Thymic Stromal Lymphopoietin (TSLP) Is Cleaved by Human Mast Cell Tryptase and Chymase

Abstract

Thymic stromal lymphopoietin (TSLP), mainly expressed by epithelial cells, plays a central role in asthma. In humans, TSLP exists in two variants: the long form TSLP (lfTSLP) and a shorter TSLP isoform (sfTSLP). Macrophages (HLMs) and mast cells (HLMCs) are in close proximity in the human lung and play key roles in asthma. We evaluated the early proteolytic effects of tryptase and chymase released by HLMCs on TSLP by mass spectrometry. We also investigated whether TSLP and its fragments generated by these enzymes induce angiogenic factor release from HLMs. Mass spectrometry (MS) allowed the identification of TSLP cleavage sites caused by tryptase and chymase. Recombinant human TSLP treated with recombinant tryptase showed the production of 1-97 and 98-132 fragments. Recombinant chymase treatment of TSLP generated two peptides, 1-36 and 37-132. lfTSLP induced the release of VEGF-A, the most potent angiogenic factor, from HLMs. By contrast, the four TSLP fragments generated by tryptase and chymase failed to activate HLMs. Long-term TSLP incubation with furin generated two peptides devoid of activating property on HLMs. These results unveil an intricate interplay between mast cell-derived proteases and TSLP. These findings have potential relevance in understanding novel aspects of asthma pathobiology.

Keywords: TSLP; VEGF-A; airway remodeling; asthma; chymase; epithelial cells; macrophage; mast cell; tryptase.

Figure 1

(A) Recombinant human lfTSLP (5 μg) was treated with recombinant human tryptase (0.5 μg) at 37 °C in the presence of heparin (1:10). Samples were withdrawn at 0, 30, 60, 120, and 240 min, and inactivated by heating for 10 min at 99 °C to stop the cleavage reaction. Each digestion mixture corresponding to 1 μg of protein was separated on 16% Tris-Tricine gel. The gel was stained with a colloidal Coomassie Brilliant Blue solution. (B) The reduction of the band intensity at ~15 kDa was quantified by densitometric analysis. The results show the mean ± SD of three independent experiments.

Figure 2

Cleavage analysis of lfTSLP by chymase. (A) Recombinant human TSLP (5 μg) was treated with chymase (0.5 μg at 37 °C). 1 μg aliquots were withdrawn at 0, 30, 60, 120, and 240 min, inactivated by heating for 10 min at 99 °C to stop the cleavage reaction and separated on 16% Tris-Tricine gel. The gel was stained with colloidal Coomassie Brilliant Blue solution. (B) Densitometric analysis of the cleavage products of TSLP generated by chymase (as shown in panel A). The progressive and marked reduction in the band intensity at ~15 kDa, and the appearance of several smaller fragments, indicated that TSLP is a substrate for chymase. The results show the mean ± SD of 3 independent experiments.

Figure 3

MALDI-MS analysis of TSLP following incubation with tryptase under strictly controlled conditions (E:S 1:1000 for 30 min at 37 °C). The signals marked with an asterisk correspond to the mono and doubly charged ions of the intact protein. Peaks at m/z 4361.19 and m/z 10,713.19 were assigned to the complementary peptides 98-132 and 1-97, respectively (marked A and B in the figure) originating from a single proteolytic cleavage between the peptide bond Met97-Lys98. The amino acid sequences of the two peptides (A and B) are shown in the inset and are underlined in red (A) or in green (B) in the upper panel of the figure. The tryptase preferential cleavage site is shown in red.

Figure 4

MALDI-MS analysis of TSLP following incubation with chymase under strictly controlled conditions (E:S 1:100 for 30 min at 37 °C). The signals marked with an asterisk correspond to the mono and doubly charged ions of the intact protein. Peaks at m/z 4164.65 and 10,916.86 were assigned to the complementary peptides 1-36 and 37-132, respectively (marked A and B in the figure) originating from a single proteolytic cleavage at Phe36. The amino acid sequences of the two peptides (A and B) are shown in the inset and are underlined in red (A) or in green (B) in the upper panel of the figure. The chymase preferential cleavage site between Phe36-Asn37 is shown in red. All other peaks in the spectrum were identified as sub-digestion products. Other fragments were observed at lower m/z, indicating further proteolytic cleavage of the two main fragments 1-36 and 37-132.

Figure 5

Tryptase and chymase preferential cleavage sites in the three-dimensional (3D) structure of TSLP. Upper panel: amino acid sequence of the mature form of TSLP. The tryptase and chymase specific proteolytic sites (Met97 and Phe36) are highlighted in red and green, respectively. Lower panel: ribbon representation of TSLP 3D structure according to Verstraete et al. [86]. The tryptase and chymase preferential cleavage sites are located within the CD and AB loops, highlighted in red and green, respectively. This representation provides insights into the spatial arrangement of the proteolytic sites within the mature form of TSLP.

Figure 6

Effects of lfTSLP and TSLP fragments generated by tryptase and chymase on mediator release from human lung macrophages (HLMs). (A) Effects of lfTSLP (30 ng/mL) (red bar) and of increasing concentrations of the two TSLP fragments generated by tryptase (TSLP_1-97 and TSLP_98-132) (blue bars) and (B) of the two fragments generated by chymase (TSLP_1-36 and TSLP_37-132) (blue bars) on VEGF-A release from HLMs. The results show the mean ± SD of eight independent experiments performed with highly purified (≥99%) HLMs from different donors. ** p < 0.01 compared to control (CTR).

Figure 7

Cleavage analysis of TSLP by tryptase and PCSK3. Recombinant human non-glycosylated TSLP (2 μg) was incubated with tryptase (0.2 μg at 37 °C) for 1 h or with PCSK3 (0.88 μg at 37 °C) for 24 h at 37 °C. Aliquots were inactivated by heating for 10 min at 99 °C to stop the cleavage reaction and separated on 16.5% Tris-Tricine gel. The gel was stained with a colloidal Coomassie Brilliant Blue solution.

Figure 8

Effects of TSLP cleavage products generated by PCSK3 and tryptase on the release of VEGF-A from human lung macrophages (HLMs). Recombinant human non-glycosylated TSLP (2 μg) was incubated with tryptase (0.2 μg at 37 °C) for 1 h or with PCSK3 (0.88 μg at 37 °C) for 24 h at 37 °C. At the end of the incubation, aliquots of untreated TSLP, tryptase-treated TSLP, and PCSK3-treated TSLP were incubated (18 h, 37 °C) with HLMs in triplicate. At the end of the incubation, the supernatants were collected and VEGF-A concentrations were evaluated by ELISA. The results show the mean ± SD of a typical experiment out of three. ** p < 0.01.