The use of small molecule high-throughput screening to identify inhibitors of the proteinase 3-NB1 interaction

Abstract

Anti-neutrophil cytoplasmic antibodies (ANCA) to proteinase 3 (PR3) are found in patients with small-vessel vasculitis. PR3-ANCA bind strongly to membrane PR3 (mPR3) that is presented by the NB1 receptor. We performed high-throughput screening using a small molecule library to identify compounds that inhibit PR3-NB1 binding. We established a human embryonic kidney (HEK293) cell-based system, where approximately 95 +/- 2% of the NB1-transfected cells expressed the NB1 receptor on the cell surface. Addition of 0.1 microg/ml human PR3 to 10(4) NB1-expressing HEK293 cells resulted in PR3 binding that was detected by immunofluorescence using a fluorescence plate reader assay. We identified 13 of 20 000 molecules that inhibited PR3 binding by >70%. Seven of 13 substances showed reproducible inhibition in four additional validation experiments. Two selected compounds (27519 and 27549) demonstrated a dose-dependent inhibition over a range from 6.25 to 100 microM as measured by the plate reader assay. We used flow cytometry as a second assay, and found that both compounds reproducibly inhibited PR3 binding to NB1-transfected HEK293 cells at 50 microM (inhibition to 42 +/- 4% with compound 27519 and to 47 +/- 6% with compound 27549 compared to the dimethylsulphoxide control). Furthermore, compounds 27519 and 27549 also inhibited binding of exogenous PR3 to human neutrophils. In contrast, the compounds did not decrease mPR3 expression on resting neutrophils, but reduced the tumour necrosis factor-alpha-mediated mPR3 increase on NB1(pos) neutrophils when present continuously during the assay. The findings suggest that small inhibitory compounds provide a potential therapeutic tool to reduce mPR3 by preventing its binding to NB1.

Fig. 1

Time–course of the NB1 transfection efficiency in human embryonic kidney (HEK293) cells; (a) 6 × 10⁵ HEK293 cells were seeded into six-well plates and were transfected with 4 µg/ml NB1 using Fugene HD transfection reagent on the following day. Membrane-expressed NB1 (mNB1) expression was evaluated from days 2 to 9 by flow cytometry (n = 3). (b) A typical fluorescence activated cell sorter (FACS) flow experiment from day 2 using NB1-transfected HEK293 cells is depicted. Staining with the isotype control is shown in grey and with the anti-NB1 monoclonal antibody (mAb) with the bold line. The dotted line represents the β-Gal-transfected control stained with the anti-NB1 mAb and is basically identical with the isotype control in NB1-transfected cells.

Fig. 2

Small compounds inhibit proteinase 3 (PR3) binding to NB1-transfected human embryonic kidney (HEK293) cells; (a) 1 × 10⁴ NB1-transfected HEK293 cells were seeded into 384-well plates and five validated compounds were tested at 50 µM for their inhibitory effect on PR3 binding to NB1 using the fluorescence plate reader (n = 5). The horizontal line shows the mean inhibition for each compound. PR3 binding to NB1-expressing HEK293 cells preincubated with dimethylsulphoxide (DMSO) control was set at 100%; (b) 1 × 10⁴ NB1-transfected HEK293 cells were incubated in 384-well plates with increasing concentrations of compounds 27519 and 27549. The non-inhibitory compound 39679 (control) and DMSO were used as controls. Exogenous PR3 (0.1 µg/ml) was added thereafter for 2 h and PR3 membrane expression was assessed in a fluorescence plate reader (n = 2).

Fig. 3

Compounds 27519 and 27549 inhibit binding of exogenous proteinase 3 (PR3) to NB1-transfected human embryonic kidney (HEK293) cells by flow cytometry. (a) The structure of the inhibitory compounds 27519 and 27549 and the control compound 39679 is depicted; (b) 1 × 10⁵ NB1-transfected HEK293 cells were incubated with 50 µM of compounds 27519 and 27549 for 30 min before the addition of 1 µg/ml PR3 for 2 h. The compound 39679 (control) served as a non-inhibitory control and dimethylsulphoxide was used as the buffer control. PR3 membrane expression was assessed by flow cytometry using the monoclonal CLB anti-PR3 antibody (n = 5). (c) A typical experiment is depicted with the isotype control (thin dotted line) and PR3 staining after treatment of cells with PR3 in the presence of the control compound (thick line) or in the presence of compound 27519 (thin line). (d) Experiments were carried out as in (b) except that a different monoclonal 81.3.3 anti-PR3 antibody was used for detection of PR3 membrane staining (n = 3). **P < 0·01.

Fig. 4

Effect of compounds 27519 and 27549 on exogenous proteinase 3 (PR3) binding to neutrophils; (a) 1 × 10⁶ neutrophils were treated for 30 min with buffer, 50 µM compound 27519, compound 27549 or the control compound 39679 (control), respectively. Cells were then incubated with 1 µg PR3 for 90 min. Membrane PR3 expression was assessed by flow cytometry using the CLB anti-PR3 monoclonal antibody (n = 5). (b) A typical flow experiment is depicted for compound 27519. **P < 0·01.

Fig. 5

Effect of compound 27519 on membrane proteinase 3 (mPR3) expression in resting and tumour necrosis factor (TNF)-α-stimulated neutrophils; (a) 1 × 10⁶ neutrophils were incubated for 30 min with buffer, 100 µM compound 27519 or the control 39679 compound (control), respectively. Cells were then treated with buffer or 2 ng/ml TNF-α. Membrane PR3 expression was assessed by flow cytometry using the CLB monoclonal antibody to PR3 (n = 8). (b) Cells were treated as in (a), but compound was re-added during all staining steps (n = 7). **P < 0·01.