+ and CD8+ subsets using magnetic beads and negative selection to prevent activation or inhibition of the cells through interactions with the Abs. Interestingly, cytotoxicity levels were greater using unfractionated CD4+ and CD8+ T cells than with either subset alone (58 ± 5% for unfractionated T cells vs 20.7 ± 5.5 for CD4+ T cells and 12.5 ± 3.5% for CD8+ T cells, p < 0.05, n = 16 for unfractionated T cells, and 8 for CD4+ and CD8+ T cells), suggesting a requirement for both subsets for optimal killing.
Because we had previously reported that inhibiting T cell activation with mAb to molecules such as LFA-1 or class II MHC also prevents APC killing (9), it was important to exclude the possibility that the anti-TWEAK and TRAIL reagents were inhibiting T cell activation. In addition, a recent study reported that TRAIL-expressing T cells that bind to DR might transmit a reverse signal and undergo increased proliferation (26). To exclude the possibility that blocking these apoptotic ligands could inhibit T cell proliferation/activation, we stimulated lupus T cells with anti-CD3 for 3 days, in the presence or absence of graded concentrations of blocking Abs or fusion protein to TRAIL, TWEAK, FasL, or a control Ig. We then measured T cell proliferation using a colorimetric method. As seen in Fig. 8 B, these molecules had no effect in inhibiting T cell proliferation, suggesting that an inhibition in T cell activation did not account for the inhibition of monocyte apoptosis by these blocking Abs.
Finally, the cytokines IL-10, IL-4, IFN-α, and IFN-γ have been reported to be involved in monocyte cytotoxicity (20, 27, 28, 29, 30). To exclude a role for soluble cytokines in inducing/blocking monocyte apoptosis in this model, we tested whether mAb to IL-4, IL-10, IFN-α, and IFN-γ could inhibit lupus T cell-mediated monocyte killing. No significant inhibition was observed using concentrations up to 100 ng/ml, as previously reported by our group for Ag-specific monocyte killing by human T cells (31) (maximum inhibition = 4.5 ± 2.2% for anti-IL-10, 5.5 ± 3.1% for anti-IL-4, 6% ± 3.4 for anti-IFN-γ, and 6% ± 4.2 for anti-IFN-α, mean ± SEM, p > 0.05 for all values when compared with isotype control). This suggests that the T cell-mediated cytotoxicity occurs largely through the proapoptotic ligands on the lupus T cells.
The present study describes the apoptotic ligands that mediate monocyte apoptosis induced by autologous lupus T cells. Our results indicate that patients with lupus show an increase in expression of TRAIL, TWEAK, and FasL in CD4+ and CD8+ T cells. Because our previously published studies show that anti-class II Abs can completely abrogate T cell-mediated autologous monocyte cytotoxicity and anti-class I was less effective (9), we propose that CD4+ T cells are required for the killing, although CD8+ T cells may play an additional role, since CD4+ T cells alone kill less efficiently than unfractionated T cells. The importance of a finely tuned cellular collaboration between CD4+ and CD8+ T cells for cytotoxicity has been described by other groups (32, 33, 34). We had previously shown that this phenomenon is specific for autologous monocytes, since SLE T cells do not kill allogeneic monocytes from normal or other lupus patients, and the killing was inhibited with mAbs to class II MHC determinants (9). The SLE patients had increased numbers of activated T cells relative to normal controls, and there was increased expression of these apoptotic ligands on the activated cells. We used CD29 as an activation marker since previous studies have reported that this marker is increased in lupus T cells (35), and it has been used by other groups to study expression of apoptosis markers on activated lupus PBMCs (3). Disease activity, measured by SLEDAI, also correlated with TRAIL, TWEAK, and FasL expression on CD4+ T cells but not with other T cell subset markers. These observations suggest that the increased expression of these molecules may be due to in vivo T cell activation. This conclusion is supported by our observation that expression of these apoptotic ligands increased following stimulation of T cells from healthy controls, as reported by others for TRAIL and FasL in human and murine T cells (10, 21, 36), and for TWEAK in murine T cells (10). Another group reported that TWEAK was not expressed on human T cells (20). The discrepancy may be related to differences in Abs used for stimulation or to variability in the degree of T cell stimulation achieved.
These studies also demonstrate that lupus T cells mediate autologous monocyte cytotoxicity using the same apoptotic ligands and receptors as those used by normal murine T cells to kill Ag-presenting macrophages. These results also indicate that TRAIL can induce cytotoxicity in a percentage of normal human monocytes, further supporting other reports that this molecule, although primarily toxic on transformed lines, has also an effect on nontransformed cells (10, 37, 38, 39, 40). Although a slight increase in cytotoxicity inhibition was observed when the blocking Abs were used in combination (data not shown), no clear synergism was found when multiple pathways were blocked. This suggests that the signals given by one of these ligands might be sufficient to inhibit apoptotic signals given by other members of the TNF family. Although FasL and TRAIL have been proposed to act in a cooperative way in inducing cytotoxicity (41), a lack of true synergism for these molecules has previously been reported in cancer lines (42). Our results confirm these findings. The apparent redundancy in apoptotic pathways involved in T cell-mediated monocyte cytotoxicity in normal states, as well as in lupus, may indicate that each of these molecules could have a different role in APC apoptosis that still remains to be characterized. It has been shown that other cell types, such as keratinocytes, can be target of multiple CTL responses that may have distinct roles in tissue injury (43). The residual monocyte death after blocking TRAIL, TWEAK, and FasL could be secondary to another apoptotic ligand or pathway as yet unidentified. Our group has previously shown that TNF-α was not involved in monocyte/macrophage cytotoxicity induced by Ag-specific CD4+ T cells (10, 31). Investigations on the role of other molecules including perforin and granzymes are currently underway, and could potentially explain the residual monocyte death induced by lupus T cells that was observed in this study.
Importantly, both control and lupus monocytes are equally susceptible to apoptosis mediated by these ligands. We found no significant difference in the susceptibility of these cells to apoptosis after treatment with TRAIL, TWEAK, or FasL recombinant proteins, suggesting that the increased susceptibility of lupus monocytes to undergo apoptosis is not associated with a decreased threshold for apoptotic signals.
Other mechanisms of monocyte death, such as complement toxicity, could contribute to increased monocyte apoptosis in lupus, as lupus sera are known to induce apoptosis of normal monocytes, possibly through complement-mediated damage (44). However, this phenomenon was observed only by a minority of lupus patients’ sera, suggesting that serum factors do not play a dominant role in the monocyte apoptosis seen in SLE patients. We considered the possibility that medications could contribute to the increase in the T cell apoptotic ligands. No effect on their expression was seen in short-term cultures. Although it is possible that this approach does not reflect chronic in vivo exposure, we also could not find any association between specific medications the lupus patients were taking and the expression of apoptotic ligands on their T cells. The monocyte apoptosis induced by autologous T cells seems to be specific for SLE, since this phenomenon was not observed in PBMC from patients with RA. This raises the possibility that the killing is mediated by a uniquely autoreactive subset present in lupus but not RA patients. Although RA T cells had higher levels of TRAIL than controls, there were no significant increases in the other apoptotic ligands, suggesting that the increase in this molecule is not sufficient to induce cytotoxicity. Multiple groups including our own have reported that T cells from patients with active lupus overexpress LFA-1 (9). We have previously reported that the T cells overexpressing LFA-1 mediate the monocyte killing in lupus (9). In other studies, we have shown that LFA-1 overexpression induces T cell autoreactivity and identical monocyte killing in vitro, and a lupus-like disease in vivo (11). The autoreactivity caused by LFA-1 overexpression is due in part to overstabilization of TCR-MHC interactions, permitting T cell activation by class II MHC molecules presenting inappropriate Ags (45). It is possible that LFA-1 overexpression in lupus T cells similarly overstabilizes TCR-MHC interactions in lupus patients, resulting in T cell activation by the monocytes with subsequent killing. An alternative interpretation is that the T cells are responding to as yet unidentified Ags presented by the lupus monocytes. However, the lack of a similar response in patients with RA, another autoimmune disease with T cell participation, argues for a different mechanism.
Our data demonstrate that monocyte apoptosis induced by lupus T cells is mediated by an increased expression of apoptotic ligands on autoreactive T cells, rather than to an increased sensitivity of SLE monocytes to these ligands. It is certainly possible that their overexpression could be due to T cell activation. Augmentation of monocyte apoptosis could account for an increase in the source of autoantigens and potentiate the immune response seen in SLE. An increase in the apoptosis of cells involved in phagocytosis, such as monocytes and macrophages, could also induce an increase in apoptotic cells by both augmenting apoptotic load and interfering with apoptotic clearance. In the absence of phagocytes, apoptotic cells eventually undergo necrosis (46, 47). Such necrosis may promote the maturation of dendritic cells that could then present the apoptotic material to T cells (48). It is known that dendritic cells, but not macrophages, efficiently cross present Ag derived from apoptotic cells to autologous cytotoxic T cells and induce class I-restricted CTLs after acquiring Ag from apoptotic cells (49). Interestingly, dendritic cells are known to engulf apoptotic monocytes (49). It has also been proposed that failure of efficient phagocytic clearance of cells dying by apoptosis results in tissue injury, and tissue protective mechanisms operating in interstitial fluid can be inactivated during apoptosis. The available evidence supports the contention that phagocyte clearance of intact cells dying by apoptosis protects tissues from inflammatory and coagulative injury initiated by cellular constituents (50). Finally, the combination of apoptotic monocytic lines cells and lupus IgG is known to be a potent IFN-α inducer (51), a molecule that has recently been shown to promote monocyte differentiation into dendritic cells in SLE (52), and to induce increased expression of TRAIL on T cells (53). These suggest additional mechanisms potentially contributing to the autoimmune response.
In conclusion, we have identified a novel mechanism by which lupus monocytes undergo accelerated apoptosis. This phenomenon could be involved in inducing an increased apoptotic load in lupus, which could provide an increased source of autoantigens. In addition, increased apoptosis of cells of the monocytic/macrophage lineage could potentially interfere with clearance of other apoptotic material, which in turn would further potentiate an autoimmune response. Having identified some of the molecules involved in this phenomenon indicates that potential therapeutic interventions could be designed.
We thank Drs. Leslie Crofford, David Fox, Gabriel Nuñez, and Stephen Weiss for very helpful discussions and critical review of the manuscript. We also thank Dr. Avi Ashkenazi and Genentech for kindly providing us with FN14:Fc, Steven Knowlton for expert secretarial assistance, and the physicians in the Division of Rheumatology (University of Michigan), in particular Drs. Brent Appleton, John Attwood, and Arturo Diaz, for their help in patient recruitment.
This work was supported by National Institutes of Health Grant 1P30-AR48310 and the Arthritis Foundation (to M.J.K.). B.C.R. was supported by Public Health Service Grants AG014783, AR42525, and AI42753, and a Merit grant from the Department of Veterans Affairs. E.A.S. was supported by Public Health Service Grant ES011196-02. This work was also supported by the Michigan Lupus Cohort at the University of Michigan. E.E.L. was a member of the Undergraduate Research Opportunity Program at the University of Michigan while these studies were performed, and was supported by a grant from the Lupus Foundation.
Abbreviations used in this paper: SLE, systemic lupus erythematosus; TWEAK, TNF-like weak inducer of apoptosis; FasL, Fas ligand; DR, death receptor; RA, rheumatoid arthritis; SLEDAI, SLE disease activity index; 6-MP, 6-mercaptopurine; PI, propidium iodide; MCF, mean channel fluorescence.