SLE patients were recruited from an ongoing prospective control program at the Department of Rheumatology, Skåne University Hospital, Lund, Sweden. Blood samples were taken at their regular visits. Healthy subjects, age-matched to the patients, were used as controls. An overview of clinical characteristics is presented in Tables 1 and 2. Disease activity was assessed using SLEDAI-2K 35. The following SLE treatments were used at the time point of blood sampling: hydroxychloroquine (n = 38), azathioprine (n = 17), mycophenolatmofetil (n = 11), rituximab (within the last 12 months, n = 5), methotrexate (n = 4), cyclosporine A (n = 3), cyclophosphamide (n = 2), chloroquine phosphate (n = 1) and intravenous immunoglobulins (n = 1). All patients fulfilled at least four American College of Rheumatology (ACR) 1982 criteria for SLE 36. The study was approved by the regional ethics board (LU 378-02). Informed consent was obtained from all participants.

The following antibodies and reagents were used in the flow cytometry analysis of the patients and the healthy volunteers: anti-CD3-Alexa 647, anti-CD4-APC-Cy7, anti-CD19-Pacific Blue, anti-CD14-PE-Cy7 (all from BioLegend, San Diego, CA, USA), anti-CD3-APC-Alexa Fluor 750, anti-CD8-PE-Cy7, anti-HLA-DR-Alexa Fluor 700, anti-CD20-PE, anti-CD38-PE-Cy5, anti-CD27-Alexa Fluor 700 (all from eBioscience, San Diego, CA, USA), propidium iodide, anti-IgD-FITC, anti-CD16-PE-Cy5, mouse IgG1-FITC (all from BD Biosciences Pharmingen, San Diego, CA, USA), anti-BDCA-1-biotin, anti-BDCA-2-PE (both from Miltenyi Biotec Inc., Auburn, CA, USA), anti-S100A8/A9-FITC (27E10, BMA Biomedicals, Rheinstrasse, Switzerland) and streptavidin Qdot-605 (Invitrogen, Carlsbad, CA, USA).

Blood was drawn into cell preparation tubes (BD Biosciences Pharmingen) and PBMCs were isolated on Lymphoprep™ according to manufacturer's instructions (Axis-Shield PoC AS, Oslo, Norway). PBMCs (1 × 10⁶ cells) were incubated with 10% mouse serum in phosphate buffered saline pH 7.2 (PBS) at a total volume of 50 μl for 20 minutes at 4°C. The cells were washed in PBS (200 g 2 minutes) and incubated with biotinylated antibodies for 20 minutes at 4°C. The cells were washed twice and then incubated for another 20 minutes at 4°C with dye-conjugated antibodies and streptavidin-Qdot 605. Finally, the cells were washed twice and resuspended in 300 μl PBS before analysis in the FACSAria (BD Biosciences Pharmingen).

T cells, monocytes and pDCs were isolated with negative isolation kits according to manufacturer's instructions (Miltenyi Biotec Inc). Neutrophils were isolated by density gradient centrifugation on Polymorphprep™ according to the manufacturer's protocol (Axis-Shield PoC AS). B cells and mDCs were isolated using a FACSAria cell sorter. B cells were labeled with Pacific Blue-conjugated CD19 antibodies and mDCs were labeled with both Alexa 700-conjugated HLA-DR antibodies and FITC-conjugated BDCA-1 antibodies. Purified cells (8 × 10⁴) were fixed with 4% paraformaldehyde and permeabilized in 0.2% TritonX-100 (Sigma, St. Louis, MO, USA) before incubated with PE-conjugated anti-S100A8/A9 antibodies (27E10, Santa Cruz Biotechnology, Santa Cruz, CA, USA) for 30 minutes. The cells were washed once in PBS and transferred to a glass slide (In Vitro, Braunschweig, Germany) by cytospin for two minutes at a speed of 1,000 g in a Shandon Cytospin 3 (Life Science International LTD, Cheshire, UK). The cells were analyzed in a LSM 510 META microscope (Carl Zeiss, Göttingen, Germany). Fluorescence was detected with pinhole settings corresponding to one airy unit.

Isolated pDCs (2 × 10⁴ cells) were cultured in 100 μl Macrophage-SFM (Invitrogen) supplemented with 20 mM HEPES (Invitrogen), 50 μg/ml Gentamicin (Invitrogen), 2 ng/ml GM-CSF (Leukine^®; Berlex, Montville, NJ, USA) and 500 U/ml IntronA (SP company, Innishannon, Ireland) and incubated for 20 h at 37°C with 5% CO₂ and 97% humidity with RNA-containing ICs prepared as described previously 37. Briefly, anti-ribonuclear protein (RNP)-positive sera were pooled and IgG was purified on a protein G column (Protein G Superose HR 10/2, Pharmacia LKB, Uppsala, Sweden). To create ICs, purified IgG at a concentration of 0.25 mg/ml was mixed with necrotic material from Jurkat cell supernatant at a concentration of 5% (v/v). The cells were washed and resuspended in PBS with anti-CD123-FITC (Miltenyi Biotech) and anti-S100A8/A9-PE (Santa Cruz Biotechnology) antibodies for 30 minutes at 4°C before analyzed by flow cytometry. The ICs used in the experiments contained undetectable amounts (< 40 ng/ml) of S100A8/A9 as measured by the in-house ELISA (data not shown). As a negative control, PE-conjugated mouse IgG1 antibodies were used.

For detection of S100A8/A9, microtitre plates (Maxisorp, Nunc, Roskilde, Denmark) were coated with monoclonal antibody MRP8/14 (27E10, BMA Biomedicals) at a concentration of 5 μg/ml, diluted in PBS at a volume of 100 μl/well, and incubated at 4°C over night. The 27E10 antibody detects a specific epitope of S100A8/A9 which is not exposed on the individual subunits. Between every following step, the plate was washed for three times in PBS containing 0.05% Tween 20. After blocking the plate with 1% BSA in PBS for 1 h, serum samples, diluted 1/100 in sample buffer (0.15 M NaCl, 10 mM HEPES (Invitrogen), 1 mM CaCl₂, 0.02 mM ZnCl₂, 0.05% Tween 20 and 0.1% BSA), were added at a final volume of 100 μl/well and incubated for 2 h at room temperature under agitation. Biotinylated anti-MRP8/14 (Abcam, Cambridge, UK), diluted 1/2000 in sample buffer, were added at a volume of 100 μl/well, and incubated at 4°C over night. Bound MRP8/14 antibody was detected with alkaline-phosphatase-labelled streptavidin (Dako, Glostrup, Denmark) diluted 1/1000 in sample buffer. After incubation for 1 h at room temperature under agitation, the enzymatic reaction was developed with 1 mg/ml disodium-p-nitrophenyl phosphate (Sigma) dissolved in 10% (w/v) dietanolamine pH 9.8 containing 50 mM MgCl₂ and the absorbance was measured at 405 nm. S100A8/A9 content of one serum sample was quantified using a commercial S100A8/A9 kit (BMA Biomedicals) and used as an internal control. The values reported are means of duplicates with the background subtracted and the concentrations were calculated from titration curves obtained from a pool of normal human serum.

Isolated PBMC from healthy donors were stained with anti-CD3-Alexa 647, anti-CD19-Pacific Blue, anti-CD14-PECy7, anti-CD16-FITC, anti-HLA-DR-Alexa 700, anti-BDCA-1-biotin and anti-BDCA-2-PE antibodies and sorted on a FACSAria before frozen at -80°C in lysis buffer. Total RNA was extracted by Purelink RNA mini Kit (Invitrogen, Carlsbad, CA, USA) and reversely transcribed to cDNA by SuperScript II Platinum synthesis system (Invitrogen) according to manufacturer's instructions. Ribosomal protein L4 (RPL4), S100A8 and S100A9 mRNA were quantified by real-time PCR using the SYBR GreenER kit (Invitrogen) in a MYIQ PCR machine (Bio-Rad, Hercules, CA, USA). The threshold cycle number and levels of each mRNA were determined using the formula 2(Rt-Et), where Rt is the threshold cycle for the housekeeping gene RPL-4 and Et the threshold cycle for the gene of interest.

Data were evaluated with analysis of variance (ANOVA) when comparing healthy controls with SLE patients or within the patient cohort when evaluating different disease manifestations. SAS version 9.2 for Windows XP (SAS Institute Inc., Cary, NC, USA) was used in the statistical evaluation. Correlations were calculated by Spearman rank correlation test without adjustment for multiple testing. All P-values were considered significant at P < 0.05.

First, we wanted to confirm abnormalities in PBMCs from SLE patients described by others to validate our patient material. Over all, SLE patients had markedly decreased cell density of PBMCs as compared to healthy controls (median SLE: 0.41 × 10⁶ (0.40 to 0.51) cells/ml and median healthy controls 1.00 × 10⁶ (0.86 to 1.12) cells/ml, P < 0.0001). A summary of all investigated leukocyte populations is shown in Table 3. SLE patients had more activated cells with increased HLA-DR expressing CD4⁺ T cells (P = 0.001), CD8⁺ T cells (P = 0.02), pro-inflammatory CD16⁺ monocytes (P = 0.003) and percentage of plasma cells (P < 0.0001), as compared to healthy controls. Altogether we have demonstrated that SLE patients have more activated leukocytes as compared to healthy controls and these results are in concordance with previous findings 35.

Detection of S100A8/A9 has previously been demonstrated on the surface of monocytes 32 as well as intracellularly in polymorphonuclear cells 28. We wanted to see if S100A8/A9 was also present on other cells since our main objective was to investigate possible aberrant expression in SLE. We could detect S100A8/A9 on naïve as well as pro-inflammatory monocytes, PMNs, B cells, myeloid dendritic cells (mDCs) and pDCs in both SLE patients and healthy controls (Figure 1). However, S100A8/A9 could not be detected on T cells where the mean fluorescence index (MFI) ratio indicated little or no cell surface S100A8/A9. Levels of the cell surface S100A8/A9 correlated well in samples from the same individual between the different cell populations (data not shown). Treatment with immunosuppressive drugs at the time of blood sampling was not statistically significantly associated with altered cell surface S100A8/A9. Cell surface S100A8/A9 was not significantly increased in SLE patients as a whole when compared to healthy controls on any cell population investigated (data not shown). We then investigated if cell surface S100A8/A9 was altered in patients with active disease (defined as SLEDAI ≥4) as compared to patients with inactive disease (SLEDAI < 4). Patients with active disease had increased cell surface S100A8/A9 on their CD16⁺ pro-inflammatory monocytes, pDCs, mDCs as well as PMNs as compared to SLE patients with inactive disease (P = 0.0005, P = 0.006, P = 0.03 and P = 0.015, respectively) and increased cell surface S100A8/A9 on their CD16⁺ pro-inflammatory monocytes and PMNs as compared to healthy controls (P = 0.0065 and P = 0.034, respectively, Figure 1). Thus we could demonstrate that cell surface S100A8/A9 was associated with disease activity and increased in patients with active disease.

Since S100A8/A9 was observed on most cell populations we wanted to know if this was due to expression of the S100A8 and S100A9 genes or due to deposition from external sources. To test the first possibility S100A8 and S100A9 mRNA levels were analyzed in FACS-sorted cells. Only low mRNA levels were found in T cells, B cells and mDCs, despite detection of cell surface S100A8/A9 on B cells and mDCs. However, clearly detectable levels of both S100A8 and S100A9 mRNA were found in monocytes, PMNs and pDCs (Figure 2A). These results confirm that S100A8/A9 is mainly produced by monocytes, PMNs and also by pDCs and the cell surface S100A8/A9 on other cell populations is most likely due to external deposition. We could confirm our flow cytometry data using confocal microscopy where we could detect membrane-associated S100A8/A9 on PMNs, monocytes and pDCs but not on T cells, mDCs or B cells (Figure 3). However, when using a non-confocal setting both mDCs and B cells had a weak S100A8/A9 staining whereas T cells still were negative (data not shown). In addition, we could detect S100A8/A9 with intracellular location in PMNs, monocytes and pDCs further supporting S100A8/A9 protein synthesis by these cells (Figure 3). Altogether, these data demonstrate that besides monocytes and PMNs, pDCs are also able to produce S100A8/A9.

If the presence of S100A8/A9 on the cell surface were due to external deposition, the level of cell surface S100A8/A9 might correlate with S100A8/A9 serum concentration. We found that S100A8/A9 serum concentrations were clearly increased in SLE as compared to healthy controls (P < 0.0001, Figure 2B), in accordance with previous published results 3334. Furthermore, we also found the most pronounced increased serum concentrations of S100A8/A9 in patients with arthritis (P = 0.016), as was previously reported 33, as well as in patients with kidney involvement (P = 0.026). There was also a statistically significant correlation between serum concentration of S100A8/A9 and SLEDAI (P < 0.0001, r = 0.49). However, serum S100A8/A9 level correlated only with cell surface S100A8/A9 in PMNs (P = 0.027, r = 0.28) and not cell surface S100A8/A9 levels on other leukocyte subpopulations. Thus, we could demonstrate that SLE patients had increased serum levels of S100A8/A9 which correlated to disease activity, but there were no strong correlations between S100A8/A9 cell surface levels and serum levels.

We also investigated whether it was possible to deposit S100A8/A9 on leukocytes in vitro. Neither recombinant S100A8/A9 nor serum containing high concentrations of S100A8/A9 (> 5,000 ng/ml) gave any increased surface staining of S100A8/A9 (data not shown). This might suggest that the S100A8/A9 binding ligands were already saturated and that other mechanisms could also be involved in the deposition of S100A8/A9 on the cell surface.

The flow cytometry data in combination with mRNA expression and confocal microscopy data strongly supported that monocytes, PMNs and also pDCs could produce S100A8/A9. Since the pDC is central in SLE pathogenesis and S100A8/A9 production is, to our knowledge, previously only described in monocytes and PMNs, we wanted to further investigate this subpopulation. When stimulating isolated pDCs with ICs in a serum-free medium the cell surface S100A8/A9 increased (Figure 2C) supporting that pDCs are able to synthesize S100A8/A9 and actively transport S100A8/A9 to the cell surface upon activation. Thus, pDCs could up-regulate cell surface S100A8/A9 in response to activation and we propose that pDCs are also able to synthesize S100A8/A9 proteins.