ADAMTS13 reduces VWF-mediated acute inflammation following focal cerebral ischemia in mice

Abstract

Background

ADAMTS13 cleaves hyperactive ultra-large von Willebrand factor (ULVWF) multimers into smaller and less active forms. It remains unknown whether VWF-mediated inflammatory processes may play a role in the enhanced brain injury due to ADAMTS13 deficiency.

Objective

We tested the hypothesis that the deleterious effect of ADAMTS13 deficiency on ischemic brain injury is mediated through VWF-dependent enhanced vascular inflammation.

Methods

Transient focal cerebral ischemia was induced by 60 minutes of occlusion of the right middle cerebral artery. Myeloperoxidase (MPO) activity and inflammatory cytokines in the infarcted region was evaluated 23 hours after reperfusion injury. Neutrophil infiltration within the infarct and surrounding areas was quantitated by immunohistochemistry.

Results

We report that ADAMTS13-deficient mice exhibited significantly enlarged infarct size, concordant with increased myeloperoxidase (MPO) activity, neutrophil infiltration, and expression of the pro-inflammatory cytokines interleukin-6 (IL-6) and tumor necrosis factor- α (TNF-α). In contrast, VWF-deficient mice exhibited significantly reduced MPO activity, neutrophil infiltration, and inflammatory cytokine induction, demonstrating a role for VWF in these inflammatory processes. Mice deficient for both ADAMTS13 and VWF exhibited an identical reduction of the same inflammatory parameters, demonstrating that the increased inflammation observed in ADAMTS13-deficient mice is VWF-dependent. Finally, the increased infarct size observed in ADAMTS13-deficient mice was completely abrogated by prior immunodepletion of neutrophils, demonstrating a causal role for acute inflammation in the enhanced brain injury that occurs in the setting of ADAMTS13 deficiency.

Conclusion

These findings provide new evidence for ADAMTS13 in reducing VWF-mediated acute cerebral inflammation following ischemic stroke.

Introduction

ADAMTS13 (A Disintegrin And Metalloprotease with Thrombospondin type I repeats-13) plays an important role in preventing microvascular thrombosis by cleaving ultra large von Willebrand factor (ULVWF) multimers, the most thrombogenic form of VWF, into smaller less active multimers, reducing potential thrombotic activity [1, 2]. VWF is a large multimeric glycoprotein that plays an important role in initial platelet adhesion to the injured vascular wall by binding to glycoprotein GPIbα on platelets [3]. Clinically, deficiency of VWF in humans causes the bleeding disorder von Willebrand disease (VWD). On the other hand, deficiency of ADAMTS13 causes thrombotic thrombocytopenic purpura (TTP), a thrombotic disorder characterized by fever, thrombocytopenia, microangiopathic hemolytic anemia, and varying degrees of renal and neurological dysfunction (extensively reviewed) [4].

Recently it has been shown that ADAMTS13 deficiency exacerbates brain injury in a murine model of acute ischemic stroke [5, 6], whereas VWF deficiency [6–8], GPIbα deficiency [9], and pharmacological blockade of VWF-GPIbα interaction [10] had the opposite effect. The protective effect of ADAMTS13 in this process was hypothesized to occur secondary to reduced platelet-dependent thrombotic potential and concomitant reduction of ischemia/reperfusion injury [5], Although it is quite evident that thrombosis plays an important role in stroke pathogenesis, studies from both animal models and human ischemic stroke have demonstrated that tissue damage is mediated in large part by a strong inflammatory response following reperfusion of cerebral vessels (extensively reviewed) [11, 12]. We and others have demonstrated that in addition to its role in hemostasis and thrombosis, the ADAMTS13-VWF axis also functions in the modulation of inflammatory responses. For example, ADAMTS13 deficiency in mice results in increased VWF-dependent leukocyte rolling, adhesion, and extravasation under acute inflammatory conditions [13]. Additionally, activated platelets bound to endothelial-associated VWF support leukocyte rolling both in vitro and in vivo [13, 14], and support neutrophil extravasation in a peritonitis model [15]. Recently, we have demonstrated that ADAMTS13 reduces vascular inflammation and the development of early atherosclerosis in mice [16]. Based on these findings regarding contributions of the VWF-ADAMTS13 axis to inflammatory pathways, we tested the hypothesis that VWF-dependent vascular inflammation contributes to the aggravated ischemic brain injury seen in setting of ADAMTS13 deficiency.

Methods

Animals

Adamts13−/− [17], Vwf−/− [18], Adamts13−/−/Vwf−/− [19] mice backcrossed >15 times to C57BL/6 were used for the experiments. Control mice were heterozygous littermates or age matched wild-type C57BL/6J mice (The Jackson Laboratory). All mice used were males between 8 and 10 weeks of age. The University of Iowa Animal Care and Use Committee approved all experiments.

Cerebral ischemia and reperfusion injury model

Transient focal cerebral ischemia was induced by 60 minutes of occlusion of the right middle cerebral artery with a 7.0 siliconized filament followed by 23 hours of reperfusion as described [20]. Briefly, mice were anesthetized with 1–1.5% isoflurane mixed with medical air. Body temperature was maintained at 37°C ± 1.0 using a heating pad. Laser Doppler flowmetry (Perimed instruments, Sweden) was used to monitor cerebral blood flow during 60 minutes of ischemia/30 minutes of reperfusion (Suppl. Table 1). Prior to sacrifice mice were evaluated for motor-deficits, which was scored on a four-point neurological scale as described [21]: 0, no observable neurological deficit (normal); 1, failure to extend left forepaw on lifting the whole body by tail (mild); 2, circling to the contralateral side but normal posture at rest (moderate); 3, leaning to the contralateral side at rest (severe); 4, no spontaneous motor activity. For morphometric measurement eight coronal serial sections were cut at 1-mm intervals from the frontal pole using a mouse Brain Matrix (Roboz surgical instrument). Coronal sections were stained with 2% triphenyl-2, 3, 4-tetrazolium-chloride (TTC) for 15 min at 37° C. Sections were scanned, digitalized and infarct areas were measured blindly using the Nikon NIS element software. To correct for brain swelling due to edema after ischemia the corrected total infarct volume (%) was calculated as follows: Corrected infarct volume (%) = {[volume of contralateral hemisphere- (volume of ipsilateral hemisphere-volume of infarct)]/ volume of contralateral hemisphere]} × 100.

Myeloperoxidase assay

Myeloperoxidase (MPO) activity was measured following 60 minutes of ischemia/23 hours of reperfusion injury. To measure the MPO activity ischemic and contralateral (control) brain hemispheres of the perfused brain were homogenized and extracted in 1% cetyltrimethylammonium bromide (Sigma-Aldrich) in 50 mM KPO₄ buffer, pH 7.0. The suspensions were sonicated for 30 s with 3 cycles of freeze-thaw in liquid nitrogen. Subsequently, the suspensions were centrifuged at 16,000 × g for 15 min and supernatant fractions were used for measurement of MPO activity and protein estimation. 100 μl of tetramethylbenzidine (TMB) solution (Sigma-Aldrich) was added to 40 μl of supernatant in a 96-well plate in duplicates. The reaction was stopped with 100 μl 1N HCl after 10 minutes and read at 450 nm. MPO activity was calculated using an MPO standard (Sigma-Aldrich) and the resultant activity was normalized as MPO units/mg of protein.

Immunohistochemical analysis

Immunostaining of neutrophils was done 23 hours following reperfusion injury on perfused brain and done as described [22]. Briefly, 9.0 μm coronal sections of fresh frozen tissue (perfused brain) were incubated with blocking reagent followed by primary antibody (rat anti-mouse neutrophil (NIMP-R14, Santa Cruz Biotech.) or rat Ig (control) in the presence of 5% goat serum overnight at 4°C followed by biotin-conjugated goat anti-rat Ig, avidin-biotin complex, and 3,3′-diaminobenzidine as substrate. Slides were counterstained with hematoxylin, dehydrated, and examined under a light microscope (Zeiss). Extravascular neutrophils were quantified by manually counting the immunoreactive cells in five predetermined cerebral regions (3 within the parietal cortex and 2 within the basal ganglia) of the ischemic hemisphere in 200× magnification field across different stereotactic levels. The mean fraction of neutrophils in the infarcted area was calculated from the ratio of immunoreactive cells per total number of cells in a predefined infarcted region. Eight coronal sections, separated by 100 μm, per mouse were analyzed from 4 mice of each genotype. In total 32 coronal sections from 4 mice/genotype were analyzed.

ELISA assay for TNF-α and IL-6

To prepare homogenates brain cortical tissue was collected from the infarcted and surrounding areas and lysed in tissue lysis buffer (20 mM Tris-Cl, pH 8.0, 10 mM NaCl, 2% Triton X-100). Supernatants from brain homogenates were used for determination TNF-α and IL-6 with commercially available mouse ELISA kits (R&D Systems) according to the manufacturer's instructions. Total protein in supernatant was calculated and the cytokine levels were normalized to pg/mg of protein.

Neutrophil depletion

Neutrophil depleting rat anti-mouse granulocyte receptor-1 (Gr-1) IgG2b (100 μg/mouse, clone RB6-8C5, BD Bioscience) was administered IV 24 hours prior to I/R injury. The RB6-8C5 clone binds to Gr-1, a glycosylphosphatidyl inositide-linked protein, which belongs to member of the Ly6 gene family. The antibody depletes neutrophils in both blood and spleen up to 48 hours after injection [23, 24]. Rat IgG 2b isotype was used as control.

Statistics

Results are reported as the mean ± SEM. Statistical comparisons were performed using Student 2-tailed unpaired t test or Analysis of variance followed by Boneferroni's multiple comparison test. Neuroscore scored on a four-point scale are reported as medium ± SD. Analysis of variance on ranks was applied to test for significant differences in the neuroscore. Treatment and genotype effects were analyzed by 2-way Analysis of variance followed by Holm-Sidak multiple comparison tests. Values of P < 0.05 were considered significant.

Results

ADAMTS13 deficiency results in an enhanced inflammatory response following ischemia/reperfusion (I/R) brain injury

We first compared infarct volume and neurological outcome in the Adamts13−/− and WT mice following 60 minutes of ischemia /23 hours of reperfusion injury. Similar to previous reports [5, 6], Adamts13−/− mice had significantly increased infarct volume (mean ± SEM: 33.2 % ± 1.7 %, P<0.01) in the ischemic brain hemisphere compared with WT mice (mean ± SEM: 21.8 % ± 2.1 %, Suppl. Fig. 1A–B). The increased infarct volume in the Adamts13−/− mice was associated with more severe neurological deficits compared with WT mice (P<0.01, Suppl. Fig. 1C). Laser Doppler measurements of cerebral blood flow were similar among groups during 60 minutes ischemia and 30 minutes of reperfusion (Suppl. Table 1).

To determine whether aggravated brain injury was associated with increased acute inflammation, we measured markers of inflammation in the infarcted and surrounded region of the perfused brain tissue following 60 minutes of ischemia /23 hours of reperfusion injury. We first measured MPO a key inflammatory enzyme secreted by activated neutrophils, and macrophages/microglial cells that can generate reactive oxygen species and is known to mediate I/R brain injury [22]. Adamts13−/− mice demonstrated significantly increased MPO activity compared with WT mice in the injured brain hemisphere, but not in the uninjured contralateral hemisphere, suggesting increased inflammatory cell infiltration into the damaged tissue (Fig. 1A). Next, in another set of experiments, we quantitated neutrophil infiltration in different cerebral regions of the injured hemisphere by immunohistochemistry. In concordance with the increased tissue MPO activity, neutrophil extravasation was significantly elevated in the Adamts13−/− mice compared with WT mice (Fig. 1B). Neutrophil extravasation was not observed in the control contralateral hemisphere (not shown). Next, we investigated the effect of ADAMTS13 deficiency on the pro-inflammatory cytokines TNF-α and IL-6. These cytokines has been shown to activate leukocytes and increases the expression of adhesion molecules on leukocyte, endothelial cells, and astrocytes [25], thereby play a role in I/R brain injury. Compared with WT mice, Adamts13−/− mice demonstrated significantly elevated levels of IL-6 and TNF-α in the supernatant fraction of brain homogenates prepared from the infarcted and surrounding areas (Fig. 1C), but not in the contralateral hemisphere (not shown). Together, these findings demonstrate that ADAMTS13 deficiency results in an enhanced acute inflammatory response following brain I/R injury, which may contribute to increased infarct size and neurological damage.

Fig. 1. ADAMTS13 deficiency promotes cerebral inflammation.

A. MPO activity from each genotype quantified by ELISA from the ipsilateral (infarct and surrounding region) or contralateral hemispheres (N = 7/group). B. The left panel shows representative coronal brain sections immunostained for extravasated neutrophils (neutrophils are stained brown and indicated by arrow), and counter stained with hematoxylin from one mouse of each genotype. The right panel shows quantification of neutrophils in the infarct region (N = 32 sections from 4 mice/genotype). Scale bar = 20 μm. C. Quantification of IL-6 and TNF-α by ELISA in the supernatant of perfused brain homogenates from the infarcted and surrounding area of each genotype (N =7 /group). Data are mean ± SEM. NS denotes non significant.

The enhanced inflammation post I/R injury associated with ADAMTS13 deficiency is VWF-dependent

VWF remains the only known substrate of ADAMTS13 in multiple experimental models [6, 13, 19]. We examined the role of VWF in promoting cerebral inflammation following 60 minutes of ischemia and 23 hours of reperfusion injury. Vwf−/− mice exhibited significantly reduced MPO activity (Fig. 2A) that was concordant with both significantly reduced neutrophil infiltration (Fig. 2B), and expression levels of the inflammatory cytokines IL-6 and TNF-α (Fig. 2C). Next, we investigated whether the enhanced inflammatory phenotype observed in Adamts13−/− mice following I/R injury is mediated by VWF. Interestingly, mice deficient for both ADAMTS13 and VWF exhibited an identical reduction of these same inflammatory parameters compared to Vwf−/− mice, demonstrating that the increased inflammation observed in Adamts13−/− mice is mediated by VWF (Fig. 2A–C). Together, these results demonstrate that VWF plays an important role in promoting brain tissue inflammation following I/R injury, and that this effect is modulated by ADAMTS13.

Fig. 2. ADAMTS13 reduces acute cerebral inflammation following I/R injury via a VWF-dependent pathway.

A. MPO activity from each genotype quantified by ELISA from the infarct and surrounding region (N = 6/group). B. The left panel shows representative coronal brain sections immunostained for extravasated neutrophils (neutrophils are stained brown and indicated by arrow), and counter stained with hematoxylin from one mouse of each genotype. The right panel shows quantification of neutrophils in the infarct region (N = 24 sections from 3 mice/genotype). Scale bar = 20 μm. C. Quantification of IL-6 and TNF-α by ELISA in the supernatant of perfused brain homogenates from the infarcted and surrounding area from each genotype (N =6 /group). Control WT mice used are same as Figure 1. Data are mean ± SEM. NS denotes non significant.

ADAMTS13 deficiency exacerbates post-I/R brain injury via a neutrophil-dependent mechanism

Neutrophil influx has been shown to correlate positively with increased ischemic damage in both human [26] and murine [22] studies. To test the hypothesis that ADAMTS13 deficiency aggravates brain injury via a neutrophil-dependent mechanism, we immunodepleted neutrophils from WT and Adamts13−/− mice 24 hours prior to I/R injury. In WT mice, neutrophil depletion resulted in a modest, but statistically significant reduction in infarct size (P<0.05) and improvement in neurological outcome, demonstrating that the brain tissue damage induced by I/R injury is at least partially dependent on neutrophils (Fig. 3A–C). Interestingly, neutrophil depletion in Adamts13−/− mice resulted in a reduction of infarct volume to the size seen in neutrophil-depleted WT mice, demonstrating that the effect of ADAMTS13 deficiency on exacerbation of ischemic brain injury is entirely neutrophil-dependent (Fig. 3B). Two-way Analysis of variance showed that the interaction of genotype and treatment with the neutrophil depleting antibody was significant (P = 0.014, F = 6.669). Neutrophil depletion also resulted in a similar abrogation of the effect of ADAMTS13 deficiency on neurological outcome (Fig. 3C). Together, these results suggest that neutrophil-dependent inflammatory pathways mediate the exacerbating effect of ADAMTS13 deficiency on tissue damage following ischemic stroke.

Fig. 3. Aggravated brain injury due to ADAMTS13 deficiency is neutrophil-dependent.

A. Representative 2% triphenyl-2, 3, 4-tetrazolium-chloride (TTC) stained coronal brain sections from one mouse of each genotype after 60 minutes of ischemia/23 hours of reperfusion injury, with and without neutrophil depletion. B. Corrected mean total infarct volumes of each genotype, with and without neutrophil depletion N = 9 –11/group). Data are mean ± SEM. C. Average neurological score of all mice from each group prior to sacrifice. Data are medium ± SD. Analysis of variance on ranks was used to test the significant difference (N = 10 to 11/group). Neutrophil-depleted Adamts13−/− mice demonstrated a two-fold reduction of infarct volume (P<0.0001) when compared with WT neutrophil-depleted mice. Two-way Analysis of variance showed that the interaction of genotype and treatment with the neutrophil depleting antibody was significant (P=0.014, F= 6.669).

Discussion

Using murine models of acute ischemic stroke, recently it has been shown that ADAMTS13 deficiency exacerbates brain injury, whereas VWF deficiency is protective [5–8]. Although not directly investigated, the mechanisms underlying these observations were thought to involve alterations in post-ischemic thrombotic potential; either decreased thrombosis in the case of VWF deficiency or enhanced post-ischemic thrombosis in the case of ADAMTS13 deficiency. In this study, we investigated an alternative mechanism by which ADAMTS13 deficiency could potentially aggravate brain injury, and focused on the hypothesis that this effect may be largely mediated via an inflammatory mechanism requiring VWF.

Using a genetic approach in mice, we first demonstrated that ADAMTS13-deficiency results in an enhanced inflammatory response following I/R brain injury. Two recent papers report confounding findings on the potential role of inflammation in ADAMTS13 deficiency-dependent exacerbated brain injury. In H&E stained sections, Zhao et al. did not find an increase in extra vascular neutrophils count in the peri-infarct regions of Adamts13−/− mice compared with WT mice after I/R brain injury [6]. In contrast, Fujioka et al. demonstrated increased inflammatory cells in the infarcted region of Adamts13−/− mice compared with WT mice after I/R brain injury; however these authors did not specifically stain for neutrophils and perform quantitative analysis [5]. The discrepancy between two studies remains unclear. We speculate that difference in results could be due to methodology used for counting neutrophils in predefined area. Zhao et al. counted neutrophils in peri-infarct region, whereas Fujioka et al. counted inflammatory cells in infarcted region. Herein, by immunohistochemistry, we demonstrate that ADAMTS13-deficiency in mice results in an increase neutrophils recruitment in the infarcted and peri-infarcted region after I/R brain injury. Our results support an important role for cerebral inflammation in mediating enhanced brain tissue damage observed in Adamts13−/− mice post I/R injury. Additionally, we show an increase in levels of MPO and pro-inflammatory cytokines (TNF-α and IL-6) in the infarcted and surrounding region of Adamts13−/− mice compared with WT mice. Our findings are in agreement with a recent study demonstrating that ADAMTS13-deficiency was associated with an increase in pro-inflammatory cytokine high-mobility group box1 and MPO activity in the infarcted region [27]. These findings are consistent with other studies in mice and humans, which have shown that inflammation plays a major role in promoting I/R brain injury [25].

Next, we found that VWF-deficiency abrogates the inflammatory response following I/R brain injury. Furthermore, we showed that the enhanced cerebral inflammation associated with ADAMTS13 deficiency is entirely VWF-dependent. There are several possibilities by which increased cerebral inflammation in Adamts13−/− mice following I/R injury could occur through a VWF-dependent mechanism. First, neutrophils have been shown to extravasate across inflamed endothelium via an interaction between endothelium bound VWF and platelets mediated via platelet GPIbα, [15, 28] and the VWF-GPIbα interaction has been shown to be critical for infarct development following I/R injury in murine model [10]. Second, neutrophils have been shown to roll on, and adhere to activated platelets bound to VWF on the surface of endothelial cells through an interaction mediated via neutrophil PSGL-1 and platelet P-selectin in vitro [14]. Activated platelets bound to endothelial associated VWF may release factors such as proinflammatory cytokines and other chemoattractants which generate signals for the recruitment and extravasation of neutrophils, resulting in exacerbation of brain injury. Third, neutrophils can also associate directly with VWF via an interaction mediated by both neutrophil PSGL-1 and integrin α_Mβ₂ independently of platelets [29]. Regarding these possibilities, we hypothesize that the increased cerebral inflammation observed in ADAMTS13-deficient mice is platelet GPIbα-dependent, as we previously have observed that increased baseline leukocyte rolling in Adamts13−/− mice is both platelet and VWF-dependent [13]. In line with this hypothesis, another recent study showed that VWF promotes neutrophil extravasation in a mouse model of peritoneal inflammation via a platelet GPIbα-dependent mechanism [15]. Future studies making use of specific tools to inhibit the VWF-GPIbα interaction without affecting platelet count will be useful to further define the role of platelet GPIbα in ADAMTS13 deficiency-dependent enhancement of acute inflammation.

Finally, we demonstrated that the aggravated I/R brain injury observed in ADAMTS13-deficient mice is completely abrogated by prior immunodepletion of neutrophils. Recently, a population-based study suggested that an elevated neutrophils count in circulation was associated with increased risk of cerebral infarction in humans [30]. Indeed the pivotal function played by neutrophils in stroke pathogenesis has been demonstrated to include: direct occlusion of cerebral microvessels, release of vasoconstrictive mediators, an increase of vascular permeability due to release of neutrophil proteases such as MMP-9, production of reactive oxygen species by myeloperoxidase and NADPH oxidase, release of other inflammatory mediators such as, platelet-activating factor [31–34]. Although our studies indicate that neutrophils exacerbates I/R brain injury in Adamts13−/− mice, it remains possible that other cells such as macrophages, astrocytes and T cells also play a role in this process.

Notably, anti-inflammatory strategies, including humanized monoclonal antibodies to ICAM-1 on activated endothelial cells or leukocyte CD18, recombinant neutrophil inhibitory factor have not shown beneficial outcome in clinical trials [35, 36]. We speculate that the failure of anti-inflammatory agents could be due to fact that other cells, such as, astrocytes, pericytes, T cells, and microglial cells contribute to inflammation following I/R brain injury. A better understanding on the role of these cells in ischemic brain injury would be required for future success of anti-inflammatory strategies in clinical trials.

In summary, we have demonstrated that ADAMTS13 deficiency exacerbates ischemic brain injury via a VWF-dependent inflammatory mechanism. Although severe ADAMTS13 deficiency clinically is only associated with TTP, these findings suggest that targeting VWF by recombinant ADAMTS13 may improve outcome in patients following acute ischemic stroke. Future studies using alternative methods to reduce ADAMTS13 activity (such as titration of anti-ADAMTS13 inhibitory antibodies) should help further elucidate whether the more modest reductions in ADAMTS13 activity observed commonly in other disease states also contribute to brain tissue inflammation and damage following acute ischemic stroke.

Supplementary Material

Suppl. Fig. 1. ADAMTS13 deficiency aggravates brain damage following I/R injury.

A. Representative 2% triphenyl-2, 3, 4-tetrazolium-chloride (TTC) stained coronal brain sections from one mouse of each genotype after 60 minutes of ischemia/23 hours of reperfusion injury. B. Corrected mean infarct volumes (%) of each genotype (N = 10–11/group). Data are mean ± SEM. C. Neurological score from each genotype before euthanasia sacrifice (N = 10–11/group). Data are medium ± SD. Analysis of variance on ranks was used to test the significant differences.

Suppl. Table 1

Measurements of cerebral Laser-Doppler blood flow (LDF) were similar among groups during 60 min ischemia and 15 min after reperfusion. LDF (%) during ischemia and reperfusion was calculated from the baseline, which was considered as 100%. Data are mean ± SEM. N= 5–6 mice/group.

Abbreviations

ULVWF

Ultra large von Willebrand factor

ADAMTS13

A Disintegrin-like And Metalloprotease with Thrombospondin type I repeats-13

MPO

Myeloperoxidase

IL-6

Interleukin 6

TNF-α

Tumor necrosis factor-α

I/R

Ischemia/Reperfusion