Meprin Metalloproteases Generate Biologically Active Soluble Interleukin-6 Receptor to Induce Trans-Signaling

Abstract

Soluble Interleukin-6 receptor (sIL-6R) mediated trans-signaling is an important pro-inflammatory stimulus associated with pathological conditions, such as arthritis, neurodegeneration and inflammatory bowel disease. The sIL-6R is generated proteolytically from its membrane bound form and A Disintegrin And Metalloprotease (ADAM) 10 and 17 were shown to perform ectodomain shedding of the receptor in vitro and in vivo. However, under certain conditions not all sIL-6R could be assigned to ADAM10/17 activity. Here, we demonstrate that the IL-6R is a shedding substrate of soluble meprin α and membrane bound meprin β, resulting in bioactive sIL-6R that is capable of inducing IL-6 trans-signaling. We determined cleavage within the N-terminal part of the IL-6R stalk region, distinct from the cleavage site reported for ADAM10/17. Interestingly, meprin β can be shed from the cell surface by ADAM10/17 and the observation that soluble meprin β is not capable of shedding the IL-6R suggests a regulatory mechanism towards trans-signaling. Additionally, we observed a significant negative correlation of meprin β expression and IL-6R levels on human granulocytes, providing evidence for in vivo function of this proteolytic interaction.

Figure 1. Meprin α and meprin β cleave a peptide of the IL-6R stalk region.

(A) Schematic representation of the IL-6 receptor (IL-6R) in its membrane bound form. Indicated are the cleavage sites that were identified by Müllberg and colleagues and the one found for ADAM17 in a peptide cleavage assay. (B) ICE logo representing the cleavage specificity of meprin α and (C) of meprin β both with a preference for negatively charged amino acids around the scissile bond. (D) Peptide cleavage assay analyzing part of the stalk region of the IL-6R comprising the previously identified cleavage sites as indicated in (A). HPLC and MALDI-TOF analyses revealed the cleavage site for meprin α and (E) for meprin β. (F) Summary of the cleavage sites identified for meprin α and meprin β. Note that both proteases cleave at the position (Q/D).

Figure 2. IL-6R cleavage by meprins generates soluble receptor independent of ADAM10/17.

(A) Human IL-6R was overexpressed in ADAM10/17 dKO HEK cells. Forty-eight hours post transfection recombinant meprin α was applied to the cell culture. After 2 hours the supernatant was harvested and ConA precipitated. A clear increase in cleaved soluble IL-6R fragment of about 50 kDa was detected with increasing concentration of meprin α. Additionally, the full-length receptor was detected in the supernatant. (B) The transfected IL-6R was also cleaved by a rather low concentration of soluble meprin α (5 nM) in a time dependent manner. Again a probable full-length version of the IL-6R was detected. (C) For soluble meprin β no cleavage product of the IL-6R was detected in the cell supernatant. However, a signal for the IL-6R with a molecular weight of the full-length receptor was observed. (D) To further address secretion of the full-length IL-6R a C-terminal antibody (CT-AB) was used, detecting the cytoplasmic part. Here, the full-length version of the IL-6R was clearly detected in cell culture supernatants at approximately 100 kDA but not the cleavage product generated by meprin α. (E) Co-transfection of the IL-6R with meprin α or meprin β revealed cleavage products generated by both proteases. As a control, recombinant meprins were added. Only meprin α and not soluble meprin β produces sIL-6R. (F) Summarizing model indicating that only soluble meprin α membrane bound meprin β shed the IL-6R. Full-length blots are shown in Supplementary Information.

Figure 3. Determination of the meprin cleavage sites in the IL-6R.

(A) Sequences of two QD (QD > AA and QD > QR) motif mutants that were anaylzed for meprin cleavage in ADAM10/17 dKO HEK. (B) IL-6R mutants QD > AA and QD > QR were co-transfected with meprin β. Surprisingly, sIL-6R was found in the supernatant 48 hours post-transfection, indicating that cleavage by meprin β does not occur between QD. (C) Structural analysis of the IL-6R (blue) fitted into the active site of membrane bound meprin β (orange/grey). It appears that type I transmembrane proteins, such as the IL-6R, require at least 20–25 amino acids in the stalk region to gain access to the active site of meprin β. The stalk of the IL-6R must form a loop to insert from the top of the active site cleft in cis-orientation that enables cleavage. Note the numerous negatively charged aspartate and glutamate residues that are present in the stalk (insert). (D) Amino acid sequences of four additional mutants that were analyzed, each with diverse parts of the stalk region absent. Note the N-glycan side chain that is present at Asn350. (E) Immunoblot analysis of the two IL-6R mutants Δ333–342 and Δ343–352 co-transfected with meprin β. (F) Immunoblot analysis of the two IL-6R mutants Δ317–352, missing most of the stalk region, and Δ353–362 missing the QD motive. For the Δ317–352 there was no cleavage fragment detectable after co-transfection with meprin β, whereas sIL-6R was seen in the supernatant of the Δ353–362 mutant upon co-expression with meprin β. This further indicates that the QD motif is not cleaved by meprin β. (G) As ADAM17 most likely cleaves in the region of the QD motif we co-expressed meprin β or ADAM17 together with the wild-type IL-6R and compared the molecular weights of the cleavage products. For meprin β a cleavage fragment was found at about 50 kDa, whereas ADAM17 generated a larger product of about 70 kDa. Thus it is highly likely that meprin β cleaves N-terminal and ADAM17 C-terminal of the glycan side chain, which accounts for the large difference in molecular weight. Full-length blots are shown in Supplementary Information.

Figure 4. Functional aspects of meprin mediated IL-6R shedding.

(A) To measure the biological activity of sIL-6R, we separated the meprin generated sIL-6R from the full-length form employing ultra-centrifugation. This resulted in a clean sample that only contains the sIL-6R. (B) Ba/F3-gp130 cell proliferation assay. The sIL-6R generated by meprin α revealed biological activity as indicated by the increase in cell proliferation in the presence of IL-6. Hy-IL-6, a fusion protein of IL-6 and the sIL-6R, was used as positive control. (C) As in (B) with sIL-6R generated by meprin β. (D) Same experimental set-up as in (B), additionally treated with Tocilizumab to block IL-6 signaling. (E) Scatter blott of a human blood sample analyzed by flow cytometry, demonstrating separation of granulocyte, monocyte and lymphocyte populations. (E) Meprin β specific antibody revealed expression on granulocytes, using pre-immune serum as control. (F) Eleven human blood samples from healthy donors were analyzed for meprin β and IL-6R levels on granulocytes. After linear regression a negative correlation between both proteins became evident. (G) On human T-cells there was no signal detected for meprin β, nevertheless there were differences in the amount of IL-6R. These are most likely due to ADAM10 cleavage. (H) As (F) but T-cells were analyzed. The linear regression showed no correlation.

Figure 5. Regulation of IL-6R shedding by meprins and ADAMs.

Soluble meprin α and membrane bound meprin β can generate sIL-6R. The cleavage however occurs at distinct sites when compared to ADAM10/17 mediated shedding. The sIL-6R can in all cases bind to IL-6 and gp130 on another cell and induce trans-signaling. Shedding of meprin β by ADAM10/17 might be a regulatory step to prevent generation of sIL-6R by membrane bound meprin β.