Mannose-Binding Lectin Promotes Local Microvascular Thrombosis After Transient Brain Ischemia in Mice
Background and Purpose—Several lines of evidence support the involvement of mannose-binding lectin (MBL) in stroke brain damage. The lectin pathway of the complement system facilitates thrombin activation and clot formation under certain experimental conditions. In the present study, we examine whether MBL promotes thrombosis after ischemia/reperfusion and influences the course and prognosis of ischemic stroke.
Methods—Middle cerebral artery occlusion/reperfusion was performed in MBL-deficient (n=85) and wild-type (WT; n=83) mice, and the brain lesion was assessed by MRI at days 1 and 7. Relative cerebral blood flow was monitored up to 6 hours after middle cerebral artery occlusion with laser speckle contrast imaging. Fibrin(ogen) was analyzed in the brain vasculature and plasma, and the effects of thrombin inhibitor argatroban were evaluated to assess the role of MBL in thrombin activation.
Results—Infarct volumes and neurological deficits were smaller in MBL knockout mice than in WT mice. Relative cerebral blood flow values during middle cerebral artery occlusion and at reperfusion were similar in both groups, but decreased during the next 6 hours in the WT group only. Also, the WT mice showed more fibrin(ogen) in brain vessels and a better outcome after argatroban treatment. In contrast, argatroban did not improve the outcome in MBL knockout mice.
Conclusions—MBL promotes brain damage and functional impairment after brain ischemia/reperfusion in mice. These effects are secondary to intravascular thrombosis and impaired relative cerebral blood flow during reperfusion. Argatroban protects WT mice, but not MBL knockout mice, emphasizing a role of MBL in local thrombus formation in acute ischemia/reperfusion.
- complement pathway, mannose-binding lectin
- complement system proteins
- infarction, middle cerebral artery
The complement system is activated in stroke1 and is regarded as a detrimental factor contributing to worsen the outcome.2,3 The complement system is not only a potent ancestral innate immune defense against invading microbes, but it is also an efficient tool promoting clearance of damaged host cells.4 Complement can be activated through the classical, alternative, and lectin pathways that all culminate in the formation of a lytic pore in cell membranes, known as the membrane attack complex, and the production of proinflammatory small protein fragments called anaphylotoxins C3a, C5a, and C4a.5 Blockade of C3a6–8 and C5a8,9 receptors is protective against ischemia/reperfusion in animals.
The classical pathway involves IgM or IgG1 antibodies attaching to the C1q complex, which is expressed in ischemic neurons.10 C1-inhibitor, a naturally occurring acute phase protein, is protective against ischemia/reperfusion in experimental models.11–14 C1-inhibitor not only blocks the classical pathway but also prevents the activation of the lectin pathway, among other effects.15 Notably, human recombinant C1-inhibitor selectively binds to mannose-binding lectin (MBL), and this mechanism is thought to mediate its protective effects in brain ischemia/reperfusion.16 Several lines of experimental17–19 and clinical17,20 evidence support that the MBL pathway exacerbates stroke brain damage. MBL knockout (KO) mice showed reduced brain damage after ischemia, and protection was lost after the administration of recombinant MBL.17
The complement and the coagulation systems crossinteract at several molecular steps.21,22 Proteins of the lectin pathway can induce thrombus formation through thrombin activation,23–27 thus exacerbating brain damage after ischemia/reperfusion.28–30 In this study, we evaluated the role of MBL in thrombus formation and whether this interaction is involved in brain damage after ischemia/reperfusion in mice.
An extended version of methods can be found in the online-only Data Supplement.
Animal work was undertaken with approval of the Ethical Committee of the University of Barcelona and in compliance with the Spanish law. Adult (3–4-month-old), male wild-type (WT) and MBL KO mice31 in the C57BL/6J background were used.17
Ischemia was produced by intraluminal 90-minute occlusion of the right middle cerebral artery (MCAO) in WT (n=83) and MBL KO (n=85) mice. Cerebral perfusion was assessed with laser Doppler flowmetry (Perimed AB, Järfälla, Sweden). A neurological score ranging from 0 (no deficit) to 20 (maximal deficit) was performed at days 1 and 7 after MCAO.
MRI was performed in a 7.0-T BioSpec animal scanner (Bruker BioSpin, Ettlingen, Germany). T2 relaxometry maps were acquired at day 1 or at days 1 and 7, and images were analyzed blinded to genotype.
An osmotic pump containing argatroban (#PI-146-0050; Enzo Life Sciences; 5 μg/kg per minute for 24 hours) or vehicle (saline containing DMSO) was implanted intraperitoneally before 90-minute MCAO. One-millimeter-thick brain sections obtained at 24 hours were stained with 1% 2,3,5-triphenyltetrazolium chloride (TTC). Infarct volume with edema correction was measured in a blinded fashion.
Optical Imaging of Cerebral Blood Flow
Relative cerebral blood flow (rCBF) was studied by laser speckle flowmetry (LSF) in ischemic and sham-operated mice. A custom LSF system allowed measuring rCBF in regions of interest (ROIs): ischemic core (MCA ROI), surrounding region (outside ROI), and the corresponding ROIs in the contralateral hemisphere. The initial CBF drop after MCAO was assessed with laser Doppler, whereas LSF was used to monitor CBF during the last 30 minutes of MCAO, for 1 hour after reperfusion, and at 4 and 6 hours. The images were averaged over 5 minutes at the 60-minute MCAO mark, at reperfusion, and 1, 4, and 6 hours later. rCBF was calculated relative to LSF baseline (15 minutes) measured before MCAO. Brain vascular anatomy was examined ex vivo after intravascular perfusion of carbon black ink in control WT and MBL KO mice.
Plasma fibrinogen concentration (collected in citrate) was analyzed by ELISA (#ab108844; Abcam, Cambridge, United Kingdom).
Mice were intracardially perfused with saline followed by paraformaldehyde 24 hours after MCAO. Immunofluorescence was performed with FITC-conjugated fibrinogen antibody (#F0111; Dakocytomation, Glostrup, Denmark; 1:100) and antiglucose transporter-1 (Glut-1) antibody (#AB1340; Merck Millipore, Billerica, MA; 1:200). Fibrin(ogen)-positive blood vessels were counted in a blinded fashion.
Mice were intracardially perfused with saline 24 hours after MCAO, and ipsilateral and contralateral tissue was dissected out. Cortex and striatum were obtained separately (n=6 per genotype and region). An antibody against native fibrinogen (#PAB13593; Abnova Corporation, Taipei, Taiwan) diluted 1:1000 was used. Antiactin antibody (#A1978; Sigma-Aldrich Química, Madrid, Spain) diluted 1:500 000 was the loading control. The ratio between fibrinogen and actin band intensities was calculated.
Differences between genotypes were assessed with Student t test or Mann–Whitney test after testing for normality (Kolmogorov–Smirnov test). Two-way ANOVA was used for comparisons by genotype and treatment, brain hemisphere, or time.
General Information About Animals Used in This Study
Ischemia was induced in 168 mice (n=83 WT; n=85 MBL KO). Five mice (3 WT and 2 MBL KO) did not enter the study because of surgical complications. Nine mice (3 WT and 6 MBL KO) were excluded because of insufficient drop in cortical perfusion (<65% of baseline) after MCAO. Ten mice (5 WT and 5 MBL KO) did not develop infarction and were excluded. Mortality within the first 24 hours was 6.0% and 11.8% in WT (n=5) and MBL KO (n=10) groups, respectively (χ2= 0.51).
MBL KO Mice Had Smaller Infarct Volumes and Neurological Deficits Compared With WT Mice
In a previous study, we found smaller infarct volumes in MBL KO than in WT mice, as assessed during postmortem with TTC.17 Here we aimed to validate these findings in larger cohorts of mice. MRI lesion volume was evaluated at 24 hours in WT (n=36) and MBL KO (n=36) mice (Figure 1A and 1B). The reduction (mean±SD) in cortical perfusion after MCAO, as assessed with laser Doppler flowmetry, was 82±10% in WT group and 81±9% in MBL KO group. The MRI lesion volume was significantly smaller (P<0.01; Student t test) in MBL KO mice (Figure 1C), which also showed better (lower) neurological scores compared with WT mice (P<0.001; Mann–Whitney test; Figure 1D).
To find out whether the early beneficial effect of MBL deficiency persisted at later time points, mice received a second MRI scan at day 7 (17 WT and 19 MBL KO mice; Figure 2A). Mortality increased similarly from day 1 to day 7 in MBL KO (37%) and WT (47%) groups (χ2=0.38). Infarct volumes at 24 hours were higher in mice that died before day 7 than in mice that survived the 7-day study period (WT group: 79.2±19.3 mm3, n=8, versus 49.3±14.7 mm3, n=9; P<0.001; MBL KO group: 68.3±11.9 mm3, n=7, versus 34.6±17.7 mm3, n=12; P<0.001). MBL-deficient mice again showed smaller infarct volumes and neurological deficits compared with the WT group (2-way ANOVA by genotype [P<0.05] and time [P<0.001]; Figure 2B and 2C). Post hoc analysis showed that the genotype effect on infarct volume was statistically significant at day 1 (P<0.05) but only showed a statistical trend (P=0.08) at day 7. The neurological scores were better in MBL KO mice than in WT mice at day 1 (P<0.001) and day 7 (P<0.05).
CBF at Reperfusion Was Improved in MBL KO Mice Versus WT Mice
We studied rCBF by laser Doppler and LSF in additional groups of ischemic (n=5 per genotype) and sham-operated (n=3 per genotype) mice. The laser Doppler measure (percentage of baseline) obtained during the first 30 minutes of MCAO was (mean±SD) 15.0±6.1% in WT group and 16.4±6.3% in MBL KO group (P=0.73). In the same mice, LSF (Figure 3A) was measured at different time points (Figure 3B). LSF measurement of rCBF (expressed as percentage of basal rCBF) in MCA ROI during MCAO was (mean±SD) 10.8±4.7% for WT group and 17.0±10.5% for MBL KO group (P=0.27). At reperfusion, rCBF recovered to a similar extent in MBL KO and WT mice (Figure 3C). However, as time progressed, rCBF decreased in WT group but not in MBL KO group (Figure 3C). Two-way ANOVA by genotype and time showed a significant genotype effect (P<0.05). In contrast, rCBF in the outside ROI of the ipsilateral hemisphere and in the corresponding (mirror) contralateral ROIs (Figure 3C) was similar for both genotypes. Nevertheless, a tendency toward contralateral CBF reduction was observed after ischemia in WT mice. This has been previously described using positron emission tomography and single photon emission CT32 and might be related to decreases in contralateral activity because of depressed transcallosal neural connections. In sham-operated mice, rCBF values in these ROIs did not differ between WT and KO mice and were stable over time (Figure 3D). To discard overt genotype differences in cerebrovascular anatomy, carbon black ink was perfused intravascularly in control mice of both genotypes (n=3 per group). Visual inspection of postmortem brain and quantification of anastomoses between the MCA and the anterior carotid artery did not show genotype differences (Figure I in the online-only Data Supplement).
MBL Deficiency Reduced Deposition of Fibrin(ogen) in Blood Vessels of Ischemic Tissue
The presence of fibrinogen increased in ischemic tissue of WT mice at 24 hours, as assessed by Western blotting (Figure 4A). This effect was significantly attenuated in cortex (P<0.01; Figure 4A) and striatum (P<0.05; not shown) of MBL KO mice (n=6 per genotype and region). Also, MBL KO mice (n=6) showed less fibrin(ogen) immunopositive vessels after ischemia (Figure 4B; P<0.05) compared with WT mice (n=6) in cortex (not shown) and striatum (Figure 4B and 4C). However, plasma fibrinogen increased after ischemia in both genotypes (Figure 4D).
Argatroban Reduced Infarct Volume and Ameliorated Neurological Deficit in WT Mice But Not in MBL KO Mice
Because the lectin pathway can induce thrombin activation,24,33 we examined whether thrombin was involved in the detrimental effects of MBL by treating the mice with the direct thrombin inhibitor argatroban and measuring infarct volume at 24 hours (Figure 5A and 5B). After MCAO, the reduction in cortical perfusion (mean±SD) as assessed by laser Doppler was 79±10% in the vehicle/WT group, 85±6% in the argatroban/WT group, 87±6% in the vehicle/MBL KO group, and 81±8% in the argatroban/MBL KO group (P=0.15). The mice treated with argatroban (n=11) showed smaller infarct volumes and better neurological scores (P<0.05) compared with the mice receiving the vehicle (n=10; Figure 5C and 5D), in agreement with previous reports.28,29 In contrast, argatroban was not beneficial in MBL KO mice (n=8 per treatment group; Figure 5C and 5D). This finding suggested that the MBL pathway promoted local thrombin activation during reperfusion.
The present study reports that MBL exacerbates acute ischemic damage by promoting local prothrombotic events in the brain vasculature hampering blood flow reperfusion. In WT mice, rCBF recovery at reperfusion was followed by progressive rCBF reductions and fibrin(ogen) deposition in brain vasculature. These effects might be attributable to secondary thrombotic events that can occur after local activation of the coagulation cascade during reperfusion.34 Regardless of similar increments in circulating levels of fibrinogen after acute stroke in WT and KO mice, we observed less local vessel fibrin(ogen) in ischemic tissue of MBL KO mice. Moreover, a significant reduction in infarct volume and greater improvement in functional outcome were seen after administering the thrombin inhibitor argatroban in WT mice only. This finding further attested to the clinical relevance of the local crosstalk between the lectin pathway of complement activation and thrombin-dependent secondary thrombosis in ischemia/reperfusion injury in mice.
Several molecular interactions between the coagulation and complement cascades illustrate the complex crosstalk between these protease systems.21 Proteins of the lectin pathway can activate the coagulation cascade, as shown in vitro23–26 and in vivo.27 MBL has no enzymatic activity and needs to bind the MBL-associated serine proteases (MASP). MBL binding to MASP-2 activates the complement by cleavage of C4 and C2.35 MASP-2 can cleave prothrombin and form thrombin,24 and MASP-1 promotes the formation of crosslinked fibrin.36 Also, the activation of MASPs can induce the formation of fibrin clots,25 and human MBL–MASPs complex can mimic thrombin and initiate coagulation.33 The procoagulant effects of lectin pathway proteins may contribute to eliminating invading pathogens by sequestering them though local activation of the coagulation cascade preventing dissemination throughout the organism.36 However, local prothrombotic events in injured brain vasculature may impair reperfusion and enhance brain damage after acute ischemic stroke.
Thrombin activates platelets, converts fibrinogen to a fibrin mesh, and cleaves factor XIII as well as performs other important functions in coagulation.37 In brain ischemia, thrombin mediates severe vascular disruption and damage to the neurovascular unit.29,30 Accordingly, the thrombin inhibitor argatroban reduces brain injury after ischemia/reperfusion in experimental animals28–30 and extends the therapeutic window of recombinant tissue-type plasminogen activator.28 In humans, a pilot study of the combined administration of argatroban and intravenous recombinant tissue-type plasminogen activator showed a greater recanalization rate compared with recombinant tissue-type plasminogen activator alone in stroke patients with proximal intracranial arterial occlusions.38 In the present study, argatroban protected WT mice, but not MBL KO mice, suggesting that the deleterious effect of MBL were mediated, at least in part, by thrombin activation.
Long-lasting protection was reported after pharmacologically targeting the MBL pathway,19 and genetic MBL deficiency in humans was associated with a better stroke outcome.17,20 However, in a previous study, the benefits of mouse MBL deficiency in acute stroke were not sustained at day 7.39 Here, the protection of MBL deficiency, as assessed by longitudinal MRI, was still seen at 7 days, but group differences in infarct volume were attenuated compared with those at day 1. Several components of the complement system have regulatory functions in stem cells,40 and neural progenitors and immature neurons express receptors for complement fragments.41 Furthermore, complement proteins can facilitate regeneration in various models of central nervous system disease.42 Therefore, the possibility that MBL may play different roles during the recovery phase cannot be excluded.
The findings suggest that MBL impairs cerebral circulation at reperfusion after MCAO by promoting local intravascular thrombotic events through a mechanism involving thrombin activation. Future clinical studies in acute stroke patients assessing the value of direct thrombin inhibitors might benefit from patient stratification by their MBL genotype.
We are indebted to the Image Platform of Institut d’Investigacions Biomèdiques August Pi i Sunyer for technical help.
Sources of Funding
Work supported by the Spanish Ministries of Health (PI10/01898 and PS09/00579) and Economy (SAF2011-30492) and Fundació Cellex Barcelona. X.d.l.R. had a PhD fellowship from the Spanish Ministry of Economy.
Guest Editor for this article was Costantino Iadecola, MD.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.113.004111/-/DC1.
- Received November 11, 2013.
- Revision received February 26, 2014.
- Accepted February 28, 2014.
- © 2014 American Heart Association, Inc.
- Klos A,
- Wende E,
- Wareham KJ,
- Monk PN
- Mocco J,
- Mack WJ,
- Ducruet AF,
- Sosunov SA,
- Sughrue ME,
- Hassid BG,
- et al
- Huang J,
- Kim LJ,
- Mealey R,
- Marsh HC Jr.,
- Zhang Y,
- Tenner AJ,
- et al
- Heydenreich N,
- Nolte MW,
- Göb E,
- Langhauser F,
- Hofmeister M,
- Kraft P,
- et al
- Orsini F,
- Villa P,
- Parrella S,
- Zangari R,
- Zanier ER,
- Gesuete R,
- et al
- La Bonte LR,
- Pavlov VI,
- Tan YS,
- Takahashi K,
- Takahashi M,
- Banda NK,
- et al
- Morris DC,
- Zhang L,
- Zhang ZG,
- Lu M,
- Berens KL,
- Brown PM,
- et al
- Chen B,
- Cheng Q,
- Yang K,
- Lyden PD
- Chen B,
- Friedman B,
- Whitney MA,
- Winkle JA,
- Lei IF,
- Olson ES,
- et al
- Takahashi K,
- Chang WC,
- Takahashi M,
- Pavlov V,
- Ishida Y,
- La Bonte L,
- et al
- Stoll G,
- Kleinschnitz C,
- Nieswandt B
- Barreto AD,
- Alexandrov AV,
- Lyden P,
- Lee J,
- Martin-Schild S,
- Shen L,
- et al
- Rahpeymai Y,
- Hietala MA,
- Wilhelmsson U,
- Fotheringham A,
- Davies I,
- Nilsson AK,
- et al