FcγRIIa Is Expressed on Natural IFN-α-Producing Cells (Plasmacytoid Dendritic Cells) and Is Required for the IFN-α Production Induced by Apoptotic Cells Combined with Lupus IgG

Preparations of IgG

IgG was prepared from plasma obtained by plasmapheresis of two SLE patients with active disease (24). Briefly, the first patient was a 16-year-old female with a disease duration of 2 years, and the second patient was a 66-year-old male with a disease duration of 3 years, both with an American College of Rheumatology index of 6. The IgG prepared from the patients was designated SLE-IgG 1 and SLE-IgG 2, respectively. The study protocol was approved by the Committee of Ethics, Faculty of Medicine, Uppsala University.

The citrated plasma was converted to serum by addition of 1 M CaCl₂ and IgG was prepared as described before (23). In short, the serum was passed through filters, 0.45 μm (Acrodisc; Gelman Sciences, Ann Arbor, MI), and endogenous DNA was eliminated by treatment with DNase I (2000 U/ml; Boehringer Mannheim, Mannheim, Germany) in 100 mM Tris-HCl (pH 7.5) and 10 mM MnCl₂ for 1 h at 37°C. The IgG was then purified on protein G-Sepharose (Amersham Pharmacia Biotech, Uppsala, Sweden). The eluted IgG was dialyzed against RPMI 1640 medium (ICN Biomedical, Costa Mesa, CA) supplemented with penicillin (60 μg/ml), streptomycin (100 μg/ml), l-glutamine (2 mM), and HEPES (20 mM) and used at a concentration of 1 mg/ml in the PBMC cultures.

Human Ig (Gammagard; Baxter, Deerfield, IL) was either used untreated or aggregated by heating in a water bath at 63°C for 60 min at a concentration of 50 mg/ml.

Preparation of Fab and F(ab′)₂

Purified SLE-IgG were treated with immobilized papain or pepsin (ImmunoPure Fab and F(ab′)₂ Preparation kits; Pierce, Rockford, IL) according to the manufacturer’s description. The Fc parts and whole IgG were removed using an immobilized protein A column and the samples were finally dialyzed in RPMI 1640 supplemented as described above. The efficiency of the enzymatic cleavage was verified by SDS-PAGE.

Culture conditions and treatment of U937 cells

Human monocytic U937 cells were cultured at 37°C in 7% CO₂ using RPMI 1640 medium supplemented as described above and with 5% FCS (Myoclone; Life Technologies, Paisley, U.K.). Mycoplasma could not be detected in the U937 cells by cultivation and staining with bis-benzimide (Hoechst no. 33258; Sigma-Aldrich, St. Louis, MO). The U937 cells were treated at a concentration of 1 × 10⁶ cells/ml by UV light (60 mJ, 254 nm) and cultured for 4 h before use as IFN-α inducers (see below). Apoptosis in U937 cells was regularly checked by staining with FITC-annexin V and propidium iodide as described previously (24).

Preparation of PBMC and IFN-α induction cultures

The PBMCs were prepared from normal blood donors by Ficoll-Hypaque (Amersham Biosciences, Uppsala, Sweden) density gradient centrifugation of buffy coats. The PBMC for flow cytometry were used immediately, otherwise cells were frozen and stored at −80°C in FCS containing 10% DMSO. The PBMC used for IFN-α induction cultures were from donors that had high IFN-α production in response to HSV and SLE sera. Frozen PBMC were thawed rapidly, washed twice, and cultured for 24 h at 5 × 10⁶ cells/ml in RPMI 1640 medium supplemented as above, with further addition of costimulatory IFN-α2b (500 U/ml; Schering-Plough, Bloomfield, NJ). All cultures were in triplicates using final volumes of 0.1 ml/well in 96-well round-bottom plates (Nunclone; Nunc, Roskilde, Denmark). As IFN-α inducers, HSV or the combination of UV-irradiated U937 cells (0.5 × 10⁶ cells/ml) and SLE-IgG (1 mg/ml) were used. The HSV was prepared and UV inactivated as described before (18) and used at a final concentration corresponding to 2 × 10⁷ PFU/ml.

The oligodeoxynucleotide (ODN) 2216 (36), with the sequence 5′- ggGGGACGATCGTCgggggG-3′ (Cybergene, Huddinge, Sweden), was used at a final concentration of 3 μg/ml. Small letters denote nucleotides with phosphorothioate backbone, capital letters denote nucleotides with phosphodiester backbone, and bold letters denote CpG dinucleotides.

For flow cytometric analysis of IFN-α-producing cells, PBMC were stimulated by HSV or by medium only (control) for 9 h in flat-bottom 24-well plates (Nunc) using 2-ml volumes and 2 × 10⁶ cells/ml. The culture medium was supplemented with IFN-α2b, as described above, and with 200 ng/ml GM-CSF (Leucomax; Schering-Plough). Brefeldin A (10 μg/ml; Sigma-Aldrich) was added 2 h before harvest of the cultures to increase intracellular concentrations of IFN-α.

Immunoassay for IFN-α

The concentration of IFN-α in culture medium was estimated by dissociation-enhanced lanthanide fluoroimmunoassay (DELFIA) with modifications as described before (37). In brief, microtiter plates were coated with the anti-IFN-α mAb LT27:293, detecting the majority of IFN-α subtypes but not the IFN-α2b. The europium-labeled LT27:297 anti-IFN-α mAb was incubated in the plates with samples or IFN-α standard for 1 h at 37°C. The standard was calibrated using the National Institutes of Health reference leukocyte IFN-α GA-23-902-530. The detection level was 2 U/ml.

Anti-FcR Abs

Abs specific for the FcγRIII/CD16 (clone 3G8), FcγRII/CD32 (clone IV.3), and FcγRI/CD64 (clone 32.2) were obtained from Medarex (Annandale, NJ). The anti-CD32 IV.3 mAb was also used as Fab fragment. The anti-CD32 mAb AT10 was purchased from BioSource International (Camarillo, CA). The Abs were dialyzed and diluted in RPMI 1640 medium supplemented as described above. The different Abs, at indicated final concentration, were added to the PBMC cultures immediately before the IFN-α inducers.

Flow cytometry

PBMC from eight healthy individuals were double stained using FITC-labeled anti-BDCA-2 mAb (7.5 μg/ml, IgG1, clone AC144; Miltenyi Biotec, Bergisch Gladbach, Germany) and PE-labeled anti-CD32 mAb (0.5 μg/ml, IgG2b, clone FLI8.26; BD PharMingen, San Diego, CA). The isotype controls were a FITC-labeled IgG1 (7.5 μg/ml, clone MOPC-31C; BD PharMingen) and PE-labeled IgG2b (0.5 μg/ml, clone DAK-GO9; DAKO, Glostrup, Denmark). The PBMC were coincubated with Abs to BDCA-2 and either anti-CD32 Abs or the IgG2b isotype control for 30 min on ice. The anti-BDCA-2-positive cells were defined by staining PBMC with its IgG1 isotype control. In addition, PBMC from four healthy individuals were double stained using unconjugated anti-CD32 (clone IV.3; Medarex) or the mouse IgG2b isotype control (DAKO), followed by PE-labeled rabbit anti-mouse IgG (DAKO). Subsequently, the cells were incubated with normal mouse serum followed by the addition of FITC-labeled anti-BDCA-2 mAb (Miltenyi Biotec).

Furthermore, PBMC from four normal blood donors were stimulated by HSV (see above) and stained by using FITC-labeled anti-CD32 mAb (0.5 μg/ml, IgG2b, clone FLI8.26; BD PharMingen) or FITC-labeled isotype control IgG2b (0.5 μg/ml, clone 27-35; BD PharMingen). In addition, HSV-stimulated PBMC from three donors were stained with the unconjugated anti-CD32 mAb IV.3, (Medarex) or the mouse IgG2b isotype control (DAKO), followed by blocking with normal mouse serum, and addition of FITC-labeled goat anti-mouse IgG (DAKO). The cells were subsequently fixed in 2% paraformaldehyde for 30 min and then stained for intracellular IFN-α using biotin-labeled LT27:295 mAb (0.5 μg/ml; produced in our laboratory) and PE-labeled streptavidin, as described before (29). The isotype control for the IFN-α staining was biotinylated IgG1 (0.5 μg/ml, clone DAK-GO1; DAKO). The cells were acquired and analyzed using a FACScan flow cytometer and CellQuest software (BD Biosciences, San Jose, CA).

FACS purification of BDCA-2-positive cells

NIPC/PDC were purified on the basis of their expression of the markers BDCA-2 and BDCA-4 (32, 38). PBMC were first enriched for BDCA-4-positive cells using the BDCA-4 cell isolation kit (Miltenyi Biotec) according to the manufacturer’s description. The BDCA-4-enriched PBMC were then stained with FITC-conjugated anti-BDCA-2 mAb (7.5 μg/ml; Miltenyi Biotec) and subsequently sorted by flow cytometry according to the forward light scatter characteristics and the BDCA-2 expression using a FACStar^Plus (BD Biosciences). Both BDCA-2-positive and -negative cells were sorted and collected using PBS as sheath fluid. All samples were handled and stored on ice until further processing.

Analysis of CD32 expression by RT-PCR

The mRNA was prepared from 1 × 10⁵ purified BDCA-2-positive cells or PBMC with 20 μl of oligo(dT) beads per cDNA reaction using a Dynabeads mRNA Direct micro kit (Dynal, Oslo, Norway) according to the manufacturer’s instructions. Reverse transcription was performed using Superscript II (Life Technologies, Paisley, U.K.) in a total volume of 40 μl following the manufacturer’s instructions. The cDNA from each cell population was stored at −20°C until used in PCR. To analyze the three different forms (a,b,c) of FcγRII, the following primer pairs were used: FcγRIIa, 5′ primer: 5′-CAGCATGGGCAGCTCTTC-3′ and 3′ primer: 5′-CACATGGCATAACGTTAC-3′ (39); FcγRIIb, 5′ primer: 5′-GGAATCCTGTCATTCTTACCTGTC-3′ and 3′ primer: 5′-CCCAACTTTGTCAGCCTCATC-3′ (40); and FcγRIIc, 5′ primer: 5′-CTCCCAGCTCTTCACCGA-3′ and 3′ primer: 5′-CACATGGCATAACGTTAC-3′ (39). Note that the same downstream primer is used to detect FcγRIIa and c. The amplification efficiency of primer pairs was verified by using serial dilutions of plasmids encoding human FcγRIIa, FcγRIIb2 (kindly provided by J. van de Winkel, Utrecht University, The Netherlands), or FcγRIIc (kindly provided by P. Morel, Pittsburgh Cancer Institute, Pittsburgh, PA) as a template. Primers used to amplify β-actin were: 5′ primer, 5′-GATTCCTATGTGGGCGAGGAG-3′ and 3′ primer, 5′-GAGGGCATACCCCTCGTAGATG-3′. Each PCR was conducted in a total volume of 15 μl with 1–2 μl of cDNA on beads, 6 pmol of each primer, 0.25 mM of each dNTP, 2 mM MgCl₂, 2 U AmpliTaq DNA polymerase, and 1× PCR buffer (Roche Molecular Systems, Branchburg, NJ). The following cycles were used: An initial denaturing step of 5 min at 94°C, followed by 35–40 cycles with denaturation for 30 s at 94°C, annealing for 30 s at 59°C (β-actin), 56.4°C (FcγRIIa), 65.5°C (FcγRIIb), 56.6°C (FcγRIIc), elongation for 30 s (β-actin and FcγRIIa,c) or 50 s (FcγRIIb) at 72°C, and a final extension for 7 min at 72°C. The expected size of PCR products are 362 bp for β-actin, 352 bp for FcγRIIa, 852 bp for FcγRIIb1, 795 bp for FcγRIIb2, 347 bp for FcγRIIc1, 361 bp for FcγRIIc2, and 309 bp for FcγRIIc3 and FcγRIIc4. PCR products were visualized with ethidium bromide in agarose gel electrophoresis.

The Fc portion of SLE-IgG is required for IFN-α induction in PBMC induced by apoptotic U937 cells and SLE-IgG

We have previously shown that the combination of SLE-IgG and apoptotic U937 cells induces IFN-α production in NIPC/PDC (24). To elucidate the role of the Fc portion of the Abs in the IFN-α response, Fab fragments and F(ab′)₂ were prepared by papain and pepsin treatment of SLE-IgG. Replacing the SLE-IgG with Fab fragment or F(ab′)₂ from the same SLE patient almost abolished the IFN-α production when PBMC were cultured with apoptotic cells as an IFN-α inducer (Fig. 1⇓A). Thus, the Fc portion of SLE-IgG seems to be of major importance for IFN-α production induced by IC consisting of SLE Abs and apoptotic material. No inhibitory effects of the Fab fragments or F(ab′)₂ were seen on the IFN-α production induced by HSV in PBMC cultures (Fig. 1⇓B).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Intact SLE-IgG, but not Fab and F(ab′)₂, combined with apoptotic U937 cells can induce IFN-α production by PBMC in vitro. The PBMC were cultured with UV-treated U937 cells, along with intact SLE-IgG (▪) or either Fab (▦) or F(ab′) ₂ (□) prepared by papain or pepsin treatment, respectively (A). As a control, the Fab (▦) and F(ab′) ₂ (□) or medium (▪) were added to cultures stimulated by HSV (B). The Fab, F(ab′)₂, and the intact SLE-IgG were used at a concentration of 0.25 mg/ml. After 24 h, the IFN-α levels (units per milliliter) in the culture medium were measured by DELFIA. Representative results from one of three experiments.

Heat-aggregated or untreated normal IgG inhibits the IFN-α production induced by apoptotic U937 cells and SLE-IgG or by CpG ODN

Because the Fc portion of the SLE-IgG appeared to be crucial for the IFN-α induction, we investigated whether multimeric IgG made by heat aggregation of IgG from normal individuals could interfere with the IFN-α response induced by UV-treated U937 cells and SLE-IgG. Heat-aggregated IgG complexes have previously been described to have both positive and negative effects on cytokine production (41, 42).

Our results (Fig. 2⇓A) show that heat-aggregated IgG caused a clear dose-dependent decrease in the IFN-α production induced by the combination of apoptotic U937 cells and SLE-IgG in cultures of PBMC, prominent even at very low concentrations. Also the untreated IgG caused an inhibition of the IFN-α production, but at higher concentrations. In addition, heat-aggregated, DNase-treated SLE-IgG inhibited the IFN-α production to the same extent as the heat-aggregated commercial IgG and did not induce IFN-α in itself (results not shown). Furthermore, both heat-aggregated and untreated IgG also inhibited the IFN-α production induced by ODN 2216 (Fig. 2⇓B), known to be a potent type I IFN inducer selectively activating natural IFN-α-producing cells (NIPC)/plasmacytoid DC (PDC) (28). The IFN-α production induced by HSV was also decreased by heat-aggregated and untreated IgG (results not shown). These results suggest that FcR binding can inhibit the IFN-α production stimulated by principally different IFN-α inducers.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Normal IgG inhibits the IFN-α production by PBMC induced by the combination of UV-treated U937 cells and SLE-IgG or by ODN 2216. The PBMC were cultured with SLE-IgG (1 mg/ml) and UV-treated U937 cells (A) or with ODN 2216 (B) in the presence of indicated concentrations of either heat-aggregated (•) or untreated (○) normal human IgG. After 24 h, the IFN-α levels in the culture medium were measured by DELFIA. Representative results from one of three experiments.

Abs to FcγRII inhibit IFN-α production induced by apoptotic U937 cells and SLE-IgG

The inhibition of the IFN-α production by heat-aggregated IgG could be due to blocking of FcR or activation of inhibitory FcR, such as FcγRIIb. We therefore examined whether mAbs to different FcR were inhibitory in this experimental system. Initially, we used mAbs to FcγRIII (CD16), FcγRII (CD32), and FcγRI (CD64) that were tested at different concentrations in the PBMC cultures. As seen in Fig. 3⇓A, the IFN-α production was clearly reduced when the IV.3 mAb to CD32 or the Fab fragment of this mAb were added to PBMC cultures with apoptotic U937 cells and SLE-IgG as an IFN-α inducer. In contrast, the mAbs to CD16 or CD64 did not alter the IFN-α production (Fig. 3⇓A), nor did the control mAb to CD19 (results not shown). Importantly, none of these anti-FcR mAbs inhibited the IFN-α production induced by HSV or by ODN 2216 in PBMC (Fig. 3⇓, B and C).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Abs to FcγRII inhibit the IFN-α production by PBMC induced by the combination of UV-treated U937 cells and SLE-IgG. The PBMC were cultured together with the combination SLE-IgG and UV-treated U937 cells (A), HSV (B), or ODN 2216 (C), and the indicated concentrations of mAbs directed against FcγRI (CD64; mAb 32.2), FcγRIIa,c (CD32; mAb IV.3 or IV.3-Fab), FcγRIIa,b,c (CD32; mAb AT10), or FcγRIII (CD16; mAb 3G8). The results with the anti-FcγRIIa,b,c mAb AT10 were obtained in a separate experiment. After 24 h, the IFN-α levels (units per milliliter) in the culture medium were measured in the supernatants by DELFIA. Representative results from one of three experiments.

The anti-CD32 Ab IV.3 binds to FcγRIIa and c, but not to b. We therefore also examined the effects of the anti-CD32 Ab clone AT10, which binds to all three forms of FcγRII (a, b, and c). We found that the AT10 Ab inhibited the IFN-α production induced by apoptotic U937 cells and SLE-IgG, but unlike the IV.3 Ab, also inhibited the IFN-α response induced by HSV (Fig. 3⇑B) and ODN 2216 (Fig. 3⇑C). These results indicate that the FcγRII is of importance for the IFN-α induction by the combination of apoptotic cells and SLE-IgG, but because the AT10 mAb inhibited the IFN-α production induced by HSV and ODN 2216, the FcγRII may also mediate a general inhibition of the IFN-α response, perhaps via FcγRIIb.

FcγRII/CD32 is expressed on NIPC/PDC

Findings in previous studies suggested that NIPC/PDC did not express surface FcγRII/CD32 (29, 32). This, along with our results above, i.e., that anti-FcγRII Abs or heat-aggregated IgG had inhibitory effects on the IFN-α production by NIPC/PDC, motivated a re-examination of the FcγRII/CD32 expression on these cells. This was conducted by flow cytometry using staining of PBMC with an Ab to the NIPC/PDC marker BDCA-2 (32, 38) and with the anti-CD32 mAb FLI8.26 that binds to all three forms of FcγRII (a, b, and c). We found that 0.52 ± 0.16% of PBMC were BDCA-2 positive and that 49.7 ± 16.3% of the BDCA-2-positive cells expressed FcγRII/CD32, although at relatively low levels (Fig. 4⇓A). Among the eight PBMC donors, a large individual variation was noted in the proportion of PBMC positive for both BDCA-2 and FcγRII/CD32, which varied between 0.09 and 0.45% (mean, 0.26%).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Expression of FcγRII on BDCA-2-positive PBMC and IFN-α-producing cells. A, Unstimulated PBMC were double stained with FITC-labeled anti-BDCA-2 Abs and PE-labeled FLI8.26 anti-CD32 Abs (upper dot plot) using a PE-labeled IgG2b isotype control (lower dot plot). B, The PBMC were also stimulated by HSV for 9 h in vitro, stained with FITC-labeled anti-CD32 mAb (upper dot plot) or FITC-labeled isotype control IgG2b (lower dot plot) and after permeablization with biotin-labeled anti-IFN-α mAb and PE-labeled streptavidin to detect intracellular IFN-α. Flow cytometric analysis was performed and indicated rectangular gates were used to define BDCA-2-positive PBMC (i.e., NIPC/PDC) that were either considered negative or positive for CD32. The percentage figures in the gates represent the proportion of cells in each gate of all analyzed PBMC. Representative results from the analysis of a total of eight (A) or four (B) different PBMC donors.

To examine whether the actual IFN-α-producing cells express FcγRII/CD32, PBMC from four different donors were stimulated by HSV and stained for intracellular IFN-α and FcγRII/CD32. The results show that most of the IFN-α-producing cells were stained for FcγRII/CD32 (Fig. 4⇑B). The staining was weak but seen as a clear shift of anti-CD32-stained cells compared with cells stained with the IgG2b isotype control.

Similar results regarding the CD32 expression were obtained when fresh PBMC or HSV-induced PBMC were stained using the IV.3 mAb instead of the FLI8.26 mAb (results not shown).

Expression of FcγRIIa, but not FcγRIIb,c in NIPC/PDC

Because we could demonstrate an expression of FcγRII on the NIPC/PDC by flow cytometry, we proceeded by determining the presence of the different types of FcγRII, i.e., the activatory FcγRIIa and FcγRIIc as well as the inhibitory FcγRIIb. This was done by enrichment for BDCA-4-positive cells by magnetic separation, followed by purification of BDCA-2-positive cells by FACS, resulting in almost 100% pure BDCA-2⁺ cells (results not shown). Subsequently, PCR was performed on cDNA prepared from these BDCA-2-positive cells and from unsorted PBMC to detect mRNA expression of the three types of FcγRII. A clear expression of FcγRIIa mRNA, but no FcγRIIb or FcγRIIc mRNA, was detected in NIPC/PDC (Fig. 5⇓A). In contrast, all three types of FcγRII mRNA were readily detected in unsorted PBMC. The expression of β-actin confirmed mRNA integrity in all samples tested. All PCR products had the expected molecular size in agarose gel electrophoresis (see Materials and Methods), as judged by comparison to molecular mass markers ranging from 100 to 1000 bp (results not shown). A similar amplification efficiency of the different primer pairs was verified by amplifying serial dilutions of plasmids encoding FcγRIIa, FcγRIIb2, or FcγRIIc (Fig. 5⇓B). The higher intensity of the FcγRIIb band is most likely explained by the larger size of the PCR product.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

Analysis of FcγRIIa, FcγRIIb, and FcγRIIc gene expression in purified NIPC/PDC. The PBMC from six donors were enriched for BDCA-4-positive cells by magnetic bead separation, subsequently stained with FITC-conjugated anti-BDCA-2 Abs, and purified by FACS. RT-PCR was performed on mRNA prepared from the BDCA-2-positive cells (first lane) and unsorted PBMC (second lane) using specific primer pairs for FcγRIIa, FcγRIIb, FcγRIIc, and β-actin. Note that the primer pair for FcγRIIb detected FcγRIIb1 and FcγRIIb2 in unsorted PBMC, whereas the primer pairs for FcγRIIa or FcγRIIc only resulted in one band, respectively (A). The efficiency of the primers detecting the different isoforms of FcγRII was verified using as templates serial dilutions of plasmids encoding each subtype (B). For RT-PCR and FACS details, see Materials and Methods.