Neutrophil Protease Cleavage of Von Willebrand Factor in Glomeruli - An Anti-thrombotic Mechanism in the Kidney

Abstract

Adequate cleavage of von Willebrand factor (VWF) prevents formation of thrombi. ADAMTS13 is the main VWF-cleaving protease and its deficiency results in development of thrombotic microangiopathy. Besides ADAMTS13 other proteases may also possess VWF-cleaving activity, but their physiological importance in preventing thrombus formation is unknown. This study investigated if, and which, proteases could cleave VWF in the glomerulus. The content of the glomerular basement membrane (GBM) was studied as a reflection of processes occurring in the subendothelial glomerular space. VWF was incubated with human GBMs and VWF cleavage was assessed by multimer structure analysis, immunoblotting and mass spectrometry. VWF was cleaved into the smallest multimers by the GBM, which contained ADAMTS13 as well as neutrophil proteases, elastase, proteinase 3 (PR3), cathepsin-G and matrix-metalloproteinase 9. The most potent components of the GBM capable of VWF cleavage were in the serine protease or metalloprotease category, but not ADAMTS13. Neutralization of neutrophil serine proteases inhibited GBM-mediated VWF-cleaving activity, demonstrating a marked contribution of elastase and/or PR3. VWF-platelet strings formed on the surface of primary glomerular endothelial cells, in a perfusion system, were cleaved by both elastase and the GBM, a process blocked by elastase inhibitor. Ultramorphological studies of the human kidney demonstrated neutrophils releasing elastase into the GBM. Neutrophil proteases may contribute to VWF cleavage within the subendothelium, adjacent to the GBM, and thus regulate thrombus size. This anti-thrombotic mechanism would protect the normal kidney during inflammation and could also explain why most patients with ADAMTS13 deficiency do not develop severe kidney failure.

Keywords: ADAMTS13; Elastase; Glomerular basement membrane; Kidney; Neutrophils; Von Willebrand factor.

Fig. 1

VWF cleavage activity in the GBM. (a) Immunoblotting exhibited the presence of endogenous VWF in the GBM. Purified VWF was used as the positive control and showed a band representing full-length VWF (FL VWF) depicted by an arrow (lane 1). The GBM sample (GBM-I diluted 1:2) showed cleavage fragments of VWF at approximately 170 kDa and 140 kDa. Reproducible results were obtained from four separate experiments. (b) VWF cleavage by the GBM was investigated by VWF multimer structure analysis. The negative control buffer incubated with exogenous VWF showed VWF multimers (lane 1) whereas, GBM-I incubated with exogenous VWF resulted in complete cleavage to VWF dimers (lane 2). Reproducible results were obtained from five separate experiments.

Fig. 2

VWF cleavage by proteases in the GBM detected by multimer structure analysis. (a) The negative control buffer showed VWF multimers (lane 1) and addition of rADAMST13 exhibited cleavage of VWF and the appearance of smaller dimers (lane 2). The cleavage activity of rADAMTS13 was specifically inhibited by either pre-incubation with anti-ADAMTS13 antibody (lane 3) or EDTA (lane 4). GBM-I, without added exogenous VWF, did not exhibit any multimers at a 1:50 dilution (lane 5), whereas incubation with exogenous VWF resulted in complete cleavage to VWF dimers (lane 6). This cleavage activity was not inhibited by pre-incubation with anti-ADAMTS13 antibody (lane 7) but was slightly inhibited by pre-incubation with EDTA (lane 8). This inhibition was more pronounced in the presence of the protease inhibitor cocktail (PI, lane 9). Reproducible results were also obtained from altogether five experiments with GBM-I, two experiments with GBM-II and two experiments with GBM-III (not all inhibitors were tested in each setting, see text). (b) Exogenous VWF incubated with control buffer showed VWF multimers (lane 1) and addition of rADAMST13 exhibited cleavage of VWF (lane 2). GBM-III incubated with exogenous VWF resulted in cleavage of VWF (lane 3). This cleavage activity was partially inhibited by pre-incubation with elastase inhibitor (lane 4), anti-PR3 antibody (lane 5) or cathepsin G inhibitor (lane 6). Reproducible results were obtained from three experiments with GBM-III and with GBM-II, respectively.

Fig. 3

VWF cleavage induced by neutrophil proteases in the GBM detected by immunoblotting. Immunoblotting was performed to investigate VWF cleavage by the GBM. The negative control buffer alone showed full-length VWF (FL VWF) along with a weak cleavage product at 176 kDa (lane 1). rADAMTS13 induced VWF cleavage by exhibiting the two cleavage products at 176 kDa and 140 kDa (lane 2). GBM-III incubated with exogenous VWF showed VWF cleavage demonstrated by the appearance of two cleavage products at 176 kDa and 140 kDa (lane 3). Pre-incubation with elastase inhibitor or anti-PR3 antibody exhibited inhibition with a weak band at 176 kDa (lanes 4 and 5, respectively). Pre-incubation with the cathepsin G inhibitor also showed similar but lesser inhibition (lane 6). Reproducible results were obtained from four experiments with GBM-II and two experiments with GBM-III.

Fig. 4

Cleavage of FRETS-VWF73 detected by mass spectrometry. MALDI-TOF mass spectrometry was performed to analyze VWF cleavage activity of GBM samples using FRETS-VWF73. A. FRETS-VWF73 incubated with buffer alone served as the negative control showing a peak at molecular mass 8314 Da. B. FRETS-VWF73 incubated with rADAMTS13 exhibited a peak at 7027 Da. C. FRETS-VWF73 incubated with neutrophil elastase exhibited a fragment peak at 6794 Da. D. FRETS-VWF73 incubated with neutrophil elastase and pre-incubated with the elastase inhibitor abolished elastase-induced cleavage. E. GBM-III incubated with FRETS-VWF73 generated two major cleavage fragments corresponding to 6795 Da (elastase/PR3-cleavage) and 6895 Da (MMP9-cleavage) and a minor fragment at 7026 Da (cathepsin G/ADAMTS13-cleavage). F. GBM-III as in (E) pre-incubated with the elastase inhibitor markedly abolished the cleavage with a minor fragment remaining at 6895 Da. G. GBM-III pre-incubated with EDTA abolished MMP9-induced cleavage but not that induced either by elastase/PR3 (6793 Da) or by cathepsin G (7014 Da). Reproducible results were obtained from both GBM-II and GBM-III from three separate experiments.

Fig. 5

GBM cleaves VWF on the surface of the glomerular endothelial cell. VWF-platelet strings were visualized by perfusion of platelets over primary glomerular endothelial cells (PGECs). A. Perfusion of PPP from a TTP patient (Patient 9 in (Tati et al., 2013)) combined with normal washed platelets onto histamine-stimulated PGECs led to the formation of VWF-platelet strings (see arrow in left panel). Neutrophil elastase perfused onto the cells cleared the strings at the same locations within 5 min (right panel). B. Perfusion of PPP from the same TTP patient over PGECs as in A (arrow in left panel indicates VWF-platelet strings) and rADAMTS13 cleared the VWF-platelet strings (right panel). C. Normal washed platelets in phosphate buffer saline (PBS, without plasma) generated VWF-platelet strings on the cell surface (arrow in left panel) and were cleared upon addition of neutrophil elastase (right panel). D–F. Normal washed platelets in PBS (without plasma) generated VWF-platelet strings on the cell surface (arrows in left panels), same conditions as in C. Upon pre-incubation with elastase inhibitor IV, cleavage activity of neutrophil elastase was abolished and strings were visible (arrow in right panel D). Similarly strings were cleared upon perfusion of GBM-III (right panel E) and pre-incubation with the elastase inhibitor abolished the cleavage (arrow in right panel F). Images are representative of three separate experiments with reproducible results from GBM-II and GBM-III. All images were acquired using a Carl Zeiss Axiovert 40CFL microscope (Carl Zeiss, Jena, Germany) equipped with a digital camera DP 450 (Deltapix, Maalov, Denmark), using a Plan Apochromat 20 × objective.

Fig. 6

Electron microscopy to detect elastase in renal tissue. The presence of elastase in the glomerulus was examined by electron microscopy. A. Ultramorphology showing an overview from a glomerulus: neutrophil (N) and glomerular basement membrane (GBM). Boxed areas are enlarged in panels B and C. B and C. Renal cortex, at higher magnification, labeled with anti-elastase (10 nm, arrows) and with anti-CD66 (20 nm, arrowheads) to label neutrophils. D. Ultramorphology showing an overview from a glomerulus: neutrophil (N) and glomerular basement membrane (GBM). Boxed areas are enlarged in panels E and F. E and F. Control antibodies showed weak or no signal in the renal tissue. Samples were examined in a JEOL JEM 1230 transmission electron microscope (JEOL, Peabody, Mass., USA) at 60 kV accelerating voltage. Images were recorded with a Gatan Multiscan 791 CCD camera using DigitalMicrograph™ software. Scale bar: 5 μm (A and D) and 500 nm (B, C, E and F).